Nutrition and Gene Expression (PDFDrive)

Nutrition
and
Gene
Expression
Edited by
Carolyn D. Berdanier, Ph.D.

James L. Hargrove, Ph.D.
Department of Foods and Nutrition
University of Georgia
Athens, Georgia
Boca Raton London New York
CRC Press is an imprint of the

Taylor & Francis CRC
Group, anPress
informa business
Boca Raton Ann Arbor London Tokyo

First published 1993 by CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
Reissued 2018 by CRC Press
© 1993 by CRC Press, Inc.

CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been
made to publish reliable data and information, but the author and publisher cannot assume responsibility for the
validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the
copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to
publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let
us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or
utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including pho-
tocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission
from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://
www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA
01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of
users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has
been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for
identification and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
Nutrition and gene expression / editors, Carolyn D. Berdanier, James

L. Hargrove.
p. cm.
Includes bibliographical references and index.
ISBN 0-8493-6961-4
1. Nutrition. 2. Genetic regulation. 3. Gene expression.
I. Berdanier, Carolyn D. II. Hargrove, James L.
QP144.G45N87 1992
612.3’9—dc20 92-16664
A Library of Congress record exists under LC control number: 92016664
Publisher’s Note
The publisher has gone to great lengths to ensure the quality of this reprint but points out that some imperfections in
the original copies may be apparent.
Disclaimer
The publisher has made every effort to trace copyright holders and welcomes correspondence from those they have
been unable to contact.
ISBN 13: 978-1-315-89601-4 (hbk)

ISBN 13: 978-1-351-07511-4 (ebk)
Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the
CRC Press Web site at http://www.crcpress.com
PREFACE
The science of nutrition has changed radically since its inception. It has
evolved from the discovery of nutrients which cured well-described diseases
to a recognition that appropriate food choices can delay the development of
a variety of degenerative diseases. In the past, nutritionists worked hard to
devise recommendations to the public for nutrient intakes that would ensure
the absence of deficiency diseases. In the future, nutritionists will make
nutrient intake recommendations not on such characteristics as sex, height,
weight or age, but on the basis of the individual's phenotypic expression for
health while suppressing that individual's phenotypic expression for disease.
We are a long way from being able to make these recommendations. However,
as a first step we are beginning to recognize the interaction of specific nutrients
with the genetic code possessed by all nucleated cells. This book and its
sequels will provide the reader with some of the insights gained by foremost
scientists in the field of nutrient-gene interactions. Some of these interactions
have been researched in depth whereas others are in their infancy. As the
reader proceeds through the book, one hopes that the imagination and dedi-
cation of the scientists who have contributed to this volume will be appre-
ciated. To these authors who struggled to meet our stringent deadline, and
who wrote of their life's work, we give our thanks.
University of Georgia Carolyn D. Berdanier

Athens, Georgia James L. Hargrove
EDITORS
Carolyn D. Berdanier, Ph.D., is a Professor of Nutrition at The Uni-

versity of Georgia in Athens, Georgia. She received a B.S. degree from the
Pennsylvania State University and the M.S. and Ph.D. degrees from Rutgers
University in Nutrition in 1966. After a post-doctoral fellowship year with
Dr. Paul Griminger at Rutgers, she served as a Research Nutritionist with
the Human Nutrition Institute which is part of ARS, a unit of the U.S.
Department of Agriculture. In 1975 she moved to the University of Nebraska
College of Medicine where she continued her research in nutrient gene in-
teractions. In 1977 she moved to the University of Georgia where she served
as Head of the Department of Foods and Nutrition. She stepped down from
this post ten years later and devoted her full time efforts to research and
teaching in her research area. Her research on the diet and genetic components
of diabetes and vascular disease has been supported by NIH, USDA, U.S.
Department of Commerce, The National Livestock and Meat Board and the
Egg Board. She is a member of the American Institute of Nutrition, the
American Society for Clinical Nutrition, The Society for Experimental Bi-
ology and Medicine and several honorary societies in science. She has served
on the Editorial Boards of the FASEB Journal, The Journal of Nutrition,
Nutrition Research and Biochemistry Archives. She is also a Contributing
Editor for Nutrition Reviews and Editor of the AIN News Notes. Current
research interests include studies on aging, the role of diet in damage to
mitochondrial DNA, and the role of specific dietary ingredients in the sec-
ondary complications of diabetes.
James L. Hargrove, Ph.D., is an Associate Professor of Nutrition at

The University of Georgia in Athens, Georgia. He received a B.S. in Zoology
from the University of Washington in Seattle, Washington in 1969, and earned
a doctorate in Physiology from Utah State University in 1975. His interest
in nutritional and hormonal effects on synthesis of mRNA and enzymes was
stimulated by post-doctoral training with Daryl K. Granner, M.D., in the
Diabetes and Endocrinology Research Center at the University of Iowa. Dr.
Hargrove joined the Department of Anatomy and Cell Biology at Emory
University School of Medicine in 1982, and was promoted to Associate
Professor. He joined the Department of Foods and Nutrition at The University
of Georgia in 1989.
Dr. Hargrove's studies in enzyme regulation and the kinetics of gene
expression have been supported by grants from the N.I.H., the American
Diabetes Association, the American Cancer Society, and the National Kidney
Foundation. He is a member of the American Institute of Nutrition and the
American Society for Biochemistry and Molecular Biology. Current research
interests include computer-based modeling of gene expression and metabolic
control, and effects of dietary protein on organ growth and gene expression.
CONTRIBUTORS
Victoria E. Allgood, Ph.D. Charles 0. Brostrom, Ph.D.

Department of Cell Biology Department of Pharmacology
Baylor College of Medicine Robert Wood Johnson
Houston, Texas Medical School
Piscataway, New Jersey
John B. Allred, Ph.D.
Department of Food Science and Margaret A. Brostrom, Ph.D.
Technology Department of Pharmacology
The Ohio State University Robert Wood Johnson
Columbus, Ohio Medical School
Piscataway, New Jersey
Roderick A. Barke, Ph.D. Hans A. Buller, M.D., Ph.D.
Department of Food Science and
Department of Pediatrics
Nutrition and Surgery
Academic Medical Center
University of Minnesota
Amsterdam, The Netherlands
St. Paul, Minnesota
David K. C. Chan, Ph.D.
Carolyn D. Berdanier, Ph.D. Department of Foods and
Department of Foods and Nutrition
Nutrition The University of Georgia
The University of Georgia Athens, Georgia
Athens, Georgia Sylvia Christakos, Ph.D.
Department of Biochemistry and
Diana F. Bowers, Ph.D. Molecular Biology
Department of Food Science and University of Medicine and
Technology Dentistry of New Jersey
The Ohio State University New Jersey Medical School
Columbus, Ohio Newark, New Jersey
John A. Cidlowski, Ph.D.
Linda J. Brady, Ph.D. Professor
Department of Food Science and Departments of Physiology,
Nutrition Biochemistry and Biophysics
University of Minnesota University of North Carolina
St. Paul, Minnesota Chapel Hill, North Carolina
Steven D. Clarke, Ph.D.
Paul S. Brady, Ph.D. Professor
Department of Food Science and Department of Food Science and
Nutrition Human Nutrition
University of Minnesota Colorado State Univerity
St. Paul, Minnesota Fort Collins, Colorado
Robert J. Cousins, Ph.D. Richard J. Grand, Ph.D.
Center for Nutritional Sciences The Floating Hospital
University of Florida New England Medical Center
Gainesville, Florida Tufts University
School of Medicine
Nicholas 0. Davidson, M.D. Boston, Massachusetts
Associate Professor
Department of Medicine James L. Hargrove, Ph.D.
University of Chicago Associate Professor
Chicago, Illinois Department of Foods and
Nutrition
Jan Dekker, Ph.D. The University of Georgia
Department of Pediatrics Athens, Georgia
Amsterdam, The Netherlands Ann Jerkins, Ph.D.
Research Fellow
Department of Nutrition
R. S. Eisenstein, Ph.D.
Harvard School of Public Health
Department of Nutritional
Boston, Massachusetts
Sciences
College of Agriculture and Life Thomas L. Jetton, Ph.D.
Science Research Associate
University of Wisconsin Department of Molecular
Madison, Wisconsin Physiology and Biophysics
Vanderbilt, Tennessee
Nick Gekakis, B.S. Donald B. Jump, Ph.D.
Student Department of Physiology
Department of Nutrition Michigan State University
Harvard School of Public Health East Lansing, Michigan
Z. Kikinis, Ph.D.
Rajbir K. Gill, Ph.D. Human Nutrition Research Center
Department of Biochemistry and on Aging
Molecular Biology United States Department of
University of Medicine and Agriculture
Dentistry of New Jersey Boston, Massachusetts
New Jersey Medical School Rolf F. Kletzien, Ph.D.
Newark, New Jersey Metabolic Diseases Research
The Upjohn Company
Howard P. Glauert, Ph.D. Kalamazoo, Michigan
Associate Professor
Department of Nutrition and Food Gerald J. Lepar, Ph.D.
Science Department of Physiology
University of Kentucky Michigan State University
Lexington, Kentucky East Lansing, Michigan
Jane B. Lian, Ph.D. Tamio Noguchi, Ph.D.
Professor Assistant Professor
Department of Cell Biology Department of Nutrition and
University of Massachusetts Physiological Chemistry
Medical Center Suita, Osaka, Japan
Worcester, Massachusetts
Antoine Puigserver
Pauline Kay Lund, Ph.D. Center de Biochemie et de
Associate Professor Biologie Moleculaire
Department of Physiology and CNRS
Pediatrics Marseille, Cedex, France
University of North Carolina
Chapel Hill, North Carolina Edmond H. H. M. Rings, M.D.
Department of Pediatrics
Ormond A. MacDougald, Ph.D.
Department of Biological
Amsterdam, The Netherlands
Chemistry
Johns Hopkins University
Kenji Sakamoto, Ph.D.
Baltimore, Maryland
Department of Nutrition
Mark A. Magnuson, M.D. Harvard School of Public Health
Associate Professor Boston, Massachusetts
Department of Molecular
Physiology and Biophysics Richard I. Schwartz, Ph.D.
Vanderbilt University Division of Cell and Molecular
Nashville, Tennessee Biology
Lawrence Berkeley Laboratory
Robert K. Montgomery, Ph.D. Berkeley, California
The Floating Hospital
New England Medical Center Neil F. Shay, Ph.D.
Tufts University School of Center for Nutritional Sciences
Medicine University of Florida
Boston, Massachusetts Gainesville, Florida
Naima Moustaid, Ph.D. Cynthia Smas, M.S.

Department of Nutrition Graduate Student
Harvard School of Public Health Department of Nutrition
Boston, Massachusetts Harvard School of Public Health
H. M. Munro, Ph.D.
Human Nutrition Research Center Gary S. Stein, Ph.D.
on Aging Professor
United States Department of Department of Cell Biology
Agriculture University of Massachusetts
Boston, Massachusetts Worcester, Massachusetts
Hei Sook Sui, Ph.D. Douglas B. Tully, Ph.D.
Associate Professor Department of Physiology
Department of Nutrition University of North Carolina
Harvard School of Public Health Chapel Hill, North Carolina
J. W. Suttie, Ph.D. Malcolm Watford, Ph.D.

Departments of Biochemistry and Department of Nutritional
Nutritional Sciences Sciences
College of Agricultural and Life Cook College
Sciences Rutgers
University of Wisconsin-Madison New Brunswick, New Jersey
Madison, Wisconsin
Takehiko Tanaka, M.D., Ph.D.

Professor Catherine Wicker-Pianquart
Department of Nutrition and Centre de Biochimie et de
Physiological Chemistry Biologie Moleculaire
Osaka University Medical School CNRS
Suita, Osaka, Japan Marseille, Cedex, France
TABLE OF CONTENTS
Chapter 1
Nutrient Receptors and Gene Expression ................................. .
James L. Hargrove and Carolyn D. Berdanier
Chapter 2
Lactose Intolerance and Regulation of Small Intestinal
Lactase Activity .......................................................... 23
Robert K. Montgomery, Hans A. Buller, Edmond H. H. M. Rings,
Jan Dekker, and Richard J. Grand
Chapter 3
Regulation of Gastrointestinal Lipase Gene Expression by
Dietary Lipids ............................................................ 55
Catherine Wicker-Pianquart and Antoine Puigserver
Chapter 4
Nutritional Control of Gastrointestinal Hormone
Gene Expression ......................................................... 91
Pauline Kay Lund
Chapter 5
Calcium Homeostasis, Endoplasmic Reticular Function, and the
Regulation of mRNA Translation in Mammalian Cells .................. 117
Margaret A. Brostrom and Charles 0. Brostrom
Chapter 6
Tissue-Specific Regulation of Glucokinase .............................. 143
Mark A. Magnuson and Thomas L. Jetton
Chapter 7
Dietary and Hormonal Regulation of L-Type Pyruvate Kinase
Gene Expression ........................................................ 169
Tamio Noguchi and Takehiko Tanaka
Chapter 8
Glucose-6-Phosphate Dehydrogenase: Diet and Hormonal
Influences on De Novo Enzyme Synthesis .............................. 187
Rolf F. Kletzien and Carolyn D. Berdanier
Chapter 9
Nutritional and Honnonal Regulation of Genes Encoding Enzymes
Involved in Fat Synthesis ................................................ 207
Hei Sook Sui, Naima Moustaid, Kenji Sakamoto, Cynthia Smas,
Nick Gekakis, and Ann Jerkins
Chapter 10
Regulation of Hepatic Gene Expression by Dietary Fats: A Unique
Role for Polyunsaturated Fatty Acids .................................... 227
Steven D. Clarke and Donald B. Jump
Chapter 11
Dietary Fat, Gene Expression, and Carcinogenesis ...................... 247
Howard P. Glauert
Chapter 12
Regulation of Acetyl CoA Carboxylase and Gene Expression ........... 269
John B. Allred and Diana F. Bowers
Chapter 13
Apolipoprotein B mRNA Editing ........................................ 297
Nicholas 0. Davidson
Chapter 14
Regulation of the 68 kDa Hepatic Camitine
Palmitoyltransferase ..................................................... 319
PaulS. Brady, Roderick A. Barke, and Linda J. Brady
Chapter 15
Long-Term Regulation of Hepatic Glutaminase and the Urea
Cycle Enzymes .......................................................... 335
Malcolm Watford
Chapter 16
Effects of Dietary Protein on Gene Expression .......................... 353
David K.-C. Chan and James L. Hargrove
Chapter 17
Vitamin D-Dependent Calcium Binding Protein, Calbindin-D:
Regulation of Gene Expression .......................................... 377
Rajbir K. Gill and Sylvia Christakos
Chapter 18
Vitamin D Regulation of Osteoblast Growth and Differentation ......... 391
Jane B. Lian and Gary S. Stein
Chapter 19
Retinoic Acid Regulation of Gene Expression in Adipocytes ............ 431
Donald B. Jump, Gerald J. Lepar, and
Ormond A. MacDougald
Chapter 20
The Role of Vitamin K in Clotting Factor Biosynthesis ................. 455
J. W. Suttie
Chapter 21
Role of Ascorbate in Regulating the Collagen Pathway ................. 483
Richard I. Schwarz
Chapter 22
Dietary Regulation of Metallothionein Expression ....................... 507
Neil F. Shay and Robert J. Cousins
Chapter 23
Iron-Dependent Regulation of Ferritin Synthesis ........................ 525
H. N. Munro, Z. Kikinis, and R. S. Eisenstein
Chapter 24
Vitamin B6 Modulation of Steroid Induced Gene Expression ............ 547
Douglas B. Tully, Victoria E. Allgood, and John A. Cidlowski
Index .................................................................... 569

Chapter I
NUTRIENT RECEPTORS AND GENE EXPRESSION
James L. Hargrove and Carolyn D. Berdanier
TABLE OF CONTENTS
I. Introduction ........................................................ 2
II. The Evolution of Genetic Control Mechanisms .................... 2
III. Elements of Genetic Control. ...................................... 4
IV. The Importance of Receptors in Nutrient-Gene

Interactions ........................................................ 6
V. Nutritional Regulation of Genes in Unicellular

Organisms ......................................................... 7
A. Nutrient-Gene Interactions in Bacteria ...................... 7
B. Nutritional Regulation of Genes in Yeast. .................. 9
VI. Nutrient Sensors in Mammals ..................................... 11
VII. Temporal and Quantitative Aspects of Gene Expression .......... 14
VIII. Conclusions and Prospects ........................................ 18
References ................................................................ 19
0-8493-6961-41931$0.00 +$.50
© 1993 by CRC Press. Inc. 1
2 Nutrition and Gene Expression
I. INTRODUCTION
The scope of potential nutrient-gene interactions is enormous. More than

40 nutrients are known to be essential to mammals, and health may be impaired
by a deficit in any one of them. 1 Normal growth, reproduction, immune
function, and lifespan require adequate intakes of foods containing these
essential dietary constituents. Just as deficient intakes can be deleterious,
excessive intakes can also be harmful, as is true for vitamins A and D,
selenium, iron, and total caloric intake. Genes are thought to play a funda-
mental role in adaptation to the day-to-day variability in nutrient supply, such
that the nutritional status and health of the individual can be relatively constant
despite the variable supply. That compensatory mechanisms exist is not in
doubt. z-s However, the means by which individual nutrients are detected by
cells, and the details of how altered nutrient supply leads to a reprogramming
of the genetic apparatus, are still lacking in many important cases.
Not only is the current explosion in genetic technology increasing our
understanding of how genetic mechanisms participate in monitoring nutrient
supply and utilization, it is also clear that genes participate in higher order
processes such as food-seeking behavior. Ingestion of nutrients quickly ac-
tivates complex neural and endocrine circuits that affect gene expression not
only in the gut and peripheral tissues, but also in the brain. Genes are activated
as a means of feedback control to deal with undersupply and oversupply of
particular nutrients, which requires the central nervous system to integrate
internal, metabolic cues and environmental conditions.
The purpose of this chapter is to provide a context for the topic of nutrient-
gene interactions, starting with the question of how the transducing systems
that mediate such effects may have arisen during evolution. Even though few
mammalian systems have been characterized that show direct effects of nu-
trients on genes, numerous examples are under study in microorganisms, and
some of these are described to illustrate some of the means by which nutrients
may affect genetic activity in the absence of hormonal mediators. Next, the
presence of nutrient-sensitive chemoreceptors in mammals is discussed, in-
cluding neural and endocrine mechanisms. To complete this general intro-
duction, temporal and quantitative aspects of gene expression in mammals
are reviewed.
II. THE EVOLUTION OF GENETIC

CONTROL MECHANISMS
Evolutionary theory postulates that present-day genetic mechanisms are

inherited from ancestral patterns that have been modified by mutation, var-
iation, and natural selection during competition for limited resources. Fore-
most among those limited resources are the nutrients, which are defined as
any chemical substances that are necessary to life processes such as cellular
Hargrove and Berdanier 3
metabolism, reproduction, motility, and sensation. Because the availability

of nutrients has always been sporadic, organisms that are best able to detect,
take up, and assimilate nutrients have the best chance of contributing their
genotype to the next generation so that it is maintained in the population.
Implicit in the theory of evolution is the idea of feedback control; living
organisms can not function optimally unless they are able to detect and use
information concerning the environment. The nutrients found in food can not
be detected unless they interact specifically with appropriate components,
such as ligand-binding proteins that may be linked to processes for their uptake
or utilization. The need to detect concentrations indicates that nutrients serve
two distinct functions: one which is a metabolic function and one which is
informational. These two functions may be related; for example, glucose both
provides energy to the pancreatic 13-cells and generates a signal that promotes
insulin secretion. The functions may also be separate, as in the use of glutamic
acid for neuronal protein synthesis or for neurotransmission.
Appropriate regulation of genes in response to diet requires two types of
controlling elements. The first is a means of sensing the level of nutrient that
is present; this is the function of sensors or receptors, which may detect one
particular nutrient, members of one class of nutrients, or hormones secreted
in response to the nutrients. We are most familiar with this aspect of nutrients
through the senses of taste and smell, by which neuronal chemoreceptors
detect the presence of sugars, amino acids, salts, and other substances. Sec-
ond, there must be a transducing mechanism by which the level of nutrient
affects cellular activities, including the activity of genes that regulate the
processes by which nutrients are used.
How might organisms have acquired the capacity to bind nutrients with
specificity and to use them in feedback control circuits? The exon-shuffling
hypothesis suggests that discrete segments of DNA encode units of structure
and function in mRNAs and proteins. These unitary elements may be recom-
bined to produce new molecules with two or more structural or functional
elements. This theory is now being applied to analyze functional elements in
genes that encode nutrient receptors and enzymes that participate in nutrient
metabolism. For instance, plasma membranes of hepatocytes and Kupffer
cells contain lectins, or receptors that bind carbohydrates. 6 The carbohydrate-
binding portion of the proteins is located at the carboxylic terminals and is
encoded by three exons at the 3' end of the gene. However, in the related
mannose-binding protein, a single exon encodes the binding function. Separate
exons encode a transmembrane domain and other structural elements.
Current theory based on the exon-shuffling hypothesis suggests that most
mammalian proteins belong to superfamilies whose members originated by
recombination of exons that encoded short functional domains. Subsequent
mutations and recombination then caused divergence among family members.
Another example of this mechanism is the gene for the vitamin D-binding
protein; the gene contains 13 exons. 7 This protein contains a single binding
site for sterols such as 25-hydroxy D3 that is thought to be encoded at the 5'
end of the gene on exons 1 and 2. The protein is multifunctional, and a region
that binds actin may be encoded on exon 10. 7 Vitamin D-binding protein is
thought to be related to albumin and a-fetoprotein. Ray et al. have proposed
that the latter proteins may have been generated from an ancestral gene that
resembled the gene for vitamin D-binding protein. 7
The rise of multicellular organisms permitted functionally distinct groups
of cells to be formed, including some that participated in storing or mobilizing
nutrient reserves. Functional specialization required that means for commu-
nication among groups of cells be developed, so that a signaling system would
permit nutrients to be withdrawn as needed by metabolically active cells.
Although secretion of metabolites and other small molecules may have served
for communication at first, finer controls eventually developed that gave rise
to paracrine and endocrine signals, intracellular signal transduction, and the
specialization of cells to coordinate the transfer of information.
Even though endocrine and neural mechanisms have supplanted nutrients
as primary signals in mammalian signaling networks, nutrient availability still
represents the primary signal for activating or inactivating most genetic net-
works in unicellular organisms. It is also likely that neural and endocrine
signals in multicellular organisms have not entirely replaced the early genetic
mechanisms, because neurotransmitters and hormones cooperate with nu-
trients in directing nutrient utilization. For these reasons, it is pertinent to
review the components of genetic feedback systems and the means by which
nutrients regulate genetic activity in unicellular organisms.
III. ELEMENTS OF GENETIC CONTROL

All information that affects gene transcription converges on regulatory
elements in chromosomes. The promoter is the segment of DNA that is
responsible for binding the enzyme, RNA polymerase II, plus general tran-
scription factors IIA-IIF, and other DNA-binding proteins that are necessary
to initiate transcription in mammals 8 (Figure lA). Because promoters are
located on the same strand of DNA as the structural genes that encode mes-
senger RNAs, they are referred to as cis-acting elements. The structural genes
are regulated by trans-acting factors that are products of separate genes, which
may be located on different chromosomes. Trans-acting factors are usually,
but not exclusively, proteins. Promoters for different genes compete for these
factors, and it has been customary to assign genes to two classes according
to promoter strength, an index of the ability to utilize transcription factors.
Genes that are active at low levels in most cells are referred to as "house-
keeping genes"; with few exceptions, such genes contain no TATA box (the
site at which RNA polymerase binds), but contain multiple binding sites for
factor Spl. s-w Although housekeeping genes were once considered to be
expressed constitutively (at relatively fixed rates), it is now apparent that
Trans-Acting Factors RNA Polymerase II A

Developmental Hormonal Tissue ®
Factorl E~ect ors Specif\city •
1
l ' '' l lntron 1 2
s;'<~------~~------~--~--11~Ex~o~n~AJ~~~~B==~~~~~c~~3'
Enhancers/Silencers Promoter Structural Gene
B
Effector Pathway Response Pathway
Receptor
v •
Signal
Transduction Altered
Metabolis
"/
Second (]) Activation
Messenger
·-~ t
DNA-Binding
~ Inactivation
or
Protein ~ Degradation
Translation
Transport
Processing
Transcription
FIGURE 1. (A) Components of a typical, mammalian gene include the promoter element,
which binds RNA polymerase II and related transcription factors; enhancer and silencer elements;
and the structural gene, which is divided into exons (the mRNA coding segments) and introns
(segments that are removed during processing of the primary transcript as mature mRNA is
formed). Trans-acting factors include proteins (and possibly RNA molecules) that modify the
rate of transcription by binding to regulatory elements in the gene. (B) Responses of genes to
nutrients require an effector pathway (left side of figure) that confers specificity of response
according to cell type and kind of nutrient, and a response pathway (right side of figure) that
may be affected at any of several levels by the nutrient or its metabolites.
most genes are subject to multiple controls. However, genes that are fully
active in only one or a few tissues tend to be transcribed at a higher basal
rate, contain a TATA element, and respond actively to signal molecules such
as hormones and second messengers. Such genes are said to be "inducible",
and the processes by which enzymes and mRNAs increase to new levels are
called induction. This process occurs in response to proteins that bind to
regulatory portions of genes that may be adjacent to the promoter, but can
also occur at long distances away in either direction. These regulatory seg-
ments are termed enhancer elements if they increase transcription, and silencer
elements if they decrease transcription. By affecting proteins that bind to these
parts of genes, nutrients and hormones may provide specificity of genetic
responses.
IV. THE IMPORTANCE OF RECEPTORS IN

NUTRIENT -GENE INTERACTIONS
In order to produce specific responses of any kind, nutrients or their

metabolites must interact with cellular components termed receptors. A re-
ceptor should be capable of binding to a nutrient or metabolite with high
specificity and with an affinity that is appropriate to physiological concen-
trations of the ligand. Binding should then initiate a change in cellular func-
tion, possibly by altering conformation of the receptor protein or by activating
a linked enzymatic activity. For instance, receptors for amino acids such as
glutamic acid and glycine in mammalian nerves are coupled to ion channels
and transducing mechanisms that alter neuronal membrane potential, calcium
metabolism, and phosphoinositide tumover. 11 • 12 Stimulation of the glutama-
tergic system not only alters electrical potential, but also activates transcription
from cellular oncogenes and can overstimulate the neuron and bring about its
death.
The simplest example of a receptor affecting gene expression would be
a protein whose binding to specific segments of DNA is altered in the presence
of the nutrient, its metabolites, or second messengers produced in response
to the nutrient (Figure lB). Numerous examples of this type of response have
been characterized in bacteria and yeast, but few examples have been dem-
onstrated in mammals. Metal ions represent an exception; the ability of zinc
and other ions to alter transcription provides a well-known example. 13 Sim-
ilarly, the ability of iron to alter the stability or translation of the mRNAs
encoding the transferrin receptor and ferritin depends upon a protein that binds
to the 3' and 5' ends of the mRNAs. 14 Although less direct than effects on
transcription, the result in each case is an altered level of a protein that affects
iron accumulation.
Receptor systems for many nutrients may be much more complex than
those for metal ions, as is true of the system in the 13 cell of the pancreas
that regulates output of insulin. 15 Evidence suggests that glucose and its
metabolites initially alter insulin synthesis at the translational level and trigger
insulin secretion through a complex system that includes effects on energy
metabolism, synthesis of cyclic adenosine 3' ,5'-monophosphate (cAMP), ion
transport, and numerous other cellular processes. One potential component
of the transducing system for glucose has been suggested to be the enzyme,
glucokinase. 15 In addition, binding of glucose or its metabolites to plasma

membranes prepared from an islet-cell tumor decreases the activity of calcium
channels over physiological concentrations. 16 Whatever intervening mecha-
nisms may be involved, the result of elevated glucose is increased secretion
of insulin from the 13 cell and decreased secretion of glucagon from the a
cell. For cells elsewhere in the body, changes in glucose levels are accom-
panied by appropriate changes in the ratio of insulin to glucagon. Therefore,
many genetic consequences of ingesting nutrients are mediated by the en-
docrine system, for which receptors and transducing systems are well defined.
Even so, the presence of nutrients often modifies the response to hormones.
V. NUTRITIONAL REGULATION OF GENES IN

UNICELLULAR ORGANISMS
A. NUTRIENT-GENE INTERACTIONS IN BACTERIA

The first genetic responses to nutrients were identified in bacteria and
fungi in the late 19th century (refer to Monod's review 17 of early studies). In
testing the ability of the mold, Aspergillus, to grow on different substances,
Duclaux found that cultures at first grew slowly when protein-containing
media were substituted for carbohydrate-containing media, but then grew
rapidly after a few days of acclimation. He showed that this mold secreted
proteases at an elevated rate after the treatment, and that the type of enzyme
produced was appropriate to the nutritional substrate.
In microorganisms, sugars, amino acids, and inorganic compounds, such
as phosphate ion, initiate or terminate transcription from specific genes by
affecting appropriate DNA-binding proteins. Unicellular organisms monitor
available nutrients, activate those genes that will provide enzymes capable of
synthesizing all needed components from the available nutrient sources, and
repress transcription from unneeded genes. Unlike mammals, bacteria and
yeast have the capacity to synthesize all of the amino acids and enzyme
cofactors from simple sugars, ammonia, and ions present in the substrate.
The idea that a regulatory segment of DNA can control genetic responses
to nutrients by affecting the rate at which gene products are synthesized
derived from studies of microbial metabolism. The classic series of papers
that culminated in this idea focused on the ability of the sugar, lactose, to
increase the activity of a permease system that transports lactose into the
bacterial cell, and also to increase the quantity of the enzyme, 13-galactosidase,
which cleaves lactose to release glucose and galactose. 18 Binding proteins for
nutrients or their metabolites serve vital roles in this process. Within the
periplasmic space that separates the bacterial cell wall from the plasma mem-
brane are located at least two dozen distinct proteins that specifically bind
sugar isomers, amino acids, or ions such as sulfate and phosphate. ' 9 •20 The
binding proteins that complex with sugars and some amino acids have been
studied by X-ray crystallography, and many of the functional groups that
interact with parts of the ligands have been identified. For instance, the
arabinose-binding protein will complex with arabinose, glucose, or fucose by
forming hydrogen bonds with sugar hydroxyl groups; the protein binds the
sugars through functional groups of amino acids in its primary structure,
including amides, hydroxyls, and carboxyl groups. The ligand is further
stabilized by stacking between hydrophobic residues that hold the sugar ring
in place. Interaction with ligands alters conformation of the binding proteins,
which are thought to interact with other proteins in the plasma membrane that
initiate active transport of the nutrient or affect flagellar motility or chemo-
taxis. At the atomic level, similar kinds of interactions might be expected to
occur in mammalian proteins that bind particular nutrients.
Intracellular binding proteins mediate the effects of sugars on bacterial
gene expression with the same high degree of specificity that is typical of
other protein-ligand interactions such as enzymatic catalysis. In the case of
the lactose operon, a protein called the lac repressor normally prevents tran-
scription from this promoter. The lac repressor is a tetrameric protein that
binds the inducing sugar with high affinity, and also has two binding sites
that recognize a specific DNA sequence present in the lac promoter. 21 When
the inducing metabolite, allolactose, binds to the repressor, the repressor
dislodges from promoter DNA, and transcription of the gene by RNA po-
lymerase may begin. Because the repressor mediates the genetic effect, it
could be considered a nutrient receptor for activation of the lac operon.
It might be argued that genes in most mammalian cells differ from those
in bacteria because they do not respond directly to nutrients, but instead are
controlled by hormones or second messengers that are generated in response
to nutrients. However, bacteria have similar mechanisms by which nutritional
status is encoded by a signal molecule. For example, dissociation of the lac
repressor by itself is ineffective in increasing transcription from the lac operon.
Also needed is a positive regulator called the catabolite activator protein
(CAP). Metabolism of glucose inactivates adenylyl cyclase and causes levels
of cAMP to decline, and the lac operon can only be activated if the complex
of cAMP and the activator protein is bound to the promoter DNA. Low levels
of cAMP indicate that glucose is being utilized, and that there is no need to
make use of lactose. The production of cAMP in mammalian liver may
indicate low glucose status in the whole organism as reflected by output of
insulin and glucagon by the pancreas. However, it should be expected that
many genetic responses of mammals might not involve unmodified nutrients,
but a signal related to nutrient utilization. For instance, glycolysis is regulated
by availability of several compounds of low molecular weight, including
glucose-6-phosphate, fructose-2, 6-(bis)phosphate, citrate, NAD+, and phos-
phorylation state of the cytosol (ratio of ATP to ADP + AMP + P;)Y It is
feasible that one or more of these compounds could serve as a metabolic
signal that activates transcription of genes whose products are needed for
integrating glycolysis with other metabolic pathways.
B. NUTRITIONAL REGULATION OF GENES IN YEAST

Because yeast are eukaryotic organisms, many aspects of their structure
and metabolism are similar to features of mammalian cells. 23 Yeast DNA is
invested with histones and other proteins typical of chromatin, which is sep-
arated from the cytoplasm by a nuclear envelope. Like mammalian cells,
yeast contain mitochondria and separate mechanisms for protein synthesis in
mitochondria and endoplasmic reticulum. Glycolysis takes place in the soluble
fraction of the cell, whereas enzymes of the tricarboxylic acid cycle are
mitochondrial. Despite numerous similarities to mammalian cells, yeast differ
in being capable of growth under anaerobic conditions, in having a cell wall
in addition to a plasma membrane, and in their ability to synthesize complex
organic cofactors and amino acids from simple precursors. Yeast resemble
bacteria in that nutrients and their metabolites directly activate or inactivate
sets of genes, with hormonal regulation being of lesser importance. The genes
for individual biosynthetic pathways are not organized in operons as in bac-
teria, but are distributed on different chromosomes and activated by trans-
acting elements.
Glucose is the preferred carbon source for yeast, and it is taken up by a
membrane carrier and quickly phosphorylated by a hexokinase or glucoki-
nase.23·24 The level of enzymes needed for utilization of glucose and other
substrates depends on glucose availability and oxygen tension. Glucose repres-
sion refers to a long-term, adaptive system that allows yeast to utilize hexoses
through the glycolytic pathway and produce ethanol. 24 When glucose, man-
nose, or fructose is available, the synthesis of enzymes that metabolize mal-
tose, sucrose, or galactose is repressed, as are enzymes of the gluconeogenic
pathway and tricarboxylic acid cycle. When glucose is not available, yeast
acquire the capacity to grow on galactose by inducing enzymes that convert
it to glucose-6-phosphate. These include galactose permease, galactokinase,
galactose-1-phosphate uridyltransferase, and uridine diphosphoglucose 4-epi-
merase, whose genes are linked on chromosome II in baker's yeast (Saccha-
romyces cerevisiae). For example, products of the GALl and GALlO genes
are needed for utilization of galactose. Transcription from these genes is
prevented by two types of negative control. 25 ·26 If glucose is available, the
genes are repressed because of inhibition exerted through elements called
operators that are present in multiple copies in the upstream activation se-
quence of these two genes (Figure 2). The mechanism by which the operators
repress transcription is not understood.
Absence of glucose is necessary but not sufficient to allow transcription
from GALl and GALlO. Transcription from these genes is stimulated by an
inducing protein, gal4p, which is a product of the GAL4 locus. Unless gal-
actose is present, however, transcription is prevented by a trans-acting re-
pressor produced from the GAL80 locus that binds to the inducer, gal4p (Figure
2). Induction is thought to occur when a metabolite that is generated when
galactose is present binds to the GAL80 gene product and relieves inhibition.
The inducer is neither galactose itself nor galactose-1-phosphateY During
G~~~~P)
~Galactose +
Activation
Repression ++ Repression
Glucose:
GAL10 GAL1
FIGURE 2. Elements of a genetic switch for regulation of galactose utilization by the yeast,
Saccharomyces cerevisiae, are shown. The switch contains an upstream activation sequence
(shaded bar) that controls transcription from two genes, GALl and GALlO (indicated by right
and left arrows, respectively). Transcription is inhibited by two major inhibitory elements:
(I) Multiple copies of the operators 01 to 06 are present, which prevent transcription when
glucose is available; (2) The inhibitory protein, ga180p, binds to a transcriptional activator unless
galactose is present. An inducer is then thought to be produced that causes ga180p to dissociate
from ga14p. When ga14p binds to multiple activating-elements (boxes 1 to 4), transcription is
stimulated by several orders of magnitude compared to the repressed rate. Other activating
elements, such as GAEl and 2, are present in this complex promoter; the figure was drawn
based on data presented by Finley et al. 25 and Yun et al. 26
the induction process, the protein, gal4p, becomes phosphorylated, and evi-
dence suggests that a protein kinase is necessary for derepression of glucose-
repressible genes. It is of considerable interest that the product of the TUPl
locus, which is needed for glucose repression, is structurally similar to mam-
malian G proteins that participate in signal transduction. 28
Although yeast do possess adenyl cyclase, the role of cAMP in glucose
repression is uncertain. However, addition of glucose to derepressed cells
results in production of cAMP and a cascade of protein phosphorylation,
which causes a transition to the repressed state. Transport of glucose is not
sufficient to generate cAMP; the cells must also contain a hexokinase or
glucokinase. It is not certain whether the sugar must be phosphorylated, or
if another function is served by the hexose kinases and/or transporters. At
least two genes encode hexose transport systems, which are homologous to
the mammalian glucose-transport proteins and contain leucine zipper motifs
that might have potential for binding to DNA. It has been postulated that they
may function as glucose sensors. 29 In addition, the glucose-induced signal
requires a RAS protein that is a member of a proto-oncogene family, and a
CDC protein that forms a complex with the RAS protein and adenyl cyclase. 30
It will be extremely interesting to learn whether or not mammalian glucose-

transport proteins and hexokinase isoenzymes may play similar roles in me-
diating effects of glucose on genetic activity in mammalian cells.
Pathways that control the biosynthesis and catabolism of amino acids
make use of similar trans-acting factors, and are subject to both positive and
negative controls that result in induction and repression. The system that
mediates nitrogen catabolite repression is analogous to the glucose-catabolite
system. When urea, glutamine, or other high-quality source of nitrogen is
present, enzymes for amino acid biosynthesis are induced while proteases and
transport systems for poorer nitrogen sources are repressed. Products that are
generated by metabolism of these compounds act as negative and positive
regulators of mRNA transcription for specific pathways. For instance, the
LEU3 gene product, leu3p, binds to upstream activating DNA sequences
found in several genes for enzymes that participate in synthesis of branched-
chain amino acids. 31 Leu3p contains functional domains for nuclear targeting,
for binding of a-isopropylmalate (the coinducer, an intermediate in leucine
biosynthesis), for binding to DNA sequences, and a separate area for activation
of transcription. In the absence of a coinducer, the inducer molecule binds
to DNA but does not activate transcription. When isopropylmalate accumu-
lates, it binds to leu3p and activates transcription from the genes that parti-
cipate in synthesis of branched-chain amino acids. These examples are but a
few of the numerous systems involved in regulating the uptake and utilization
of nutrients by microorganisms. Do any related systems occur in mammals?
VI. NUTRIENT SENSORS IN MAMMALS
Although few nutrient receptors have been characterized in vertebrates,

there is no question that transducing mechanisms exist for water, salts, sugars,
and amino acids. Receptors for these molecules are present on neurons that
mediate the senses of taste and smell. It is true that many responses to nutrients
result secondarily from release of hormones and transmitters from endocrine
cells or neurons. Nevertheless, numerous cells in the digestive tract, vagus
nerve, and central nervous system respond to nutrients either by secreting
hormones into the bloodstream or generating action potentials, and these cells
must have receptors for nutrients or their metabolites. 32- 36 The hormones and
action potentials convey information to the brain and other organs concerning
the levels of specific nutrients in the gut or the bloodstream.
At least three structural elements in the gastrointestinal tract respond to
nutrients. These elements include the afferent, sensory fibers of the vagus
nerves, the enteroendocrine cells of the gastric and intestinal mucosa, and
probably sensory fibers from the enteric nervous system. Evidence for glu-
coreceptors has come from recordings made from nerve bundles and single
nerve fibers. Sensory fibers are ends of unmyelinated nerves that possess
swellings, mitochondria, and vesicles. 32 In some cases, they are found in
proximity to endocrine cells, and it is uncertain whether responses are direct
or mediated by locally secreted hormones. Both pathways are possible, be-

cause receptors for peptides such as cholecystokinin are found on vagal af-
ferent neurons. Vagal sensory fibers in the gut are often inactive in the absence
of substrate, but an infusion of glucose leads to production of action potentials
within a few seconds. This very short latency is evidence that at least some
effects may be due to direct interactions of nutrients with neurons. Similar
results have been obtained after infusion of the intestine with solutions con-
taining amino acids. 32 In both cases, the responses were relatively specific
and did not occur after stimulation with KCl or by mechanical means. It is
interesting that nerve fibers in the hepatic vagus and area postrema are in-
hibited rather than stimulated by glucose. 33 -36 These fibers are not silent, but
generate action potentials spontaneously. It appears that the brain receives
distinct patterns of information concerning nutrients from different parts of
the viscera and brainstem.
The cells of the diffuse endocrine system in the intestinal mucosa are also
thought to be sensitive to individual nutrients. For instance, secretion of
cholecystokinin occurs in the presence of amino acids and some fatty acids
but not glucose. 37 In contrast, glucose is a strong stimulus to secretion of
gastric inhibitory peptide, which stimulates insulin secretion. 38 The plasma
concentration of hormones from the enteroendocrine system increases sev-
eralfold after a mixed meal, and the hormones are then carried to target sites
elsewhere in the body. 37- 40 Receptors for the hormones have been characterized
in liver, gall bladder, pancreas, and brain (including the area postrema). Thus,
neural and endocrine signals initiated by nutrients in the gut are relayed to
other parts of the body, where appropriate responses occur that may include
altered genetic activity. For instance, ingestion of a protein-containing meal
results in increased transcription from the gene for cholecystokinin in endo-
crine cells of the duodenum. 41
Regulatory circuits for detection of nutrients can be completed starting
with the sensory system in the gut, which conveys information to the brain
by neuronal and endocrine means, and modulates behavior, metabolism, and
function of the autonomic nerves (Figure 3). Sensory information from the
vagus nerves impinges on the receptive field in the brainstem, including the
nucleus of the tractus solitarius (NTS). 42 .43 Information is relayed to the
parabrachial nucleus and then to hypothalamic nuclei and the median emin-
ence, with potential effects on secretion of hormones from the pituitary. 44 .45
Humoral circuits follow similar paths, because chemosensitive neurons and
binding sites for hormones are found in the area postrema. 43 This region is
supplied by fenestrated capillaries through which nutrients and peptides readily
pass. The area postrema is specialized to sample the chemical content of the
bloodstream and is probably important for feeding behavior as well as car-
diovascular function. Neurons in the area postrema project to the NTS and
also to the parabrachial nucleus, both of which project to hypothalamic nuclei.
Although the median eminence is also supplied by fenestrated capillaries, it
Metabolic Behavioral
Integration Integration
Pituitary
+ C rt
o ex
""";'"r;".""/
+.'
Hypothalamus Autonomic
Integration
·~"'"'"h"' "'
~----~
Receptors in
Area Postrema
Blood-Borne Sensory Afferent

Nutrients and Nerve Fibers of Vagus
Hormones
Nutrient Sensors in
Abdominal Organs
FIGURE 3. Neural and endocrine systems for sensing nutrients communicate between the gut,
brain, and peripheral organs. Receptors for glucose and amino acids are thought to be present
on sensory endings of afferent fibers in the vagus nerve (right side of figure). These fibers pass
to the nucleus of the tractus solitarius, which is a major relay for chemical sensation and
baroreflexes; it sends information to the hypothalamus and also controls efferent activity of the
sympathetic and parasympathetic nervous system. Nutrient receptors are also present in endocrine
cells of the gut and glands such as the pancreas, which secrete numerous peptide hormones in
response to dietary constituents (left side of figure). Blood passing to the liver, peripheral organs,
and brain contains nutrients and hormones that may alter gene expression. The blood-brain barrier
prevents much of this information from reaching the brain directly, but circumventricular organs,
such as the area postrema, lack this barrier. They contain receptors for nutrients and hormones
and relay information to other parts of the brain. Small molecules such as glucose pass through
the blood-brain barrier and may directly stimulate neurons in the hypothalamus. These interactions
serve to integrate autonomic function, metabolic activity, and behavior.
contains secretory nerve endings that are not thought to convey information
to other parts of the brain.
Some neurons in the brain are sensitive to nutrients, particularly glucose
and amino acids. Glucose inhibits the rate at which action potentials are
generated by neurons in the area postrema and in the ventromedial hypo-
thalamus, indicating a possible chemosensory function. 34 Amino acids in the
bloodstream are capable of affecting nerve function in the circumventricular
organs; for instance, large doses of glutamate cause cell death in the area
postrema. 46 In contrast, neurons that are protected by the blood-brain barrier
are spared, but utilize amino acids as transmitters. Receptors for excitatory
amino acids in the circumventricular organs could function as nutrient sensors,
but the gating function of the blood-brain barrier makes this less likely for
neurons that are sited deeper in the brain.
No matter how indirect the route, information concerning nutrient status
alters genetic activity in the brain. Messenger RNAs corresponding to neu-
ropeptide Y, cholecystokinin, and pituitary hormones are affected by fasting
and dietary composition. 44 .45 Recent work by Oomura and colleagues47 has
identified a striking effect of feeding or glucose administration on release of
a protein from ependymal cells that line the third ventricle of the lateral
hypothalamus. Owing to its growth-promoting activity, the protein is named
fibroblast growth factor (FGF), but this may be misleading. The concentration
of the protein in the cerebrospinal fluid (CSF) increases as much as 10,000-
fold when glucose concentrations in the CSF increase 2- to 4-fold after a
meal. FGF suppressed the activity of glucose-sensitive neurons in the lateral
hypothalamic area, but not in the ventromedial hypothalamus, and suppressed
feeding. When antibodies to FGF were injected bilaterally into the third
ventricle, nocturnal food intake increased. This evidence suggests that epen-
dymal cells may relay information concerning nutrient status to neurons in-
volved in ingestive behavior, and possibly to neurons that produce releasing
factors involved in output of pituitary hormones that respond to nutrient status.
VII. TEMPORAL AND QUANTITATIVE ASPECTS

OF GENE EXPRESSION
Regulatory pathways that require activation of transcription are excellent

devices for relatively long-term control (hours to days), but are relatively
sluggish and not suited to short-term controls. To provide a specific example,
it is interesting to compare genetic and nongenetic responses in the liver to
hormones that increase glucose production. The classic studies of Sutherland
and colleagues established that glucagon and epinephrine activated glycogen
phosphorylase48 and glucose production49 by increasing the level of the second
messenger, cAMP, within minutes of being added to liver slices, and that the
response persisted for at least an hour in the continued presence of hormone
(Figure 4A and B). However, glucagon and cAMP also activate glucose
. ,• ... is
G)
.: 110 .~ B
>
A E
~ 100
•
l!
Add Hormone f 6
·;: ~
Q
90
0 g
c 80
•
'04
Gl
..
Gl
70 ••• 0
E
••
LL
>.
0
••
.
Gl
2
.
.<:
60 Minus Hormone 0
0.. u
0
.<: 50 "'
00
D.. 0 10 20 30 40 0 20 40
50 60 80
Time (Minutes) Time (Minutes)
2 40 '2
~·;: c 0
Oi
::I
•••••••••••••••••••••••• 30
:.::
.. 1:
0 .5 ~
~Q
. •.•. •.•. -~· · ·-· ... 1:
>-o
~0
Ill
;: . •.•. 20 .J:J
:a
0
1:
·-
. •·····
•••·•.((-- Enzyme Activity
0..111
C( 10 w.,
o...,
................ ·········•
z
a: Gl
E t1:
0 ~~~-.--~---r--~--.---~-.--~---.--~--+o
0 2 4 6 8 10 12
Time (Hours)
FIGURE 4. The time scale of genetic responses is much slower than metabolic control exerted
by phosphorylation/dephosphorylation in response to the same initiating stimulus. (A) Response
of hepatic glycogen phosphorylase to addition of epinephrine (arrow) or glucagon; 48 (B) output
of glucose from liver slices after addition of glucagon; 49 (C) response of the gene for PEPCK
in glucose-fed rats to addition of cAMP or starvation gives rise to elevated mRNA and enzyme
(only the increase in enzyme level, and not the total concentration, is plotted). 50 •51 Note that
phosphorylase is maximally activated within 5 min after addition of hormone, compared to
several hours needed for transcription to fully alter enzyme levels. Panels A through C were
drawn from data in the references cited.
production for a longer period by increasing synthesis of phosphoenolpyruvate

carboxykinase and its mRNA. Normally, this occurs in response to fasting.
The time course over which this enzyme responds is relatively long because
the half-life of the enzyme is about 5 to 7 h, and a full day is required for
the enzyme to respond to onset of starvation (Figure 4C). 50 In contrast, the
half-life of the mRNA is less than 1 h, and levels of the RNA shift to new
values within 2 to 3 h after injection of a cAMP analog that permeates cell
membranes. 51
The slow time course of enzyme induction can be understood through a

principle of kinetics that applies to gene products as well as nutrients and
drugs. Namely, elimination constants (or biological half-lives) are very im-
portant determinants of the time course of biological effects. 52 The theory
suggests that if a stimulus elevates the rate of transcription for a gene, the
time required to obtain a proportional elevation in mRNA concentration in
the cytoplasm is proportional to the half-life of the mRNA. Figure 5 depicts
the basis for the simplest kinetic model that can describe this result. 53 In this
model, the gene serves as a template for catalyzing conversion of nucleotides
into mRNA, and the mRNA serves as a template for converting amino acids
into specific proteins. Factors that promote transcription, processing, and
transport of mRNA from the nucleus to the cytoplasm contribute to a rate of
mRNA formation, and this is considered to be constant during specific in-
tervals unless a new stimulus increases or decreases the rate. Therefore,
synthesis of mRNA would follow zero-order kinetics. In contrast, degradation
of most mRNAs, proteins, and other biological molecules is proportional to
the amount present in the cells of interest. By definition, this is first-order
decay, and the amount of decay equals the product of the first-order decay
constant and the quantity of mRNA present (Figure 5). When this is true,
product will accumulate according to a simple, exponential function that can
be expressed as follows:
c (I)
This equation predicts that, at equilibrium, the concentration of a product

(C) equals the ratio between its rate constant for synthesis (ks) and the rate
constant for decay (kd). If one of the kinetic parameters changes, the con-
centration will approach a new steady-state with a period defined by the rate
constant for decay. The exponential term, 1 - e-kd', is a fraction that varies
between the values of 0 and I as a function of time and the decay constant.
What is the message buried in this equation? In any system to which these
assumptions apply, the time course for change is governed more by the rate
of product elimination than by the rate at which it is synthesized or introduced!
The half-lives of different mRNAs and intracellular proteins range from
abut half an hour to several days, 54 •55 and the time required to fully alter gene
expression equals at least five half-lives. The maximum period required may
be much longer, particularly if amino acid availability limits the rate of protein
synthesis. This kinetic model is fully valid only when applied to transitions
between steady states, and assumes that all mRNA enters the translated pool,
and that the supply of energy and activated amino acids does not limit protein
synthesis. Despite its limitations, this model makes an important prediction
that is supported by experimental data. Half-lives of mRNAs and proteins are
unrelated, and a delay that is proportional to protein half-life should occur
Transcription
ks1
.:....-.......
~
NUCLEOTIDES mRNA Pool
kd1 Rt Rt
mRNA Degradation
Translation
AMINO ACIDS Protein Pool
kd2Pt Pt
Protein Degradation
FIGURE 5. Basis for a quantitative model that shows why gene products accumulate slowly
in response to altered rates of transcription. Specific mRNAs are synthesized from nucleotides
at relatively constant rates unless regulatory molecules alter their transcription; the rate is not
proportional to mRNA concentration. In contrast, the rate of mRNA decay is usually proportional
to mRNA concentration. This gives rise to a model based on zero-order synthesis and first-order
decay, which predicts that a period equal to the half-life of an mRNA or protein must pass before
induction can be 50% complete. Supporting data are discussed elsewhere. 53
before a change in mRNA concentration can produce an equal change in the

concentration of the protein it encodesY This results from the fact that dif-
ferent decay constants apply to the separate pools of mRNA and protein
(Figure 5). This is a physical principle that produces a biological delay between
altered transcription and altered product concentration.
How much time is required to complete each step involved in generating
mRNAs and proteins that are encoded by mammalian genes? The periods
required for the mammalian enzyme, phosphoenolpyruvate carboxykinase
(PEPCK), will be used as an example of a gene product that is regulated
relatively rapidly. Transcription from the gene for PEPCK attains a maximum
within 30 min after treatment of hepatoma cells with cAMP analogues and
declines at a similar rate after addition of insulin. 56 ·57 The rate of transcription
of this gene by RNA polymerase II has been estimated to be about 1000 nt/
min in the presence of insulin and 2500 nt/min in the presence of cAMP. 58
A period of about 2 to 5 min is required to complete transcription for this 6
kb gene, which generates a mature transcript of about 2.6 kb. Transcription

of other class II genes has been estimated to proceed at about 3000 nt/min. 59
Processing and transport of mRNA from the nucleus delay the accumulation
of product in the cytoplasm. In the case of the mRNA for PEPCK, the first
mature mRNA appeared within 10 to 20 min after addition of cAMP, 60 in-
dicating that processing of the primary transcript required no more than 10
min. The concentration of this mRNA was maximal in nuclei by 45 min after
treatment, but it continued to increase in the cytoplasm for about 3 h. The
period required to translate mRNAs is similar to the time needed to complete
transcription; transit time of ribosomes on typical mRNAs is about 2 to 3
min. 61 A minimum of 12 his needed to attain a new, steady-state concentration
of PEPCK. 50 This implies that decay constants are the major determinants of
the period required to attain new, steady-state levels of gene products, a
conclusion that is supported by data derived from studies of numerous mRNAs
and proteins. 53 This explains why genetic processes are so well suited for
long-term control of metabolism, and why short-term regulation requires
activation or inactivation of preexisting enzymes. For instance, many struc-
tural proteins, such as components of the cytoskeleton, the contractile ap-
paratus, the nuclear matrix, or the extracellular matrix, are very stable once
they have become part of mature structures. Short-term regulation of these
stable proteins is usually achieved through a variety of covalent modifications.
Even when the final product is stable, however, newly synthesized proteins
may be unstable, and this pool may be subject to rapid regulation.
VIII. CONCLUSIONS AND PROSPECTS
The energy-yielding macronutrients are the substrates for numerous met-

abolic reactions, and metal ions and vitamins are cofactors for the enzymes
that catalyze the interconversion of nutrients. Examples of nutrient-gene in-
teractions that occur in microorganisms demonstrate that transcription of genes
for biosynthetic reactions tends to be stimulated by precursors but inhibited
by products of the enzyme reactions, whereas transcription of genes for ca-
tabolic reactions is often stimulated by presence of substrate. Similar principles
apply to the regulation of genes in mammals, with the added complexities of
hormonal and other, higher-order regulation. Even though the majority of
genetic controls are positive in mammals, negative controls are important in
preventing futile cycles, such as may occur in glycolysis and gluconeogenesis.
Negative controls also terminate production of cholesterol when adequate
levels are available. In order for feedback and feed-forward controls to be
effective, the level of substrate or product must be sensed by means of
receptors for nutrients or signals related to nutrient use. This chapter has
provided examples of some of these systems, and explained the utility of
genes in controlling processes that take place over several hours or days, as
in adaptation to novel diets or periods of fasting. A clearer picture of the
relationship between structural elements in genes and functional elements in

proteins that participate in nutrient binding and feedback mechanisms is now
emerging, and should suggest unifying principles of great importance to the
science of nutrition.
Much current work dealing with nutrient-gene interactions emphasizes
effects that are regulated by nutrients but secondary to hormonal secretion.
At the time of this writing, there is no example that shows the pathway by
which a macronutrient controls activity of a mammalian gene independently
of hormonal controls. A few approaches to this challenging problem are
discussed in this volume, and results of current studies using DNA transfection
and gene transfer suggest that examples will soon be available for glucose,
amino acids, and fatty acids. 61 •62 One problem is that cells that monitor nutrient
availability are often scarce and scattered among other cells, making them
difficult to study, and few established cell lines exhibit these properties. It
would be very desirable to understand the complete transduction pathway by
which carbohydrates, lipids, and amino acids affect synthesis of mammalian
gene products and to identify the receptors and metabolites involved. Equally
worthwhile would be to learn how nutrients interact with hormones in affecting
gene expression, as exemplified by the phenomenon of glucose-dependent
repression of synthesis of gluconeogenic enzymes and stimulation of enzymes
involved in synthesis of fatty acids and triglycerides. In the following chapters,
expert investigators will describe current knowledge in some of these areas
and studies that are seeking answers to these important questions.
REFERENCES
I. Shils, M. E. and Young, V. R., Modern Nutrition in Health and Disease, 7th edition,
Lea and Febiger, Philadelphia, 1988.
2. Young, V. R. and Marchini, J. S., Mechanisms and nutritional significance of metabolic
responses to altered intakes of protein and amino acids, with reference to nutritional
adaptation in humans, Am. J. Clin. Nutr., 51,270, 1990.
3. Eisenstein, R. S. and Harper, A. E., Relationship between protein intake and hepatic
protein synthesis in rats, J. Nutr., 121, I 581, 1990.
4. Shay, N. F., Nick, H, R., and Kilberg, M. S., Molecular cloning of an amino acid-
regulated mRNA (amino acid starvation-induced) in rat hepatoma cells, J. Bioi. Chern.,
265, 17844, 1990.
S. Munro, S. and Pelham, H. R. B., An Hsp70-like protein in the ER: identity with 78 kd
glucose-regulated protein and immunoglobulin heavy chain binding protein, Cell, 46,
291, 1986.
6. Hoyle, G. W. and Hill, R. L., Structure of the gene for a carbohydrate-binding receptor
unique to rat Kupffer cells, J. Bioi. Chern., 266, 1850, 1991.
7. Ray, K., Wang, X., Zhao, M., and Cooke, N. E., The rat vitamin D binding protein
(Gc-globulin) gene. Structural analysis, functional and evolutionary correlations, J. Bioi.
Chern., 266, 6221, 1991.
8. Wingender, E., Transcription regulating proteins and their recognition sequences, Crit.
Rev. Eukaryotic Gene Expr., 1, 11, 1990.
9. Boyer, T. G. and Maquat, L., Minimal sequence requirements for the initiation of
transcription from an atypical, TATATAA box-containing housekeeping promoter, J. Bioi.
Chern., 265, 20524, 1990.
10. Dynan, W. S. and Tjian, R., Control of eukaryotic messenger RNA synthesis by
sequence-specific DNA-binding proteins, Nature (London), 316, 774, 1985.
II. Betz, H., Ligand-gated ion channels in the brain: the amino acid receptor superfamily,
Neuron, 5, 383, 1990.
12. Verdoorn, T. A., Burnashev, N., Monyer, H., Seeburg, P. H., and Sakmann, B.,
Structural determinants of ion flow through recombinant glutamate receptor channels,
Science, 252, 1715, 1991.
13. Durnam, D. M. and Palmiter, R. D., Transcriptional regulation of the mouse metal-
lothionein-1 gene by heavy metals, J. Bioi. Chern., 256, 5712, 1981.
14. Thiel, E. C., Regulation of ferritin and transferrin receptor mRNAs, J. Bioi. Chern.,
265, 4771, 1990.
15. Matschinsky, F., Glucokinase as glucose sensor and metabolic signal generator in pan-
creatic [3-cells and hepatocytes, Diabetes, 39, 647, 1990.
16. Hoenig, M., Lee, R. J., and Ferguson, D. C., Glucose inhibits the high-affinity (Ca+ 2 -
Mg+2)-ATPase in the plasma membrane of a glucose-responsive insulinoma, Biochim.
Biophys. Acta, 1022, 333, 1990.
17. Monod, J., The phenomenon of enzymatic adaptation and its bearing on problems of
genetics and cellular differentiation, Growth, II, 223, 1947.
18. Jacob, F. and Monod, J., On the regulation of gene activity, Cold Spring Harbor Symp.
Quant. Bioi., 26, 193, 1960.
19. Quiocho, F. A., Vyas, N. K., Sack, J. S., and Vyas, M. N., Atomic protein structures
reveal basic features of binding of sugars and ionic substrates, and calcium cation, Cold
Spring Harbor Symp. Quant. Bioi., 52, 453, 1987.
20. Vyas, N. K., Vyas, M. N., and Quiocho, F. A., Sugar and signal-transducer binding
sites of the Escherichia coli galactose chemoreceptor protein, Science, 242, 1290, 1988.
21. Pace, H. C., Lu, P., and Lewis, M., lac repressor: crystallization of intact tetramer
and its complexes with inducer and operator DNA, Proc. Nat!. Acad. Sci. U.S.A., 87,
1870, 1990.
22. Pilkis, S. J., El-Maghrabi, M. R., and Claus, T. H., Hormonal regulation of hepatic
gluconeogenesis and glycogenolysis, Annu. Rev. Biochem., 57, 755, 1988.
23. Strathern, J, N., Jones, E. W., and Broach, J. R., Eds., The Molecular Biology of
the Yeast Saccharomyces. Metabolism and Gene Expression, Cold Spring Harbor Lab-
oratory, New York, 1982.
24. Entian, K.-D., Glucose repression: a complex regulatory system in yeast, Microbial.
Sci .. 3, 366, 1986.
25. Finley, R. L., Jr., Chen, S., Ma, J., Byrne, P., and West, R. W., Jr., Opposing
regulatory functions of positive and negative elements in UASG control transcription of
the yeast GAL genes, Mol. Cell. Bioi., 10, 5663, 1990.
26. Yun, S.-J., Hiraoka, Y., Nishizawa, M., Takio, K., Titani, K., Nogi, Y., and Fukusawa,
T., Purification and characterization of the yeast negative regulatory protein GAL80, J.
Bioi. Chern., 266, 693, 1991.
27. Bhat, P. J. and Hopper, J, E., The mechanism of inducer formation in ga/3 mutants
of the yeat galactose system is independent of normal galactose metabolism and mito-
chondrial respiratory function, Genetics, 128, 233, 1991.
28. Williams, F. E. and Trumbly, R. J., Characterization of TUP!, a mediator of glucose
repression in Saccharomyces cerevisiae, Mol. Cell. Bioi., 10, 6500, 1990.
29. Lewis, D. A. and Bisson, L. F., The HXTI gene product of Saccharomyces cerevisiae
is a new member of the family of hexose transporters, Mol. Cell. Bioi., II, 3804, 1991.
30. Van Aelst, L., Jans, A. W. H., and Thevelein, J. M., Involvement of the CDC25
gene product in the signal transmission pathway of the glucose-induced RAS-mediated
cAMP signal in the yeast, Saccharomyces cerevisiae, J. Gen. Microbial., 137, 341, 1991.
31. Zhou, K. and Kohlhaw, G., Transcriptional activator LEU3 of yeast. Mapping of the
transcriptional activation function and significance of activation domain tryptophans, J.
Bioi. Chern., 265, 17409, 1990.
32. Mei, N., Intestinal chemosensitivity, Physiol. Rev., 65, 211, 1985.
33. Niijima, A., Glucose-sensitive afferent nerve fibers in the liver and their role in food
intake and blood glucose regulation, J. Autonomic Nervous Syst., 9, 207, 1983.
34. Oomura, Y. and Yoshimatsu, H., Neural networks of glucose monitoring system, J.
Autonomic Nervous Sysr., 10, 359, 1984.
35. Vigas, M., Tatar, P., Jurcovicova, J., and Jezova, D., Glucoreceptors located in
different areas mediate the hypoglycemia-induced release of growth hormone, prolactin,
and adrenocorticotropin in man, Neuroendocrinology, 51, 365, 1990.
36. Ewart, W. R., Medullary integration of afferent information from the gastrointestinal
tract, in Brain-Gut Interactions, Tache, Y. and Wingate, D., Eds., CRC Press, Boca
Raton, FL, 1991, 109.
37. Rehfeld, J, F., Cholecystokinin, in Handbook of Physiology, Seeton 6: The Gastroin-
testinal System, Vol. II, Schultz, S., Makhlouf, G. M., and Rauner, B. B., Eds., 1989,
337.
38. Brown, J. C., Buchan, A.M. J., Mcintosh, C. H. S., and Pederson, R. A., Gastric
inhibitory peptide, Brain-Cut Interactions, Tache, Y. and Wingate, D., Eds., CRC Press,
Boca Raton, FL, 1991, 109.
39. Taylor, I. L., Pancreatic polypeptide family: pancreatic polypeptide, neuropeptide Y,
and peptide YY, Brain-Gut Interactions, Tache, Y. and Wingate, D., Eds., CRC Press,
Boca Raton, FL, 1991,109.
40. Miselis, R. R., Shapiro, R. E., and Hyde, T. M., The area postrema, in Circumven-
tricular organs and body fluids, Vol. II, Gross, P. M., Ed., CRC Press, Boca Raton,
FL, 1987, 185.
41. Liddle, R. A., Carter, J.D., and McDonald, A. R., Dietary regulation of rat intestinal
cholecystokinin gene expression, J. Clin. Invest., 81, 2015, 1988.
42. Czyzyk-Krzeska, M. F., Bayliss, D. A., Seroogy, K. B., and Millhorn, D. E., Gene
expression for peptides in neurons of the petrosal and nodose ganglia of the rat, Exp.
Brain Res., 83, 411, 1991.
43. Adachi, A., Kobashi, M., Miyoshi, N., and Tsukamoto, G., Chemosensitive neurons
in the area postrema of the rat and their possible functions, Brain Res. Bull., 26, 137,
1991.
44. Brady, L. S., Smith, M.A., Gold, P. W., and Herkenham, M., Altered expression
of hypothalamic neuropeptide mRNAs in food-restricted asnd food-deprived rats, Neu-
roendocrinology, 52,441, 1990.
45. Chua, S. C., Jr., Liebel, R. L., and Hirsch, J., Food deprivation and age modulate
neuropeptide gene expression in the murine hypothalamus and adrenal gland, Mol. Brain
Res., 9, 95, 1991.
46. Phelix, C. F. and Hartle, D. K., Systemic glutamate induces degeneration of a sub-
population of serotonin-immunoreactive neurons in the area postrema of rats, Neurosci.
Lett., 117, 31, 1990.
47. Oomura, Y., Sasaki, K., Suzuki, K., Muto, T., Li, A., Ogita, A.-I., Hanai, K.,
Tooyama, I., Kimura, H., and Yanaihara, N., A new brain glucosensor and its phys-
iological significance, Am. J. Clin. Nutr., 55, 278S, 1992.
48. Sutherland, E. W., The effect of the hyperglycemic factor and epinephrine on enzyme
systems of liver and muscle, Ann. N.Y Acad. Sci., 54, 693, 1951.
49. Rail, T. W., Sutherland, E. W., and Wosilait, W. D., The relationship of epinephrine
and glucagon to liver phophorylase, J. Bioi. Chern., 218, 483, 1956.
50. Hopgood, M. F., Ballard, F. J., Reshef, L., and Hanson, R. W., Synthesis and
degradation of phosphoenolpyruvate carboxylase in rat liver and adipose tissue. Changes
during a starvation-re-feeding cycle, Biochern. J.. 134, 445, 1975.
51. Beale, E. G., Hartley, J. L., and Granner, D. K. N6, 0 2 '-dibutyryl cAMP and glucose
regulate the amount of messenger RNA coding for hepatic phosphoenolpyruvate car-
boxykinase (GTP), J. Bioi. Chern., 257, 2022, 1982.
52. Goldstein, A., Aronow, L., and Kalman, S.M., Principles of Drug Action, 2nd ed.,
Harper and Row, New York, 1976, 280.
53. Hargrove, J. L., Hulsey, M.G., and Beale, E. G., The kinetics of mammalian gene
expression, BioEssays, 13, 667, 1991.
54. Moore, R. E., Goldsworthy, T. L., and Pitot, H. C., Turnover of 3' -po1yadenylate-
containing RNA in livers from aged, partially hepatectomized, neonatal, and Morris
5123C hepatoma-bearing rats, Cancer Res., 40, 1449, 1980.
55. Shapiro, D. J., Blume, J. E., and Nielsen, D. A., Regulation of messenger RNA
stability in eukaryotic cells, BioEssays, 6, 221, 1987.
56. Granner, D., Andreone, T., Sasaki, K., and Beale, E., Inhibition of transcription of
the phosphoenolpyruvate carboxykinase gene by insulin, Nature, 305, 549, 1983.
57. Sasaki, K., Cripe, R., Koch, S., Andreone, T. L., Petersen, D. D., Beale, E. G.,
and Granner, D. K., J. Bioi. Chern., 259, 15242, 1984.
58. Sasaki, K. and Granner, D. K., Regulation of phosphoenol-pyruvate carboxykinase
gene transcription by insulin and cAMP: reciprocal actions on initiation and elongation,
Proc. Nat/. Acad. Sci. U.S.A., 85, 2954, 1988.
59. Sehgal, P., Derman, E., Molloy, G. R., Tamm, I., and Darnell, J. E., 5,6-Dich1oro-
1-f3-D-ribofuranosylbenzimidazole inhibits initiation of nuclear heterogeneous RNA chains
in HeLa cells, Science, 194, 431, 1976.
60. Chrapkiewicz, N. B., Beale, E. G., and Granner, D. K., Induction of the messenger
ribonucleic acid coding for phosphoenolpyruvate carboxykinase in H4IIE cells. Evidence
for a nuclear effect of cAMP, J. Bioi. Chern., 257, 14428, 1982.
61. Palmiter, R. D., Quantitation of parameters that determine the rate of ovalbumin syn-
thesis, Cell, 4, 189, 1975.
62. Thompson, K. S. and Towle, H. C., Localization of the carbohydrate response element
of the rat L-type pyruvate kinase gene, J. Bioi. Chern., 266, 8679, 1991.
63. McGrane, M. M., Yun, J. S., Patel, Y. M., and Hanson, R. W., Metabolic control
of gene expression: in vivo studies with transgenic mice, Trends Biochern. Sci., 17, 40,
1992.
Chapter 2
LACTOSE INTOLERANCE AND REGULATION OF

SMALL INTESTINAL LACTASE ACTIVITY
Robert K. Montgomery, Hans A. Buller, Edmond H. H. M. Rings,

Jan Dekker, and Richard J. Grand
TABLE OF CONTENTS
I. Introduction: Milk Drinking and Lactose Intolerance ............. 24
II. Definition of Clinical Symptoms of Lactose Intolerance .......... 26

A. Diagnosis .................................................. 27
B. Fecal pH and Reducing Substances ........................ 27
C. Lactose Absorption Test and Lactose Breath
Hydrogen Test. ............................................ 27
D. Lactose Activity in Intestinal Biopsy ...................... 27
E. Comparison of Tests ....................................... 28
F. Treatment of Lactose Intolerance .......................... 28
III. Influence of Lactose Intolerance on Nutrition ..................... 29
IV. Lactase ''Deficiency'' in Human Populations ..................... 29
V. Lactase Genetics and Development ............................... 31

A. Adaptation vs. Inheritance ................................. 31
B. Lactase Development ...................................... 32
C. Lactase Distribution and Hormonal Effects ................ 33
VI. Cellular and Molecular Biology of Lactase ....................... 34

A. Structure and Function .................................... 35
B. Glycosylation, Processing, and Transport ................. 36
C. Molecular Structure of Lactase ............................ 38
D. Comparison of Lactase Amino Acid Sequences
from Three Species ........................................ 40
VII. Regulation of Lactase Decrease ................................... 40

A. Molecular Biology of Lactase Decline ..................... 42
B. Molecular Biology of Lactase Persistence
in Humans ................................................. 44
VIII. Conclusions ....................................................... 44
IX. Summary ......................................................... 46

0-8493-6961-41931$0.00 +$.50
© I 993 by CRC Press. Inc 23
Acknowledgments ........................................................ 47
References ................................................................ 47
I. INTRODUCTION: MILK DRINKING AND

LACTOSE INTOLERANCE
The use of milk as food for adults has its origins in prehistory. The
average North American or Northern European has been taught to think of
milk (usually cow's milk) as the perfect food. This attitude is so pervasive
that an inability to drink milk is considered a serious abnormality. Although
the terminology of milk or lactose intolerance and lactase deficiency is well
established in the literature, they are misnomers and misleading, as will be
discussed below. Milk intolerance can be due to allergies to milk proteins or
to the inability to digest milk sugar or lactose; the remainder of this review
will be confined to the latter: lactose intolerance. Lactose, the major carbo-
hydrate in milk, is hydrolyzed to glucose and galactose by the small intestinal
enzyme, lactase-phlorizin hydrolase (E.C. 3.2.1.23; 3.2.1.62), subsequently
referred to as lactase. Lactose intolerance is due to reduced levels of lactase
in the small intestine. The clinical aspects of lactose intolerance and nutrition, 1
and the population distribution of high and low lactase levels, have been
published in detaiF and will be reviewed only briefly. The major part of the
discussion will focus on recent progress in understanding the structure and
function of lactase and its regulation.
It is now clear that although milk intolerance is a frequent problem in
human adults, it is not a disease, nor an abnormality, but the normal condition
for most humans. 3 •4 Although milk intolerance was first described in 190 I , 5
a clear understanding of the digestion and absorption of lactose depended on
the availability of reliable tests. The development of the glucose oxidase
method by Dahlqvist6 provided a rapid, accurate method for assaying lactase
in tissue samples, including intestinal biopsies. In combination with mea-
surements of blood glucose levels following ingestion of a standardized lactose
dose, this method enabled investigators to link intestinal lactase activity levels
with lactose absorption and lactose intolerance. 7 Numerous studies using these
methods have delineated the distribution of lactase along the small intestine,
during human development, and in different human populations throughout
the world (summarized in References 2 and 8 through 10).
Montgomery et al. 25
Milk intolerance was first described by Jacobi et al., 5 who suggested a

relationship between carbohydrate ingestion and infantile diarrhea. Mendel
and Mitche1Jl 1 observed a decline in lactase activities from high levels in
neonatal calves, through a slow decrease, to low adult levels. The same pattern
was confirmed in other animals, which led to the hypothesis that lactase levels
were the result of milk consumption, and that continuation of milk ingestion
would maintain lactase enzyme at newborn levels. However, as early as
1907, Plimmer 12 examined this hypothesis in adult rabbits and rats, and re-
ported no adaptation of lactase activity to the presence of milk or lactose in
the diet. Neither prolonged ingestion of lactose in humans 13 nor exclusion of
lactose from the dieti 4 influences lactase activity, strongly suggesting that the
enzyme is not directly regulated by availability of substrate. However, because
of the frequency and variable course of lactose intolerance, this subject has
remained controversial, and a great deal of effort has been devoted to deter-
mining if lactose induction of lactase activity occurs and to examining the
mechanism of lactase decline.
Lactose is a disaccharide unique to milk. In the lactating mammary gland,
membrane-bound galactosyl transferase interacts with the soluble milk pro-
tein, alpha-lactalbumin, to form lactose synthetase. Lactose synthetase cat-
alyzes the reaction of UDP-galactose with glucose to form lactose. Although
the lactose group also occurs as part of oligosaccharides, it is found as a free
disaccharide only in milk. It is not clear, however, what advantage lactose
confers, that it should be the sugar in milk. In the milks which have been
analyzed, the percentage of lactose remains remarkably constant; it varies by
approximately 50%, from a high of 7% in human milk to a low of about 4%
in the dog and others. 15 In contrast, the fat content of milk can vary by 30-
fold, with human milk having a low fat content and that of aquatic mammals
having a very high fat content, presumably because of the high energy re-
quirements necessitated by their environment. Although suggestions (dis-
cussed below) of possible selective advantages conferred by lactose have been
advanced, there are few data to support the hypotheses. Furthermore, there
are significant examples of milks which do not contain lactose: the pinnipeds
which have been examined (and probably other aquatic mammals) neither
synthesize lactose in their milk nor have lactase activity, 16 most likely a recent
evolutionary adaptation to their aquatic habitat. A number of Australian mar-
supials that have been examined do not have neutral beta-galactosidase (lac-
tase) in their intestine 17 and secrete little or no lactose in their milk, 18 probably
an evolutionarily primitive stage of mammalian development. Unhydrolyzed
lactose is not absorbed in any significant quantity, but is passed through the
small intestine unchanged. The symptoms grouped as "lactose intolerance"
are due to the osmotic effects of lactose on the intestine and the formation
of gas when lactose is digested by the colonic flora.
II. DEFINITION OF CLINICAL SYMPTOMS OF

LACTOSE INTOLERANCE
A variety of terms has been used to describe clinical symptoms induced

by the ingestion of milk and/or milk products. Typical complaints include
abdominal pain, cramps or distention, nausea, flatulence, and diarrhea; in
children and adolescents, vomiting may predominate. While patients com-
monly identify these symptoms as a consequence of milk intolerance, they
can be based either on the inability to digest lactose or sensitivity to milk
proteins. 4 Lactose intolerance is characterized by symptoms, as described
above, after the ingestion of lactose-containing foods or a test dose either of
lactose in water or of milk. The term lactose malabsorption is reserved for
those patients in whom the intestinal malabsorption of lactose has been in-
vestigated using an appropriate test of lactose absorption (lactose absorption
test) or malabsorption (lactose breath hydrogen test). 19 Lactase deficiency is
defined only when low ( <2 SD below the mean) or, very rarely, no level of
lactase activity has been found in a small intestinal biopsy sample appropri-
ately assayed. 6 ·7•20 Lactase deficiency is either a primary or secondary event.
Primary lactase deficiency occurs as a developmental process in premature
infants or as a rare clinical syndrome. 10 •21 It also appears as ''late onset lactase
deficiency" (a misnomer, as discussed below) in the majority of the world's
population around the age of 5 years. Secondary lactase deficiency is found
following mucosal injury. 21
Several factors account for the variability of symptoms produced by lac-
tose ingestion in people with lactose intolerance accompanying low lactase
activity; these include the osmolarity and fat content of the food in which
lactose is ingested, the rate of gastric emptying, sensitivity to intestinal dis-
tension produced by the osmotic load of unhydrolyzed lactose in the upper
small bowel, the rate of intestinal transit, and the response of the colon to
the lactose load. In general, the higher the osmolarity of gastric contents and
the higher the fat content of the diet containing lactose, the slower the gastric
emptying, and the lesser the symptoms induced by lactose. Consumption of
a meal simultaneously with ingestion of lactose-containing beverages was
demonstrated to significantly reduce symptoms in low lactase individuals,
probably due to delayed gastric emptying. 22 Different individuals appear to
have more or less sensitivity to abdominal distension, and complain differently
when ingested lactose stimulates an influx of water into the lumen of the
small intestine. Thus, those with greater tolerance will report fewer symptoms.
These subjective responses are difficult to quantify. Intestinal transit also is
influenced by the quality of the diet and individual motility patterns. Ac-
cordingly, some lactose intolerant people experience very rapid movement of
lactose to the cecum, while others have slower motility. If responses to
lactulose (a nonabsorbable disaccharide) simulate those to lactose, transit time
from the stomach to the cecum may be as rapid as 10 min. 23 Finally, fecal
flora are known to adapt to ingested carbohydrate, and it is clear that non-
absorbed lactose can be salvaged by the colonic flora by fermentation to short
chain fatty acids, hydrogen, methane, water, carbon dioxide, and the pro-
duction of energy. 24 If lactose is provided slowly over a long period of time
in many ''intolerant'' people, the flora may adapt to the load and symptoms
produced by gas and acid in the colon may be reduced or eliminated. 1 This
mechanism of lactose tolerance in people with low lactase levels accounts for
the discrepancy in some studies in which "lactase deficient patients" are not
lactose intolerant. 25 Colonic distension may also vary among lactose intolerant
people, providing another variable in the expression of symptoms after lactose
ingestion.
A. DIAGNOSIS
The diagnosis of lactose malabsorption and its pathogenesis is based on
a combination of clinical findings and the results of appropriate tests. ' 9·26.29
The use of screening tests for lactose malabsorption in the diagnosis of lactose
intolerance has received wide attention. 19 ·26
B. FECAL pH AND REDUCING SUBSTANCES

The presence of low fecal pH and reducing substances indicates lactose
malabsorption, but these tests are only valid when lactose has been ingested,
intestinal transit time is rapid, stools are collected fresh and assayed imme-
diately, and when bacterial metabolism of colonic carbohydrate is incom-
plete.26 In general, confirmation oflactose malabsorption is best accomplished
using more specific tests.
C. LACTOSE ABSORPTION TEST AND LACTOSE

BREATH HYDROGEN TEST
The capacity for lactose absorption can be measured using the lactose
absorption test. ' 9 In adults, it has a sensitivity and specificity of 75 and 96%,
respectively. However, in children, it is cumbersome, invasive, and time-
consuming, and has largely been replaced by the lactose breath hydrogen
test. 28 ·30 Although this test really measures lactose nonabsorption rather than
lactose hydrolysis and monosaccharide uptake, its sensitivity ( 100%) and
specificity (100%) 19 in adults with "late onset lactase deficiency" is actually
superior to those for the lactose absorption test, and it is simple and nonin-
vasive. 19 ·25-30 Breath hydrogen testing is discussed in detail elsewhere. 31
D. LACTASE ACTIVITY IN INTESTINAL BIOPSY

The assay of lactase activity in small bowel biopsy samples establishes
the presence of lactase deficiency and has been used to define populations
with "late onset lactase deficiency". 20 However, when lactase deficiency
accompanies intestinal injury, the lesion may be focal or patchy; consequently,

intestinal biopsy samples may not yield an abnormal result. Furthermore, this
test is invasive, time-consuming, and assays are available only in a few
centers. Normal values have been published. 20
E. COMPARISON OF TESTS
Studies are available which have compared intestinal histology, lactase
activity, and breath hydrogen test results in children with chronic diarrhea or
abdominal pain. 25 •27 In patients with abnormal intestinal histology, approxi-
mately 75% had abnormal lactose breath hydrogen values (sensitivity 75%).
However, in patients with normal histology, the lactose breath hydrogen test
was only 54% specific, reflecting the presence of a patchy villus lesion. Very
similar values were found when lactase activities on biopsy were compared
with results of lactose breath hydrogen tests. 25 •29 Accordingly, an abnormal
lactose breath hydrogen test is of value in identifying those patients who will
be symptomatic after lactose ingestion. In children less than 5 years of age,
an abnormal lactose breath hydrogen test always signifies abnormal intestinal
mucosa, which usually needs further definition with a small intestinal biopsy.
A normal lactose breath hydrogen test does not rule out an intestinal mucosal
abnormality and should not be used to avoid an intestinal biopsy in the
diagnosis of suspected mucosal disease (e.g., gluten sensitive enteropathy)Y
F. TREATMENT OF LACTOSE INTOLERANCE

The treatment of lactose intolerance includes four general principles: (1)
reduction or restriction of dietary lactose, (2) substitution of alternative nu-
trient sources to avoid reduction in energy and protein intake, (3) regulation
of calcium intake, and (4) use of a commercially available enzyme substitute.
When lactose restriction is necessary, the patient must be instructed to read
labels of commercially prepared foods, as hidden lactose may be difficult to
identify. Calcium is supplemented in the form of calcium carbonate ("Turns"
and "OsCal'' are popular and effective). In infants, liquid calcium gluconate
is readily tolerated and available. 32
Commercially available "lactase" preparations are actually bacterial or
yeast ~-galactosidases. When added to lactose-containing food or ingested
with meals containing lactose, these are effective in reducing symptoms and
breath hydrogen values in many lactose intolerant subjectsY However, they
are not capable of completely hydrolyzing all dietary lactose, and the results
achieved in individual patients are variable. Live-culture yogurt, which con-
tains endogenous ~-galactosidase, is a useful alternative source of both cal-
cium and calories and may be well tolerated by a number of lactose intolerant
patients. 34 However, yogurts which contain milk products added back after
fermentation may produce symptoms. While consumption of yogurt alone by
low lactase individuals reduced symptoms, consumption of yogurt together
with additional lactose did not reduce symptoms. 22
Montgomery et a!. 29
III. INFLUENCE OF LACTOSE INTOLERANCE

ON NUTRITION
The discrepancy between measured intestinal lactase levels by either breath

test or biopsy has given rise to a wide variety of interpretations of the capacity
of populations to tolerate milk intake. A study by Bayless et al. 35 reported
that 59% of low lactase subjects were intolerant of 240 ml of milk and had
a minor milk intake. Other studies 36 have claimed that the great majority of
low lactase subjects can tolerate one glass of milk and have a normal milk
intake. A study by Lisker eta!. 37 correlated milk drinking habits with intestinal
lactase activity as judged by a lactose tolerance test. These authors concluded
that intestinal lactase activity was important in determining the extremes of
milk consumption: either four or more glasses of milk in a day or none, but
that it had little influence on the intermediate pattern. A number of individuals
with measured low lactase could tolerate ingestion of amounts of milk less
than 250 ml at one time. Possible explanations for varying tolerance are
discussed above.
Low lactase activity may influence the absorption of calcium, an important
nutrient, in two ways. Milk is a major source of dietary calcium, and no or
reduced milk intake due to low lactase activity and lactose intolerance may
result in calcium deficiency. On the other hand, several studies suggest that
lactose in the intestine may enhance the uptake of calcium. Studies by Kocian
et al. 38 indicated that while calcium absorption in lactose-intolerant subjects
was reduced, calcium retention and the total absorbed were normal, suggesting
that compensatory slowing of intestinal transit or other compensatory mech-
anisms allowed normal levels of calcium absorption. A more recent study by
Gries sen et al. 39 also indicated that low lactase individuals did not have
decreased calcium absorption when ingesting milk containing either lactose
or glucose.
IV. LACTASE "DEFICIENCY" IN HUMAN

POPULATIONS
Since lactose is a key nutrient, lactase plays a critical role in the nutrition
of mammalian neonates. As mentioned above, human congenital lactase de-
ficiency, present from birth, is extremely rare and inherited as an autosomal
recessive gene. 10 In adulthood, milk drinking is considered normal among
populations of Northern European extraction and a few other groups. How-
ever, for most humans and all other adult mammals, significant milk ingestion
results in mild to severe gastrointestinal symptoms, caused by the inability
to digest lactose, 2 due to low levels of intestinal lactase. The initial human
studies were performed in adult patients of Northern European origin who
had elevated lactase. The data obtained led to the assumption that high lactase
TABLE 1 TABLE 2
Distribution of Lactase Distribution of Lactase Phenotypes in
Phenotypes in Selected Selected European Populations
Populations in the
United States Low lactase
Country Population (%)
Low lactase
Sweden Swedes
Population (%)
Netherlands Dutch 0
Austria Austrians 20
Northern European 7
France French 32
Whites 22
Southern French 44
Blacks 65
Italy Northern Italians 50
American Indians 95
Southern Italians 72
Vietnamese 100
Sicilians 71
Data from Reference 4.

Data from Reference 4.
activity throughout life was normal. Subsequent observations by Dahlqvist et

al. in 1963 7 of the almost total absence of lactase activity in small intestinal
biopsies of healthy adults with normal mucosal histology, and the findings
of ethnic distribution of lactose malabsorption in studies by Cuatrecasas et
al. 8 and Bayless and Rosensweig9 led to the designation of these patients as
having lactase "deficiency". Numerous studies have delineated the detail of
the population patterns of lactase ''deficiency''. There is a definite geographic
pattern, which has been worked out in considerable detait.2 In Europe, there
is a general north to south gradient of increasing frequency of low lactase
activity, although there are specific areas where this generalization does not
hold, presumably because of migration patterns. 40 (Selected population ex-
amples are shown in Tables 1 and 2.)
The general pattern of lactase distribution was established early in the
analysis of digestive enzymes. The first investigators studied adult and ju-
venile mammals, as well as birds, reptiles, and amphibians. Their findings
were summarized by PlimmerY "Neither frog nor fowl have lactase in their
intestine, and we may conclude that animals lower than mammals do not
possess this ferment.'' These early studies also indicated that lactase was
distributed differently in different regions of the intestine and that juvenile
mammals exhibited lactase activity, while mature mammals did not. 11 Sub-
sequent workers have confirmed these findings and provided a more detailed
picture of the comparative biology of lactase and the regional distribution of
lactase in the small intestine.
That a similar pattern prevails in the human population has been known
for some time, yet the common terminology still implies that low lactase
levels in adult humans represent an abnormality (e.g., see discussion in Ref-
erence 4). Extensive population studies in the late 1960s and early 1970s
(summarized in Reference 4) showed that lactase is ''deficient'' in the majority

of the world's adult population. Unfortunately, the term "adult lactase de-
ficiency" has become established in the literature, even though it is now clear
that low lactase activity is not a deficiency, but the normal adult human
condition (as for all mammals). Although population genetic analysis indicates
that elevated lactase activity is inherited as a single autosomal dominant gene, 4
a generally accepted mechanism of regulation has not yet been described.
The extensive population studies of the distribution of elevated and low
lactase levels in adults have demonstrated that while the majority of human
adults have low lactase activities, there are specific groups with elevated
lactase levels. Today, the populations with elevated lactase activity also use
milk from cattle as a major food source. Since the primitive human pattern
was clearly one of low lactase activity in adulthood, there must have been
some selective pressure to maintain and increase the number of people with
elevated lactase levels in adulthood. It has been suggested that in the case of
nomadic herdsman with little else available as a food source, there was strong
selection in favor of the ability to utilize milk. 4 This argument is less persuasive
for Northern European populations, which have other resources, but which
express elevated lactase activity in nearly 100% of those studied. It is also
somewhat puzzling that lactose should be the carbohydrate in milk instead of
some other, suggesting that there may be some selective advantage to lactose.
Flatz4 combined these arguments and suggested that lactose might enhance
the absorption of calcium and that this conferred an advantage in Northern
Europe where reduced sunshine decreased vitamin D levels and thereby cal-
cium absorption. There are some data to support this argument, but they do
not make a wholly convincing case. It also necessitates an explanation of
how the sea lion and other marine mammals maintain calcium balance without
ingesting lactose at any time in their lives. 41 .42
V. LACTASE GENETICS AND DEVELOPMENT
A. ADAPTATION VS. INHERITANCE

In order to explain the regional and ethnic distribution of lactose mal-
absorption in adults, two conflicting basic explanations have been suggested.
The first hypothesis assumed lactase induction by lactose and argued that in
areas where milk and milk products are not constituents of the traditional diet
of young children and adults, such as in Africa and Asia, the decline in lactase
activity was due to the absence of lactose. 43 The second hypothesis suggested
that lactase "deficiency" is genetic in origin and due to homozygosity for
an autosomal recessive gene. 44
Although early studies showed that rats and rabbits did not adapt their
levels of lactase enzyme after ingestion of lactose-rich feedings, the adaptive
theory remained widely accepted until the 1970s. It was frequently observed
that patients, who on biopsy were lactase deficient, did not report symptoms
of lactose intolerance. 45 .4 6 Another major source of confusion was the dis-
appearance of symptoms of lactose malabsorption after prolonged periods of
lactose ingestion in lactose intolerant patients. In addition to clinical obser-
vations, the induction hypothesis was supported by studies which showed that
lactase activity increased in adult rats after very high lactose intake. 47•48 How-
ever, the changes in lactase activity were relatively small compared to the
magnitude of decrease that occurs around the weaning period, and high glu-
cose intake produced comparable effects on lactase activity. In the studies by
Lebenthal et a!., 48 lactase activity was higher during prolonged nursing than
in controls, but the decline of lactase activity around weaning was not pre-
vented by these experimental conditions.
Kretchmer4 1 described Nigerian medical students with clinically proven
lactose intolerance who remained unable to digest lactose, when tested by
lactose tolerance tests, after 6 months of daily lactose intake (50 g). Never-
theless, the students gradually adjusted to this large lactose intake and did
not show any symptoms of intolerance. It is now clear that the explanation
for this adaptive response was the emergence of fecal flora capable of fer-
menting the nonabsorbed lactose to short chain fatty acids. 49 Thus, there is
no compelling evidence for adaptive changes in small intestinal lactase activity
in humans.
The genetic hypothesis was first put forward in 1966 when Bayless and
co-workers showed that 70% of a group of black Americans examined in
Baltimore were lactose malabsorbers, while only 8% of the white subjects
were malabsorbers. 9 Many subsequent studies correlated racial and ethnic
distribution of lactase enzyme levels and lactose intolerance in adults (sum-
marized in References 2, 4, 50). The necessary genetic analysis was provided
by Sahi, who showed Mendelian inheritance of lactose malabsorption as an
autosomal recessive trait in a large family study from Finland. 51 Similar studies
performed subsequently support the genetic hypothesis. 52 The available data
indicate that persistence of high levels of lactase enzyme is probably an
autosomal dominant trait. 52 The persistence in adulthood of the capacity for
lactose digestion in a few human groups is probably a recent evolutionary
development, while the majority of the world's population, and all placental
mammals studied, show a reduction of lactase activity in adulthood. It has
been hypothesized that elevated lactase levels in adults emerged over a span
estimated to be a minimum of 10,000 years in several loci around the world,
probably coincident with the development of dairying. 50
B. LACTASE DEVELOPMENT
In rats and rabbits, the two mammals which so far have been studied in
greatest detail, the specific activity of lactase (expressed in units per mg
protein) exhibits a similar developmental pattern. 53 Lactase activity is un-
detectable until a few days before birth; then there is a late gestational rise,
with a peak in specific activity shortly after birth, and a fall during weaning
to the low levels which are seen in adulthood. 53 This pattern is partially a
reflection of the short gestation and immaturity of these mammals. The guinea
pig, which has a gestation of 68 days and is born in a more mature state,
shows detectable lactase activity at 30 days of gestation and a steady increase
until shortly before birth when levels fall slightly. 54 Guinea pigs are also
reported to show very little decrease in lactase levels from the time of birth
to adulthood. 55 In rats, the initial appearance of enzyme activity coincides
with the initial morphogenesis of the enterocyte, which occurs about three
days prior to birth. 53 ·56
In contrast to these patterns, the human intestine begins to synthesize
lactase at approximately the lOth week of gestation, but levels remain low
until 27 to 32 weeks, when enzyme levels increase rapidly. 5 7 In view of the
relatively low lactase levels in infants born prematurely (28 to 32 weeks), it
is of interest that few develop clinical signs of carbohydrate intolerance. It
has been demonstrated 58 - 60 that lactose is malabsorbed in these infants and
reaches the colon where it is salvaged by colonic flora.
In the majority of the world's population who develop "late onset lactase
deficiency'' in mid-childhood, the pattern is similar to that in other mammals;
the late gestational rise is followed by persistence of lactase specific activity
until approximately 5 to 7 years with a fall thereafter to low adult levels. 20
In the Caucasian population, especially those peoples from or derived from
Northern Europe, and certain localized clusters of other racial origins, lactase-
specific activity rises late in gestation and remains at, or slightly below, this
level throughout adult life-in other words, the maintenance of a juvenile
trait (summarized in References 2, 4).
C. LACTASE DISTRIBUTION AND HORMONAL EFFECTS

Intestinal cell fractionation methods, pioneered by Miller and Crane, 61
coupled with the Dahlqvist assay, 6 demonstrated that lactase, as well as other
intestinal enzymes, was associated with the microvillus membrane of the
absorptive epithelial cells of the small intestine. Nordstrom et a!. 62 first de-
scribed the characteristic crypt to villus increase in lactase activity, which is
typical of a number of digestive enzymes and has been termed the "vertical"
activity gradient. 63 Studies from Kretchmer's laboratory first described the
rapid increase in lactase-specific activity prior to birth, peak around birth,
and decline at weaning, which represents the developmental pattern of rat
lactaseY Shortly after, studies from the laboratory of Auricchio 64 described
the pattern in human intestine. These studies also delineated the pattern of a
peak of activity in the mid intestine, with decreased activity proximally and
distally, the characteristic distribution of lactase in the intestine, which has
been termed the "horizontal" gradient of enzyme activity by Gordon. 63 New-
comer and McGill 45 showed that both normal and deficient human adults
displayed this pattern, with the "normal" subjects having markedly elevated
activities along the intestine, except at the most proximal and most distal
points It should be emphasized that in adult humans as well as adult rats and
rabbits, even though lactase activity levels are decreased significantly from
levels found in sucklings, there are still measurable levels of activity. At the
time of weaning, lactase activity is not decreased to zero nor "switched off".
Animal studies indicate that thyroxine is a major regulator of the devel-
opmental pattern of lactase at the time of weaning in rats. 65 Hypophysectomy
and thyroidectomy retard the decrease in lactase activity during the third
postnatal week, while thyroxine replacement restores the normal pattern. 65
Glucocorticoids enhance rat lactase activity in the first weeks of life. 66 Recent
detailed studies in rats demonstrated cooperative effects of thyroxine and
glucocorticoid hormones in modulating the postnatal development of lactase. 67
Malo and Menard68 observed an increase in lactase-specific activity following
injection of epidermal growth factor (EGF); Foltzer-Jourdainne and Raul 69
have recently reported on the effects of EGF on intestinal enzymes. In hy-
pophysectomized animals, the decrease is retarded but still occurs. 65 Studies
of transplanted fetal intestine have shown that the decrease in lactase activity
occures according to the age of the tissue, not of the host. 70 Thus, the un-
derlying pattern appears to be programmed or intrinsic to the tissue, not
dependent on hormonal changes.
There are few data available on hormonal regulation of lactase in human
intestine. Experiments with explants of 12- to 14-week gestation human in-
testine showed that hydrocortisone induced an increase in lactase levels. 71
However, the major increase in human lactase occurs in the last several weeks
of fetal life. A late gestational upsurge in fetal serum cortisol levels has been
described in humans, 72 but whether or not this has a regualtory role in humans
is currently unknown.
VI. CELLULAR AND MOLECULAR BIOLOGY

OF LACTASE
One hypothesis to explain the decrease in lactase-specific activity, which

occurs at weaning, that has received considerable attention is structural mod-
ification of the enzyme such that its activity is decreaased. This idea has
prompted numerous investigations of the structure of lactase.
Despite a sizeable literature devoted to the measurements of the activity
of lactase in various animal and human populations, and many studies of
lactose intolerance in humans, there remain important gaps in our knowledge
of the synthesis, glycosylation, and processing of lactase, although an overall
pattern has emerged in the last several years. A number of mechanisms have
been proposed to explain the decrease in lactase-specific activity at weaning,
although no consensus has yet emerged. In addition, there still remain a

number of important, unresolved questions, such as the number of active sites
on the enzyme and the structure of the active enzyme on the microvillus
membrane.
A. STRUCTURE AND FUNCTION

Lactase exhibits at least three characteristic enzyme activities: lactase (!3-
D-galactoside galactohydrolase, EC 3.2.1.23), phlorizin hydrolase (phlorizin
glucohydrolase, EC 3.2.1.62), and glycosylceramidase (EC 3.2.1.45-46). 73 - 78
In addition, microvillus membrane lactase is one of three intestinal epithelial
cell enzymes with 13-galactosidase activity. The enterocyte also contains a
lysosomal acid 13-galactosidase, which in addition to its main function also
hydrolyzes lactose, and a cytosolic 13-galactosidase, which has no specificity
for lactose. 74 •75
Lactase has been studied primarily in relation to its hydrolysis of lactose,
although the enzyme also hydrolyzes a number of other less well known
substrates. Earlier publications76 - 78 indicated that lactase also exhibited gly-
cosylceramidase activity. However, in those studies, the enzyme was isolated
using column chromatography, and the possibility that more than one protein
had been copurified was never excluded. Recent work from our laboratory
clearly demonstrates the presence of multiple enzyme specificities (not only
lactase, but also glycosylceramidase activities) on the single protein immu-
noprecipitated with a monoclonal antibody from intestinal epithelium in young
and adult rats. 73 Additional studies of immunoprecipitated human lactase have
also demonstrated enzymatic activities against lactose and phlorizin, as well
as several glycolipids. The strong affinity of this enzyme for glycosylcer-
amides, in combination with the finding of appreciable amounts of active
lactase enzyme in the adult intestine, suggests a possible role for lactase in
adulthood in the digestion of glycolipids. 73
Although multiple substrate specificity for lactase has been well estab-
lished, the number and location of the active sites is still unclear. Heat
inactivation data suggested that there were two sites. 79 The active sites of
lactase have been studied using conduritol-!3-epoxide (CBE) as an affinity
label. 8° CBE is an inhibitor that binds covalently to amino acids in the groups
that are essential for the hydrolysis of glucosides. CBE, first introduced by
Legler, 81 has been used in the study of a variety of 13-glucosidases81 •82 as well
as the 13-glucosidase, sucrase-isomaltase. 82 •83 Preliminary results showed that
CBE inactivates lactase, and all other known galactosidase and glucosidase
activities, in human and rat lactase immunoprecipitated with a monoclonal
antibody. These findings are the first to identify the inactivation of a galac-
tosidase by CBE. Stoichiometric studies showed that CBE bound to lactase
in a molar ratio of two moles of CBE to one mole of lactase from both rat
and human intestine, strongly indicating the existence of two distinct active
sites. 80
B. GLYCOSYLATION, PROCESSING, AND TRANSPORT

As with other membrane glycoproteins,x4 glycosylation influences intra-
cellular transport and degradation of lactase, 85 but whether or not glycosylation
plays a role in intracellular targeting of lactase to the MVM or in the devel-
opmental changes in expression of lactase activity on the MVM remains
unclear.
In a compositional analysis of chromatographically purified rat lactase,
Birkenmeier and Alpers 79 found that the enzyme was a glycoprotein containing
17% carbohydrate by weight. The carbohydrate composition suggested the
presence of both N-linked and 0-linked oligosaccharides. Recent studies 86
have shown that the initial intracellular form of lactase could readily be labeled
with [2- 3 H] mannose, suggesting that this form represents the high mannose-
type glycosylated precursor. In follow-up studies, 87 confirmation was obtained
by the susceptibility of this 205-kDa intracellular precursor peptide to both
endo-H and N-glycanase. These high mannose oligosaccharides are then mod-
ified by a series of glycosidases and glycosyltransferases located in the ER
and Golgi complex, as described for many glycoproteins. 88 Further 0-linked
glycosylation may take place in the trans-Golgi complex.
Following final modification of the carbohydrate composition in the
Golgi, rat lactase is transported by an as yet unknown mechanism, and
inserted into the MVM as a complex glycosylated, high molecular mass
precursor of 220 kDa. 86 Studies of human biopsies in organ culture and
Caco-2 cells have documented a similar pattern, although the human protein
is somewhat larger. 89. 91 Danielsen et al. 92 presented evidence that human
lactase was cleaved intracellularly to the mature 160-kDa form prior to its
arrival on the cell surface. Based on these data, Mantei et al. 93 hypothesized
that the intracellular cleavage step produced lactase and another undiscovered
enzyme from a large synthetic precursor, as discussed below. In the rat, the
220-kDa form is proteolytically cleaved on the microvillus membrane in two
separate steps: first, from 220 to 180 (a transient intermediate observed by
metabolic labeling) and then from 180 to the final130-kDa form. 86 A similar
analysis in humans did not identify an intermediate step, but direct processing
from a 240-kDa precursor to the 160-kDa cell surface form. 90 •92 The proteo-
lytic processing, most likely by integral membrane proteases, is independent
of the presence of luminal (pancreatic) proteases, 86 in contrast to sucrase-
isomaltase where the role of pancreatic proteases has been clearly estab-
lished. 94 •95 Recent studies in rat intestine by Yeh et at.96 support the model
of final cell surface processing of lactase by integral membrane proteases.
Additional support for this model is provided by the data of Nairn et al. ,97
who showed that transfection of COS cells with a full length human lactase
eDNA resulted in the synthesis and transport to the cell surface of both a
high mannose and a complex glycosylated high molecular weight form of
lactase, with full enzymatic activity. Addition of exogenous trypsin cleaved
Intracellular MVM
ER
cc
195 kDa 205
Nascent High
Mannose
'I' High Mannose & Complex Processed < Active Site
i' Processed High Mannose • 0-Linked
FIGURE 1. Representational model of the synthesis and processing of rat lactase-phlorizin

hydrolase. See text for details. (Data from Reference 86.)
the cell surface enzyme to a 160-kDa form very similar to the mature form
found in normal intestine. The authors feel, however, that because the COS
cells are not epithelial cells, the processing of lactase may be aberrant. A
schematic diagram summarizing the data from studies in the rat on lactase
synthesis and processing is shown in Figure 1.
Studies of intestinal microvillus membrane composition have demon-
strated a change at the time of weaning from high sialic acid to high fucose
content. 98 Rat intestinal MVM glycoproteins undergo sialylation during the
suckling period, while fucosylation begins at weaning, indicating develop-
mental alterations in glycosylation. Concurrent with this change, the enzymes
mediating fucosylation and sialylation also change in activity. 99 Based on
lectin binding studies, Nsi-Emvo et al. 100 proposed that the postweaning de-
cline of lastase activity is due to changes in glycosylation with age, leading
to formation of an unprocessed, inactive precursor. These data remain to be
confirmed.
So far, no other studies have reported an inactive, high molecular weight
precursor form of lactase in adult intestine, although a recent study from
Gray's laboratory describes the accumulation of an inactive 100-kDa pro-
cessed form of the enzyme. 101 On the contrary, lactase immunoprecipitated
with a monoclonal antibody shows comparable protein forms in suckling and
adult animals. 102 When the immunoprecipitated enzyme is analyzed, suckling
and adult forms show identical specific activities as well as Km values. 73
Recent studies of the age-specific changes in glycosylation of rat lactase
indicate that the core structure, consisting of both N-linked and 0-linked
oligosaccharides, remains constant during development, although terminal
sugars shift from predominantly sialic acid during the suckling period to fucose
in adulthood. This alteration in glycosylation of the protein occurs in a dif-
ferent pattern from the postweaning decline in lactase-specific activity. Con-
sequently, age-dependent changes in glycosylation are unlikely to result in
an inactive precursor nor account for the decrease in lactase-specific activity
observed during development. 103
C. MOLECULAR STRUCTURE OF LACTASE

Recently, the primary amino acid structures of human and rabbit lactase
have been deduced from the sequence of the cloned cDNAs. 93 The two showed
almost 80% homology at both the nucleotide and amino acid level. Human
lactase is encoded by an mRNA 6274 nucleotides in length. An unexpected
finding was that the mRNA is much longer than required to code for the
microvillus form of the mature active protein. The amino terminus of human
lactase starts at amino acid 869, beginning the sequence AFfFPS .... 127
The subsequent sequence represents the luminal portion of the enzyme, in-
cluding the active site(s), with the 3' portion coding for the hydrophobic,
presumptive transmembrane anchor sequence. It is presently unknown what
function, if any, is represented by the nearly 2 kb of mRNA which do not
code for lactase. Mantei et a!. 93 suggested that it directed the synthesis of a
separate, active enzyme which still remains to be discovered. Because of the
difficulty in estimating sizes of such large glycoproteins on SDS-PAGE, it
is not clear if the large precursor forms of lactase reported in several studies
may be of sufficient size to encompass the complete initial synthetic product.
Ifthe size is estimated from the 1927 amino acids which code for the complete
human sequence from initiation site to stop site, and taking 110 daltons as
an average amino acid, the estimated size of the nonglycosylated precursor
is approximately 210 kDa. This is reasonably close to the size of 215 kDa
estimated from immunoprecipitated nascent human lactase analyzed by SDS-
PAGE. 90 Whether the proteolytic cleavage of the amino terminal fragment of
the lactase molecule occurs intracellularly or at the cell surface, if it is no
longer attached when lactase reaches the cell surface, this fragment may
rapidly disappear into the intestinal lumen and be lost. The complete nucleo-
tide sequence for rat lactase has recently been published. 104 Rat lactase mRNA
is of a similar size to that reported for human and rabbit. 104 • 105 The three
eDNA sequences reported to date show a high degree of homology and
structural similarity (Figure 2), including the location of possible glycosylation
sites.
Analysis of the derived amino acid sequence reveals a fourfold symmetry,
including a high degree of homology among all four regions, suggesting the
possibility that the large enzyme molecule arose through gene duplication. 93
Montgomery et at. 39
r
oc
r
F.bb
l
F.c
r
~F.·
r; . ~;l;i
D EDDCCCYC
J.... l :r: :r: n:r::r: :r: :r:

Rabbit~:.,
CCH'
1;¢1:, ;,.·
CCCC E H'CE' CDC D
, .. • 1 .7 •••
EC H'Cr
,, ,.,:I;, ••
DC EDD EC H' E'
:~D EDDCCCYC
~0N
Rat ~il ~ , , Xl~:r:~

II II r ' " " :r:DC' :r:'~'l, EDD
, rEC H'CE' :r:
H·
CCCH' CCCC E H'CE' CDC D EC D EDDCCCYC
AA 500 1000 1500 2000
I cm.,IOO AA
H• Conserved AA at active site (residue)
Y• N-Glycosylation site
FIGURE 2. Comparison of derived amino acid sequences from human, rabbit, and rat lactase-
phlorizin hydrolase. Region II begins at approximately residue 360; region III at approximately
882; and region IV at approximately 1370. (Prepared by Dr. 1. Dekker; data from References
93 and 104.)
Studies of chromosome loss in cell hybrids indicate that the gene for human
lactase is located on chromosome 2; 106 there are no data available on the
chromosomal location of rat or rabbit lactase. Hybridization of the human
lactase probe to human DNA digested with a panel of restriction enzymes in
a search for restriction fragment length polymorphisms (RFLP) revealed a
common Mspl polymorphism. This RFLP may provide a marker for the lactase
gene in genetic analysis. 107
It has recently been shown that intestinal isomaltase exhibits sequence
homology with lysosomal alpha-glucosidase, suggesting a common evolu-
tionary origin. ~ A similar relationship might be postulated for lactase and
10
lysosomal enzymes with similar specificities. However, Mantei et al. 93 com-

pared their human lactase sequence with that of lysosomal glucocerebrosidase
and found no homology. Although the cytosolic and lysosomal beta-galac-
tosidases from the enterocyte are not well characterized, the complete nu-
cleotide sequence of human lysosomal beta-galactosidase from placenta has
recently been published. 109 Comparison of this sequence with that of human
lactase by dot plot and sequence alignment (our unpublished data) showed
no homology at either the nucleotide or amino acid level. A provocative recent
study found that a newly cloned ~-glucosidase (cellobiase) from Clostridium
thermocellum showed extensive homology to human lactase, particularly at
the putative active sites. 110 The Clostridium enzyme is homologous to one of
the duplicated regions of human lactase, providing support for the idea that
the enzyme structure is the result of duplication of the gene. This provides a
link between a bacterial enzyme and the three mammalian enzymes which
have been analyzed, but does not address the issue of intermediate steps or
how the two may be related.
D. COMPARISON OF LACTASE AMINO ACID SEQUENCES

FROM THREE SPECIES
Although functional data are currently lacking, alignment of the derived
amino acid sequences from the three species so far available, human, rabbit,
and rat, demonstrates some intriguing similarities, as shown in Figure 2. A
number of important amino acids are conserved in all three species, suggesting
that they play important structural or functional roles in the protein. All of
the sequences show the same fourfold symmetry, suggesting that the postu-
lated gene duplication was a relatively ancient event in mammals. Based on
the Clostridium data, the active site is the segment containing histidine and
glutamic acid which is in the same relative position in the Clostridium enzyme
and the human segment IV. These positions are conserved in all three species.
In addition, aspartic acid residues, which may also be important components
of the active sites, are highly conserved, as shown by their consistent position
in all three lactases. The potential N-glycosylation sites are also largely con-
served, although there are some exceptions, such as position 1033 in the rat.
All three species also show very similar cytoplasmic tails, with an inner
positive segment, a membrane spanning hydrophobic region, and an outer
negative region. Each has a tyrosine residue, potentially a phosphorylation
site, on the cytoplasmic portion, and a cysteine residue, potentially a pal-
mitoylation site, in the membrane spanning region. In region I, there are two
areas which are not homologous to the other three regions, but which do show
homology to the corresponding regions I in the other lactases. These areas
contain conserved cysteine residues in the same positions in each species,
suggesting that they may be important in folding and processing of the protein.
The conservation of these regions may have important implications for un-
derstanding the fate of the large molecular mass synthetic precursor or its
function. If this region is included in the precursor form which is inserted
into the microvillus membrane, 86 •89 •97 it may contain targeting information.
On the other hand, there appear to be amino acids homologous to the active
site of the Clostridium enzyme in both regions I and II; so, there may be a
second, as yet unidentified, enzyme produced, as suggested by Mantei eta!. 93
The other three regions also contain cysteine residues, which are conserved
in all three species, and, thus, may be important in protein folding.
VII. REGULATION OF LACTASE DECREASE
The major focus of interest in lactase has been the mechanism which
regulates the decrease in lactase activity at weaning. As discussed above, the
Montgomery eta/. 41
most obvious explanation would be that lactase levels depend on the level of
the substrate lactose, so that when lactose decreased in the diet, lactase
decreased also. However, it is now clear that enzyme activity levels do not
depend upon ingested lactose. The available data on lactase structure, func-
tion, size, and glycosylation provide no compelling evidence for changes in
any of these attributes which occur with weaning and account for the decrease
in enzyme-specific activity. On the contrary, the available evidence indicates
that the same protein is present before and after weaning, or in lactase suf-
ficient and deficient humans. Those individuals with low levels of activity
have less lactase protein. Although a number of other mechanisms have been
suggested and investigated, no consensus has been reached and the subject
is still controversial.
Tsuboi eta!. 110 suggested that the decline in the specific activity of lactase
was dependent upon a change in the rate of enterocyte turnover, and that its
synthesis was constant, resulting in a decrease in specific activity. At the
time of weaning, as shown by others and confirmed by Tsuboi et a!., llO cell
migration rates do increase while lactase-specific activity is decreased. How-
ever, a recent study by Yeh showed that corticoid injection induced both an
increase in cell migration and an increase in lactase levels, not a decrease,
in suckling rats, 111 suggesting that a change in enterocyte migration rate is
not sufficient to account for the decrease in lactase activity. In contrast, Jonas
et al. 112 found that lactase synthesis fell in parallel with the decline seen in
lactase-specific activity during development, indicating a decrease in lactase
synthesis as a mechanism. Smith and James 113 suggested that a two-stage
process was involved: early in the postnatal period, the fall in the specific
activity of lactase was due to increased cell turnover, and later was dependent
upon a decreased rate of synthesis. Another mechanism which has received
little attention is increased degradation of lactase. Studies of dietary effects
on adult lactase levels suggest that these may occur at least partially through
modifications of lactase degradation. 114 It is known that the level of luminal
proteases, which modulate lactase turnover, increases with weaning. 115 How-
ever, there is currently no direct evidence of increased turnover in the decline
in lactase-specific activity.
The studies described all measured lactase-specific activity, which may
lead to deceptive conclusions. The studies in our laboratory 112 indicated sub-
stantial synthesis of lactase even in adult rats, suggesting that the whole
concept of lactase decline at weaning needed reexamination. In fact, our
recent work 73 shows that total lactase activity remains nearly constant from
day 12 to adulthood during development of rat intestine, although the specific
activity decreases over time. This is due to a substantial increase in total
cellular protein and a probable reduction in the number of lactase enzyme
molecules per cell. These findings are in agreement with studies by Ekstrom
et a!., 116 demonstrating significant total lactase activity in pigs past the time
of weaning, as well as observations by Moog of significant postweaning
levels of lactase in mice. 117 This persistence of lactase activity into adulthood
challenges previous concepts of the regulation of the enzyme and alters the
interpretation of previous data in the literature suggesting a marked loss of
enzyme activity during maturation. It also implies that this enzyme might
maintain important function after the weaning period.
Lactase protein levels reflect the synthetic rate of the protein, which in
tum is dependent upon the steady-state level of mRNA. There is presently
no convincing evidence of a significant role for posttranslational mechanisms,
such as modifications of glycosylation, as an overall mechanism for the
developmental pattern of lactase activity. Increased degradation may play a
major role, but there is currently little supportive evidence for such a step.
As described below, the primary regulation of lactase levels in the small
intestine is probably transcriptional, as has been found for most other mam-
malian proteins. 118
A. MOLECULAR BIOLOGY OF LACTASE DECLINE

The cloning of human, rabbit, and rat lactase probes has made it possible
to examine lactase mRNA levels and extend studies of regulatory mechanisms
to the level of gene transcription.
The first such study was reported by Sebastio et al. who presented an
examination of the molecular basis of lactase regulation in a group of surgical
patients in Naples. 119 They found high enzyme activity with high mRNA,
low enzyme activity with high mRNA, and high enzyme activity with low
mRNA in different individuals. Based on this lack of correlation between
enzyme and mRNA levels, the authors concluded that lactase enzyme levels
must be regulated posttranscriptionally. Freund et al. 120 have recently pre-
sented similar data from a rat model in which they found discordant changes
in lactase mRNA levels and enzyme activities. They suggest that this was
due to posttranscriptional regulation of lactase enzyme levels and that regu-1
lation of lactase levels in jejunum and ileum occurs by different mechanisms.
In contrast, a comparison of the level of rat lactase mRNA and total
lactase enzyme activity during development indicated that the two changed
coordinately, 105 as shown in Figure 3. The pattern of regional distribution
along the intestine was not completely coordinated with enzyme activity, but
appeared to be affected by the level of pancreatic proteases. These findings
are consistent with a fundamental transcriptional regulation, modified by
specific regional factors.
Two recent studies in which lactase synthesis was examined suggest that
human intestinal lactase levels are determined by the rate of synthesis of the
enzyme. Sterchi et al. 121 recently reported a careful examination of lactase
synthesis in 18 lactose tolerant and 14 intolerant patients. Their data indicate
that the lactose intolerant subjects had reduced levels of lactase enzyme, due
TOTAL LACTASE ACTIVITY AN 0

RELATIVE LACTASE mRNA LEVELS
B Total Activity
~ mRNA
?d 14d 21d AM LF
FIGURE 3. Coordinate changes in total lactase mRNA and total lactase enzyme activity during
development of rat intestine. The time shown along the horizontal axis refers to postnatal age
in days. AM refers to adult males; LF refers to lactating females. The increased lactase activity
seen in LF is probably a posttranscriptional event. (Data from Reference 105.)
to a greatly decreased rate of lactase synthesis compared to the tolerant sub-

jects. The structure of lactase was identical in both groups. They also identified
some specific defects in intracellular transport and processing in some of the
individuals with reduced lactase activity. Witte et a!., 122 identified markedly
reduced synthesis in three deficient patients with low lactase enzyme levels
and decreased conversion of precursor to mature lactase in a fourth deficient
patient. Together, these studies suggest that (1) the major determinant of
lactose "intolerance" is reduced synthesis of lactase enzyme, and (2) within
racial groups which commonly display high lactase activity, there are indi-
viduals with specific defects in synthesis and processing. The overall pattern
would be consistent with a decrease in levels of lactase mRNA in those
individuals with reduced levels of lactase enzyme synthesis.
A further possible mechanism is suggested by the recent report by Maiuri 123
of both complete absence and mosaic expression of lactase immunoreactivity
in the intestinal epithelium in a group of adult patients with hypolactasia.
This pattern is very similar to that demonstrated in young mouse intestine by
Schmidt et a!. , 124 when the crypt cells are still polyclonal in origin. In mice,
the crypts become monoclonal, that is, derived from a single lineage stem
cell type, by two weeks of age. Such data are not available on human crypt
cell lineages.
TABLE 3
Distribution of Sucrase/Lactase Activity
Ratios and Presence of Lactase mRNA
( +) in Human Intestinal Biopsy Samples
in Three Racial Groups
(Number of subjects)
Data Orientals Blacks White
Ratio> 2 12 4 6
Ratio< 2 0 2 7
Total 12 6 13
Lactase mRNA ~ 12 4 6
and ratio> 2
Lactase mRNA + 0 2 7
and ratio< 2
Note: A ratio less than 2 signifies elevated or "nor-

mal" lactase enzyme activity. The patients with
ratios below two and positive for lactase mRNA
are the same individuals.
B. MOLECULAR BIOLOGY OF LACTASE PERSISTENCE

IN HUMANS
Data from our laboratory indicate that human lactase expression is a direct
consequence of the corresponding concentration of lactase mRNA. 125 We
correlated lactase-specific activity levels and expression of lactase mRNA in
biopsies from adult human orientals, blacks, and whites. Because of the
variability in enzyme levels obtained by biopsy, the ratio of sucrase to lactase
activity has been shown to be a useful criterion of lactase deficiency. 46 All
of the Orientals examined had low levels of lactase activity (and high ratios)
and no detectable lactase mRNA. Two groups of black and white patients
could be distinguished. One had high lactase levels and a ratio of sucrase to
lactase specific activities less than two, and readily detectable lactase mRNA.
The second group had ratios greater than two and did not have readily de-
tectable lactase mRNA. There was an exact correlation between high levels
of lactase mRNA and high levels of lactase enzyme expression (Table 3 and
Figure 4). Similar studies by Lloyd et al. reached the same conclusions. 126
Therefore lactase in humans also appears to be regulated at the level of
transcription. 125
VIII. CONCLUSIONS
Although the literature still frequently refers to "lactose intolerance" and
"lactase deficiency" as if these were abnormalities, it is clear that low levels
Oriental Block White
23456789 10 H 12 B 14 15
Lactose mRNA
Actin mRNA
60 r--------------------------------.
6
LACTASE/ACTIN mRNA
FIGURE 4. Comparison of lactase activity and mRNA levels in biopsies from human intestine.
(A) The Northern blot shows results from 5 Oriental, 3 black, and 7 white patients. Four 11-g of
total human intestinal RNA were prepared and hybridized to an appropriately labeled human
eDNA probe (h-LPH-cDNA-1 ). 125 The blots at the bottom of the figure represent hybridization
to a human 13-actin probe. (B) Lactase mRNA content/actin mRNA content were obtained by
densitometry and a ratio calculated; this was compared to lactase specific activity for each patient:
(•) Oriental patients, (6) black patients, (O) white patients. (Data from Reference 125; reprinted
with permission of the Rockefeller University Press, New York.)
of lactase, and the associated intolerance for significant intake of lactose, are
the normal condition for most of the world's human population. An extensive
literature has correlated the occurrence of lactose intolerance with decreased
levels of lactase in the small intestine. However, it is worth reiterating that,
with the exception of frank disease states, humans (as well as most mammals
examined) have measurable levels of lactase in their small intestine. As dis-
cussed above, "lactase" may actually have other functions in the adult in-
testine, explaining why it is still present. In this sense, rats and rabbits are
useful models for studying the decline in lactase levels in humans. The cur-
rently available evidence suggests that the decreased expression of lactase
protein is a reflection of decreased expression of lactase mRNA.
The overall pattern in the human population thus appears similar to that
described in rats, where enzyme levels and mRNA levels are coordinate in
development. 105 Elevated lactase levels in specific population groups, which
have been correlated with the presence of an autosomal dominant gene, appear
due to an elevated transcription of lactase mRNA. The molecular control of
this elevated expression remains to be elucidated.
The regulation of the decline in lactase levels is probably a complex
combination of factors, including decreased mRNA transcription, the resulting
decreased enzyme synthesis, and increased enzyme degradation. The unique
and more interesting question is what mechanism controls the persistence in
some human groups of high juvenile levels of lactase into adulthood.
IX. SUMMARY
The small intestinal enzyme lactase-phlorizin hydrolase, which hydrolyzes
lactose, the sole carbohydrate in milk, plays a critical role in the nutrition of
the mammalian neonate. Milk is also an important source of nutrients, es-
pecially calcium, for many adults, particularly those of Northern Europe and
North America. Lactose intolerance, which can preclude significant milk
intake, is common in adult humans, usually due to low levels of small intestinal
lactase. Low lactase levels result from either intestinal injury, or, in the
majority of the world's adult population, alterations in the genetic expression
of lactase. Although the mechanism of decreased lactase levels has been the
subject of intensive investigation, no concensus has yet emerged. Recent
studies have begun to define the cellular and molecular biology of this enzyme.
In animals and humans, a glycosylated precursor is proteolytically cleaved
to yield the mature enzyme on the microvillus membrane of the enterocyte,
bound to the lipid bilayer only by a hydrophobic anchor sequence. The enzyme
hydrolyzes lactose, phlorizin, and glycosylceramides. A decline in lactase-
specific activity occurs at the time of weaning in most mammalian species;
in the majority of humans, who have low lactase activity as adults, the decline
Montgomery eta!. 47
occurs at approximately 5 to 7 years of age. In a few human groups, the

elevated juvenile level of lactase-specific activity persists throughout adult-
hood. These developmental patterns of lactase expression are influenced by
a number of factors, but most likely the fundamental regulation is at the level
of gene transcription.
ACKNOWLEDGMENTS
Portions of this chapter are reprinted from the FASEB Journal with per-
mission of the Federation of American Societies for Experimental Biology,
Rockville, MD 20852.
Work in the authors' laboratories has been supported by National Institutes
of Health research grant DK-32658; a grant from the Whitaker Health Sciences
Fund; grant P30 DK 34928 to the Center for Gastroenterology Research on
Absorptive and Secretory Processes, New England Medical Center Hospitals;
a grant from Nutricia, Zoetermeer, The Netherlands; and a NATO collabo-
rative research grant.
REFERENCES
I. Paige, D. M. and Bayless, T. M., Eds., Lactose Digestion, The Johns Hopkins Uni-
versity Press, Baltimore, 1981.
2. Scrimshaw, N. S. and Murray, E. B., The acceptability of milk and milk products in
populations with a high prevalence of lactose intolerance, Am. J. Clin. Nutr., 48, 1083,
1988.
3. Johnson, J, D., Kretchmer, N., and Simoons, F. J,, Lactose malabsorption: its biology
and history, in Advances in Pediatrics, Vol 21, Schulman, 1., Ed., Year Book Medical
Publishers, Chicago, 197, 1974.
4. Flatz, G., Genetics of lactose digestion in humans, in Advances in Human Genetics,
Harris, H. and Hirschhorn, K., Eds., Plenum Press, New York, I, 1987.
5. Jacobi, A., Milk sugar in infant feeding, Trans. Am. Pediatr. Soc., 13, 150, 1901.
6. Dahlqvist, A., Method for assay of intestinal disaccharidases, Anal. Biochem., 7, 18,
1964.
7. Dahlqvist, A., Hammond, J, B., Crane, R. K., Dunphy, J, V., and Littman, A.,
Intestinal lactase deficiency and lactose intolerance in adults, Gastroenterology, 45, 488,
1963.
8. Cuatrecasas, P., Lockwood, D. H., and Caldwell, J, R., Lactase deficiency in the
adult, a common occurrence, Lancet, I, 14, 1965.
9. Bayless, T. M. and Rosensweig, N. S., A racial difference in the incidence of lactase
deficiency: a survey of milk intolerance and lactase deficiency in healthy males, lAMA,
197, 968, 1966.
10. Savilahti, E., Launiala, K., and Kuitunen, P., Congenital lactase deficiency: a clinical
study on 16 patients, Arch. Dis. Child, 58, 246, 1983.
11. Mendel, L. B. and Mitchell, P. H., Chemical studies on growth: the inverting enzymes
of the alimentary tract, especially in the embryo, Am. J. Physiol., 20, 81, 1907.
12. Plimmer, R. H. A., On the presence of lactase in the intestine of animals and on the
adaptation of the intestine to lactose, J. Physiol. (London). 35, 81, 1906.
13. Gilat, T., Russo, S., Gelman-Malachi, E., and Aldor, T. A. M., Lactase in man, a
non-adapwble enzyme, Gastroenterology, 62, 1125, 1972.
14. Kogut, M.D., Donnell, G. N., and Shaw, K. N. F., Studies of lactose absorption in
patients with galactosemia, J. Pediatr., 71, 75, 1967.
15. Palmiter, R. D., What regulates lactose content in milk, Nature (London), 221, 912,
1969.
16. Kretchmer, N. and Sunshine, P., Intestinal disaccharidase deficiency in the sea lion,
Gastroenterology, 53, 123, 1967.
17. Crisp, E. A., Czolij, R., and Messer, M., Absence of beta-galactosidase (lactase)
activity from intestinal brush borders of suckling macropods: implications for mechanism
of lactose absorption, Comp. Biochem. Physiol., 88B, 923, 1987.
18. Jenness, R., Regehr, E. A., and Sloan, R. E., Comparative biochemical studies of
milks. II. Dialyzable carbohydrates, Comp. Biochem. Physiol., 13, 339, 1964.
19. Newcomer, A. D., McGill, D. B., Thomas, P. J,, and Hofmann, A. F., Prospective
comparison of indirect methods for detecting lactase deficiency, N. Engl. J. Med., 293,
1232, 1975.
20. Welsh, J.D., Poley, J. R., Bhatia, M., and Stevenson, D. E., Intestinal disaccharidase
activities in relation to age, race, and mucosal damage, Gastroenterology, 75, 847, 1978.
21. Mobassaleh, M., Montgomery, R. K., Biller, J, A., and Grand, R. J,, Development
of carbohydrate absorption in the fetus and neonate, Pediatrics, 75, 160, 1985.
22. Martini, M. C. and Savaiano, D. A., Reduced intolerance symptoms from lactose
consumed during a meal, Am. J. Clin. Nutr., 47, 57, 1988.
23. Bond, J, H. and Levitt, M.D., Investigation of small bowel transit time in man utilizing
pulmonary hydrogen (H 2 ) measurements, J. Lab. Clin. Med., 85, 546, 1975.
24. Levitt, M. D., Production and excretion of hydrogen gas in man, N. Engl. J. Med.,
281, 122, 1969.
25. Hyams, J, S., Stafford, R. J,, Grand, R. J., and Watkins, J, B., Correlation of
lactose breath hydrogen test, intestinal morphology, and lactase activity in young children,
J. Pediatr., 97, 609, 1980.
26. Newcomer, A. D., Screening tests for carbohydrate malabsorption, J. Pediatr. Gas-
troentero/. Nutr., 3, 6, 1984.
27. Barr, R. G., Perman, J, A., Schoeller, D. A., and Watkins, J, B., Breath test in
pediatric gastrointestinal disorders: new diagnostic opportunities, Pediatrics, 62, 393,
1978.
28. Douwes, A. C., Fernandes, J,, and Degenhardt, H. J,, Improved accuracy of lactose
tolerance test in children, using expired H2 measurement, Arch. Dis. Child., 53, 939,
1978.
29. Barr, R. G., Levine, M.D., and Watkins, J, B., Recurrent abdominal pain of childhood
due to lactose intolerance: a prospective study, N. Engl. J. Med., 300, 1449, 1979.
30. Ostrander, C. R., Cohen, R. S., Hopper, A. 0., and Stevenson, D. K., Breath
hydrogen analysis: a review of the methodologies and clinical applications, J. Pediatr.
Gastroenterol. Nutr., 2, 525, 1983.
31. Perman, J, A., Clinical application of breath hydrogen measurements, Can. J. Physiol.
Pharmacal., 69, Ill, 1991.
32. Queen, P. M. and Henry, R. R., Growth and nutrient requirements of children, in
Pediatric Nutrition: Theory and Practice, Grand, R. J., Sutphen, J. L., and Dietz, W. H.,
Eds., Butterworths, Boston, 1987, 341.
33. Rosado, J, L., Solomons, N. W., Lisker, R., and Bourges, H., Enzyme replacement
therapy for primary adult lactase deficiency, Gastroenterology, 87, 1072, 1984.
34. Newcomer, A. D. and McGill, D. B., Clinical importance of lactase deficiency, N.
Engl. J. Med., 310, 42, 1984.
35. Bayless, T. M., Rothfeld, B., Massa, C., Wise, L., Paige, D. M., and Bedine, M.S.,
Lactose and milk intolerance: clinical implications, N. Engl. J. Med., 292, 1156, 1975.
36. Stephenson, L. S. and Latham, M. C., Lactose intolerance and milk consumption: the
relation of tolerance to symptoms, Am. J. Clin. Nutr., 27, 296, 1974.
37. Lisker, R., Aguilar, L., and Zavala, C., Intestinal lactase deficiency and milk drinking
capacity in the adult, Am. J. Clin. Nutr., 31, 1499, 1978.
38. Kocian, J., Skala, I., and Bakos, K., Calcium absorption from milk and lactose-free
milk in healthy subjects and patients with lactose intolerance, Digestion, 9, 317, 1973.
39. Griessen, M., Cochet, B., Infante, F., Jung, A., Bartholdi, P., Donath, A., Loizeau,
E., and Courvoisier, B., Calcium absorption from milk in lactase-deficient subjects,
Am. J. Clin. Nutr., 49, 377, 1989.
40. Cavalli-Sforza, L. T., Strata, A., Barone, A., and Cucurachi, L., Primary adult
lactose malabsorption in Italy: regional differences in prevalence and relationship to
lactose intolerance and milk consumption, Am. J. Clin. Nutr., 45, 748, 1987.
41. Kretchmer, N., Memorial lecture: lactose and lactase-a historical perspective, Gas-
troenterology, 61, 805, 1971.
42. Biiller, H. A. and Grand, R. J,, Lactose intolerance. Annu. Rev. Med., 41, 141, 1990.
43. Bolin, T. D. and Davis, A. E., Primary lactase deficiency: genetic or acquired?, Am.
J. Dig. Dis., 15, 679, 1970.
44. Rosensweig, N. S., Adult lactase deficiency: genetic control or adaptive response?.
45. Newcomer A. D. and McGill, D. B., Distribution of disaccharidase activity in the small
bowel of normal and lactase-deficient subjects, Gastroenterology, 51, 481, 1966.
46. Welsh, J. D., Isolated lactase deficiency in humans: report on 100 patients, Medicine,
49, 257, 1970.
47. Bolin, T. D., Pirola, R. C., and Davis, A. E., Adaptation of intestinal lactase in the
rat. Gastroenterology, 60, 432, 1969.
48. Lebenthal, E., Sunshine, P., and Kretchmer, N., Effect of prolonged nursing on the
activity of intestinal lactase in rats, Gastroenterology, 64, 1136, 1973.
49. Bond, J, H. and Levitt, M. D., Quantitative measurement of lactose absorption, Gas-
troenterology, 70, 1058, 1976.
50. Simoons, F. J., The geographic hypothesis and lactose malabsorption, Dig. Dis. Sci.,
23, 963, 1978.
51. Sahi, T., Isokoski, M., Jussila, J., Launiala, K., and Pyoriilii, K., Recessive inher-
itance and adult type-lactose malabsorption, Lancet, 2, 823, 1973.
52. Sahi, T. and Launiala, K., More evidence for the recessive inheritance of selective
adult type lactose malabsorption, Gastroenterology, 73, 231, 1977.
53. Doell, R. G. and Kretchmer, N., Studies of small intestine during development. I.
Distribution and activity of I)-galactosidase, Biochim. Biophys. Acta., 62, 353, 1962.
54. Bailey, D. S., Cook, A., McAllister, G., Moss, M., and Mian, N., Structural and
biochemical differentiation of the mammalian small intestine during fetal development,
J. Cell Sci.. 72, 195, 1984.
55. Blaxter, K. L., Lactation and the growth of the young. in Milk: The Mammary Gland
and Its Secretion, Vol. II, Kon, S. K. and Cowie, A. T., Eds., Academic Press, New
York. 1961. 305.
56. Montgomery, R. K., Kothe, M. J. C., Biiller, H. A., and Grand, R. J., Rat lactase
mRNA appears concurrently with development of columnar epithelial cells in fetal in-
testine, Gastroenterology, 98, A423(Abstr.). 1990.
57. Antonowicz, I., Chang, S. K., and Grand, R. J,, Development and distribution of
lysosomal enzymes and disaccharidases in human fetal intestine. Gastroenterology, 67,
51, 1974.
58. MacLean, W. C. and Fink, B. B., Lactose malabsorption by premature infants: mag-
nitude and clinical significance, J. Pediatr., 97, 383, 1980.
59. Kien, C. L., Liechty, E. A., Myerberg, D. Z., and Mullett, M. D., Dietary carbo-
hydrate assimilation in the premature infant: evidence for a nutritionally significant bac-
terial ecosystem in the colon, Am. J. Clin. Nutr., 46, 456, 1987.
60. Murray, R. D., Boutton, T. W., Klein, P. D., Gilbert, M., Paule, C. L., and
MacLean, W. C., Comparative absorption of [13C] glucose and [13C] lactose by pre-
mature infants, Am. J. Clin. Nutr., 51, 59, 1990.
61. Miller, D. and Crane, R. K., Digestive function of the epithelium of the small intestine.
II. Localization of disaccharide hydrolysis in the isolated brush border portion of intestinal
epithelial cells, Biochim. Biophys. Acta, 52, 293, 1961.
62. Nordstrom, C., Dahlqvist, A., and Josefsson, L., Quantitative determination of en-
zymes in different parts of the villi and crypts of rat small intestine, J. Histochem.
Cytochem., 15, 713, 1968.
63. Gordon, J. I., Intestinal epithelial differentiation: new insights from chimeric and trans-
genic mice, J. Cell Bioi., 108, 1187, 1989.
64. Auricchio, S., Rubino, A., and Murset, G., Intestinal glycosidase activites in the
human embryo, fetus and newborn, Pediatrics, 35, 944, 1965.
65. Yeh, K. Y. and Moog, F., Intestinal lactase activity in the suckling rat: influences of
hypophysectomy and thyroidectomy, Science, 183, 77, 1974.
66. Koldovsky, 0. and Sunshine, P., Effect of cortisone on developmental pattern of the
neutral and acid beta-galactosidase of the small intestine of the rat, Biochem. J., 117,
467, 1970.
67. Yeh, K. Y., Yeh, M., and Holt, P.R., Thyroxine and cortisone cooperate to modulate
postnatal intestinal enzyme differentiation in the rat, Am. J. Physiol., 260, G371, 1991.
68. Malo, C. and Menard, D., Influence of epidermal growth factor on the development
of suckling mouse intestinal mucosa, Gastroenterology, 83, 28, 1982.
69. Foltzer-Jourdainne, C. and Raul, F., Effect of epidermal growth factor on the expres-
sion of digestive hydrolases in the jejunum and colon of newborn rats, Endocrinology,
127, 1763, 1990.
70. Montgomery, R. K., Sybicki, M. A., and Grand, R. J., Autonomous biochemical
and morphological differentiation in fetal rat intestine transplanted at 17 and 20 days of
gestation, Dev. Bioi., 87, 76, 1981.
71. Arsenault, P. and Menard, D., Influence of hydrocortisone on human fetal small
intestine in organ culture, J. Pediatr. Gastroenterol. Nutr., 4, 893, 1985.
72. Murphy, B. E. P., Human fetal serum cortisol levels related to gestational age: evidence
of midgestational fall and a steep late gestational rise, independent of sex or mode of
delivery, Am. J. Obstet. Gynecol., 144, 276, 1982.
73. Buller, H. A., Wassenaer, A. G. van, Raghavan, S., Montgomery, R. K., Sybicki,
M.A., and Grand, R. J., New insights into the lactase and glycosylceramidase activities
of rat microvillus membrane lactase-phlorizin hydrolase, Am. J. Physiol., 257, G616,
1989.
74. Asp, N. G. and Dahlqvist, A., Human small intestine 13-galactosidases: specific assay
of three different enzymes, Anal. Biochem., 47, 527, 1972.
75. Gray, G. M. and Santiago, N. A., Intestinall3-galactosidases. I. Separation and char-
acterization of 3 enzymes in normal human intestine, J. Clin. Invest., 48, 716, 1969.
76. Brady, R. 0., Gal, A. E., Kanfer, J. N., and Bradley, R. M., The metabolism of
glucocerebrosides, III, Purification and properties of glucosyl- and galactosylceramide
cleaving enzyme from rat intestinal tissue, J. Bioi. Chern., 240, 3766, 1965.
77. Kobayashi, T. and Suzuki, K., The glycosylceramidase in the murine intestine, J. Bioi.
Chern., 256, 7768, 1981.
Montgomery et at. 51
78. Leese, H. J. and Semenza, G., On the identity between the small intestinal enzymes
phlorizin hydrolase and glycosylceramidase, J. Biol.Chern .. 248, 8170, 1973.
79. Birkenmeier, E. and Alpers, D. H., Enzymatic properties of rat lactase-phlorizin hy-
drolase, Biochirn. Biophys. Acta, 350, 100, 1974.
80. Haringsma, J., Biiller, H. A., Matsudaira, P. T., Montgomery, R. K., and Grand,
R. J., Human lactase-phlorizin hydrolase, function and active sites, Eur. J. Clin.Invest.,
20 A26 (Abstr.), 1990.
81. Legler, G., Untersuchungen zum wirkungsmechanismus glycosidspaltender enzyme. III.
Markierung des aktiven zentrums einer beta-glucosidases aus aspergillus wentii mit con-
duritol-b-epoxid, Z. Physiol. Chern .. 349, 767, 1968.
82. Dinur, T., Osiecki, K. M., Legler, G., Gatt, S., Desnick, R. J,, and Grabowski,
G. A., Human acid 13-glucosidase: isolation and amino acid sequence of a peptide con-
taining the catalytic site, Proc. Nat!. Acad. Sci. U.S.A.. 83, 1660, 1986.
83. Quaroni, A. and Semenza, G., Partial amino acid sequences around the essential car-
boxylate in the active sites of the sucrase-isomaltase complex, J. Bioi. Chern .• 251, 3250,
1976.
84. Rousset, M., Trugnan, G., Brun, J, L., and Zweibaum, A., Inhibition of the post-
translational processing of microvillar hydrolases is associated with a specific decrease
of sucrase-isomaltase and an increased turnover of glucose in CaCo-2 cells treated with
monensin, FEES Lett.. 208, 34, 1986.
85. Danielsen, E. M. and Cowell, G. M., Biosynthesis of intestinal microvillar proteins:
processing of N-linked carbohydrate is not required for surface expression, Biochern. J ..
240, 777. 1986.
86. Biiller, H. A., Montgomery, R. K., Sasak, W. V., and Grand, R. J., Biosynthesis,
glycosylation, and intracellular transport of intestinal lactase-phlorizin hydrolase in rat,
J. Bioi. Chern .. 262, 17206, 1987.
87. Biiller, H. A., Rings, E. H. H. M., Montgomery, R. K., Sasak, W. V., and Grand,
R. J,, Further studies of glycosylation and intracellular transport of lactase-phlorizin
hydrolase in rat small intestine, Biochern. J., 263, 249, 1989.
88. Hubbard, C. S. and Ivatt, R. J,, Asparagine-linked oligosaccharides, Annu. Rev.
Biochern .. 50, 555, 1981.
89. Hauri, H. P., Sterchi, E. E., Bienz, D., Fransen, J, A.M., and Marxer, A., Expres-
sion and intracellular transport of microvillus membrane hydrolases in human intestine,
J. Cell Bioi.• 101, 838, 1985.
90. Nairn, H. Y., Sterchi, E. E., and Lentze, M. J,, Biosynthesis and maturation of lactase-
phlorizin hydrolase in the human small intestinal epithelial cells, Biochern. J.. 241,
427, 1987.
91. Skovbjerg, H., Danielsen, E. M., Noren, 0., and Sjostrom, H., Evidence for bio-
synthesis of lactase-phlorizin hydrolase as a single-chain high molecular weight precursor,
Biochirn. Biophys. Acta, 798, 247, 1984.
92. Danielsen, E. M., Skovbjerg, H., Noren, 0., and Sjostrom, H., Biosynthesis of
intestinal microvillar proteins, intracellular processing of lactase-phlorizin hydrolase,
Biochern. Biophys. Res. Cornrn .• 122, 82, 1984.
93. Mantei, N., Villa, M., Enzler, T., Wacker, H., Boll, W., James, P., Hunziker, W.,
and Semenza, G., Complete primary structure of human and rabbit lactase-phlorizin
hydrolase: implications for biosynthesis, membrane anchoring and evolution of the en-
zyme, EMBO J., 7, 2705, 1988.
94. Hauri, H. P., Quaroni, A., and Isselbacher, K. J,, Biogenesis of intestinal plasma
membrane: posttranslational route and cleavage of sucrase-isomaltase, Proc. Nat!. Acad.
Sci. U.S.A., 76, 5183, 1979.
95. Montgomery, R. K., Sybicki, M.A., Forcier, A. G., and Grand, R. J., Rat intestinal
microvillus membrane sucrase-isomaltase is a single high molecular weight protein and
fully active enzyme in the absence of luminal factors, Biochirn. Biophys. Acta, 661, 346,
1981.
96. Yeh, K., Yeh, M., Pan, P., and Holt, P. R., Post-translational cleavage of rat intestinal
lactase occurs at the luminal side of the brush border membrane, Gastroenterology, 1OJ,
312, 1991.
97. Nairn, H. Y., Lacey, S. W., Sambrook, j. F., and Gething, M.-J., Expression of a
full-length eDNA coding for human intestinal lactase-phlorizin hydrolase reveals an un-
cleaved, enzymatically active, and transport-competent protein, J. Bioi. Chern., 266,
12313, 1991.
98. Srivastava, 0. P., Steele, M. I., and Torres-Pinedo, R., Maturational changes in
terminal glycosylation of small intestinal microvillar proteins in the rat, Biochirn. Biophys.
Acta, 914, 143, 1987.
99. Chu, S. W. and Walker, W. A., Developmental changes in the activities of sialyl- and
fucosyltransferases in rat small intestine, Biochirn. Biophys. Acta, 883, 496, 1986.
100. Nsi-emvo, E., Launay, J, F., and Raul, F., Is adult-type hypolactasia in the intestine
of mammals related to changes in the intracellular processing of lactase, Cell Mol. Biol.,
33, 335, 1987.
101. Quan, R., Santiago, N. A., Tsuboi, K. K., and Gray, G. M., Intestinal lactase, shift
in intracellular processing to altered, inactive species in the adult rat, J. Bioi. Chern.,
265, 15882, 1990.
102. Quaroni, A., Pre- and postnatal development of differentiated functions in rat intestinal
epithelial cells, Dev. Bioi., Ill, 280, 1985.
103. Biiller, H. A., Rings, H. H. M., Pajkrt, D., Montgomery, R. K., and Grand, R. j.,
Glycosylation of lactase-phlorizin hydrolase in rat small intestine during development,
104. Duluc, 1., Boukamel, R., Mantei, N., Semenza, G., Raul, F., and Freund, J.,
Sequence of the precursor of intestinal lactase-phlorizin hydrolase from fetal rat, Gene,
103, 275, 1991.
105. Biiller, H. A., Kothe, M. J, C., Goldman, D. A., Grubman, S. A., Sasak, W. V.,
Matsudaira, P. T., Montgomery, R. K., and Grand, R. J., Coordinate expression of
lactase-phlorizin hydrolase mRNA and enzyme levels during development, J. Bioi. Chern.,
265, 6978, 1990.
106. Kruse, T. A., Bolund, L., Grzeschik, K. H., Ropers, H. H., Sjostrom, H., Noren,
0., Mantei, N., and Semenza, G., The human lactase-phlorizin hydrolase gene is located
on chromosome 2, FEBS Lett., 240, 123, 1988.
107. Kruse, T. A., Bolund, L., Byskov, A., Sjostrom, H., Noren, 0., Mantei, N., and
Semenza, G., Mapping of the human lactase-phlorizin hydrolase gene to chromosome
2, Cytogenet. Cell Genet., 51, 1026, (Abstr.), 1989.
108. Hoefsloot, L. H., Hoogeveen-Westerveld, M., Kroos, M. A., Beeumen, J, van,
Reuser, A. J. J., and Oostra, B. A., Primary structure and processing of lysosomal
alpha-glucosidase; homology with the intestinal sucrase-isomaltase complex, EMBO J.,
7, 1697, 1988.
109. Oshima, A., Tsuji, A., Nagao, Y., Sakuraba, H., and Suzuki, Y., Cloning, sequenc-
ing, and expression of eDNA for human beta-galactosidase, Biochem. Biophys. Res.
Comm., 157, 238, 1988.
110. Tsuboi, K. K., Kwong, L. K., D'Harlingue, A. E., Stevenson, D. K., Kerner, J, A.,
and Sunshine, P., The nature of maturational decline of intestinal lactase activity, Biochim
Biophys. Acta, 840, 69, 1985.
Ill. Yeh, K., Yeh, M., and Holt, P. M., Intestinal lactase expression and epithelial cell
transit in hormone-treated suckling rats, Am. J. Physiol., 260, 0379, 1991.
Montgomery eta/. 53
112. Jonas, M. M., Montgomery, R. K., and Grand, R. J., Intestinal lactase synthesis
during postnatal development in the rat, Pediatr. Res., 19, 956, 1985.
113. Smith, M. W. and James, P. S., Cellular origin of lactase decline in post weaned rats,
Biochim. Biophys. Acta, 905, 503, 1987.
114. Goda, T., Bustamante, S., and Koldovsky, 0., Dietary regulation of intestinal lactase
and sucrase in adult rats: quantitative comparison of effect of lactose and sucrose, f.
Pediatr. Gastroenterol. Nutr., 4, 998, 1985.
115. Seetharam, B., Yeh, K. Y., and Alpers, D. H., Turnover of intestinal brush border
proteins during postnatal development in rat, Am. f. Physiol., 239, 0524, 1980.
116. Ekstrom, K. E., Benevenga, N.J., and Grummer, R. H., Changes in the intestinal
lactase activity in the small intestine of two breeds of swine from birth to 6 weeks of
age, J. Nutr., 105, 1032, 1975.
117. Moog, F., Denes, A. E., and Powell, P.M., Disaccharidases in the small intestine of
the mouse: normal development and influence of cortisone, actinomycin D, and cyclo-
heximide, Dev. Bioi., 35, 143, 1973.
118. Johnson, P. F. and McKnight, S. L., Eukaryotic transcriptional regulatory proteins,
Anna. Rev. Biochem., 58, 799, 1989.
I 19. Sebastio, G., Villa, M., Sartorio, R., Guzzetta, V., Poggi, V., Auricchio, S., Boll,
W., Mantei, N., and Semenza, G., Control of lactase in human adult-type hypolactasia
and in weaning rabbits and rats, Am. J. Hum. Genet., 45, 489, 1989.
120. Freund, J. N., Duluc, I., and Raul, F., Lactase expression is controlled differently in
the jejunum and ileum during development in rats, Gastroenterology, 100, 388, 1991.
121. Sterchi, E. E., Mills, P.R., Fransen, J. A.M., Hauri, H. P., Lentze, M. J., Nairn,
H. Y., Ginsel, L., and Bond, J., Biogenesis of intestinal lactase-phlorizin hydrolase in
adults with lactose intolerance, J. Clin. Invest., 86, 1329, 1990.
122. Witte, J., Lloyd, M., Lorenzsonn, V., Korsmo, H., and Olsen, W., The biosynthetic
basis of adult lactase deficiency, J. Clin. Invest., 86, 1338, 1990.
123. Maiuri, L., Raia, V., Potter, J., Swallow, D., Ho, M. W., Fiocca, R., Finzi, G.,
Cornaggia, M., Capella, C., Quaroni, A., and Auricchio, S., Mosaic pattern oflactase
expression by villous enterocytes in human adult-type hypolactasia, Gastroenterology,
100, 359,1991.
124. Schmidt, G. H., Wilkinson, M. M., and Ponder, B. A. J., Cell migration pathway
in the intestinal epithelium: an in situ marker system using mouse aggregation chimeras,
Cell, 40, 425, 1985.
125. Escher, J. C., de Koning, N. D., Engen, C. G. J. van, Arora, S., Biiller, H. A.,
Montgomery, R. K., and Grand, R. J., Molecular basis of lactase levels in adult
humans, J. Clin. Invest., 89, 480, 1992.
126. Lloyd, M., Mevissen, G., Fischer, M., Olsen, W., Goodspeed, D., Genini, M., Boll,
W., Semenza, G., and Mantei, N., Regulation of intestinal lactase in adult hypolactasia,
J. Clin. Invest., 89, 524, 1992.
127. Biiller, H. A., unpublished data.
Chapter 3
REGULATION OF GASTROINTESTINAL LIPASE

GENE EXPRESSION BY DIETARY LIPIDS
Catherine Wicker-Pianquart and Antoine Puigserver
TABLE OF CONTENTS
I. Introduction ....................................................... 56
II. Lipid Digestion within the Gastrointestinal Tract ................. 57

A. Preduodenal Lipases ....................................... 57
B. Pancreatic Lipases ......................................... 59
C. Sequential Events in Fat Digestion and
Absorption ................................................ 60
1. Intragastric Lipolysis ............................... 60
2. Intraduodenal Lipolysis ............................ 60
3. Intestinal Absorption of Dietary and
Biliary Lipids ...................................... 61
D. Significance of Gastric and Pancreatic Digestion .......... 61
III. Adaptative Response of Gastrointestinal Lipases to

Dietary Fat. ....................................................... 62
A. Preduodenal Adaptation ................................... 62
1. Gastrointestinal Transit. ............................ 62
2. Prepyloric Lipases ................................. 63
B. Pancreatic Lipase (and Colipase) Adaptation .............. 63
I. Pancreatic Secretion after Intraduodenal
Perfusate or Meal Intake ........................... 64
a. Intraduodenal Perfusates .................... 64
b. Response to Meal Intake ................... 64
2. Effect of Type of Fat, Chain Length, and
Degree of Saturation ............................... 65
a. Type of Fat. ................................ 65
b. Chain Length ............................... 65
c. Effect of Degree of Saturation ............. 67
3. Response of Pancreatic Lipase and
Colipase to Amount of Dietary Fat ................ 68
a. Lipase-Specific Activity .................... 68
b. Colipase-Specific Activity .................. 68
c. Lipase Biosynthesis ........................ 71
0-8493-6%1-41931$0.00 +$.50
© 1993 by CRC Press, Inc. 55
4. Time Course of the Variations in Lipase

and Colipase Activities and Biosynthesis
in Pancreatic Tissue ................................ 72
a. Lipase Activity and Biosynthesis ........... 72
b. Colipase Activity ........................... 73
C. Possible Regulators of Pancreatic Lipase
Adaptation to Dietary Fat ................................. 73
1. Nutritional Regulators .............................. 73
2. Hormonal Regulators .............................. 75
D. Mechanisms Involved in the Regulation of
Pancreatic Lipase and Colipase by Dietary Lipids ......... 76
1. Outline of the Expression of Pancreatic
Enzymes ........................................... 76
a. General Scheme of Eukaryotic
Gene Expression ........................... 76
b. Synthesis and Subsequent
Processing of Pancreatic Secretory
Proteins .................................... 78
2. Induction of Lipase and Colipase mRNAs
in Response to High Fat Diet ...................... 78
a. Lipase mRNA .............................. 79
b. Colipase mRNA ............................ 80
3. Transcriptional Regulation of Pancreatic
Lipase Gene Expression ........................... 80
IV. Perspectives and General Conclusion ............................. 80
References ................................................................ 83
I. INTRODUCTION
Triacylglycerol is the main component of dietary lipids which are con-

sumed by most mammalian species. It represents 95 to 99% of lipid intake.
This corresponds to a daily intake in human of 100 to 150 g of triglyceride,
vs. 2 to 10 g phospholipids, 0.2 to 0.8 g cholesterol, and about 16 mg
liposoluble vitamins. 1 Intestinal absorption of triglycerides requires prior hy-
drolysis into monoglycerides and fatty acids. This is ensured by the combined
action of prepyloric and pancreatic lipases. The contribution made by each
Wicker-Planquart and Puigserver 57
enzyme to the digestion of triglycerides is influenced by diet and physiological

state of the organism. Other enzymes, phospholipase A2 and carboxyl ester
hydrolase, contribute to lipid digestion2 (see Table 1). Pancreatic phospho-
lipase A2 (MW = 14 kDa) hydrolyzes alimentary and biliary phospholipids,
producing lysophospholipids and fatty acids while pancreatic carboxyl ester
hydrolase (MW = 64 to 100 kDa) hydrolyzes steryl and lipovitamin esters,
which are solubilized in the aqueous phase in the intestinal content, producing
cholesterol, free vitamins, and fatty acids.
We will focus this chapter on the enzymes responsible for triglyceride
lipolysis (preduodenal and pancreatic lipases) and the effects of dietary tria-
cylglycerol on lipolytic activity, with emphasis on the adaptation of pancreatic
lipase to dietary lipid intake. Effects of other dietary components, as well as
neural or hormonal control of lipases, will not be developed. Finally, the
mechanisms of pancreatic lipase and colipase adaptation to dietary lipids
(transcriptional, pretranslational, translational, posttranslational) will be dis-
cussed.
II. LIPID DIGESTION WITHIN THE

GASTROINTESTINAL TRACT
A. PREDUODENAL LIPASES
As early as in 1946, Schonheyder and colleagues demonstrated that the
human stomach was able to digest dietary lipids and reported the presence of
a lipase enzyme in gastric juice. 3 Preduodenallipases have been detected in
most animal species and are widely distributed. Their synthesis and secretion
differ from one species to another. Depending on the species considered,
preduodenallipase is almost entirely produced by the tongue as in the rat and
mouse, 4 •5 pharynx as in the calf and the sheep, 6 ·7 or stomach as in the man,
rabbit, dog, hog, horse, cat, guinea pig, baboon, and macaque. 3 •8 - 11
In the rat, lingual lipase is located in the acinar cells of serous von Ebner
glands and in the lingual mucous gland. 12 ·13 By means of specific polyclonal
antibodies to human and rabbit gastric lipase, human gastric lipase has been
demonstrated to be exclusively located in the chief cells of the fundic mucosa
of the human stomach, 14 - 16 whereas the cells producing rabbit gastric lipase
are restricted in the upper base of the gastric fundic gland. 10 • 17 Lipase activity
in human gastric content is 15 to 75 U/mP 8 or 4700 U/g of fresh mucosa in
men under 50 years. 15 Gastric lipolytic activity decreases in aged persons, to
reach 700 U/g of fresh mucosa in subjects over 60 years of age. 15
Most of the biochemical studies have been performed with human and
rabbit gastric lipases. Preduodenal lipases are acid enzymes (the pH optima
for preduodenal lipases generally fall in the range of 2 to 62 ), which may
work in extreme pH conditions. They can act in the stomach, where the pH
is 4.5 to 5.5 after a meal intake, to initiate the hydrolysis and digestion of
dietary fat. Human gastric lipase does not hydrolyze a tributyrin emulsion in
TABLE 1
Classification of Digestive (Phospho)lipases IJl
01:1
Main site of action

Stomach (prepyloric lipases) Small intestine (postpyloric lipases)
Proposed Generic Names
Acid lipases Colipase-dependent lipases Bile salt-dependent lipases Phospholipases A2
Tissue and Cellular Origin
Tongue (Von Ebner glands) Exocrine pancreas (acinar cells) Human milk Exocrine pancreas (acinar cells)
Pharynx Exocrine pancreas (acinar cells) Intestine (Paneth cells)
Stomach (chief cells)
Other Names
Lingual lipase Pancreatic lipase (EC 3.1.1.3) Bile salt stimulated lipase (BSSL) Phospholipase A2 (EC 3 .I. 1.4)
Salivary lipase Carboxylic ester hydrolase (CEH)
Pharyngeal lipase Cholesterol esterase (CE) =<:
~
Gastric lipase Carboxyl ester lipase (CEL) ~·
Pregastric lipase Monoglyceride lipase c;·
;::
Nonpancreatic lipase Nonspecific lipase
>:l
Preduodenal lipase Lysophospholipase ;:::
l:l...
CJ
Substrate Specificity ~
~
sn-3 Position of glycerides sn-l And sn- 3 positions of Cholesterol esters sn-2 Position of L-o.-glycero- ~
(also sn-l and sn-2 positions glycerides Fat-soluble vitamin esters phospholipids ~
Glycerides ~
of glycerides)
Glycerophospholipids
"'"'
c·
;:::
Primary, secondary, and tertiary
carboxylic esters
Adapted from Reference 2.
the absence of bile salts or alimentary proteins. The presence of most of the
bile salts (taurocholate, glycocholate, taurodeoxycholate, glycodeoxycholate,
or bile salts mixture), which must be added prior to the enzyme in order to
prevent irreversible lipase denaturation, activates human gastric lipase on the
tributyrin substrate. 2 This can be explained by the fact that human gastric
lipase activity is restricted to a triacylglycerol!water interfacial tension range
of 8 to 13 dyn!cm. Interfacial tension at tributyrin/water interface is 15 dyn!cm,
which rapidly inactivates human gastric lipase. The fact that the enzyme has
a single disulfide bridge 19 can explain the rapid interfacial denaturation of
human gastric lipase. Naturally occurring bile salts like taurodeoxycholate in
the concentrations prevailing in the upper small intestine reduce the interfacial
tension to about 9 dyn/cm. Other amphiphiles, like most alimentary proteins,
which decrease the tributyrin/water interfacial tension to values between 8 to
13 dynlcm, also allow the expression of lipase activity.
Prepyloric lipases can hydrolyze a variety of lipid substrates including
milk fat globule triacylglycerols. They are essential for the digestion of milk
fat in the newborn. It was thought until recently that intragastric digestion
was specific for short and medium chain triglycerides and had very little effect
on long chain triglycerides, the major constituents of dietary fat. However,
using purified human gastric lipase, Gargouri et a!. 19 have shown that human
gastric lipase catalyzes the hydrolysis of both long and short chain triacyl-
glycerols at comparable rates. Sn-3 positions of ester bonds in triglycerides
are preferentially hydrolyzed by the enzyme. An important role for gastric
lipolysis is to prepare triacylglycerols for further intraduodenal digestion.
B. PANCREATIC LIPASES
Pancreatic lipase (triacylglycerol acyl hydrolase, EC 3.1.1.3) is, along
with phospholipase A2 , the predominant pancreatic lipolytic enzyme. It amounts
to 4.5% of the total proteins in pancreatic juice. 20 It is synthesized and secreted
by pancreatic acinar cells and is present in duodenum aspirate. The optimum
pH for pancreatic lipase is at around pH 8 to 9. 21 Pancreatic lipase irreversibly
loses its activity below pH 5.0. 21
Pancreatic function is immature in neonates of most species (man, dog,
rat). However, a pancreatic lipase activity, although weak, is detectable during
the last week of fetal development in the rat. 22 Lipase level is still reduced
in the rat during the period from birth to weaning, lipase concentration reach-
ing 40% of that in adult at the end of weaning. 22 However, in healthy adults,
pancreatic lipase is considered to be of prime importance in the digestion of
dietary lipids.
Pancreatic lipase (52 kDa for the pig enzyme) hydrolyzes long chain
triacylglycerols (preferentially by attacking the external chains at positions
sn-1 and sn-3 of glycerides), producing first diacylglycerol, then 2-mono-
acylglycerol and free fatty acids. Pancreatic lipase, similarly to the gastric
enzyme, displays almost no activity when substrate is in the monomeric state. 23
Formation of substrate aggregates or emulsified particles which occurs by

increasing the concentration of substrate above saturation results in a marked
lipolysis by pancreatic lipase. Activation of the enzyme by adsorption to the
interfaces is about 10 3 -fold. 24 Presence of various amphiphiles (bile salts,
phospholipids) in the duodenum inhibits in vivo pancreatic lipase action due
to a desorption of the enzyme from the amphiphile coated interface. Pancreatic
lipase requires the presence of a cofactor, the co lipase, to restore its activity. 25
This cofactor is a small protein (about 10 kDa). cosecreted with lipase in
pancreatic juice. Its primary role is to adsorb to bile-salt covered interface
and to act as an anchor for lipase. 26 Co lipase binds lipase in a 1: 1 complex
either in the presence or in the absence of bile salts. Lipase/colipase binding,
as well as the binding of colipase to bile salts. is enhanced by free fatty
acids. 27·28 Another function for colipase is to stabilize lipase, by protecting it
from proteolytic digestion. Concentrations of lipase and colipase in the gut
are closely equimolar. They are about 2 X 10- 6 M in the small intestine of
man after the ingestion of a liquid test meaP and 2 X 10- 7 M in the pig
under basal conditions. 30
C. SEQUENTIAL EVENTS IN FAT DIGESTION

AND ABSORPTION
With the major mechanical steps of physical-chemical behavior of fat
during duodenal digestion now understood, 31 •32 it is possible to describe more
accurately the sequential levels of dietary lipid digestion.
1. Intragastric Lipolysis
Lipids are mixed and dispersed with the other nutrients in the stomach.
Hydrolysis of alimentary lipids, catalyzed by one of the preduodenallipases
secreted from the serous glands of the tongue or from the gastric mu-
cosa,2·3.11 takes place in the stomach. It has been shown that these lipases are
particularly stable under the acidic conditions prevailing in the stomach (pH 1
to 3) and are activated by various dietary proteins. 33 During in vivo gastric
lipolysis, triacylglycerol is hydrolyzed to primarily a,f3-diacylglycerol and
free fatty acid. Production of free fatty acids facilitates emulsification of
triglycerides and diglycerides. 34 The amount of hydrolysis products (digly-
cerides and free fatty acids) increases over time following ingestion to reach
a maximum corresponding to 10 to 30% 17 to 65% 35 of total hydrolysis. The
lack of complete hydrolysis is most likely due to preduodenal lipolysis in-
hibition by long chain free fatty acids. 33 ·36
2. Intraduodenal Lipolysis
Pancreatic lipase and colipase are discharged into the duodenum and
associate first to bile lipid micelles and vesicles prior to adsorption to emul-
sified particles from the stomach. 37 Triglyceride-rich emulsion particles con-
taining 2 to 3% triglycerides are thus covered by a mixed lipid monolayer
composed of phospholipids, free cholesterol, fatty acids, and vitamins. 38

When hydrolysis of diglycerides and triglycerides by pancreatic lipase occurs,
monoglycerides and fatty acids are produced, which accumulate and form
multi lamellar bilayers (lamellar liquid crystal) at the emulsion surface. 32 •39-42
Increased surface pressure due to the accumulation of hydrolysis products
leads to a phase separation of core and surface lipids. i.e., coexistence of
large liquid-crystalline structures and small unilamellar vesicles at the emul-
sion-water interface generating from the multilamellar liquid-crystal. Unila-
mellar vesicles represent primary dispersed product of fat digestion. 32 The
vesicles are heterogenous in size (hydrodynamic radius, Rh = 200 to 600 A
by quasielastic light scattering) and in lipid composition. 32 In the presence of
bile salts, a two-phase unilamellar vesicle-plus-saturated mixed micelle system
will form. 40
3. Intestinal Absorption of Dietary and Biliary Lipids

According to Hemell et al. , 32 saturated mixed micelles of bile salt and
lipolytic products provide the ideal thermodynamic conditions for highest
efficiency in absorption rates of lipids. However, a vesicle-mediated absorp-
tion could also occur in particular physiological conditions. The maximal
absorption capacity of the rat small intestine is about 500 J.Lmol fatty acid per
hour. 43 Lipid absorption and transport is a rapid process, occurring in 14 min
in the rat. 44
D. SIGNIFICANCE OF GASTRIC AND PANCREATIC DIGESTION

In physiological conditions, pancreatic levels of lipase are significantly
higher (about 10 times) than those required to ensure complete digestion of
dietary triglycerides (100 to 150 g daily in human). 1 However, digestion of
triacylglycerols in the stomach is not negligible, since gastric lipase hydrolyzes
between 10 to 65% of dietary lipids. 15 •35 In adults rats, in which gastric
lipolysis depends exclusively on oral lipases, when saliva is diverted from
the stomach by ligation of the esophagus about 8.6% bovine milk triglycerides
are hydrolyzed within 1 h, the reaction products being diglycerides (7. 9%),
monoglycerides (0.7%), and free fatty acids (3.2%) compared to 18.7% hy-
drolysis in control animals (the reaction products being diglycerides [16.4%],
monoglycerides [2.3%] and free fatty acids [7.3%]). 45 Absence of gastric
lipolysis leads to a marked decrease in fat hydrolysis in the duodenum and
ileum since free fatty acids amount to 77.3% of total fatty acid in control
animals vs. 54.3% in experimental animals. There is a synergy of hydrolysis
between gastric and pancreatic lipases: prehydrolysis of triglycerides by pre-
duodenal lipase enhances pancreatic lipase action. 45 Partially digested tri-
glycerides are hydrolyzed at much higher rates than triglycerides by pancreatic
lipase. Furthermore, the binding of colipase to lipase and to bile salts is
enhanced by the presence of free fatty acidsY·28
The gastric phase of fat digestion is also thought to play an essential role
in milk digestion in preterm or fullterm neonates, where bile salt levels 46 and
pancreatic lipase secretions 47 -49 are quite low. Functional immaturity of ex-
ocrine pancreatic function, although associated with excretion of up to 25%
of ingested fat, does not appear to have a major clinical impact. 50 •51 It has
been shown that 14 to 60% of milk lipid is hydrolyzed to diglyceride, mon-
oglyceride, and free fatty acid in the first 30 to 60 min following ingestion. 35
Subsequent hydrolysis is achieved by a milk lipase (optimum pH = 8.5),
which acts in the intestine and requires bile salts to be active. This bile salt
stimulated lipase (BSSL) is identical to carboxyl ester hydrolase found in
pancreatic juice. 52 In vivo, the enzyme hydrolyzes monoglyceride to free
glycerol and fatty acid. Prior intragastric digestion greatly enhances BSSL
activity where 30% hydrolysis by BSSL is achieved, compared to 5% hy-
drolysis without prior exposure. 35 Thus, concerted action of the two enzymes
(preduodenallipase and BSSL) explains the efficiency of fat digestion in the
suckling neonate.
Isolated lipase deficiency, 53 - 55 co lipase deficiency, 56 and combined lipase-
co1ipase deficiencies 57•58 have been reported. These genetic diseases and pan-
creatic lipase deficiency caused by cystic fibrosis or chronic alcoholism are
commonly associated with fat excretion in the stools54 •58 but are not related
to growth deficit or malnutrition. Dietary fat of 50 to 70% in severe pancreatic
insufficiency59 or in isolated lipase and colipase deficiency 58 is indeed ab-
sorbed. In these situations. gastric enzyme plays an essential role in the
digestion, acting first in the stomach and then in the upper small intestine
where the postprandial pH is still acidic. 60
III. ADAPTIVE RESPONSE OF GASTROINTESTINAL

LIPASES TO DIETARY FAT
A. PREDUODENAL ADAPTATION
1. Gastrointestinal Transit
Inhibitory effect of fat in the small intestine on the rate of gastric emp-
tying has been reported. 61 However, the importance of the recent dietary
history in small bowel transit time has been emphasized. 62 Both gastric
emptying and mouth to cecum transit time are significantly faster after con-
sumption of a high fat diet for 14 days than after consumption of a low fat
diet in human 62 (half-time for gastric emptying = 98 vs. 147 min, mouth to
cecum transit time = 240 vs. 360 min, respectively). According to the
authors, the changes observed in gastrointestinal transit time could be related
to the release or sensitivity to humoral transmitters such as cholecystokinin
(CCK) and peptide YY (PYY). These peptides have been shown to be
released by the presence of lipid in the small intestine and to delay gastric
and small intestinal transit. 63 - 65
2. Prepyloric Lipases
In contrast with pancreatic lipase, which level is very low in the new-
bom,4749 the concentration of human gastric lipase at birth is quite comparable
to that found in the adult. 18 In the rat, lingual lipase activity was found to
raise exponentially immediately after birth. 66 This suggests that ingestion of
milk might be of major importance in eliciting preduodenal lipase activities
during the first days of postnatal life.
Only a few studies have clearly established that preduodenallipases were
able to undertake an adaptive response to dietary fat. The first preliminary
observation was made in 1978 by Hamosh, 4 who showed that rat lingual lipase
content in animals fed a high fat diet (22% com oil) for 2 weeks was 1.4-
fold higher than that of animals fed a similar purified diet containing 4% com
oil. A more detailed examination of the effect of dietary fat on lingual lipase
activity established that even 10% fat in the diet (composed of a mixture of
half sunflower oil and half lard), when given to rats for 3 weeks, was able
to induce the specific activity of lingual lipase by a 1.23-fold factor, the
maximum increase (1.30-fold) being reached when rats were fed a 20% fat
dietY Adding more fat to the diet (up to 30%) did not further enhance lingual
lipase activity. Maximum values of enzyme-specific activities were measured
after 1 week of feeding a 20% fat diet, and no additional change in lingual
lipase specific activity occurred when rats were fed the 20% fat diet for
4 weeks. Whatever the type of fat added in the diet (lard, sunflower oil, olive
oil, peanut oil, soybean oil, com oil, salmon oil, or butter), no significant
differences in lipase induction factor could be measured, i.e., the adaptive
response of lingual lipase to fat was not affected either by the chain length,
by the number, or the position of the double bonds in the fatty acid molecule,
even when dietary fat was provided in a moderate amount (10% by weight)Y
Evidence of an adaptive response to dietary fat has also been described for
the gastric enzyme in the rabbit. 68 Similar findings concerning the amount of
fat inducing a maximum increase in gastric lipase activity (6% fat in the diet
for fundus lipase and 12% for cardia lipase) or the lack of a specific effect
of the fatty acid composition of dietary triglyceride molecules on lipase ad-
aptation were reported. 68 However, feeding rabbits a 12% dietary fat diet for
1 week did not change the specific activity of lipase in the cardia or fundus
mucosae, as compared with feeding a 2.7% dietary fat diet. A 2-week period
was necessary to achieve adaptation of gastric enzyme.
B. PANCREATIC LIPASE (AND COLIPASE) ADAPTATION

Adaptation of pancreatic lipase to the lipid content of the diet is clearly
established in a number of species, including the rat, 69- 75 dog, 76 chicken, 77
and hog. 78 ·79 The response of pancreatic lipase may be influenced by a number
of factors such as the mode of meal intake (intestinal perfusate, liquid, or
solid meal), the type of dietary triglycerides, the degree of saturation of fatty
acid chains, the other components of the diet, the species studied, and the
age of animal. This can explain the differences measured in the extent of
induction of pancreatic lipase.
In the rat, pancreatic lipase concentration is high at birth but falls after
1 d. 48 Emergence of lipase is quite slow since it represents 40% of the value
in adult rats just before weaning. 22 This suggests that an increase in pancreatic
lipase before weaning is not a direct consequence of milk fat intake during
suckling. Low pancreatic lipase secretions have also been reported in the
human infant. 49 . 51 and in the neonatal dog. 35 According to Snook, 48 lipase
accumulation before weaning occurs mainly in response to a developmental
clock that may be affected only indirectly by the type of food eaten. In fact,
weaning rats on a high lipid diet (50% com oil) does not increase the rate of
lipase accumulation beyond that observed in rats weaned on to a high starch
diet during the first 25 d after birth. 70 By contrast, 30-d-old rats on the high
fat diet have higher lipase content than 30-d-old rats on the high starch diet. 70
Thus, adaptation of pancreatic lipase to nutritional components only occurs
after the usual weaning period (about 3 weeks in the rat).
1. Pancreatic Secretion after Intraduodenal Perfusate or Meal Intake

a. lntraduodenal Perfusates
Intestinal perfused fat digestive products have various effects on pan-
creatic secretion. 80- 82 Undigested triglycerides (com oil) do not stimulate more
pancreas output than a control perfusion with 0.15 M NaCl. However, when
the same amount of triglyceride is incubated with lipase-active pancreatic
juice, bile, and saline, prior to perfusion, pancreatic bicarbonate and protein
secretion are significantly enhanced. Triglyceride hydrolysis is thus required
to stimulate pancreatic secretion. 80 Monoglyceride itself is a luminal stimulus
of pancreatic secretion. Ability of fatty acids to stimulate pancreatic secretion
has been related to their chain length. 80 •82 Fatty acids of longer than eight
carbons in chain length increase output of pancreatic bicarbonate and protein.
The pancreatic response to fatty acids is dose-related over fatty acids con-
centrations ranging from 5 to 160 mM. The mode of dispersion of fatty acid
can also affect pancreatic response since oleate dispersed as a micellar solution
of fatty acid salt (soap) or in polysorbate-SO is a more efficient stimulant than
when dispersed in monolein-bile stabilized micelles, or than nondispersed
oleate. 80
b. Response to Meal Intake

The secretory response of the pancreas may be quite different, depending
on the mode of lipid administration (intraduodenal perfusions or meals). In
the healthy human, increased amounts of fat in the diet (10 to 40%) are
associated with enhanced rates of interdigestive and postprandial exocrine
pancreatic enzyme secretion and lipase output. 83 In long-term adaptated dogs,
consumption of a diet rich in olive oil ( = rich in monounsaturated fatty acids)
induces no change of the flow of pancreatic secretion throughout the entire

postprandial period, whereas a diet rich in sunflower oil ( = high polyun-
saturated fatty acid content) significantly enhances the volume of pancreatic
juice secretion during the first 3 h after the meal, as well as bicarbonate levels
and protein output during the first 2 h after eating. 84 Lipase activity and
specific activity did not significantly differ between the two high fat-fed
groups. However, due to the enhanced volume of secreted pancreatic juice
throughout the experimental period, lipase production was significantly higher
in the group fed the sunflower diet than in the group fed the olive oil diet.
This suggests that the degree of unsaturation of dietary fat can affect lipase
production.
Pancreatic enzyme response to perfusates and to meals is thought to occur
via a hormone-mediated pathway. Oleic acid is known to be a potent releaser
of CCK. 63 ·64 Some other hormones, such as secretin, also mediate pancreatic
output stimulation. 85 ·86 According to Ballestra et al., 84 the lack of response
of bicarbonate and protein content in pancreatic juice to the olive oil meal
might be due to the fact that oleic acid also releases other peptides from the
gut such as pancreatic peptide (PP) 87 and PYY, 65 which inhibit CCK- and
secretin-stimulated pancreatic secretion. However, a direct effect of lipid on
pancreatic secretion cannot be excluded since a fat-rich diet (34% fat = 17%
soybean oil + 17% coconut fat) is able to stimulate pancreatic secretion as
well as lipase secretion of an isolated perfused rat pancreas, 88 indicating an
action independent of altered hormonal stimulation and neural reflexes.
2. Effect of Type of Fat, Chain Length, and Degree of Saturation

a. Type of Fat
Comparison of the response of pancreatic lipase to type of fat, changes
in chain length, and degree of saturation of fatty acids reported from different
studies is difficult, because of varied experimental procedures, controls, and
dietary compositions. We have tentatively compared three studies, 70 •73 ·74 per-
formed in the rat, where lipase assays were measured in tissue homogenates
and under conditions where the presence of colipase and bile salts was un-
necessary for full lipase activity/0 •74 or in the presence of an excess of crude
exogenous colipase. 73 Table 2 reports induction factors for lipase (and for
colipase, when measured) in response to high triglyceride fat diets. Diets rich
in phospholipids (20%) significantly increase lipase-specific activity in pan-
creatic tissue and juice, whereas a diet enriched with cholesterol does not
induce any change in pancreatic lipase content in the rat. 89
b. Chain Length
According to Sabb et al., 74 no influence of chain length (from C 12 to C 18)
was seen on lipase activity in weanling rats (40 to 60 g) fed high fat diets
(29% lipid, i.e., 67% kcal as fat) for 1 week since coconut oil, the dietary
~
TABLE 2 ~
Fatty Acid Composition of Diets, Induction Factors for Lipase and Colipase in Response
to the Ingestion of Fat Diets
Induction Factors
Main FA• Unsaturated Lipase Colipase
Diet (%) FA(%) Deschodt-Lanckman et al. 70 Sabb et al. 74 Saraux et al. 73 Saraux et al. 73
Tricapry lin 98% c.,o <1 1.8

Medium chain 99% C 8 ,o-Cw,o <1 - - 1.5 1.3
triglycerides
Tristearin 2% c,.,o <2 1.8
96% CIS,o
Coconut oil 20% C 8 ,o-Cmo <15 - 1.8 2.0 1.8
45% c12o
Lard 29% c,.o 45 1.7 2.1 2.0 2.0
23% CIS,o
33% c,"' <:
Olive oil 80 3.6 1.5 ::::
......
14% c,"'" ..,
65% c,., :::;:
82 4.0 1.9 - -
c:;·
Com oil 14% c,.,o ;::
33% CIS,! $:)
;::
49% C,u ~
Sunflower oil 7% c,.,o 91 3.4 - 2.4 2.0 C'l
~
20% CIS,! ;::
~
69% CIS,2
~
"\:i
FA stands for fatty acid.
..,
~
"'"'c:;·
;::
Wicker-Planquart and Puigsen'er 67
fat with the greatest percentage (45%) of cl2 chain length used in their study,
stimulated lipase similarly to the other fats (corn oil, lard, safflower oil,
butter, olive oil) (Table 2). Regarding the effects of lard (99% fatty chain
superior to cl2:0: 2% cl4:0> 29% cl6:0> 23% CI8:0> 2% cl6:1> 33% cl8:1> 10%
cl8:2) and coconut oil (20% c8:0 to CIO:O> 45% cl2:0> 18% cl4:0> 10.9% cl6:0>
9.8% C 18 ,0 ) in adult rats (180 g) fed 40% fat diet for 3 weeks, Saraux et al. 73
observed that lipase was stimulated to the same extent (about twofold). How-
ever, lipase- and colipase-specific activities were less efficiently increased
(only 1.5- and 1.3-fold, respectively) by diets containing medium chain tri-
glycerides (Table 2). The fact that absorption of the medium chain fatty acids
resulting from hydrolysis of medium chain triglycerides occurs via the portal
route, the fatty acids being rapidly transported to the liver and oxidized/0
might explain their relative inaction with respect to inducing maximal pan-
creatic lipase content. Unfortunately, in Sabb's and Saraux's studies, all
experimental diets, except the medium chain triglyceride one, contained vary-
ing amounts of unsaturated lipids, which, in turn, may influence induction
factor values for pancreatic lipase. Indeed, Deschodt-Lanckman et al. 70 re-
ported that a 50% (w:w) fat diet with tricaprylin as the lipid source, given to
adult rats (180 to 220 g) for 5 d, poorly increased lipase content, when
compared to other lipid diets. This effect seemed to be related to the degree
of saturation rather than to the chain length of the dietary fat since tricaprylin
(98% C 8 ,0 ) and tristearin (96% C 180) diets led to the same induction of lipase-
specific activity in pancreas homogenates (l.8-fold)1° (Table 2).
c. Effect of Degree of Saturation

Only a few studies have dealt with the adaptative response of the pancreas
as a function of the degree of unsaturation of triglycerides/0 ·73 •74 •79 reaching
opposite conclusions. In the rat, as far as high fat diets are concerned, un-
saturated fats were better inducers than saturated according to Deschodt-
Lanckman et al. (using 50% fat dief0 ), whereas for Saraux et al. (using 40%
fat diets 73 ) and for Sabb et al. (using 29% fat diets, i.e., 67% kcal as faf 4 ),
the degree of saturation was found to have no impact on lipase 73 ·74 and
colipase. 73 On the other hand, in Sabb's study, when rats were fed diets
containing 17.4% lipid (i.e., 40% kcal fat), only the highly unsaturated saf-
flower oil (polyunsaturated/saturated ratio = 7. 9) increased pancreatic lipase
activity 1.63-fold compared to the low fat control diet, 5% lipid (i.e., 11%
kcal as corn oil). By contrast, 17.4% corn oil, lard, butter, olive oil, and
coconut oil diets did not significantly change lipase levels. 74 In the pig, only
one study concerning the degree of saturation of dietary lipids is available. 79
This study reported that pancreatic lipase was highly affected by the nature
of dietary lipids since lipase activity was 1.6- and 2. 75-fold enhanced in
pancreata of pigs fed 21% lard and 21% sunflower oil, respectively, as com-
pared to the control diet (3.5% lard + 3.5% sunflower oil). Together, these
results suggest that pancreatic lipase is sensitive to the degree of unsaturation

of triacylglycerol fatty acids and is affected only below the level of maximal
enzyme response.
3. Response of Pancreatic Lipase and Colipase to

Amount of Dietary Fat
a. Lipase-Specific Activity
The effect of varying the proportion of fat to carbohydrate in the diet
while keeping the diet isonitrogenous and isocaloric is reported in Table 3
(from Reference 75). Indeed, colipase, but not lipase, is particularly respon-
sive to protein intake. 90 It is therefore necessary to investigate adaptation of
lipase and colipase to dietary fat keeping the protein intake at a constant level.
Reduction in sunflower oil (0 to 1%, by weight), as compared to the control
diets (3 or 5%), resulted in nonsignificant changes in lipase activity. Lipase
rose significantly between 5 and 10% fat in the diet (1.4-fold) and even more
between 10 and 25% dietary lipid (2.5-fold, as compared to the control diet).
Between 25 and 30% dietary fat, lipase activity was at its maximum level.
Thus, the amount of sunflower oil (polyunsaturated/saturated ratio = 5.19)
necessary to induce an enzyme response is around 10% (by weight), corre-
sponding to 25% dietary calories as fat. This amount depends on the type of
fat ingested (Table 4). When highly saturated butter is given to animals
(polyunsaturated/saturated ratio = 0.03), more than 34% of the fat content
(55% of total energy) must be provided to enhance lipase activity in the
tissue. 71 Examination of previous investigations on specific tissue content in
pancreatic lipase in response to changes in dietary lipids at the expense of
carbohydrate67•71 •73 - 75 (Table 4) seems to indicate that the threshold of fat
content above which there is a significant induction of pancreatic lipase is
lower for unsaturated fat than for saturated fat. However, whatever the type
of lipid in the diet, the lipase activity in the tissue rises as the level of dietary
fat increases, and maximal lipase induction factors vary from 1.8 71 to 2.8. 74
The four- 70 and sixfold91 increases in lipase activity reported in other studies
performed in the rat may be due to the fact that rats were not fed isocaloric
diets, which could have altered the amounts of nutrients ingested. In the pig,
inductibility of pancreatic lipase varies from 3 when specific activity was
measured in pancreatic homogenates 79 to 7 when the amount was determined
in pancreatic juice. 78 Once more, differences in experimental procedures may
explain these different inductions.
b. Colipase-Specific Activity
As far as the existence of an adaptation of pancreatic colipase to nutritional
lipids is concerned, contradictory findings have been reported. No adaptation
of colipase to dietary lipids was observed in either rats or mice by Vander-
meers-Piret eta!. ,92 whereas several studies have described colipase adaptation
TABLE 3 ~
r,
Specific Tissue Content in Lipase and Colipase Activity and Relative Synthesis of Lipase in
~....
Pancreatic Lobules in Response to Changes in Lipid in the DieC5
~
IS"
;::
Enzyme-specific activity in ~
Sunflower oil %Energy pancreatic tissue (U/mg) Lipase relative synthesis §

(%, by weight) provided as fat Lipase Colipase Lipase/Colipase (% of total synthesis) ::::t
~
;::
~
0 0 207 ± 27 230 ± 29 0.90 ± 0.05 6.9 ± 0.9
"'tl
:;;:
I 2.5 190 ± 30 193 ± 29 0.99 ± 0.10 7.8 ± 0.4
3 7.4 186 ± 25 175 ± 30 1.08 ± 0.20 7.6 ± 0.6 o:o·
~
"'
5 12 219 ± 39 224 ± 42 0.99 ± 0.20 7.4 ± 0.3
~
10 25 259 ± 51 253 ± 44 1.01 ± 0.04 9.8 ± 0.8 ~
....
20 49 372 ± 38 280 ± 36 1.33 ± 0.10 15.3 ± 1.7
25 62 462 ± 130 396 ± 90 1.16 ± 0.10 13.6 ± 0.9
30 74 447 ± 64 332 ± 65 1.37 ± 0.17 15.6 ± 2.0
~
~
TABLE 4 -...1
0
Experimental Procedures and Type of Dietary Fat Related to the Amount of Lipids Necessary to Induce
Minimal and Maximal Lipase Response in Rat Pancreatic Tissue
Experimental conditions and resulting lipase activity adaptation

Wicker and
Authors Snook71 Saraux et al. 73 Armand et al. 67 Sabb et al. 74 Puigserver 75
Experimental procedures Dietary fat Butter Lard Sunflower oil/lard Com oil Sunflower oil
Polyunsaturated/saturated 0.03 0.13 2.6 5.0 5.19
(from Ref. 67) adaptation 2 weeks 3 weeks 3 weeks I week IOd
period
Minimum amount of % By weight of dietary fat 34.6 23 20 23.2 10
fat-inducing lipase % Dietary calories as fat 55 45.4 39.3 54 25
adaptation Resulting induction factor 1.2 1.57 1.89 2.4 1.38
for lipase
Maximum induction of % Dietary calories as fat 77 73.2 52.7 67 62
lipase activity Maximum inducation factor 1.83 2.15 2.15 2.85 2.48
<:
:::
~·
c;·
:::::
:::,
:::::
:::::..
,a
~
tl1
-8
~
t;
c;·
:::::
to dietary fat in the raf 3 •75 •93 •94 as well as in the pig. 78 These conflicting
findings may be explained by quantitative and qualitative differences in dietary
protein ingestion. 73
The effect on rat colipase activity of changing relative proportions of
dietary fat and carbohydrate while keeping the diet isonitrogenous (22% pro-
tein, the daily protein intake being around 4 g per day) and isocaloric79 is
described in Table 3. In contrast with lipase, colipase activity was found to
exhibit a significant enhancement (1 .3-fold) in the pancreas of rats fed the
lipid-free diet, as compared to rats fed the standard diet. Between 3 and 25%
fat in the diet, a 2.3-fold increase in colipase-specific activity was measured
in the pancreas, which was lower than that of pancreatic lipase. Colipase
activity then slightly declined between 25 and 30% dietary fat but was still
significantly higher (I .9-fold) than that in control rats. Due to the differential
adaptation of lipase and its cofactor, the colipase/lipase ratio progressively
decreased from 0.93 in control animals to a minimal value of 0.73 in 30%
lipid-fed rats. Nonparallel adaptations of lipase and colipase, where colipase/
lipase ratio dropped from 0.78 to 0.51, have also been reported in the rat94
when dietary fat increased from 3 to 19.5% and in the pig, 78 where the ratio
decreased from 2.35 in 25% fat-fed pigs to 1.77 in 5% fat-fed animals. In
hyperlipidic diet-fed rats, intrinsic colipase level is therefore not high enough
to allow full activation of pancreatic lipase when bile salts are present.
Saraux et a!. 73 showed that lipid intake modulates co lipase in the rat only
when the protein intake is between 3.5 and 6.0 g per day. When protein
ingestion is below 3.5 g, colipase does not adapt to dietary lipids (from 2 to
40%, by weight), while above 6.0 g protein intake, colipase level is maximal,
even with 2% lipid in the diet. Between 3.5 and 6.0 g, lipid intake modulates
colipase and lipase in a parallel way, colipase/lipase ratio being always around
0. 7. In their conditions, a correlation coefficient of co lipase with fat intake
of 0.64 was reported. The correlation of colipase with protein intake was
even stronger (0.85). 73 The particular sensitivity of colipase to nutritional
proteins might explain that their lipidic diets did increase colipase content to
the same extent as lipase since, unfortunately, their experimental diets were
not isonitrogenous, an increase in the protein content of the diet (from 17 to
37%) accompanying the rise in fat (from 0. 7 to 45%, by weight, corresponding
to 1.7 to 73.2% of total energy). Nevertheless, these findings suggest that
there is a cooperative effect on colipase activity between lipid and protein in
the rat in the range of 3.5 to 6.0 g daily protein intake. 73 •
c. Lipase Biosynthesis
There is little information on the role of dietary lipids in the regulation
of the earlier steps in the secretory pathway of acinar cells, especially their
effect on protein synthesis. Data concerning lipase biosynthesis, as influenced
by dietary lipids, are quite scarce. 71 •75 •95 Snook, 71 when directly measuring
protein synthesis in vitro by the rate of incorporation of 14C-valine into pan-
creas slices from rats fed a high fat diet (providing more than 55% of the
dietary calories as fat) or a control diet, reported that slices from rats fed the
high fat diet incorporated more label into pancreatic proteins. We found that
the increase in the synthesis of total pancreatic secretory proteins was about
twofold in magnitude 75 by incubating pancreatic lobules from rats fed high
fat (25 and 30% sunflower oil) and standard (3% sunflower oil) diets in the
presence of 35 S-methionine for 30 min and measuring incorporation of labeled
methionine into the pancreatic secretory proteins.
The increase in pancreatic lipase content by the high fat diets was directly
caused by an enhanced synthesis of the enzyme, as indicated by the good
correlation between the increases in lipase-specific activity and relative syn-
thesis. The specific enhancement in lipase fractional synthetic rate (expressed
as the percentage of radioactivity incorporated into pancreatic lipase, com-
pared to radioactivity incorporated into the total mixture of exocrine proteins)
was about twofold in the high fat diet fed animals (Table 3), a plateau in
lipase biosynthesis being reached when rats were given the 20% fat diet.
4. Time Course of the Variations in Lipase and Colipase Activities and

Biosynthesis in Pancreatic Tissue
a. Lipase Activity and Biosynthesis
Time studies over a 5-d period of high lipid diet ingestion 70 ·95 have shown
that lipase content was significantly enhanced (by 2-fold 70 and l.4-fold 95 ) in
the pancreas of fat-fed rats as early as the second day of dietary induction
(Figure l). Further consumption of the high lipid diet resulted in a further
increase in lipase-specific activity (3-fold 70 and 1.9-fold95 ). Adaptive response
of lipase activity is rapid, although several days of induction are necessary
for lipase to reach a maximal steady-state level. Long-term (2 to 4 weeks)
dietary fat feedings lead to increases in lipase activity similar to those measured
in 1-week fed rats. 67 Lipase dietary adaptation is reversible since shifting
lipid-adapted rats (on a 50% com oil diet) to a standard diet (4% com oil)
results in a decrease in lipase activity, the kinetics of lipase decay being almost
exactly opposite to those previously observed in rats shifted from a low fat
to high fat diet. 70
Ingestion of a 25% sunflower oil diet has an even more rapid effect on
total protein and lipase synthesis, as measured by incorporation rates of 35 S-
labeled methionine into pancreatic proteins. 95 Increase in the rate of total
pancreatic protein synthesis is almost immediate (1.6-fold on the first day of
adaptation) and persists throughout the ingestion period of the lipid-rich diet. 95
When measuring rates of synthesis of individual exocrine proteins, the relative
synthetic rate of lipase increases by a 1.4-fold factor, as early as on the first
day of diet intake. This corresponds to a calculated induction of 2.2-fold of
lipase absolute synthesis. This rise in lipase synthesis can be easily visualized
on following the separation of radioactive labeled pancreatic proteins by using
PAGE electrophoresis and gel fluorography (Figure 2). Enhancement in lipase
biosynthesis is progressive, and a 2.2-fold increase in fractional rate of lipase
synthesis is reached on the 5th day of induction.
+
+
•+
+ +
1 400
l
f
..!..
300
+
•
I
1
200
•
• 100
.i..:I 0
0 1
D A Y 8
J
•
FIGURE 1. Short-term dietary adaptation oflipase (empty blocks) and colipase (shaded blocks)
activities in rat pancreatic tissue, as a function of the ingestion period (0 to 5 d) of a high fat
diet (25% sunflower oil). 95 Each point represents the mean value of five determinations. The
vertical bars indicate the SE. The P values, calculated from Student's t-test comparing 25%
lipidic diet to the control diet (3% lipid), are <0.05( + ), <0.01( + + ), <0.001( + + + ).
b. Colipase Activity
Short-term adaptation of colipase does not resemble that of lipase. Co-
lipase responds more slowly than lipase; its content in the pancreas decreases
initially and then increases gradually in response to dietary fat (Figure 1).
Coenzyme level on day 5 is quite comparable to that of rats fed the standard
diet. 95 Increasing the adaptation period to 10 d leads to a twofold induction
in colipase activity. 75
Although the alterations in food composition change the lipase and co-
lipase content, the mechanisms for the changes are probably distinct, as
suggested by the differential short-term adaptation responses of both pan-
creatic proteins.
C. POSSIBLE REGULATORS OF PANCREATIC LIPASE

ADAPTATION TO DIETARY FAT
1. Nutritional Regulators
The role of triglyceride hydrolysis products on pancreatic secretion is
now well documented. 80- 82 We have seen that intraluminal free fatty acids
and monoglycerides (but not undigested triglycerides) stimulated pancreatic
secretion in man, dog, and rat. 80- 82 However, contradictory findings have been
DAY S
0 2 3 5
A
L
PCP
SP
FIGURE 2. Fluorographic pattern of [35 S]-methionine-labeled pancreatic secretory proteins as

influenced by a 0 to 5 day ingestion period of fat diet (25% sunflower oil)!5 Pancreatic lobules
prepared from rats fed the lipid diet for 0 to 5 d were incubated in a Krebs-Ringer buffer in the
presence of ["S]-methionine (3.7 MBq/ml) for 30 min and homogenized. Radiolabeled proteins
(200,000 cpm per lane) were separated by PAGE electrophoresis and the gel was impregnated
with a fluorographic reagent, prior to autoradiography. Radioactive glycoprotein, amylase, lipase,
procarboxypeptidases A and B, and serine proteases are indicated as G, A, L, PCP, and SP,
respectively.
reported, as far as the effect of intravenous administration of lipid on pan-

creatic secretion is concerned: intravenous fat has been reported either to exert
no stimulation of pancreatic secretion in dog 96 •97 and human98 or to increase
the lipase content in dogs 99 and rat. 100 The first findings (i.e., no effect of
intravenous lipid on pancreatic secretion) suggest an indirect action of tri-
glyceride hydrolysis products, through the release of a second messenger from
the intestinal mucosa. By contrast, the latter results (i.e., increase of lipase
consecutive to intravenous fat) indicate that the hydrolysis products have to
pass through the intestinal wall to induce pancreatic response. Other inter-
mediates in triglyceride metabolism, such as ketones, 101 have also been thought
to play a role in the regulation of lipase. The circulating levels of ketones in
the blood are increased during adaptation to a high fat diet. When infused
into rats for 5 days, they lead to a 40% raise in lipase activity. 101 However,
their role in lipase induction seems to be relatively limited. Furthermore, the
direct effects of 13-hydroxybutyrate (13-HBA) on a primary culture of pan-
creatic acinar cell is unclear. 13-HBA affects lipase activity only in cells from
low fat fed rats ( 11% energy as fat corresponding to 5% fat) cultured in low
glucose medium for 48 h, when the antecedent diet is a semipurified diet,
prepared with well-defined components. 102 The magnitude of lipase induction
in these conditions is relatively small (33%). On the other hand, when the
antecedent diet, providing 10% energy as fat, is prepared from crude com-
ponents, no increase in cellular lipase activity was observed. 102 According to
the authors, 102 their results support a role for the ketones in the regulation of
pancreatic lipase; however, ketones are not the sole mediator and might act
in conjunction with other food components.
2. Hormonal Regulators
Hormonal factors are thought to be involved in lipase response to dietary
fat. Fat hydrolysis products are assumed to act on the intestinal mucosa (more
likely on the proximal part of the small intestine 103 ), leading to a secretion
of peptide hormones. In fact, the presence of lipid in the small intestine leads
to a release of CCK and secretin in man, 63 dog, 64 •85 •86 and ratHl4 but not in
the pig. 105 CCK does not seem to be the mediator of lipase adaptation since
intravenous administration of caerulein, a synthetic analog of CCK, does not
modify lipase synthesis. 106 On the other hand, secretin is a possible inter-
mediate between ingested fat and pancreatic lipase regulation since the hor-
mone exerts stimulatory effects on lipase concentration in pancreatic tissue 107
and on total protein 108 and lipase 109 synthesis. On the other hand, secretin
does not alter the lipase content of cultured pancreatic cells. 110 Injection of
secretin for 15 days increases the pancreatic weight, RNA amount, and lipase
content in the rat pancreas. 107 These findings have been confirmed and studied
more extensively by others. 108 • 109 Infusion of synthetic secretin in conscious
rats (16 clinical units per kg and h) elicits total protein synthesis after a 2-
to 3-h lag period, a maximal twofold enhancement being obtained after a 12-
h infusion. 108 When measuring rates of synthesis of individual exocrine pro-
teins by following the incorporation of a mixture of 15 14 C-labeled amino
acids into proteins, a pronounced increase (by a 2.5 factor) in the relative
synthetic rate of lipase was observed. 109 This leads to a calculated 4. 9- and
4.1-fold change with time, after 12- and 24-h stimulation, respectively, in
lipase absolute synthetic rate. These values are very close to those obtained
from high fat fed rats/ 5 ·95 as far as relative and absolute synthesis rates of
lipase are concerned. Therefore it can be concluded that secretin remains a
very possible mediator in lipase adaptation. Whether other hormones in the
secretin family such as glucagon, vasoactive intestinal polypeptide (VIP), and
gastric inhibitory polypeptide (GIP) also have a stimulatory effect on lipase
synthesis has not yet been demonstrated.
Finally, other nonpeptidic hormones, such as steroid ones, are not re-
sponsible for nutritional changes of lipase synthesis.
D. MECHANISMS INVOLVED IN THE REGULATION OF

PANCREATIC LIPASE AND COLIPASE BY DIETARY LIPIDS
1. Outline of the Expression of Pancreatic Enzymes
a. General Scheme of Eukaryotic Gene Expression
Numerous studies in many laboratories have contributed to our present
understanding of gene organization and the control of their expression (for
review, see References Ill through 113). All eukaryotic cells follow the same
sequence of operations to express one gene which codes for a protein. The
essential steps of pancreatic gene expression are summarized in Figure 3.
Control of gene expression can be exerted at any one of these essential steps.
The first level of control is that of gene transcription. Protein-coding
genes are transcribed by RNA polymerase II (RNA pol II or B). The sequences
required for accurate and efficient initiation of transcription ( = promoters)
are dispersed upstream from the start site of transcription, generally locating
between about -40 and -200 bp. Some promoter elements, such as TATA,
GC, and CCAAT boxes, are common to many genes transcribed by pol II.
These sequences interact with general transcription factors and RNA poly-
merase II leading to the formation of stable preinitiation complexes. Less
common regulatory signals may confer inducible transcription, for example
in response to a specific hormone, and are also binding sites for specific
transcription factors. There are multiple DNA binding factors which can act
on gene transcription as positive or negative regulatory elements, vary in
abundance or in binding ability, and compete with other factors for DNA
binding sites. The frequency of initiation of mRNA synthesis depends on all
these interactions of transcription factors with gene promoters in a sequence-
or conformation-specific manner.
Another class of transcriptional controlling elements is that of enhancers
which, in contrast with that of promoters, are not located in the vicinity of
regulated genes but can function over large distances of more than 20 kilo-
bases. They can enhance the transcription from a gene more than 100-fold.
Stimulation is possible regardless of their orientation and their position from
the gene they are turning on. Enhancer-like sequences have been found for
the genes of chymotrypsinogen, 114 elastase, 115 and other pancreatic exocrine
enzymes. 116 Enhancers are often responsible for tissue- or development-spe-
cific control of gene expression. These regulations also occur via the inter-
action of DNA binding proteins with consensus sequences.
The eukaryotic gene is constituted of discrete regions on the DNA mol-
ecule (exons) which contain all coding information for the final gene product,
separated by extra DNA stretches (introns) which do not carry any coding
information. The DNA gene is integrally transcribed into heterogenous nuclear
RNA (hn RNA) in the cell nucleus.
Although transcriptional control in eukaryotes is fundamentally important,
several posttranscriptional steps must be correctly realized in order to maintain
the steady state levels of functional mRNAs in the cytoplasm. The primary
specific start site termination site

recognition sites
DNA gene
l
ex on intron exon
Transcription
hn RNA
l
Processing -cap structure
NUCLEUS -poly A tailing
-adenylate methylation
-splicing
ACXNAIR. mature mRNA
------------------------------------------------ ~ Transport-------------------------------
CELL mature mRNA
~ mRNA
translationJ
CYI'OPLASM mRNA / Pre(pro)protein

degradation Processing:
-segregation of nascent
polypeptide chain
-splicing of signal peptide
~
(Pro)protein
Intracellular transport
-rough endoplasmic reticulum
-transitional elements
-Golgi vesicles
-condensing vacuoles
-zymogen granules
Storage (zymogen granules)
I
Discharge at cell surface
(exocytosis)
~
LUMEN (Pro)protein
I
Activation
Proteint
FIGURE 3. Essential steps of gene expression of pancreatic secretory proteins.
RNA transcript is converted to mRNA in the nucleus through several mod-

ifications: a cap of 7-methylguanosine and a tail of polyadenosine tract are
added to the 5' and 3' end of RNA, respectively. Internal adenine residues
(l out of 400 Ap) are methylated and introns are eliminated through RNA
splicing at consensus sequences. Mature mRNA is then transported to the
cytoplasm (about 5% of the mass of mRNA) where it will be translated.
The events of pancreatic protein synthesis are carried out on the ribosomes
which are attached on the rough endoplasmic reticulum (RER). Protein syn-
thesis is a complex, multistep process. which can be affected by a variety of

physiological and pharmacological stimuli (for review, see Reference 117).
Translational regulation occurs mainly at the initiation step of protein syn-
thesis.
The rate of mRNA turnover, or degradation, can also significantly affect
the rate of protein synthesis. 118 The half-lives of mRNAs in eukaryotes vary
widely, from seconds to several hours, or even days. Poly(A) tract in con-
junction with poly( A)-binding protein has been shown to protect mRNA chains
against exonucleolytic attack. The nucleotide sequences in the 5' noncoding
segment of mRN A may also be involved with the decay process. The influence
of the coding region on mRNA turnover has also been demonstrated. The
stabilization of effect appears to involve ribosome interaction.
b. Synthesis and Subsequent Processing of Pancreatic Secretory Proteins

Pancreatic enzyme level can also be regulated at a posttranslational stage.
The general pathway and the time course of intracellular transport of secretory
proteins are essentially due to the work of Jamieson and Palade. 119 Pancreatic
proteins are synthesized as large precursor proteins (presecretory protein) on
membrane-bound ribosomes. During synthesis, proteins are vectorially trans-
ported across the microsomal membrane where the signal peptide is spliced. 120
Newly synthesized secretory proteins, once segregated in the cisternal space
of the RER, pass successively from the vesicular space of the RER into small
vesicles to be conveyed to the Golgi complex and to condensing vacuoles.
Glycosylation of some proteins on asparagine residues is initiated in the RER
and completed in the Golgi apparatus. Condensing vacuoles progressively
concentrate to form mature zymogen granules. The pathway for all pancreatic
enzymes is essentially the same. Secretory proteins are stored within mature
zymogen granules until their discharge from the apical portion of the acinar
cell upon hormonal or nutritional stimulation. A mean transit time of 30 min
has been estimated in the guinea pig. Activation of pancreatic proteins occurs
in the gut, by cleavage of trypsinogen activation peptide by enterokinase, and
subsequent removal of activation peptides of other pancreatic proteins by
trypsin.
2. Induction of Lipase and Colipase mRNAs in Response

to High Fat Diet
We have shown that lipase accumulation in pancreatic tissue, consecutive
to fat absorption, resulted from a selective stimulation of enzyme synthe-
sis. 75 •95 The question arises then as to the molecular level of the mechanisms
of enzyme induction.
A major problem encountered in the purification of mRNA from the
pancreatic tissue is the presence of high levels of pancreatic ribonuclease,
especially in the rat where the mean content is about 260 J..Lg ribonuclease
per gram wet weight of pancreas. 121 Methods of isolation of undegraded
mRNA have been reported, 122 which allow. with due precaution. satisfactory
preparation of pure, biologically active material. Quantitation of specific
mRNA expression has been then rendered possible thanks to the establishment
of lipase and colipase eDNA clones. The cloning and characterization of
pancreatic colipase eDNA has been performed in man, 123 · 124 rat, 125 and dog, 126
those of pancreatic lipase eDNA in man, 127 dog. 128 and rat recently. 129 We
have chosen, in the assay of lipase and colipase mRNAs, to hybridize mRNA
immobilized on filters to the corresponding synthetic labeled eDNA. This is
a more sensitive method than the translational one in a cell-free system.
a. Lipase mRNA
The effects of a 10-d ingestion of 10% or 25% sunflower oil on lipase
mRNA level result in a 2.2- and 3.9-fold increase in the lipase messenger,
respectively. 130 These data support a pretranslational mechanism by which
triglycerides regulate pancreatic adaptation.
When we examined the kinetics of the cellular mRNA concentration to
dietary stimulation, two distinct responses were obtained. 95 · 131 Lipase mRNA
increased in a biphasic manner in adapted rats. 95 This may have been the
result of the presence of a rather large amount of sodium bicarbonate (18.5%
by weight) in the experimental diet (25% sunflower oil). Indeed, we had
emphasized in another study 130 the possible interference of sodium bicarbonate
with the lipase response, since sodium bicarbonate had been shown to be
responsible for a variety of side effects, such as sodium overload. impaired
absorption of calcium, and an increase in plasma CCK level. On the other
hand, lipase mRNA was found to be maximally induced as soon as the second
day of lipid diet intake (by twofold) when no sodium bicarbonate was present
in the 25% sunflower diet and did not further increase up to 10 d. 131 Never-
theless, all these data suggest that the mechanism of lipase adaptation by
dietary lipids is mostly pretranslational. Changes in lipase mRNA levels
consecutive to lipid diet ingestion probably result from an increased tran-
scriptional rate of lipase gene. At present, we cannot exclude yet other mech-
anisms, such as RNA processing. regulation of lipase messenger transport,
and changes in mRNA stability.
The rise in lipase mRNA was consistent with the increase in the measured
values for lipase relative synthesis. 130 · 131 However, when the total increase
in pancreatic protein synthesis was taken into account. lipase absolute syn-
thesis rates increased more than the mRNA level. It is therefore possible that
a control mechanism acting at the translational level favors the translation of
lipase mRNA among the different pancreatic mRNAs. There is now a great
deal of evidence that such translational discriminations between different
mRNAs exist. 117• 132
The effect of a high fat diet on lipase mRNA levels in pancreatic tissue
allows us to eliminate the steroid hormones as possible regulators in diet
induced lipase activity. In fact, adrenalectomy plus castration 133 or adrenal-

ectomy alone, 134 although increasing lipase content in rat pancreas, do not
alter lipase mRNA levels. 135 This suggests a translational control mechanism
by these hormones.
b. Colipase mRNA
The level of colipase mRNA, as measured by dot-blot hybridization with
the corresponding eDNA probe, moderately increases, by a 1.5-fold factor,
in the pancreas of high fat adapted rats (25% sunflower oil), a plateau being
reached on the 3rd day of diet ingestion. 125 This finding suggests that ad-
aptation of pancreatic colipase to dietary lipid is genetically regulated. It is
nevertheless difficult to correlate colipase mRNA level and cofactor activity
in pancreatic tissue, since alterations in cofactor synthesis and secretion may
occur. The magnitude of the rise in colipase cytoplasmic mRNA concentra-
tion, although significant, is less than that in lipase mRNA under the same
experimental conditions. 131 This can explain the differences in induced activity
levels of both proteins in adapted rats. 75 •95
3. Transcriptional Regulation of Pancreatic Lipase Gene Expression

Even if there is good evidence that the control of pancreatic gene expres-
sion in the rat is exerted in some cases at the translational leveP 36 and that
the stability of some mRNAs may vary according to nutritional situations, 137
transcription is most likely a key step in this regulation.
Measurements of specific gene transcription can be directly realized with
isolated nuclei from control and adapted animals. In the run on transcription
assay, nuclei are incubated with radioactive UTP, and 32 P-labeled RNAs in
nuclei are purified and measured after hybridization to nitrocellulose filters
to which an excess of plasmid DNA containing the gene sequences had been
bound. This assay predominantly estimates the elongation of in vivo initiated
RNA chains and directly indicates the relative number of polymerases on the
specific gene.
The time course response of synthesis of pancreatic lipase mRNA to a
fat diet ingestion (25% sunflower oil) has been determined by run-on
transcription 131 (Table 5). A progressive induction in lipase mRNA synthesis
occurs as a function of time of adaptation to the fat diet. This increase in
lipase gene transcription is rapid, beginning on the first day of diet intake.
The stimulation of lipase transcription can entirely explain messenger accu-
mulation in the cytoplasm during dietary adapation.
IV. PERSPECTIVES AND GENERAL CONCLUSION

In the present chapter, we have seen that both prepyloric and pancreatic
lipases are of prime importance in the hydrolysis of dietary triglycerides.
These enzymes act sequencially, and prior hydrolysis of triglycerides by
TABLE 5
Changes in Lipase Relative Synthesis, mRNA Levels, and
Transcriptional Rates, Induced by a Lipid-Rich Diet 131
Days on diet
Changes 0 2 3 5
Biosynthesis' 6.1 ± 0.5 6.7 ± 0.6 7.8 ± 0.8 8.9 ± 0.5 9.6 ± 0.7
mRNA• I 1.4 ± 0.08 1.79 ± 0.28 1.83 ± 0.16 1.99 ± 0.33
Transcriptional rate' 3.3 4.7 ND ND 5.7
Relative synthesis is expressed as a percentage of radioactivity incorporated into lipase

compared to that incorporated into total exocrine proteins.
mRNA content is normalized to that in the pancreas of rats fed the control diet (day 0).
Transcriptional rates are normalized to that of actin, included as an internal standard in the
hybridization experiments.
preduodenallipase enhances pancreatic lipase action. Both lipases show rapid

adaptation to the amount of fat ingested. Such nutritional regulations are
thought to occur via the release of gastrointestinal peptide(s). It is believed
that secretin largely controls the pancreatic adaptation to dietary lipids. Lipase
accumulation in pancreatic tissue has been shown to result from a selective
stimulation of enzyme synthesis. The increases in lipase synthesis and cofactor
activity in the pancreas of rats adapted to a lipid-rich diet are due to rises in
both messengers. As far as colipase is concerned, maximal change in mRNA
level occurs on the third day after alteration in dietary composition, thus
preceding changes in tissue levels of the cofactor. Adaptation of lipase mRNA
to a lipid-rich diet is faster than that of colipase, since changes in messenger
concentration are observed as early as I d after dietary manipulation. This
suggests that different effectors might be involved in lipid adaptation of each
pancreatic protein. The magnitude of lipase mRNA response is also greater
than that of its cofactor, explaining the nonparallel changes in lipase and
colipase activity following fat ingestion. Finally, the mechanism by which
lipids alter the expression of pancreatic lipase is essentially transcriptional,
as directly demonstrated by nuclear run-on assays.
Investigations into the structure-function relationships of pancreatic lipase
gene, as well as DNA-mediated transfection techniques to introduce the func-
tional lipase gene into cells, will help in studying lipase gene regulation by
nutrients. The only available structure .:>fa pancreatic lipase gene and DNA
sequence of the 5' flanking region has been performed in the dog. 138 Unfor-
tunately, analysis of canine lipase gene by DNA transfer methods in pancreatic
acinar cells, such as AR4-2J or 266-6 for the presence of regulatory signals
in the vicinity of the canine structural gene, failed to reveal these regulating
stretches of DNA. 138 1t is possible that these cell lines, derived from pancreatic
acinar cells, one from rat (AR4-2J) and the other from mouse (266-6), are
not suitable for the study of the canine gene. Another possibility is that
regulatory elements are positioned in another region of the gene. It would be
of interest to test the 5' flanking region of the rat gene, when rat lipase gene
cloning will be performed, and analyze the upstream region to help to define
genetic elements that mediate the induction of the lipase gene by nutritional
agents and secretin.
In the case of other pancreatic genes, it has been possible, when con-
structing expression recombinant plasmids containing the 5' flanking region
of rat chymotrypsin gene linked to the coding sequence of a reporter gene
such as chloramphenicol acetyltransferase (CAT), to obtain specific expres-
sion of chymotrypsin gene recombinant in the rat exocrine pancreas tumor
line cells AR4-2J, but not in HIT cells, a transformed line from hamster
pancreatic endocrine cells. 114 Careful examination of the DNA sequences
upstream for nine rat pancreatic genes (amylase, carboxypeptidases Al and
A2, chymotrypsin B, elastases I and II, ribonuclease, trypsins I and II) coupled
with deletion experiments, have led to the identification of a consensus se-
quence: GTCACCTGTGCTITTCCCTG located -89 to -213 bp from the
cap site; this confers preferential expression of these genes in pancreatic
exocrine cells. 116 These regulatory sequences are similar to other enhancers
discovered so far in that they work whether they are located before or after
the genes and whether they are in the normal orientation or are inverted. 116
Enhancers are believed to work through the binding of pancreatic nuclear
protein, which leads to a change in the structure of the chromosome in the
vicinity of the gene. Indeed, two teams independently have demonstrated
binding of pancreatic nuclear protein with the enhancer consensus sequence
of mouse amylase gene 139 and rat chymotrypsin gene. 140 The specific factor
for amylase gene consensus element isolated from rat pancreatic nuclei in-
teracts also with elastase I gene consensus sequence, as indicated by com-
petition experiments for protein binding to the amylase sequences. 139 Site-
directed mutagenesis of the amylase sequence results in concomitant loss of
protein binding and enhancer activity, which indicates that the nuclear protein
is a positive regulator, controlling the activity of the amylase gene enhancer. 139
Another factor, isolated from 266-6 and HeLa cells, having the DNA-binding
specificity of MLTF (major late transcription factor of adenovirus), has been
shown to bind to the chymotrypsin sequence in vitro. 140 Site-directed muta-
genesis of chymotrypsin sequence into a perfect MLTF site results in the loss
of enhancer activity in acinar cells, but not of the binding of MLTF-like
factor. 140 This finding seems to indicate that MLTF-like factor acts as a
negative regulator of chymotrypsin enhancer function. Such transcriptional
repressions have also been described in other systems. 14 I. 142
Finally, another exciting new area of active research is the generation of
transgenic mice which can be analyzed and tested under a variety of physi-
ological circumstances. Such studies have already led to pancreas-specific
expression of amylase/CAT construct in transgenic lines and to the study of
insulin dependence of the hybrid amylase gene 143 or to the study of islet-
specific expression of human insulin and its regulation by glucose and amino
acids. 144
All these techniques constitute powerful tools which will be exploited in
the future and will open up other sets of experimental possibilities for the
study of lipase and colipase gene regulation.
REFERENCES
1. Lairon, D. and Borel, P., La regulation nutritionnelle des enzymes de Ia digestion des
lipides, Cah. Nutr. Diet., 24, 413, 1989.
2. Gargouri, Y., Moreau, H., and Verger, R., Gastric lipases: biochemical and physio-
logical studies, Biochem. Biophys. Acta, 1006, 255, 1989.
3. Schonheyder, F. and Volquartz, K., The gastric lipase in man, Acta Physiol. Scand.,
11,349,1946.
4. Hamosh, M., Rat lingual lipase: factors affecting enzyme activity and secretion, Am. J.
Physiol., 235, E416, 1978.
5. Field, R. B. and Scow, R. 0., Purification and characterization of rat lingual lipase, J.
Bioi. Chern., 258, 14563, 1983.
6. Sweet, B. J., Matthews, L. C., and Richardson, T., Purification and characterization
of pregastric esterase from calf, Arch. Biochem. Biophys., 234, 144, 1984.
7. Bernback, S., Hernell, 0., and Bliickberg, L., Purification and molecular character-
ization of bovine pregastric lipase, Eur. J. Biochem., 148, 233, 1985.
8. Cohen, M., Morgan, R. G. H., and Hofmann, A. F., Lipolytic activity of human
gastric and duodenal juice against medium and long-chain triglycerides, Gastroenterology,
60, 1, 1971.
9. Perret, J. P., Lipolyse gastrique chez le lapereau. Origine et importance physiologique
de Ia lipase, J. Physiol. (Paris). 78. 221, 1982.
10. DeNigris, S. J., Hamosh, M., Kasbekar, D. K., Lee, T. C., and Hamosh, P., Lingual
and gastric lipases: species differences in the origin of prepancreatic digestive lipases and
in the localization of gastric lipase, Biochim. Biophys. Acta, 959, 38, 1988.
II. Moreau, H., Gargouri, Y., Lecat, D., Junien, J, L., and Verger, R., Screening of
preduodenallipases in several mammals, Biochem. Biophys. Acta, 959, 247, 1988.
12. Roberts, I. M. and Jaffe, R., Lingual lipase: immunocytochemical localization in the
rat von Ebner gland, Gastroenterology, 90, 1170, 1986.
13. Roberts, I. M., Nochomovitz, L. E., Jaffe, R., Hanel, S. I., Rojas, M., and Agostini,
R. A., Immunocytochemical localization of lingual lipase in serous cells of the developing
rat tongue, Lipids, 22, 764, 1987.
14. Abrams, C. K., Hamosh, M., Lee, T. C., Ansher, A. F., Collen, M. J., Lewis,
J. H., Benjamin, S. B., and Hamosh, P., Gastric lipase: localization in the human
stomach, Gastroenterology, 95, 1460, 1988.
15. Moreau, H., Laugier, R., Gargouri, Y., and Verger, R., Human gastric lipase is
entirely of gastric fundic origin, Gastroenterology, 95, 1221, 1988.
16. Moreau, H., Bernadac, A., Gargouri, Y., Benkouka, F., Laugier, R., and Verger,
R., lmmunocytolocalization of human gastric lipase in chief cells of the fundic mucosa,
Histochemistry, 91,419, 1989.
17. Moreau, H., Bernadac, A., Tn!tout, N., Gargouri, Y., Ferrato, F., and Verger, R.,
Immunocytochemical localization of rabbit gastric lipase and pepsinogen, Eur. J. Cell
Bioi., 51, 165, 1990.
18. Fredrickzon, B. and Hernell, 0., Role of feeding on lipase activity in gastric contents,
Acta Paediatr. Scand., 66, 479, 1977.
19. Gargouri, Y., Pieroni, G., Riviere, C., Sauniere, J, F., Lowe, P. A., Sarda, L., and
Verger, R., Kinetic assay of human gastric lipase on short- and long-chain triacylglycerol
emulsions, Gastroenterology, 91, 919, 1986.
20. Schick, J,, Verspohl, R., Kern, H., and Scheele, G., Two distinct adaptative responses
in the synthesis of exocrine pancreatic enzymes to inverse changes in protein and car-
bohydrate in the diet, Am. J. Physiol., 247, G611, 1984.
21. Verger, R., Pancreatic lipases, in Lipases, Borgstrom, B. and Brockman, H. L., Eds.,
Elsevier, Amsterdam, 1984, 84.
22. Saraux, B. and Girard-Globa, A., Development of pancreatic enzymes in fetal and
suckling rats with emphasis on lipase and colipase, J. Dev. Physiol., 4, 121, 1982.
23. Sarda, L. and Desnuelle, P., Action de la lipase pancreatique sur les esters en emulsion,
Biochem. Biophys. Acta, 30, 513, 1958.
24. Semeriva, M. and Desnuelle, P., Pancreatic lipase and colipase. An example of het-
erogeneous biocatalysis, Adv. Enzymol., 48, 319, 1979.
25. Maylie, M.-F., Charles, M., Gache, C., and Desnuelle, P., Isolation and partial
identification of a pancreatic colipase, Biochem. Biophys. Acta, 229, 286, 197!.
26. Chapus, C., Sari, H., Semeriva, M., and Desnuelle, P., Role of colipase in the
interfacial adsorption of pancreatic lipase at hydrophobic interfaces, FEBS Lett., 58, 155,
1975.
27. Patton, J, S., Donner, J,, and Borgstrom, B., Lipase-colipase interactions during gel
filtration high and low affinity binding situations, Biochem. Biophys. Acta, 529, 67,
1978.
28. Sauve, P. and Desnuelle, P., Interactions of pancreatic colipase with taurodeoxycholate-
oleate mixtures above the critical micellar concentrations, FEBS Lett., 122, 91, 1980.
29. Borgstrom, B. and Hildebrand, H., Lipase and colipase activities of human small
intestinal contents after a liquid test meal, Scand. J. Gastroenterol., 10, 585. 1975.
30. Patton, J, S., Albertsson, P.-A., Erlanson, C., and Borgstrom, B., Binding of porcine
pancreatic lipase and colipase in the absence of substrate studied by two-phase partition
and affinity chromatography, J. Bioi. Chern., 253, 4195, 1978.
3!. Staggers, J. E., Hernell, 0., Stafford, R. J,, and Carey, M. C., Physical-chemical
behavior of dietary and biliary lipids during intestinal digestion and absorption. I. Phase
behavior and aggregation states of model lipid systems patterned after aqueous duodenal
contents of healthy adult human beings, Biochemistry, 29, 2028, 1990.
32. Hernell, 0., Staggers, J, E., and Carey, M. C., Physical-chemical behavior of dietary
and biliary lipids during intestinal digestion and absorption. II. Phase analysis and ag-
gregation states of luminal lipids during duodenal fat digestion in healthy adult human
beings, Biochemistry, 29, 2041, 1990.
33. Gargouri, Y., Pieroni, G., Lowe, P. A., Sarda, L., and Verger, R., Human gastric
lipase. The effect of amphiphiles, Eur. J. Biochem., 156, 305, 1986.
34. Borgstrom, B., Importance of phospholipids, pancreatic phospholipase A2 and fatty acid
for the digestion of dietary fat, Gastroenterology, 78, 954, 1980.
35. Iverson, S. J., Kirk, C. L., Hamosh, M., and Newsome, J., Milk lipid digestion in
the neonatal dog: the combined actions of gastric and bile salt stimulated lipases, Biochim.
Biophys. Acta, 1083, 109, 1991.
36. Hamosh, M., Fat digestion in the newborn: role of lingual lipase and preduodenal
digestion, Pediatr. Res., 13, 615, 1979.
37. Lairon, D., Nalbone, G., Lafont, H., Leonardi, J,, Domingo, N., Hauton, J. C.,
and Verger, R., Possible roles of bile lipids and colipase in lipase adsorption, Biochem-
istry, 17, 5263, 1978.
38. Miller, K. W. and Small, D. M., The phase behavior of triolein, cholesterol and
emulsions, J. Colloid Int. Sci., 89, 466, 1982.
39. Carey, M. C., Small, P. M., and Bliss, C. M., Lipid digestion and absorption, Annu.
Rev. Physiol., 45, 651, 1983.
40. Hofman, A. and Borgstrom, B., Physical-chemical state of lipids in intestinal content
during their digestion and absorption, Fed. Proc., 21, 43, 1962.
41. Patton, J, S. and Carey, M. C., Watching fat digestion, Science, 204, 145, 1979.
42. Lindstrom, M., Ljusberg-Wahren, H., and Larsson, K., Aqueous lipid phases of
relevance to intestinal fat digestion and absorption, Lipids, 16, 749, 1981.
43. Clark, B. S., Lawergren, B., and Martin, J, V., Regional intestinal absorptive ca-
pacities for triolein: an alternative to markers, Am. J. Physiol., 225, 574, 1973.
44. Tso, P., Pitts, V., and Granger, D. N., Role of lymph flow in intestinal chylomicron
transport, Am. J. Physiol., 249, G21, 1985.
45. Plucinski, T. M., Hamosh, M., and Hamosh, P., Fat digestion in rat: role of lingual
lipase, Am. J. Physiol., 237, E541, 1979.
46. Watkins, J. B., Ingall, D., and Szczepanik, P., Bile salt metabolism in the newborn,
New Engl. J. Med., 288, 431, 1973.
47. Robberecht, P., Deschodt-Lanckman, M., Camus, J., Bruylands, J., and Christophe,
J., Rat pancreatic hydrolases from birth to weaning and dietary adaptation after weaning,
Am. J. Physiol., 221, 376, 1971.
48. Snook, J. T., Effect of diet on development of exocrine pancreas of the neonatal rat,
Am. J. Physiol., 221,1388,1971.
49. Zoppi, G., Andreotti, G., Pajno-Ferrara, F., Njai, D. M., and Gaburro, D., Exocrine
pancreas function in premature and full term neonates, Pediatr. Res., 6, 880, 1972.
50. Fomon, S. J., Ziegler, E. R., Thomas, L. N., Jensen, R. L., and Filer, L. J.,
Excretion of fat by normal full term infants fed various milks and formulae, Am. J. Clin.
Nutr., 23, 1299, 1970.
51. Katz, L. and Hamilton, J. R., Fat absorption in infants of birth weight less than 1300 g,
J. Pediatr., 85, 608, 1974.
52. Blackberg, L., Lombardo, D., Hernell, 0., Guy, 0., and Olivecrona, T., Bile-salt-
stimulated lipase in human milk and carboxyl ester hydrolase in pancreatic juice. Are
they identical enzymes?, FEES Lett., 136, 284, 1981.
53. Sheldon, W., Congenital pancreatic lipase deficiency, Arch. Dis. Child., 39, 268, 1964.
54. Larbre, F., Hartman, E., Cotton, J. B., Mathieu, M., Charrat, A., and Moreau,
P., Diarrhee chronique par absence de lipase pancreatique, Pediatric, 24, 807, 1969.
55. Figarella, C., DeCaro, A., Leupold, D., and Poley, J., Congenital pancreatic lipase
deficiency, J. Pediatr., 96, 412, 1980.
56. Hilderbrand, H., Borgstrom, B., Bokassy, A., Erlanson-Aibertsson, C., and Helin,
I., Isolated colipase deficiency in two brothers, Gut, 23, 243, 1982.
57. Fayez, K., Ghishan, J., Moran, R., Durie, P. R., and Greene, H. L., Isolated
congenitallipase-colipase deficiency, Gastroenterology, 86, 1580, 1984.
58. Ligumsky, M., Granot, E., Branski, D., Stankiewicz, H., and Goldstein, R., Isolated
lipase and colipase deficiency in two brothers, Gut, 31, 1416, 1990.
59. Ross, C. A. C. and Sammons, H. G., Nonpancreatic lipase in children with pancreatic
fibrosis, Arch. Dis. Child., 30, 428, 1955.
60. Abrams, C. K., Hamosh, M., VanHubbard, S., Dutta, S. K., and Hamosh, P.,
Lingual lipase in cystic fibrosis: quantitation of enzyme activity in the upper small intestine
of patients with exocrine pancreatic insufficiency, J. Clin. Invest., 73, 374, 1984.
61. Cooke, A. R., Control of gastric emptying and motility, Gastroenterology, 68, 804,
1975.
62. Cunningham, K. M., Daly, J., Horowitz, M., and Read, N. W., Gastrointestinal
adaptation to diets of differing fat composition in human volunteers, Gut, 32, 483, 1991.
63. Hopman, W. P. N., Jansen, J, B. M. J,, and Lamers, C. B. H. W., Comparative

study of the effects of equal amounts of fat, protein, and starch on plasma CCK in man,
Scand. J. Gastroenterol., 20, 843, 1985.
64. Konturek, S. J., Tasler, J., Bilski, J., Jong, A. J,, Jansen, J. B. M. J,, and Lamers,
C. B., Physiological role and localization of CCK release in dogs, Am. J. Physiol., 250,
G391, 1986.
65. Aponte, G. W., Fink, A. S., Meyer, J. H., Takemoto, K., and Taylor, I. L., Regional
distribution and release of peptide YY (PYY) with fatty acids of different chain length,
Am. J. Physiol., 249, G745, 1985.
66. Hamosh, M. and Hand, A. R., Development of secretory activity in serous cells of the
rat tongue. Cytological differentiation and accumulation of lingual lipase, Develop. Bioi.,
65, 100, 1978.
67. Armand, M., Borel, P., Cara, L., Senft, M., Chautan, M., Lafont, H., and Lairon,
D., Adaptation of lingual1ipase to dietary fat in rats, J. Nutr., 120, 1148, 1990.
68. Borel, P., Armand, M., Senft, M., Andrey, M., Lafont, H., and Lairon, D., Gastric
lipase: Evidence of an adaptive response to dietary fat in the rabbit, Gastroenterology,
100, 1582, 1991.
69. Bucko, A. and Kopec, Z., Adaptation of enzyme activity of the rat pancreas on altered
food intake, Nutr. Dieta, 10, 276, 1968.
70. Deschodt-Lanckman, M., Robherecht, P., Camus, J., and Christophe, J,, Short-
term adaptation of pancreatic hydrolases to nutritional and physiological stimuli in adult
rats, Biochimie, 53, 789, 1971.
71. Snook, J. T., Dietary regulation of pancreatic enzymes in the rat with emphasis on
carbohydrate, Am. J. Physiol., 221, 1383, 1971.
72. Gidez, L. I., Effect of dietary fat on pancreatic lipase levels in the rat, J. Lipid. Res.,
14, 169, 1973.
73. Saraux, B., Girard-Gioba, A., Ouagued, M., and Vacher, D., Response of the ex-
ocrine pancreas to quantitative and qualitative variations in dietary lipids, Am. J. Physiol.,
243, GIO, 1982.
74. Sabb, J, E., Godfrey, P.M., and Brannon, P.M., Adaptive response ofrat pancreatic
lipase to dietary fats: effects of amount and type of fat, J. Nutr., 116, 892, 1986.
75. Wicker, C. and Puigserver, A., Effects of inverse changes in dietary lipid and carbo-
hydrate on the synthesis of some pancreatic secretory proteins, Eur. J. Biochem., 162,
25, 1987.
76. Behrman, H. R. and Kare, M. R., Adaptation of canine pancreatic enzymes to diet
composition, J. Physiol., 205, 667, 1969.
77. Wills, J, R. and Hinners, S. W., Effect of dietary lipids upon the lipase, pancreatic
cholesterol esterase, plasma triglyceride and total plasma lipids level in the chick, Poult.
Sci., 47, 1732, 1968.
78. Mourot, J, and Corring, T., Adaptation of the lipase-colipase system to dietary lipid
content in pig pancreatic tissue, Ann. Bioi. Anim. Biochim. Biophys., 19, 119, 1979.
79. Simoes Nunes, C., Adaptation of pancreatic lipase to the amount and nature of dietary
lipids in the growing pig, Reprod. Nutr. Develop., 26, 1273, 1986.
80. Meyer, J, H. and Jones, R. S., Canine pancreatic responses to intestinally perfused fat
and products of fat digestion, Am. J. Physiol., 226, 1178, 1974.
81. Malagelada, J.-R., DiMagno, E. P., Summerskill, W. H. J,, and Go, V. L. W.,
Regulation of pancreatic and gall bladder function by intraluminal fatty acids and bile
acids in man, J. Clin. Invest., 58, 493, 1976.
82. Singer, M. V., Pancreatic secretory response to intestinal stimulants: a review, Scand.
J. Gastroenterol., 22, 1, 1987.
83. Boivin, M., Lanspa, S. J., Zinsmeister, A. R., Go, V. L. W., and DiMagno, E. P.,
Are diets associated with different rates of human interdigestive and postprandial pan-
creatic enzyme secretion? Gastroenterology, 99, 1763, 1990.
84. Ballestra, M. C., Manas, M., Mataix, F. J., Martinez-Victoria, E., and Seiquer,
I., Long-term adaptation of pancreatic response by dogs to dietary fats of different degrees
of saturation: olive and sunflower oil, Br. J. Nutr., 64, 487, 1990.
85. Kim, Y. C., Faichney, A., and Ky, L., Endogenous release of secretin by sodium oleate
in dog, Gastroenterology, 78, 1195, 1980.
86. Faichney, A., Chey, W. Y., Kim, Y. C., Lee, K. Y., Kim, M. S., and Chang, T. M.,
Effect of sodium oleate on plasma secretin concentration and pancreatic secretion in dogs,
87. Fink, A. S., Taylor, I. L., Luxemburg, M., and Meyer, J. H., Pancreatic polypeptide
release by intraluminal fatty acids, Metabolism, 32, 1063, 1983.
88. Sommer, H. and Kasper, H., Dietary effects on pancreatic exocrine function. Exper-
iments on the isolated perfused rat pancreas, Ann. Nutr. Metab., 25, 38I, 1981.
89. Belleville, J., Prost, J., and Gillet, M., Effets des regimes riches en triglycerides et
en phospholipides sur les activites de !a lipase et de la phospholipase A du sue pancreatique
et du pancreas de rat, Arch. Int. Physiol. Biochim., 86, 631, 1978.
90. Ouagued, M., Saraux, B., Girard-Globa, A., and Bourdel, G., Differential regulation
of lipase and colipase in the rat pancreas by dietary fat and proteins, J. Nutr., 110, 2302,
1980.
91. Robberecht, P., Deschodt-Lanckman, M., Camus, J., Bruylands, J., and Christophe,
J., Rat pancreatic hydrolases from birth to weaning and dietary adaptation after weaning,
Am. J. Physiol., 221, 376, 1971.
92. Vandermeers-Piret, M. C., Vandermeers, A., Wijns, W., Rathe, J., and Christophe,
j., Lack of adaptation of pancreatic colipase in rats and mice, Am. J. Physiol., 232,
E131, 1977.
93. Girard-Globa, A. and Simond-Cote, E., Nutritional and circadian variations in lipase
activity and colipase saturation in rat pancreas, Ann. Bioi. Anim. Biochim. Biophys.,
17, 539, 1977.
94. Lairon, D., Lacombe, C., Borel, P., Corraze, G., Nibbelink, M., Chautan, M.,
Chanussot, F., and Lafont, H., Beneficial effects of wheat germ on circulating lipo-
proteins and tissue lipids in rats fed a high fat, cholesterol-containing diet, J. Nutr.,
117, 838, I 987.
95. Wicker, C. and Puigserver, A., Changes in mRNA levels of rat pancreatic lipase in
the early days of consumption of a high-lipid diet, Eur. J. Biochem., I80, 563, I989.
96. Stabile, B. E., Borzatta, M., Stubbs, R. S., and Debas, H. T., Intravenous mixed
aminoacids and fats do not stimulate exocrine pancreatic secretion, Am. J. Physiol., 246,
G274. I984.
97. Burns, G. P. and Stein, T. A., Pancreatic enzyme secretion during intravenous fat
infusion, J. Parenteral Enteral Nutr., I I, 60, 1987.
98. Edelman, K. and Valenzuela, J. E., Effect of intravenous lipid on human pancreatic
secretion, Gastroenterology, 85, 1063, I983.
99. Konturek, S. J., Tasler, J., Cieszkowski, M., Jaworek, J.,and Konturek, J., Intra-
venous amino acids and fat stimulate pancreatic secretion, Am. J. Physiol., 236, E678,
I979.
100. Lavan, M., Bazin, R., and Herzog, J., Comparative effects of oral and parenteral
feeding on pancreatic enzymes in the rat, J. Nutr., 104, 1432, I974.
101. Bazin, R., Lavan, M., and Herzog, J., Pancreatic lipase and ketogenic conditions,
Biomedicine, 28, 160, 1978.
102. Hirschi, K. K., Sabb, j. E., and Brannon, P. M., Effects of diet and ketones on rat
pancreatic lipase in cultured acinar cells, J. Nutr., 121, 1129, 1991.
I 03. Simoes Nunes, C., Lack of pancreatic enzyme adaptation to diet carbohydrates and lipids
after proximal small intestine bypass in the pig, Digestion, 25, I08, I982.
I04. Douglas, B. R., Woutersen, R. A., Jansen, J. B. M. J., De Jong, A. J. L., and
Lamers, C. B. H. W., The influence of different nutrients on plasma cholecystokinin
levels in the rat, Experimentia, 44, 2I, I 988.
105. Corring, T. and Chayvialle, J, A., Diet composition and the plasma levels of some
peptides regulating pancreatic secretion in the pig, Reprod. Nutr. Develop., 27, 967, 1987.
106. Schick, J., Kern, H., and Scheele, G., Hormonal stimulation in the exocrine pancreas
results in coordinate and anticoordinate regulation of protein synthesis, J. Cell Bioi., 99,
1569, 1984.
107. Solomon, T. E., Petersen, H., Elashoff, I., and Grossman, M. J., Interaction of
caerulein and secretin on pancreatic size and composition in the rat, Am. J. Physiol.,
235, E714, 1978.
108. Rausch, U., Vasiloudes, P., Riidiger, K., and Kern, H. F., In-vivo stimulation of rat
pancreatic acinar cells by infusion of secretin. I. Changes in enzyme content, pancreatic
fine structure and total rate of protein synthesis, Cell Tissue Res., 242, 633, 1985.
109. Rausch, U., Vasiloudes, P., Riidiger, K., and Kern, H. F., In-vivo stimulation of rat
pancreatic acinar cells by infusion of secretin. II. Changes in individual rates of enzyme
and isoenzyme biosynthesis, Cell Tissue Res., 242, 641, 1985.
I 10. Kemper, K. A. and Brannon, P. M., Effects of secretin on enzymes of cultured
pancreatic acinar cells, FASEB J., 2, A731, 1988.
Ill. Brown, D. D., Gene expression in eukaryotes, Science, 211,667, 1981.
112. Darnell, J, E., Variety in the level of gene control in eukaryotic cells, Nature, 297, 365,
1982.
113. Mitchell, P. J. and Tjian, R., Transcriptional regulation in mammalian cells by sequence-
specific DNA binding proteins, Science, 245, 371, 1989.
114. Walker, M., Edlund, T., Boulet, A., and Rutter, W., Cell specific expression con-
trolled by the 5' flanking region of insulin and chymotrypsin genes, Nature, 306, 557,
1983.
115. Swift, G., Hammer, R., MacDonald, R., and Brinster, R., Tissue specific expression
of the pancreatic elastase I gene in transgenic mice, Cell. 38, 639, 1984.
116. Boulet, A. M., Erwin, C. R., and Rutter, W. J., Cell-specific enhancers in the rat
exocrine pancreas, Proc. Nat!. Acad. Sci. U.S.A., 83, 3599, 1986.
117. Austin, S. A. and Kay, J. E., Translational regulation of protein synthesis in eukaryotes,
Essays Biochem., 18, 79, 1982.
118. Ross, R. J., Messenger RNA turnover in eukaryotic cells, Mol. Bioi. Med., 5, I, 1988.
119. Jamieson, J.D. and Palade, G. E., Synthesis, intracellular transport and discharge of
secretory proteins in stimulated pancreatic exocrine cells, J. Cell Bioi., 50, 135, 1971.
120. Blobel, G. and Dobberstein, B., Transfer of proteins across membranes. Presence of
proteolytically processed and unprocessed nascent immunoglobulin light chains on mem-
brane-bound ribosomes of murine myeloma, J. Cell Bioi., 67, 835, 1975.
121. Barnard, E. A., Biological function of pancreatic ribonuclease, Nature, 221, 340, 1969.
122. Chirgwin, J, M., Przybyla, A. E., MacDonald, R. J., and Rutter, W. J., Isolation
of biologically active ribonucleic acid from sources enriched in ribonuclease, Biochem-
istry, 18, 5294, 1979.
123. Lowe, M. E., Rosenblum, J. L., McEven, P., and Strauss, A. W., Cloning and
characterization of the human colipase eDNA, Biochemistry, 29, 823, 1990.
124. Renaud, W. and Dagorn, J, C., eDNA sequence and deduced amino acid sequence of
human preprocolipase, Pancreas, 6, 157, 1991.
125. Wicker, C. and Puigserver, A., Rat pancreatic colipase mRNA: nucleotide sequence
of a eDNA clone and nutritional regulation by a lipidic diet, Biochem. Biophys. Res.
Commun., 167, 130, 1990.
126. Fukuoka, S.-1., Taniguchi, Y., Kitagawa, Y., and Scheele, G., Full length eDNA
sequence encoding canine pancreatic colipase, Nucleic Acids Res., 18, 5549, 1990.
127. Lowe, M. E., Rosenblum, J. L., and Strauss, A. W., Cloning and characterization of
human pancreatic lipase eDNA, J. Bioi. Chern., 264, 20042, 1989.
128. Kerfelec, B., LaForge, K. S., Puigserver, A., and Scheele, G., Primary structures of
canine pancreatic lipase and phospholipase A2 messenger RNAs, Pancreas, I, 430, 1986.
129. Wicker-Planquart, C. and Puigserver, A., Primary structure of rat pancreatic lipase
mRNA, FEBS Lett., 296, 61, 1992.
130. Wicker, C., Scheele, G. A., and Puigserver, A., Pancreatic adaptation to dietary lipids
is mediated by changes in lipase mRNA, Biochirnie, 70, 1277, 1988.
131. Wicker, C. and Puigserver, A., Expression of rat pancreatic lipase gene is modulated
by a lipid-rich diet at a transcriptional level, Biochern. Biophys. Res. Cornrnun., 166,
358, 1990.
132. Kern, H. F., Rausch, U., and Scheele, G., Regulation of gene expression in pancreatic
adaptation to nutritional substrates or hormones, Gut, 28, 89, 1987.
133. Beaudoin, A. R., Grondin, G., St-Jean, P., Vacherear, A., Cabana, C., and Grossman,
A., Steroids and secretory function of the exocrine pancreas, Endocrinology, 119, 2106,
1986.
134. Duan, R.-D. and Erlanson-Aibertsson, C., The anticoordinate changes of pancreatic
lipase and colipase activity to amylase activity induced by adrenalectomy in normal and
diabetes rats, Int. J. Pancreatol., 6, 271, 1990.
135. Duan, R.-D., Poensgen, J., and Erlanson-Albertsson, C., Adrenalectomy anticoor-
dinately changes the synthesis of rat pancreatic enzymes, Digestion, 43, 138, 1989.
136. Wicker, C., Puigserver, A., Rausch, U., Scheele, G., and Kern, H., Multiple-level
caerulein control of the gene expression of secretory proteins in the rat pancreas, Eur.
J. Biochern., 151,461, 1985.
137. Puigserver, A., Wicker, C., and Gaucher, C., Aspects moleculaires de !'adaptation
des enzymes pancreatiques et intestinales au regime alimentaire, Reprod. Nutr. Dev., 25,
787, 1985.
138. Mickel, F. S., Weidenbach, F., Swarovsky, B., LaForge, K. S., and Scheele, G. A.,
Structure of the canine pancreatic lipase gene, J. Bioi. Chern., 264, 12895, 1989.
139. Howard, G., Keller, P. R., Johnson, T. M., and Meisler, M. H., Binding of a
pancreatic nuclear protein is correlated with amylase enhancer activity, Nucleic Acids
Res., 17, 8185, 1989.
140. Meister, A., Weinrich, S. L., Nelson, C., and Rutter, W. J,, The chymotrypsin
enhancer core. Specific factor binding and biological activity, J. Bioi. Chern., 264, 20744,
1989.
141. Nir, U., Walker, M.D., and Rutter, W. J., Regulation of rat insulin I gene expression:
evidence for negative regulation in nonpancreatic cells, Proc. Natl. Acad. Sci. U.S.A.,
83, 3180, 1986.
142. Zhang, Z. X., Kumar, V., Rivera, R. T., Chisholm, J,, and Biswas, D. K., Cis-
acting negative regulatory element of prolactin gene, J. Bioi. Chern., 265, 4785, 1990.
143. Osborn, L., Rosenberg, M. P., Keller, S. C., Ting, C.-N., and Meisler, M. H.,
Insulin response of a hybrid amylase/CAT gene in transgenic mice, J. Bioi. Chern., 263,
16519, 1988.
144. Selden, R. F., Skoskiewicz, M. J,, Burke Howie, K., Russell, P. S., and Goodman,
H. M., Regulation of human insulin gene expression in transgenic mice, Nature (London),
321,525, 1986.
Chapter 4
NUTRITIONAL CONTROL OF GASTROINTESTINAL

HORMONE GENE EXPRESSION
Pauline Kay Lund
TABLE OF CONTENTS
I. Introduction ....................................................... 92
II. General Considerations ........................................... 93
III. Gastrin Gene Expression .......................................... 95

A. Background ............................................... 95
B. Fasting Reduces Gastrin mRNA Abundance ............... 95
C. Specific Nutrients Increase Gastrin
mRNA Abundance ........................................ 97
D. Model Systems to Analyze the Molecular Basis
for Nutrient Effects on Gastrin Gene Expression .......... 98
IV. Cholecystokinin (CCK) Gene Expression ......................... 98

A. Background ............................................... 98
B. Fasting Reduces CCK mRNA Abundance in
Duodenum ................................................. 99
C. Refeeding and Specific Nutrients Increase CCK
mRNA Abundance in Duodenum ......................... 100
D. Model Systems to Analyze Nutrient Effects on
CCK Gene Expression ................................... 10 I
V. Somatostatin Gene Expression ................................... 102

A. Background .............................................. 102
B. Fasting Elicits Region-Specific Effects on
Gastrointestinal Somatostatin mRNA ..................... 102
C. Nutrient Effects on Gastric and Intestinal
Somatostatin Gene Expression ............................ 103
D. 21Model Systems to Analyze Nutrient Effects on
Somatostatin Gene Expression ............................ 104
VI. Glucagon!Enteroglucagon ........................................ 105

A. Background .............................................. 105
B. Nutrient Control of Glucagon Gene Expression
in Pancreas ............................................... 106
0-8493-6961-41931$0.00 + $.50
© 1993 by CRC Pres>, Inc. 91
C. Nutrient Control of Glucagon Gene Expression

in the Intestine ........................................... 108
D. Nutrient Dependent and -Independent Changes in
Intestinal Proglucagon Gene Expression after
Proximal Small Bowel Resection ......................... 108
E. Model Systems to Analyze Nutrient Control of
Proglucagon Gene Expresion ............................. 109
VII. Conclusions and Future Directions ............................... 110
Acknowledgments ....................................................... 111
References ............................................................... 111
I. INTRODUCTION
The gastrointestinal (GI) hormones are a large and complex group of

peptides that are synthesized by endocrine cells in the mucosal lining of the
gut. 1 They regulate the functioning of the GI tract and the secretion of elec-
trolytes and digestive enzymes that are necessary for optimal digestion and
absorption of ingested nutrients. Radioimmunoassays of plasma and extracts
of gastrointestinal mucosa have established that the secretion of different GI
hormones is regulated by specific nutrients. ,_ 3 Until recently there has been
little emphasis on whether these effects are linked to, or mediated by, effects
on GI hormone gene expression. An understanding of nutrient regulation of
GI hormone gene expression is relevant to understanding the factors that
underlie normal development and functioning of the GI tract and for defining
the role of specific nutrients and GI hormones in the etiology, pathophysi-
ology, and treatment of GI disease. The cloning and characterization of the
genes and complementary DNAs (cDNAs) encoding many of the GI hormones
have provided the potential for investigations of nutritional control of GI
hormone gene expression. Progress in this area will be reviewed with partic-
ular emphasis on gastrin, cholecystokinin (CCK), somatostatin, and entero-
glucagon. These hormones represent only a subset of the large array of hor-
mones expressed in the GI tract. However, each of these hormones has a
relatively well-defined physiological role, and collectively there is more in-
Lund 93
formation available about their regulation by nutrients than is available for

other GI hormones.
II. GENERAL CONSIDERATIONS

To date, most studies of nutritional regulation of GI hormone gene expres-
sion have focused on quantification of the steady-state concentrations of the
particular mRNA in total RNA or total messenger RNA (mRNA) extracted
from the entire wall of a given region of the gut or from gut mucosa. The
extracted RNA is then analyzed by Northern blot hybridization, dot or slot
blot hybridization, or solution hybridization using a radiolabeled DNA or
RNA probe that is complementary to the GI hormone mRNA of interest.
There are some general features of the GI endocrine system which pose
particular challenges in understanding nutritional control of gene expression
using these or other methods:
Heterogeneity-In contrast to "classical endocrine glands" that are com-
prised of primarily endocrine cells, the GI endocrine system is a diffuse and
heterogeneous system. Multiple endocrine cell types are dispersed within the
mucosal layer of the gut wall. The majority of cell types in the mucosa are
"non-endocrine" epithelial cells and supporting mesenchymal cells. In ad-
dition to mucosa, the gut wall contains underlying submucosa, muscle layers,
and serosa. From a practical standpoint this means that the mRNAs encoding
an individual GI hormone represent a relatively small fraction of the
total mRNA expressed in the entire gut wall or mucosa. Highly sensitive
hybridization methods must therefore be employed to analyze GI hormone
mRNAs. The heterogeneity of cell types within the gut wall also complicates
the interpretation of nutrient dependent changes in steady-state concentrations
of GI hormone mRNAs in vivo. Such changes may reflect a true alteration
in the level of expression of the gene encoding the GI hormone or a change
in the stability of the mRNA. Alternatively, nutrient-dependent changes in
steady-state levels of GI hormone mRNAs may be secondary to changes in
relative abundance of other, more predominant mRNAs derived from the
"non-endocrine" cell types. One way to distinguish these possibilities is by
use of quantitative or semiquantitative in situ hybridization histochemistry in
parallel with the hybridization analyses on extracted RNA. Quantitative in
situ hybridization could establish if altered levels of steady-state mRNA reflect
altered levels of mRNA in individual GI endocrine cells. To date, this has
been done infrequently, probably because of the inherent difficulties in de-
veloping quantitative in situ hybridization methods.
Turnover and renewal-Cells within the mucosal lining of the GI tract
have a high rate of turnover and renewal and these processes are highly
susceptible to nutritional status. 4 •5 Both the amount and the type of ingested
nutrient exert profound effects on the mass of mucosal cells within a given
gut region by altering the rate of stem cell division and the rate of renewal
of the mucosal lining. Such alterations in mucosal mass are termed adaptive
growth of the GI tract. For example, fasting results in hypoplasia and atrophy
of the gut mucosa while hyperphagia results in mucosal hyperplasia. 4 •5 It is
not clear whether alterations in nutrient load or type affect the rate of turnover
of GI endocrine cells to the same degree as other mucosal cell types. Indeed,
there is evidence that some GI endocrine cells tum over much more slowly
than other epithelial cells. 6 Changes in the concentrations of GI hormone
mRNAs may therefore be secondary to differential effects of nutrient status
on relative mass of endocrine and "non-endocrine" cell types and may not
reflect the effect of nutrient on gene expression per se. Histological analyses
of the mucosa performed in parallel with quantification of GI hormone mRNAs
could help discern if apparent alterations in the levels of GI hormone mRNAs
are secondary to alterations in· relative mass of endocrine cells. To date, such
combined studies have been performed very infrequently.
Chemical and mechanical stimuli-Ingested nutrients present a complex
stimulus to the GI tract. Mechanical stimulation results from distention of the
gut by food. Chemical stimulation occurs via nutrients within the lumen of
the GI tract or via altered concentrations of absorbed nutrients in the circu-
lation. Gut responses to these nutritional stimuli can involve long neural
reflexes mediated by the central nervous system and short neural reflexes
mediated by the intrinsic neuronal plexuses in the gut wall. Mechanical and
chemical stimuli may stimulate secretion of GI hormones into the circulation
or local secretion into the extracellular fluid. Nutrient effects on GI hormone
gene expression therefore may be mediated by direct effects of nutrients on
the GI endocrine cells, by indirect effects of neuronal reflexes, or by indirect
effects of other GI hormones acting in an endocrine and paracrine manner.
Thus, careful experimental design is required to elucidate precisely which of
these multiple nutrient-mediated events underlies a nutrient-dependent change
in GI hormone gene expression.
Hormone expression-A given GI hormone may be expressed in mul-
tiple regions of the gut or accessory tissues. A nutrient-mediated effect ob-
served in one gut region or cell type may not necessarily reflect effects in
other regions or cells.
Cis- and trans-acting factors-Cell-specific and regulated expression of
most genes appear to be determined by cis-acting regulatory elements com-
prised of DNA sequences that flank the transcribed portions of the gene or
lie within the structural gene. 7 Trans-acting factors, usually proteins, bind to
these DNA regulatory elements and alter the rate of gene transcription. 7 It is
increasingly clear that steady-state levels of many mRNAs are regulated by
posttranscriptional mechanisms, probably via effects on RNA binding factors
that influence mRNA stability and turnover. 8 Ultimately, an understanding
of nutrient regulation of GI hormone-gene expression and synthesis will re-
Lund 95
quire characterization of the cis-elements, trans-acting factors, and posttran-

scriptional events that mediate the nutrient effects. Even though many of the
genes for gastrointestinal hormones were cloned in the early 1980s, remark-
ably little progress has been made in characterizing the promoters and reg-
ulatory elements that determine basal, cell-specific, and regulated expression
in enteroendocrine cells or GI tract in vivo. This likely relates to the limited
availability of useful model culture systems for analyses of GI hormone-gene
expression. It is difficult to maintain gut regions or slices in culture. The
diffuse nature of the GI endocrine system creates problems in isolation and
culture of homogeneous GI endocrine cell populations. Relatively few cell
lines express GI hormones and those that do are often phenotypically very
different from enteroendocrine cells. Progress towards development and use
of model systems for the analyses of GI hormone-gene expression will be
described.
III. GASTRIN GENE EXPRESSION
A. BACKGROUND
Gastrin is synthesized primarily in G cells of the antrum of the stomach
and exists as two major forms of 34 (G34) and 17 (G 17) amino acids. 1 The
two forms derive from differential processing of a single precursor encoded
by a single gene and mRNA. 1 Gastrin has two major physiological roles, the
stimulation of gastric acid secretion and stimulation of mucosal growth in the
acid-secreting portion of the stomach. 1 Since gastric acid secretion is a major
determinant, if not causative factor in peptic ulcer disease, understanding the
molecular basis of nutritional effects on gastrin synthesis and release has
implications for understanding the etiology and pathophysiology of peptic
ulcer.
Fasting results in decreases in plasma gastrin, decreases in antral gastrin
concentrations, and a decrease in antral G cell density. 1•9 •10 These effects are
reversed by refeeding. 1 •9 Gastrin is released into the circulation in response
to food ingestion. 1 The chemical composition of food influences gastrin re-
lease. Proteins, digestion products of proteins, and individual amino acids
are particularly strong stimulants of gastrin secretion. 1 • 11 • 12 Neural reflexes
also influence gastrin release. 1 Recent studies, conducted primarily in rat and
dog, suggest that the effects of fasting and nutrient-induced changes in cir-
culating and antral gastrin concentrations are mediated at the level of gastrin
mRNA (Figure 1). These findings therefore indicate that nutrient status in-
fluences gastrin gene transcription of mRNA stability.
B. FASTING REDUCES GASTRIN mRNA ABUNDANCE

Fasting results in parallel decreases in antral gastrin concentrations and
antral gastrin mRNAY Effects are rapid and are observed within hours of
96 Nutrition and Gene E xpression
circulating
metabolites
and nutri en ts
(? )
l l' Ur ~ l
nm lr o l
I'.' I
Gastrin
I f-A
Ch:
pep tones
cell amino acid s
amin es
~ G RP
FIGURE l. Schematic to summarize nutrient effects on the gastrin cell that are implicated in
regulation of gastrin gene ex.pression: + = positive effect; - = negative effect; ? = indirect
evidence for an effect. Peptones, amino acid s. and amines appear to act via the lumen either by
cell surface receptors or after uptake into the G cell as indicated.
fasting . Effects are also progressive since a continued decrease in gastrin

mRNA occurs with prolonged fasting . 13 It is possible that decreases in relative
numbers of G cells contribute to the progressive decrease in antral gastrin
mRNA observed after a prolonged fast. 9 However, alterations in relative mass
of G cells are not likely to be rapid enough to account for a decline in antral
gastrin mRNA within 12 hours of fasting . 13 It seems likely, therefore, that
fasting alters the rate of gastrin gene transcription or gastrin mRNA stability
although this has not been tested directly . The mechanism underlying these
changes is not well defined. Fasting leads to a decline in gastric pH , in part
due to the loss of buffering capacity of food . Low gastric pH is known to
inhibit gastrin secretion. 1 Achlorhydria induced by pharmacological inhibition
of gastric acid secretion or fundectomy leads to an increase in antral gastrin
mRNA . 14 • 15 Direct effects of hydrogen ions on the G cell to decrease gastrin
mRNA during a fast cannot be excluded. It seems likely, however, that
neuronal reflexes or endocrine or paracrine consequences of reduced gastric
pH contribute to the fasting-induced decrease in antral gastrin mRNA.
Somatostatin is a primary candidate as mediator of fasting-induced de-
creases in gastrin mRNA . In vivo, reduced gastric pH stimulates somatostatin
Lund 97
release from gastric D cells, and somatostatin decreases gastrin release. 1 • 16 . 18

Infusion of a somatostatin analogue can prevent the increases in antral gastrin
rnRNA induced by achlorhydria. 15 Recent studies in isolated dog antral mu-
cosa provide direct evidence that somatostatin inhibits gastrin gene expression.
In this system, somatostatin has no effect on basal gastrin gene transcription
but does decrease gastrin gene transcription in response to a number of stimuli
and also increases gastrin mRNA turnover. 19 ·20 Definitive evidence that so-
matostatin mediates fasting-induced decreases in antral gastrin mRNA will
require additional studies, such as immunoneutralization of somatostatin in
fasted rats in vivo and analyses of effects on antral gastrin gene transcription
and mRNA.
C. SPECIFIC NUTRIENTS INCREASE GASTRIN

mRNA ABUNDANCE
Refeeding of fasted rats results in rapid increases in gastrin rnRNA. 21
Studies where different nutrients were directly instilled into the stomach
indicate that the effects of refeeding on gastrin rnRNA depend on the type
of nutrient. Gastric instillation of peptones or amino acids, such as phenyl-
alanine, elicits the greatest increases in gastrin mRNA and effects are apparent
within 15 min. 13 More modest increases in antral gastrin rnRNA occur in
response to gastric instillation of fat but not in response to glucose or saline. 13
The rapid induction of antral gastrin mRNA by gastric instillation of
peptones, amino acids, and fat suggests effects on gastrin gene transcription
or mRNA stability (Figure I). As with fasting, the precise mediators of these
effects are not well defined. Decreased release of somatostatin and thereby
a reduced negative influence of somatostatin on gastrin synthesis may mediate
the nutrient dependent increases in antral gastrin mRNA. Compared with
effects on gastrin mRNA, gastric instillation of nutrients has reciprocal effects
on gastric somatostatin mRNA and peptide. 13 The effects of nutrient on gastric
somatostatin or somatostatin mRNA are, however, slower than effects on
gastrin mRNA. 13 Thus, if reduced somatostatin does mediate nutrient effects
on gastrin mRNA, a decreased release of somatostatin, locally in the stomach,
may precede nutrient dependent changes in somatostatin synthesis.
As well as somatostatin, direct effects of nutrients on gastrin gene expres-
sion cannot be excluded (Figure 1). Peptone, amino acids, and amines that
derive from decarboxylation of amino acids can all stimulate gastrin release
from isolated G cells in culture suggesting direct effects of these nutients on
G cell function. 12 To date, however, the direct effect of nutrients on gastrin
gene expression in isolated G cells has not been evaluated.
Gastrin release in vivo and in vitro is stimulated by acetylcholine released
from vagal efferents and by gastrin releasing peptide (GRP; the mammalian
counterpart of the amphibian peptide bombesin) released from efferent nerve
endings in proximity to the G cell. 1 To date, little is known about the role
of neural influences in nutrient effects on gastrin gene expression (Figure 1).
D. MODEL SYSTEMS TO ANALYZE THE MOLECULAR BASIS

FOR NUTRIENT EFFECTS ON GASTRIN GENE EXPRESSION
It it clear that additional studies are required to establish definitively that
the effects of nutrients on antral gastrin mRNA are mediated at the level of
gene transcription and to elucidate the precise stimuli and molecular mech-
anisms that underlie these effects. Progress has been made in the development
of a number of model systems to address these issues.
Isolated G cells represent one useful model system to study nutrient effects
on gastrin gene expression. 22 Isolated G cells offer the advantage that they
are not transformed cells but are limited by the small numbers of cells that
can be isolated and by their limited viability. The development of highly
sensitive techniques to measure gene expression, such as quantitative poly-
merase chain reaction (PCR), should facilitate analyses of gastrin gene expres-
sion in isolated G cells.
Isolated antral mucosa maintained in short-term tissue culture also rep-
resents a useful model system to analyze gastrin gene expression. This system
has the advantage that it has been used successfully to measure gastrin gene
transcription in response to a number of stimuli and retains some of the
paracrine and neural influences that may influence nutrient effects. 19 •20 The
major limitation of this system is its short-term viability.
Recently, several cell lines have been used to analyze gastrin gene expres-
sion and have led to the characterization of the gastrin promoter and cis-
elements that mediate changes in gastrin gene transcription in response to
hormones or neuropeptides. 23 - 27 In GH4 pituitary cell lines and a number of
insulinoma cell lines, somatostatin inhibits gastrin gene expression, and GRP
may have modest stimulatory effects on gastrin gene expression. 24 ·27 These
findings provide indirect evidence that somatostatin or GRP may be involved
in fasting or nutrient effects on gastrin gene expression. The applicability of
such model cell lines to nutrient effects on gastrin gene transcription may be
limited since most cell lines are neoplastic as well as divorced from the normal
G cell environment. The progress made using these cell lines to define putative
gastrin cis-regulatory elements should facilitate the development of other
systems to assess nutrient effects on gastrin gene expression. One potentially
promising approach is to generate transgenic animals with putative gastrin
cis-elements linked to a readily assayable reporter gene. Such transgenic
models should help elucidate the cis-regulatory elements that mediate nutrient
effects on gastrin gene expression in vivo.
IV. CHOLECYSTOKININ (CCK) GENE EXPRESSION

A. BACKGROUND
CCK is structurally related to gastrin but derives from a different
Lund 99
gene. 28 - 3° CCK is synthesized by endocrine cells in the mucosa of the small

intestine as well as expression in neurons within the central and peripheral
nervous system. 1 It has long been recognized that CCK is a stimulant of
pancreatic enzyme secretion and gallbladder contraction. 31 •32 More recent
evidence suggests that CCK exerts trophic effects on the exocrine pancreas. 33
Understanding the nutritional effects on CCK gene expression has relevance
for understanding the role of nutrients in normal development and functioning
of the exocrine pancreas as well as the pathophysiology of pancreatic disease
such as pancreatitis.
CCK is secreted into the circulation after ingestion of nutrients although
there are species differences in the responses. In man, fat, fatty acids, protein,
and amino acids are potent stimulants of CCK secretion and glucose is a less
potent stimulant. 34 In dogs, intraduodenal fatty acids and amino acids stimulate
CCK secretion whereas intact proteins are poor stimulants. 35 •36 In the rat,
ingested protein appears to be the major stimulant of CCK secretion whereas
amino acids, carbohydrates, and fats are poor stimulants. 37•38
CCK secretion is controlled by negative feedback. Active proteases in
the intestinal lumen inhibit CCK secretion. CCK secretion is stimulated by
inactivation of intestinal proteases or by binding of proteases to substrates or
trypsin inhibitor. 38 In other species there is emerging evidence for feedback
control of CCK secretion via intestinal proteases. 39 Recent studies have elu-
cidated one mechanism underlying the feedback control of CCK secretion in
the rat. In this species, the rate of CCK secretion is tightly linked to the
amount of protein within the duodenal lumen via a novel CCK releasing
peptide, termed monitor peptide. 39-41 Rat monitor peptide has been sequenced
and cloned. 4 Monitor peptide possesses trypsin inhibitor activity and is co-
secreted from exocrine pancreas into the intestinal lumen together with other
digestive enzymes. 39-41 Monitor peptide appears to compete with ingested
protein for binding to trypsin. In the presence of ingested protein, monitor
peptide remains free from trypsin to stimulate CCK secretion from duodenal
endocrine cells. In the absence of ingested protein, monitor peptide forms a
tight complex with trypsin so that the CCK releasing activity of monitor
peptide is inhibited. 39-41 These findings raise interesting possibilities that as
well as CCK secretion, CCK gene expression may be influenced by monitor
peptide or other luminal protease inhibitors (Figure 2).
Recent studies in rat indicate that some of the nutrient effects on CCK
secretion are mediated at the level ofCCK mRNA and CCK gene transcription
(summarized in Figure 2). It is probably important to note that findings in
the rat may not be directly extrapolated to other species including man since
there are species differences in nutrient effects on CCK secretion. 33 - 38
B. FASTING REDUCES CCK mRNA ABUNDANCE

IN DUODENUM
Prolonged fasting produces a gradual and progressive decline in CCK
mRNA. 42 •43 It is not clear whether the decrease in CCK mRNA represents
GASTRI N
GASTIU
FIGURE 2. Summary of evidence for nutrient mediated effects on expression of gastrin,

somatostatin (SOM), secretin. enteroglucagon (enterogluc), and CCK genes: + = positive effect
on gene expression; - = negative effect; ? = indirect evidence for an effect; TI = monitor
peptide.
effects on CCK gene transcription, mRNA stability, or abundance of CCK

cells relative to other mucosal cell types. The mechanism underlying the
fasting-induced decrease in CCK mRNA is not known but is not paralleled
by effects on duodenal somatostatin mRNA or peptide. 43 If monitor peptide
influences CCK gene transcription, fasting-induced decreases in CCK mRNA
may relate to reduced availability of free monitor peptide for stimulation of
CCK producing cells (Figure 2). This possibility has not been tested.
C. REFEEDING AND SPECIFIC NUTRIENTS INCREASE CCK

mRNA ABUNDANCE IN DUODENUM
After only one day of refeeding of fasted rats, CCK mRNA levels are
restored to normal. 43 When elemental diet is given to rats by perfusion into
the intestinal lumen, there is no effect on CCK secretion or CCK mRNA. 44
In contrast, when a soybean trypsin inhibitor is perfused into the intestinal
lumen concomitantly with the elemental diet, there is a rapid increase in CCK
secretion and in CCK mRNA. 44 The enhanced CCK secretion is presumably
because the trypsin inhibitor competes with the endogenous monitor peptide
for binding to trypsin and thereby "allows" the monitor peptide to stimulate
CCK producing cells. The increase in CCK mRNA in response to soybean
trypsin inhibitor is mediated at the level of CCK gene transcription. 31 The
precise stimuli and mechanisms underlying this effect are unknown. Clearly,
monitor peptide is a primary candidate, but effects of monitor peptide on
CCK gene transcription have not been reported. Direct stimulatory effects of
Lund 101
protein or other nutrients on the CCK producing cell also cannot be excluded
(Figure 2).
Another candidate mediator of nutrient-mediated effects on CCK gene
transcription is GRP which is known to stimulate CCK secretion and is
localized in efferent nerve endings throughout the gastrointestinal tract. 1 Evi-
dence againt GRP as a mediator of increased CCK gene transcription in
response to luminal nutrient stems from studies demonstrating that intravenous
bombesin (the amphibian counterpart of mammalian GRP) transiently in-
creases plasma levels of CCK but does not influence CCK mRNA levels in
duodenum. 45 This finding is of interest in that it indicates that increases in
CCK secretion are not obligatorily linked to increases in CCK mRNA. 45 On
the other hand, a role for GRP in nutrient effects to increase CCK mRNA
cannot entirely be excluded since intravenous bombesin may not necessarily
elicit the same effects as activation of GRP efferents by nutrients within the
intestinal lumen.
The soybean trypsin inhibitor model is a well defined model to study
effects of luminal protein on CCK mRNA since overall nutrient status of the
animals is not significantly affected. 44 Using this model Kanayama and Liddle
have demonstrated recently that somatostatin can inhibit the induction of CCK
mRNA in response to soybean trypsin inhibitor. 46 This raises the possibility
that, as with gastrin, a component of nutrient effects to stimulate CCK gene
transcription may be mediated by removal of a negative influence of so-
matostatin (Figure 2). To date, however, there are no data to directly support
this possibility. Thus, the precise mediators of nutrient effects on CCK gene
transcription are not defined (Figure 2).
D. MODEL SYSTEMS TO ANALYZE NUTRIENT EFFECTS ON

CCK GENE EXPRESSION
In contrast to gastrin there has been little progress to date in developing
model cell culture systems to study regulation of the CCK gene. Isolation
and culture of enriched populations of enteric CCK endocrine cells has not
yet been reported, and CCK gene expression has not been studied in short-
term culture of intestinal mucosa. Recent studies have indicated that chimeric
genes containing the rat CCK 5' flanking region and a chloramphenicol
acetyltransferase gene are active when transfected into a number of cell
lines. 47•48 These cell lines have been used as model systems to define cis-
elements that determine basal expression of the CCK gene and increased CCK
expression induced in response to elevated cyclic AMP or phorbol esters. 47.48
Specific nuclear proteins appear to bind to these cis-elements and therefore
are candidate trans-acting factors for regulation of CCK gene transcription. 47
It is not clear that findings in these systems will relate to control of the CCK
gene in enteroendocrine cells. Progress in understanding nutritional control
of the CCK gene in endocrine cells of the small intestine thus awaits devel-
opment of useful model systems and progress in characterizing the CCK cis-
regulatory elements.
V. SOMATOSTATIN GENE EXPRESSION
A. BACKGROUND
Somatostatin is a tetradecapeptide originally isolated from the hypothal-
amus. 49 It is now established that the GI tract and pancreas are major sites
of somatostatin synthesis. 1•50•51 In stomach and pancreas, like brain, the pre-
dominant form of somatostatin is a tetradecapeptide whereas a 28 amino acid
form predominates in the intestine. 5 1-53 These two forms of somatostatin derive
from differential processing of a precursor encoded by a single somatostatin
gene and mRNA. 54- 56 Somatostatin is synthesized in D cells of the pancreatic
islets and D cells within the mucosa of the stomach, small intestine, and large
intestine. 1 •50- 53 •57•58 In some species, somatostatin is also synthesized by neu-
rons within the intrinsic neuronal plexuses of the gut. 1•58
Somatostatin has widespread actions that have been reviewed elsewhere
and are beyond the scope of this chapter. 1 •50 In general, somatostatin is an
inhibitory peptide. In the GI tract, somatostatin acts to inhibit gastric acid
secretion, pancreatic exocrine secretion, and the secretion of hormones from
pancreatic islet and GI endocrine cells. 1 Available evidence suggests that rather
than acting in an endocrine manner, gastrointestinal somatostatin may act in
a paracrine manner to modulate target cell function, possibly via release of
somatostatin from cytoplasmic processes projecting from the D cells onto
target cells. 59
A number of studies suggest effects of nutrients on somatostatin concen-
trations in plasma and somatostatin content of different regions of the GI
tract, especially stomach and pancreas. 1.3.so The interpretation of such findings
is complicated by the multiple cellular sites of somatostatin biosynthesis. In
addition, since somatostatin acts in a paracrine manner, tissue content of
somatostatin may remain unchanged even in the face of local alterations in
somatostatin synthesis or secretion by specific cell populations. As well as
providing information about nutrient regulation of somatostatin gene expres-
sion, analyses of nutrient effects on somatostatin mRNA content in different
regions of GI tract may therefore aid in the interpretation of nutrient effects
on D cell function.
B. FASTING ELICITS REGION-SPECIFIC EFFECTS ON

GASTROINTESTINAL SOMATOSTATIN mRNA
Within 12 h of fasting there are increases in somatostatin mRNA in the
gastric antrum that are maintained with prolonged fasting. 13 • 14 Since short
duration fasting does not alter the relative number of gastric D cells, 10 these
observations suggest that fasting increases somatostatin gene transcription or
Lund 103
mRNA stability. In contrast to somatostatin mRNA, fasting does not signif-

icantly alter the concentrations of immunoreactive somatostatin in the an-
trum.14 Measurement of somatostatin mRNA may therefore provide an in-
dication of alterations in gastric D cell function that are not readily apparent
by the measurement of peptide. Fasting induces no significant change in
somatostatin mRNA or peptide within the duodenum, suggesting that fasting
induces region-specific effects on somatostatin gene transcription/mRNA sta-
bility within the GI tract. 43
While the mechanism underlying fasting-induced increases in antral so-
matostatin mRNA is not defined, it likely relates to increased acidity and
reduced pH of antral contents. Effects of acidification of antral contents on
antral somatostatin mRNA have not been tested directly but reduced gastric
pH stimulates somatostatin secretion from gastric D cells. 1·16 Antral somato-
statin mRNA is decreased in response to elevation of gastric pH by fundectomy
or inhibition of acid secretion from parietal cells using omeprazole, a blocker
of the parietal cell hydrogen/potassium ATPase. 14 ·15 Thus, a number of lines
of evidence suggests that reduced gastric pH may mediate fasting-induced
increases in antral somatostatin gene expression (Figure 2). If so, this could
provide an explanation for the region-specific effects of fasting on somato-
statin mRNA. Pancreatic juice and bile serve to neutralize gastric acid as it
enters the duodenum so that duodenal D cells would not be subject to the
same low pH as antral D cells.
C. NUTRIENT EFFECTS ON GASTRIC AND INTESTINAL

SOMATOSTATIN GENE EXPRESSION
To date, few studies have addressed the role of specific nutrients in
regulation of somatostatin gene expression. Refeeding of fasted rats with solid
food, or gastric instillation of peptone, amino acids, or oil, results in a rapid
decline in antral somatostatin mRNAY· 14 Specific nutrients within the stom-
ach therefore appear to inhibit somatostatin gene expression or reduce mRNA
stability. 13 ·14 As with fasting, effects of nutrients on antral somatostatin mRNA
are more pronounced than effects on peptide. 14 An increase in gastric pH as
a result of the buffering capacity of nutrients may account for their effects
on antral somatostatin mRNA. Indirect effects via gastrin release or other
events that may be modulated by the nutrients cannot be excluded. 14
Duodenal somatostatin mRNA is not altered by refeeding of fasted rats. 43
Perfusion of the intestinal lumen with soybean trypsin inhibitor does, however,
increase duodenal somatostatin mRNA (Figure 2). 44 It is not clear whether
this is a direct effect on duodenal D cells or an indirect effect due to stimulation
of CCK secretion, secretion of other hormones, or secretion of pancreatic
enzymes. Somatostatin appears able to exert "autoregulation" of its own
synthesis in the duodenum since intravenous infusion of somatostatin inhibits
the trypsin inhibitor induced increase in duodenal somatostatin mRNA. 46 This
may be a direct effect of somatostatin on D cells since similar autoregulation

of somatostatin synthesis and secretion is observed in isolated canine D cells. 60 •61
Alterations in duodenal somatostatin mRNA induced by trypsin inhibitor or
somatostatin are not accompanied by effects on peptide, again pointing to the
potential advantages of analyses of somatostatin mRNA as a measure of D
cell function.
Intravenous bombesin produces rapid, quite dramatic yet transient in-
creases in duodenal somatostatin mRNA, implicating the neuropeptide GRP
as one potential mediator of nutrient effects on somatostatin gene expression
(Figure 2). 45 It is not clear whether the effect of bombesin is specific for
duodenal D cells or also influences gastric D cells.
D. MODEL SYSTEMS TO ANALYZE NUTRIENT EFFECTS ON

SOMATOSTATIN GENE EXPRESSION
Accumulating evidence, therefore, suggests a role of nutrients in mod-
ulating somatostatin gene expression and suggests differential effects of nu-
trient status in different regions of the GI tract (Figure 2). The complexity
of the somatostatin system in the GI tract suggests that additional approaches
to nutrient control of somatostatin gene expression in vivo are warranted. The
paracrine mode of action of somatostatin suggests that highly localized changes
in somatostatin gene expression may occur in subsets of D cells within a
given region of gut. Such changes may be missed or underestimated by
analyses of gene expression in a whole organ or gut region. The multiplicity
of cell types that express the somatostatin gene in the gastrointestinal tract
also raises the possibility for cell-type specific effects of nutrients on so-
matostatin gene expression, but few or no in vivo studies have addressed this
possibility. Interesting information on cell specific regulation of somatostatin
gene expression may emerge from a combination of histological methods,
such as in situ hybridization, with other approaches used to date.
The cis-acting elements that dictate somatostatin gene expression have
been studied extensively in cultured cell lines. A cyclic AMP responsive
octanucleotide TGACGTCA (CRE) is important for basal and cyclic AMP
regulated expression of the somatostatin gene in a number of cell lines. 62 This
element is bound by a CRE binding protein CREB which belongs to a growing
family of structurally related transcription factors. The CRE together with
other sequences that lie within 65 base pairs of the transcription initiation site
is sufficient for specific expression in pancreatic islet cell lines. 63 Islet cell-
specific expression of the somatostatin gene also appears to depend on other
DNA binding proteins in addition to the more ubiquitously expressed CREB. 63
As with gastrin and CCK, progress in defining the cis-elements and trans-
acting factors that determine somatostatin gene expression in gastric and
enteroendocrine D cells is hampered by lack of suitable systems. Isolated
gastric D cells in or short-term cultures of antral mucosa represent promising
systems but suffer from the same limitations as described for gastrin. 60 •61
Lund 105
1-30
II G.OCAOON G.P-1 (1-37) IP-11 I PROGLUCAGON
G.ICENTIN 69
L_G._~_7~--~~~~I ____G._P-1
__ 1 __ ~
GUT
I OXYNTOM:XlU.IN
PANCREAS
FIGURE 3. Schematic of mammalian proglucagon and products of posttranslational processing

in gut and pancreas.
Isolated fetal rat intestinal cells in culture (FRIC cells) are another promising
system which has already been used to demonstrate positive regulation of
somatostatin gene expression by cyclic AMP and phorbol esters. 64 Nonethe-
less, progress in elucidating nutrient effects on gastrointestinal somatostatin
gene expression will require a combination of information derived from these
in vitro systems and in vivo approaches such as transgenic animals.
VI. GLUCAGON/ENTEROGLUCAGON
A. BACKGROUND
The proglucagon gene is a particularly interesting and complex system
with regard to nutrient control of expression. Mammals possess a single
glucagon gene and mRNA, but the encoded precursor, proglucagon, is a
polyprotein precursor that contains three glucagon related peptides as well as
other peptides (Figure 3). 1•65 -68 A common proglucagon precursor is processed
very differently in the pancreas and intestine to result in different biologically
active peptides with different biological activities. 69 - 73 The structure of the
common pancreatic and intestinal proglucagon is shown in Figure 3. Proglu-
cagon contains the 29 amino acid pancreatic glucagon in the midregion with
amino acid extensions at amino and carboxyl termini. A long carboxyl terminal
extension contains two glucagon-like peptides (GLPs) with structural ho-
mology to glucagon, and these have been termed GLP-1 and GLP-11. Spacer
or intervening peptides (IPs) separate glucagon from GLP-1 and GLP-1 from
GLP-11.
In the pancreas, proglucagon is synthesized in pancreatic A cells. 1·74
Products of posttranslational processing of proglucagon in the pancreas are a
29 amino acid pancreatic glucagon, a 30 amino acid peptide comprised of
the amino terminus of proglucagon, and a "Major proglucagon fragment"
(Figure 3). 69- 73 The only product of pancreatic proglucagon processing with
defined biological activity is glucagon. 74 The primary biological role of pan-
creatic glucagon is well established. Glucagon regulates metabolic pathways
in liver and adipose tissue and maintains euglycemia during fasting or nutrient
perturbation. 74 Pancreatic glucagon is secreted into the circulation in response
to fasting, decreases in plasma glucose, and increases in plasma amino acids
and neural influences. 71 ·74 Pancreatic glucagon secretion is suppressed by high
plasma glucose. 71 •74
In the intestine, proglucagon is processed very differently than in the
pancreas. The products of intestinal proglucagon are enteroglucagons (alter-
nately named glucagon-like immunoreactants [GUs]), GLP-I (73-77), GLP-
11, and IP-11 (Figure 3). 69- 73 Enteroglucagons or GUs include glicentin and
oxyntomodulin (Figure 3). Glicentin is composed of the first 69 amino acids
of proglucagon and includes the entire sequence of glucagon in the midregion
flanked by amino acid extensions at both amino and carboxyl termini (Figure
3). Oxyntomodulin consists of glucagon with a hexapeptide extension at the
carboxyl terminus (Figure 3). At physiological concentrations, glicentin and
oxyntomodulin appear not to share the biological actions of pancreatic glu-
cagon but appear to act primarily to inhibit gastric acid secretion and thus
are enterogastrones. 1 •71 •75 Of the other intestinal pro glucagon derived peptides,
only GLP-I has been extensively characterized in terms of biological activity.
GLP-I 7-37 is the primary form of GLP-I in intestine and is a potent stimulant
of insulin secretion and insulin gene expression. 71 •76 •77 The biological roles
of GLP-11 and IP-11 are not defined. Based on observations that plasma
concentrations of enteroglucagons are elevated in most situations of increased
cell proliferation of the intestinal mucosa, the enteroglucagons (glicentin/
oxyntomodulin) or other products of intestinal proglucagon processing are
candidate enterotrophic hormones. 68 •78
Products of intestinal proglucagon-derived peptides are secreted in re-
sponse to different stimuli than pancreatic products. Available evidence sug-
gests that all of the intestinal products of proglucagon are secreted in paral-
lel. 69- 72 Plasma enteroglucagon levels are reduced by fasting and elevated
after food ingestion. 71 Intrajejunal glucose or carbohydrate and fatty acids
are especially strong stimulants of enteroglucagon secretion. 1•79 The secretion
of enteroglucagons in response to luminal glucose but suppression of pan-
creatic glucagon by high plasma glucose indicates that the pancreatic A cell
and intestinal L cell have very different responses to glucose.
To date, the regulation of glucagon gene expression by nutrients has not
been studied extensively. The molecular basis for nutrient control of proglu-
cagon gene expression is likely to be both complex and interesting given the
different secretory responses of two different proglucagon expressing cell
types, the A cell and the L cell, in response to nutritional stimuli.
B. NUTRIENT CONTROL OF GLUCAGON GENE EXPRESSION

IN PANCREAS
In rats after a 3-d fast we find no significant change in pancreatic pro-
glucagon mRNA abundance as assessed by Northern blot hybridization of
Lund 107
A. PROGLUCAGON mANA-PANCREAS AND ILEUM
.
""' • PANCREAS
"" • • INTESTINE
% ..
FED FASTED
B. GLUCAGON ir-PANCREAS AND ILEUM
pM/t
.
""' • PANCREAS
• INTESTINE
FED FASTED
C. ENTEROGLUCAGON RIA - PLASMA
pt.l/g
FED FASTED
FIGURE 4. Proglucagon mRNA and peptide in ad libitum fed and fasted rats. (A) Proglucagon
mRNA in pancreas and small intestine. Proglucagon mRNA was analyzed in poly A+ RNA by
Northern hybridization and abundance assessed as concentration per f.Lg poly A+ RNA. Shown
is the abundance in pancreas or intestine of fasted animals expressed as % of abundance of fed
animals; (B) Glucagon immunoreactivity in pancreas and intestine of fed and fasted rats; (C) Plasma
enteroglucagon immunoreactivity in fed and fasted rats. * = significant difference in fasted
animals vs. fed animals (p < 0.05).
total mRNA extracted from pancreas (Figure 4). In contrast, Chen et a!.,
using in situ hybridization histochemistry, observed a significant increase in
pancreatic proglucagon mRNA in response to a 4-d fast that induced signif-
icant glucopenia. 80 It is possible that the different durations of fasting account
for these different results. However, it also seems possible that the different
results reflect the different methodologies utilized. Our inability to detect a
change in pancreatic proglucagon mRNA by Northern hybridization may be
due to the fact that pancreatic A cells represent a very small proportion of the
total pancreatic mass. Alterations in proglucagon mRNA in such small cell

populations may be masked by changes in other more abundant mRNAs. This
illustrates the potential advantages of a combined approach using both hy-
bridization analyses of extracted mRNA and histological in situ hybridization
analyses. The increase in A cell proglucagon mRNA in response to fasting
indicates that low plasma glucose or neural and hormonal consequences of
fasting result in increases in proglucagon gene transcription or mRNA sta-
bility. 8° Chen eta!. also found that hyperglycemia reduces the abundance of
proglucagon mRNA in pancreatic A cells suggesting negative effects of hy-
perglycemia on pancreatic proglucagon gene transcription or mRNA stabil-
ity. 80 Of interest, however, are other observations that pancreatic proglucagon
mRNA is elevated in rats with streptozotocin-induced, insulin-dependent di-
abetes, despite hyperglycemia. 81 Insulin administration and restoration of
euglycemia rapidly reduces pancreatic proglucagon mRNA in streptozotocin
diabetic rats. 81 Together, these findings suggest that insulin is involved in the
control of pancreatic pro glucagon gene expression by glucose. 81
C. NUTRIENT CONTROL OF GLUCAGON GENE EXPRESSION

IN THE INTESTINE
To date, there is virtually no information on nutrient control of glucagon
gene expression in the intestine. This is surprising given the potential relevance
of intestinal proglucagon-derived peptides in regulating normal, adaptive, or
aberrant growth of intestinal mucosa. 78 In recent studies we observed an
increase in proglucagon mRNA abundance in intestine or rats after a 3-d fast
(Figure 4) even though plasma enteroglucagon is reduced by fasting (Figure
4) and there is no parallel decline in enteroglucagon immunoreactivity in
intestine (Figure 4). It seems likely that the fasting-induced increase in in-
testinal pro glucagon mRNA derives from a reduction in mass of nonendocrine
cells in intestinal mucosa, since fasting induces significant mucosal hypo-
plasia. 4 • 15 We are currently addressing this possibility using in situ hybridi-
zation histochemistry. These difficulties in data interpretation attest to the
need for caution in the interpretation of nutrient-dependent changes in steady-
state GI hormone mRNAs in the intestine as discussed at the beginning of
this chapter.
Effects of specific nutrients on intestinal proglucagon gene expression
have not been reported, but luminal glucose and fats represent likely candi-
dates given their effects on enteroglucagon secretion (Figure 2).
D. NUTRIENT -DEPENDENT AND INDEPENDENT CHANGES IN

INTESTINAL PROGLUCAGON GENE EXPRESSION AFTER
PROXIMAL SMALL BOWEL RESECTION
In recent studies, we and others have focused attention on intestinal
proglucagon gene expression in a model of altered growth of intestinal mu-
cosa, the rat after proximal small bowel resection. In the resection model
Lund 109
there is compensatory hyperplasia of the mucosa of the remnant ileum to

restore digestive and absorptive capacity. 4 ·78 This model is of relevance to
man since resection is used as treatment for morbid obesity, and some studies
indicate that resection is associated with enhanced carcinogenesis in the gut.
Many studies have established that the adaptive growth of ileum after resection
is correlated with elevated plasma levels of enteroglucagon, 4 •78 supporting the
possibility that enteroglucagons or other proglucagon-derived peptides are
enterotrophic hormones. We and Taylor et a!. have established that adaptive
hyperplasia is associated with increases in intestinal proglucagon mRNA. 68 •82 •83
Such increases do not appear to reflect increased L cell number and therefore
likely reflect an increase in intestinal proglucagon gene transcription or mRNA
stability. 84 It is well established that resection-induced adaptive growth after
resection requires luminal nutrient, 4 ·5 as does the resection-induced increases
in plasma enteroglucagon. 85 Thus, we analyzed the role of luminal nutrient
in resection-induced changes in intestinal proglucagon mRNA. Of interest
are our observations that there are nutrient-independent increases in intestinal
proglucagon mRNA within hours of resection and preceding the adaptive
growth response. 82 Such observations are consistent with one or more intes-
tinal proglucagon derived peptides as initiators of or early markers of adap-
tation. However, sustained increases in intestinal proglucagon mRNA after
resection require luminal nutrient, suggesting an interplay between nutrient-
independent and nutrient-dependent factors in regulating intestinal proglu-
cagon gene expression and adaptive growth. 82 •83 In follow-up studies, we are
investigating the role of specific nutrients in resection-induced changes in
expression of the proglucagon gene expression and other intestinal hormones
and growth factors. 81 •88 Such studies should have relevance for understanding
the role of nutrient effects on GI hormone gene expression in regulating normal
and aberrant growth of intestinal mucosa.
E. MODEL SYSTEMS TO ANALYZE NUTRIENT CONTROL OF

PROGLUCAGON GENE EXPRESSION
A number of useful model systems are available for analyses of nutrient
control of pancreatic glucagon gene expression. The in vitro perfused pancreas
represents one useful model. 86 Pancreatic islets can be isolated from exocrine
pancreas and maintained in culture for analyses of islet hormone biosyn-
thesis. 70 These systems have been used to analyze insulin gene expression
and represent promising systems for future studies of nutrient control of
glucagon gene expression. They offer the advantage that many of the paracrine
mechanisms that may mediate nutrient effects remain intact. A number of
established insulinomalislet cell lines express the proglucagon gene and have
been used to analyze the cis-elements that determine islet-specific expression
of the proglucagon gene. 87 . 89 Using such systems three DNA elements in the
5' flanking region of the glucagon gene have been shown to underlie islet-
specific expression. 89 Using these islet cell lines, it also has been established
that pancreatic proglucagon gene expression is upregulated by protein kinase

C-dependent pathways and not by cyclic AMP-dependent pathways, even
though the glucagon 5' flanking region contains a cyclic AMP response
element. 87 Such findings raise the possibility that nutrient control of pancreatic
proglucagon gene expression is mediated by effects on protein kinase C. This
possibility can be tested in transgenic animals with the kinase C responsive
cis-elements linked to a reporter gene. Progress has been made towards de-
velopment of such models since transgenic mice have been developed that
show pancreas specific expression of a reporter gene driven by 900 base pairs
of the proglucagon 5' flanking region. 90
As with other GI hormones, the availability of model systems to analyze
nutrient control of intestinal proglucagon expression is more limited. Drucker
and Brubaker have used fetal rat intestinal cells in culture (FRIC cells) as a
model system. 90 Using this system they have demonstrated that control of
proglucagon gene expression differs in fetal intestine and islet cells. In FRIC
cells, protein kinase C-dependent pathways stimulate enteroglucagon secretion
but not gene expression. 91 In contrast to islet cells, proglucagon gene expres-
sion is upregulated by cyclic AMP-dependent pathways in FRIC cells. 90 These
findings suggest that nutrient effects on intestinal pro glucagon gene expression
may involve cyclic AMP-dependent pathways. Furthermore, differences in
pancreatic and intestinal proglucagon gene expression may be mediated by
protein kinase C and cyclic AMP-dependent pathways, respectively. Addi-
tional studies in the FRIC cell system and studies of transgenic mice with the
proglucagon cis-elements linked to a reporter gene should help elucidate
mechanisms that underlie the nutrient control of intestinal proglucagon gene
expression. It is noteworthy, however, that 900 base pairs of proglucagon 5'
flanking region are not sufficient to drive expression of a reporter gene in
intestinal L cells of transgenic mice, even though this construct was active
in A cells and in brain. 90 Thus, additional studies are required to define the
cell-specific elements that underlie proglucagon gene expression in L cells
before nutrient regulation can be addressed in transgenic mice.
VII. CONCLUSIONS AND FUTURE DIRECTIONS

In the past few years, progress has been made towards understanding
nutrient control of expression of the genes encoding a number ofGI hormones.
Nonetheless it is probably clear that progress has been hampered by the
complexity of the GI endocrine system and the paucity of useful enteroen-
docrine cell lines. The utility of the transgenic approach has been alluded to
above. With the development of some transgenic animals, and the increased
use of these models, it seems likely that substantial and rapid progress will
be forthcoming. In addition, potentially useful model enteroendocrine cell
lines have been developed recently. 91 STC-1 cell lines derive from enteroen-
docrine tumors that spontaneously developed in mice generated by cross-
Lund 111
breeding of two transgenic lines. 91 The transgenic animals used for cross-
breeding were ones with an insulin promoter used to drive the SV40 large T
antigen gene and ones with the insulin promotor linked to mouse polyoma
small T antigen. The SV40 transgenics develop pancreatic insulinoma but not
enteroendocrine tumors. 91 The polyoma transgenics do not develop tumors in
pancreas or gut. 91 On crossbreeding, however, a large proportion of animals
develop enteroendocrine tumors that express a spectrum of GI hormones.
STC-1 cell lines derived from the enteroendocrine tumors express a wide
range of GI hormone. The molecular basis for the development of enteroen-
docrine tumors in these crossbred mice is of itself very interesting since
enteroendocrine tumors are relatively rare. For those working in the field of
GI hormone expression, the cell lines derived from these tumors may represent
a very important and unique tool for future studies.
ACKNOWLEDGMENTS
This work was supported by NIH grant DK 40247. Thanks are expressed
to Dr. Rodger Liddle for extremely useful input and discussion. Blanche Jones
provided invaluable editorial and secretarial assistance.
REFERENCES
I. Walsh, J, H., Gastrointestinal hormones, in Physiology of the Gastrointestinal Tract,
Johnson, L. R., Ed., Raven Press, New York, 1987, 181.
2. Brown, J, C., Mcintosh, C. H. S., and Pederson, R. A., The gastrointestinal peptides
and nutrition, Can. J. Physiol. Pharmacal., 61, 282, 1982.
3. Yamada, T., Gut hormone release induced by food ingestion, Am. J. Clin. Nutr., 42,
1033, 1985. (DB 44)
4. Dowling, R. H., Small bowel adaptation and its regulation, Scand. J. Gastroenterol.,
74, 53, 1982.
5. Lipscomb, H. L. and Sharp, J, G., Effects of reduced food intake on morphometry
and cell production in the small intestine of the rat, Virchows Arch., 41, 285, 1982.
6. Fujimoto, S., Kimoto, K., Yamashita, S., Kawai, K., and Hattori, T., and Fugita,
S., Tritiated thymidine autoradiographic study on origin and renewal of gastrin cells in
antral area of hamsters, Gastroenterology, 79, 785, 1985.
7. Mitchell, P. J. and Tjian, R., Transcriptional regulation in mammalian cells by sequence-
specific DNA binding proteins, Science, 245, 371, 1989.
8. Raghowr, R., Regulation of messenger RNA turnover in eukaryotes, Trends Biochem.
Sci., 12, 358, 1987.
9. Lichtenberger, L. M., Lechago, J,, and Johnson, L. R., Depression of antral and
serum gastrin concentration by food deprivation in the rat, Gastroenterology, 68, 1473,
1975.
10. Schwarting, H., Koop, H., Gellert, G., and Arnold, R., Effect of starvation on
endocrine cells in the rat stomach, Regul. Pep!., 14, 33, 1986.
II. Taylor, I. L., Byrne, W. J., Christie, D. L., Ament, M. E., and Walsh, J. H., Effect
of individual L-amino acids on gastric acid secretion and serum gastrin and pancreatic
polypeptide release in humans, Gastroenterology, 83, 272, 1982.
12. Lichtenberger, L. M., Importance of food in the regulation of gastrin release and
formation, Am. J. Physiol., 243, G429, 1982.
13. Wu, V., Sumii, K., Tari, A., Sumii, M., and Walsh, J. H., Regulation of rat antral
gastrin and somatostatin gene expression during starvation and after refeeding, Gastro-
enterology, 101, 1552, 1991.
14. Wu, S. V., Sumii, K., Tari, A., Mogard, M., and Walsh, J. H., Regulation of gastric
somatostatin gene expression, Metabolism, 39, 125, 1990. (DB41)
15. Brand, S. J. and Stone, D., Reciprocal regulation of antral gastrin and somatostatin
gene expression by omeprazole-induced achlorhydria, J. Clin. Invest., 82, 1059, 1988.
16. Schusdziarra, V., Harris, V., Conlon, J. M., Arimura, A., and Unger, R., Pancreatic
and gastric somatostatin release in response to intragastric and intraduodenal nutrients
and HCI in the dog, J. Clin. Invest., 62, 509, 1978.
17. Bloom, S. R., Mortimer, C. H., and Thorner, M. 0., Inhibition of gastrin and gastric
acid secretion by growth hormone release inhibitory hormone, Lancet, I, 1106, 1974.
18. Harty, R. F., Maico, D. G., and McGuigan, J. E., Somatostatin inhibition of basal
and carbachol stimulated gastrin release in rat antral organ culture, Gastroenterology,
81' 707, 1981.
19. Karnik, P. S., Monahan, S. J., and Wolfe, M. M., Inhibition of gastrin gene expression
by somatostatin, J. Clin. Invest., 83, 367, 1989.
20. Karnik, P. S. and Wolfe, M. M., Somatostatin stimulates gastrin mRNA turnover in
dog antral mucosa, J. Bioi. Chern., 265, 2550, 1990.
21. Wu, S. V., Sumii, K., and Walsh, J. H., Studies of regulation of gastrin synthesis and
post-translational processing by molecular biology approaches, Ann. N.Y. Acad. Sci.,
597, 17, 1990. (DB 18)
22. Lichtenberger, L. M., Forssman, W. G., and Ito, S., Functional responsiveness of
an isolated and enriched fraction of rodent gastrin cells, Gastroenterology, 79, 447, 1980.
23. Theill, L. E., Wiborg, 0., and Vuust, J., Cell-specific expression of the human gastrin
gene: evidence for a control element located downstream of the TATA box, Mol. Cell.
Bioi., 7, 4329, 1987.
24. Brand, S. J. and Wang, T. C., Gastrin gene expression and regulation in rat islet cell
lines, J. Bioi. Chern., 263, 16597, 1988. (DB 106)
25. Wang, T. C. and Brand, S. J., Islet cell-specific regulatory domain in the gastrin
promoter contains adjacent positive and negative DNA elements, J. Bioi. Chern., 265,
8908, 1990. (DB63)
26. Merchant, J. L., Demediuk, B., and Brand, S. J., A GC-rich element confers epi-
dermal growth factor responsiveness to transcription from the gastrin promoter, Mol.
Cell. Bioi., II, 2686, 1991. (DB61)
27. Godley, J. M. and Brand, S. J., Regulation of the gastrin promoter by epidermal growth
factor and neuropeptides, Proc. Nat/. Acad. Sci. U.S.A., 86, 3036, 1989. (DB62)
28. Takahashi, Y., Fukushige, S., Murotsu, T., and Matsubara, K., Structure of human
cholecystokinin gene and its chromosomal location, Gene, 50, 353, 1986.
29. Deschenes, R. J., Lorenz, L. J., Haun, R. S., Roos, B. A., Collier, K. J., and
Dixon, J., Cloning and sequence analysis of a eDNA encoding rat preprocholecystokinin,
Proc. Nat!. Acad. Sci. U.S.A., 81, 726, 1984.
30. Gubler, U., Chua, A. 0., Hoffman, B. J., and Collier, K. J., Cloned eDNA to
cholecystokinin mRNA predicts an identical preprocholecystokinin in pig brain and gut,
Lund 113
31. Ivy, A. C. and Oldberg, E., A hom10ne mechanism for gallbladder contraction and
evacuation, Am. J. Physiol., 85, 381, 1928.
32. Harper, A. A. and Raper, H. S., Pancreozymin, a stimulant of secretion of pancreatic
enzymes in extracts of small intestine, J. Physiol., 102, 115, 1943.
33. Niederau, C., Luthen, R., Niederau, M., Strohmeyer, G., Ferrell, L. D., and Grendell,
J, H., Effects of long term CCK stimulation and CCK blockade on pancreatic and
intestinal growth, morphology and function, Digestion, 46, 217, 1990.
34. Liddle, R. A., Goldfine, I. D., Rosen, M. S., Taplitz, R. A., and Williams, J. A.,
Cholecystokinin bioactivity in human plasma, J. Clin. Invest., 75, 1144, 1985.
35. Konturek, S. J., Radecki, T., Thor, P., and Dembinski, A., Release of cholecystokinin
by amino acids, Proc. Soc. Exp. Bioi. Med., 143, 305, 1973.
36. Meyer, J, H. and Jones, R. S., Canine pancreatic responses to intestinally perfused fat
and products of fat digestion, Am. J. Physiol., 226, 1178, 1974.
37. Liddle, R. A., Green, G. M., Conrad, C. K., and Williams, J. A., Proteins but not
amino acids, carbohydrates, or fats stimulate cholecystokinin secretion in the rat, Am.
J. Physio/., 251, G243, 1986.
38. Liddle, R. A., Goldfine, I. D., and Williams, J. A., Bioassay of plasma cholecystokinin
in rats: effects of food, trypsin inhibitor, and alcohol, Gastroenterology, 87, 542, 1984.
39. Iwai, K., Fushiki, T., and Fukuoka, S.-1., Pancreatic enzyme secretion mediated by
a novel peptide: monitor peptide hypothesis, Pancreas, 3, 720, 1988.
40. Iwai, K., Fukuoka, S.-1., Fushiki, T., Tsujikawa, M., Hirose, M., Tsunasawa, S.,
and Sakiyama, F., Purification and sequencing of a trypsin-sensitive cholecystokinin
releasing peptide from rat pancreatic juice. Its homology with pancreatic secretory trypsin
inhibitor, J. Bioi. Chern., 262, 8956, 1987.
41. Fukuoka, S. and Scheele, G. A., Rapid and selective cloning of monitor peptide, a
novel cholecystokinin-releasing peptide, using minimal amino acid sequence and the
polymerase chain reaction, Pancreas, 5, I, 1990.
42. Greenstein, R. J., Isola, L., and Gordon, J., Differential cholecystokinin gene expres-
sion in brain and gut of the fasted rat, Am. J. Med. Sci., 299, 32, 1990.
43. Kanayama, S. and Liddle, R. A., Influence of food deprivation on intestinal chole-
cystokinin and somatostatin, Gastroenterology, 100, 909, 1991. (DB58)
44. Liddle, R. A., Carter, J.D., and McDonald, A. R., Dietary regulation of rat intestinal
cholecystokinin gene expression, J. Clin. Invest., 81, 2015, 1988. (DB40)
45. Kanayama, S. and Liddle, R. A., Regulation of intestinal cholecystokinin and so-
matostatin mRNA by bombesin in rats, Am. J. Physiol., 261, G71, 1991.
46. Kanayama, S. and Liddle, R. A., Somatostatin regulates duodenal cholecystokinin and
somatostatin messenger RNA, Am. J. Physiol., 258, 1990, G358. (DB22)
47. Haun, R. S. and Dixon, J, E., A transcriptional enhancer essential for the expression
of the rat cholecystokinin gene contains a sequence identical to the - 296 element of the
human c-fos gene, J. Bioi. Chern., 265, 15455, 1990.
48. Monstein, H.-J. and Folkesson, R., Phorbo\12-myristate-13-acetate (PMA) stimulates
a differential expression of cholecystokinin (CCK) and c-fos mRNA in a human neuro-
blastoma cell line, FEBS, 293, 145, 1991.
49. Brazeau, P., Vale, W., Burgos, R., Ling, N., Butcher, M., Riveir, J., and Guillemin,
R., Hypothalamic polypeptide that inhibits the secretion of immunoreactive pituitary
growth hormone, Science, 179, 77, 1973.
50. Newman, J. B., Lluis, F., and Townsend, C. M, Jr., Somatostatin, in Gastrointestinal
Endocrinology, Thompson, J. C., Greeley, G. H., Jr., Rayford, P. L., and Townsend,
C. M., Jr., Eds., McGraw-Hill, Toronto, 1987,286.
51. Patel, Y. C. and Reichlin, S., Somatostatin in hypothalamus, extrahypothalamic brain,
and peripheral tissues of the rat, Endocrinology, 102, 523, 1978.
52. Pradayrol, L., Jornvall, H., Mutt, V., and Ribet, A., N-terminally extended somato-
statin: the primary structure of somatostatin-28, FEBS, 109, 55, 1980.
53. Trent, D. F. and Weir, G. C., Heterogeneity of somatostatin-like peptides in rat brain,
pancreas, and gastrointestinal tract, Endocrinology, 108, 2033, 1981.
54. Shen, L.-P. and Rutter, W., Human somatostatin I: sequence of the eDNA, Proc. Nat!.
Acad. Sci. U.S.A., 224, 168, 1984.
55. Funckes, C. L., Minth, C. D., Deschenes, R., Magazin, M., Tavianini, M. A.,
Sheets, M., Collier, K., Weith, H. L., Aron, D. C., Roos, B. A., and Dixon, J. E.,
Cloning and characterization of a mRNA-encoding rat preprosomatostatin, J. Bioi. Chern.,
258, 8781, 1983.
56. Goodman, R. H., Aron, D. C., and Roos, B. A., Rat pre-prosomatostatin: structure
and processing by microsomal membranes, J. Bioi. Chern., 258, 5570, 1983.
57. Hofler, H., Childers, H., Montminy, M. R., Goodman, R. H., Leehan, R. M.,
DeLellis, R. A., Tischler, A. S., and Wolfe, H. J., Localization of somatostatin mRNA
in the gut, pancreas and thyroid gland of the rat using antisense RNA probes for in situ
hybridization, Acta Histochem., 34, 101 (Suppl.), 1987.
58. Vinik, A. I., Gaginella, T. S., O'Dorisio, T. M., Shapiro, B., and Wagner, L., The
distribution and characterization of somatostatin-like immunoreactivity in epithelial cells,
submucosa, and muscle of the rat stomach and intestine, Endocrinology, I 09, 1921, 1981.
59. Larrson, L.-1., Goltermann, N., De Magistris, L., Rehfeld, J. F., and Schwartz,
T. W., Somatostatin cell processes as pathways for paracrine secretion, Science, 205,
1393, 1979.
60. Chiba, T., Park, J., and Yamada, T., Biosynthesis of somatostatin in canine fundic
D cells, J. Clin. Invest., 81, 282, 1988.
61. Park, J., Chiba, T., Yokotani, K., DelValle, J., and Yamada, T., Somatostatin
receptors on canine fundic D-cells: evidence for autocrine regulation of gastric somato-
statin, Am. J. Physiol., 257, 0235, 1989.
62. Montminy, M. R., Sevarino, K. A., Wagner, J. A., Mandel, G., and Goodman,
R. H., Identification of a cyclic-AMP-responsive element within the rat somatostatin
gene, Proc. Nat/. Acad. Sci. U.S.A., 83, 6682, 1986.
63. Powers, A. C., Tedeschi, F., Wright, K. E., Chan, J. S., and Habener, J. F.,
Somatostatin gene expression in pancreatic islet cells is directed by cell-specific DNA
control elements and DNA-binding proteins, J. Bioi. Chern., 264, 10048, 1989.
64. Brubaker, P. L., Drucker, D. J., and Greenberg, G. R., Synthesis and secretion of
somatostatin-28 and -14 by fetal rat intestinal cells in culture, Am. J. Physiol .. 258,
0974, 1990.
65. Habener, J. F., Goodman, R. H., and Lund, P. K., Complementary DNAs encoding
precursors of glucagon and somatostatin, in Biogenetics of Neurohormonal Peptides,
Hakanson, R. and Thorell, J., Eds .. Academic Press, San Diego, 1985, 47.
66. Drucker, D. J. and Brubaker, P. L., Glucagon biosynthesis in fetal rat intestine,
Biomed. Res., 9 (Suppl. 3), 29, 1988.
67. Novak, U., Wilks, A., Buell, G., and McEwen, S., Identical mRNA for preproglucagon
in pancreas and gut, Eur. J. Biochem., 164, 553, 1987.
68. Hynes, M. A., Shiow-Lian, J., Ulshen, M. H., Simmons, J. G., and Lund, P. K.,
Characterization, localization and regulation of extra pancreatic proglucagon mRNAs,
Biomed. Res., 9 (Suppl. 3), 147, 1988.
69. Moody, A. J., Holst, J., Thim, L., and Jensen, S., Relationship of glicentin to
proglucagon and glucagon in the porcine pancreas, Nature (London), 289, 514, 1981.
70. Patzelt, C. and Chiltz, E., Conversion of proglucagon in pancreatic alpha cells: the
major endproducts are glucagon and a single peptide, the major proglucagon fragment,
that contains two glucagon-like sequences, Proc. Nat!. Acad. Sci. U.S.A., 81, 5007,
1984.
71. Holst, J. J., Orskov, C., Olsen, J., Buhl, T., Schjoldager, B., and Kofod, H.,
Secretion and effects of the naturally occurring products of proglucagon, Biomed. Res.,
9 (Suppl. 3), 181, 1988.
Lund 115
72. Mojsov, S., Heinrich, G., Wilson, I. B., Ravazzola, M., Orci, L., and Habener,
J. F., Preproglucagon gene expression in pancreas and intestine diversifies at the level
of post-translational processing, J. Bioi. Chern., 261, 11880, 1986.
73. Orskov, C., Holst, J., Knuhtsen, S., Baldissera, F. G. A., Poulsen, S. S., and Nielsen,
0. V., Glucagon-like peptides GLP-1 and GLP-2, predicted products of the glucagon
gene are secreted separately from pig small intestine but not pancreas, Endocrinology,
119, 1467, 1986.
74. Unger, R. and Oric, L., Physiology and pathophysiology of glucagon, Physiol. Rev.,
56, 778, 1976.
75. Jarrousse, C., Audousset-Puech, M.-P., Dubrasquet, M., Niel, H., Martinez, J.,
and Bataille, D., Oxyntomodulin (glucagon-37) and its C-terminal octapeptide inhibit
gastric acid secretion,. FEES Lett., 188, 81, 1985.
76. Holst, J. J., Orskov, C., Vagn Hielsen, 0. V., and Schwartz, T. W., Truncated
glucagon-like peptide I, an insulin-releasing hormone from the distal gut, FEES Lett.,
211, 169, 1987.
77. Drucker, D. J., Philippe, J., Mojsov, S., Chick, W. L., and Habener, J. F., Glucagon-
like peptide I stimulates insulin gene expression and increases cyclic AMP levels in a rat
islet cell line, Proc. Nat/. Acad. Sci. U.S.A., 84, 3434, 1987.
78. Bloom, S. R. and Polak, J. M., The hormonal pattern of intestinal adaptation: a major
role for enteroglucagon, Scand. J. Gastroenterol., 74 (Suppl.), 93, 1982.
79. Holst, J. J., Christiansen, J., and Kohl, C., The enteroglucagon response to intrajejunal
infusion of glucose, triglycerides, and sodium chloride, and its relation to jejunal inhibition
of gastric acid secretion in man, Scand. J. Gastroenterol., II, 297, 1976.
80. Chen, L., Komiya, I., Inman, L., O'Neill, J., Appel, M., Alam, T., and Unger,
R. H., Effects of hypoglycemia and prolonged fasting on insulin and glucagon gene
expression, J. Clin. Invest., 84, 711, 1989.
81. Chen, L., Komiya, I., Inman, L., McCorkle, K., Alam, T., and Unger, R. H.,
Molecular and cellular responses of islets during perturbations of glucose homeostasis
determined by in situ hybridization histochemistry, Proc. Nat/. Acad. Sci. U.S.A., 86,
1367, 1989.
82. Rountree, D. B., Ulshen, M. H., Selub, S., Fuller, C. R., Bloom, S. R., Ghatei,
M. A., and Lund, P. K., Rapid, nutrient independent increases in proglucagon and
ODC mRNAs after jejuno-ileal resection, Gastroenterology, in press.
83. Taylor, R. G., Verity, K., and Fuller, P. J., Ileal glucagon gene expression: ontogeny
and response to massive small bowel resection, Gastroenterology, 99, 724, 1990.
84. Buchan, A.M. J., Griffiths, C. J., Morris, J. F., and Polak, J. M., Enteroglucagon
cell hyperfunction in rat small intestine after gut resection, Gastroenterology, 88, 8, 1985.
85. Sagor, G. R., Al-Mukntar, M. Y. T., Ghatei, M. A., Wright, N. A., and Bloom,
S. R., The effect of altered luminal nutrition on cellular proliferation and plasma con-
centrations of enteroglucagon and gastrin after small bowel resection in rat, Br. J. Surg.,
69, 14, 1982.
86. Grodsky, G. M. and Fanska, R. E., The in vitro perfused pancreas, Methods Enzyrnol.,
39, 364, 1975.
87. Philippe, J., Drucker, D. J., Chick, W. L., and Habener, J. F., Transcriptional
regulation of genes encoding insulin, glucagon, and angiotensinogen by sodium butyrate
in a rat islet cell line, Mol. Cell Bioi., 7, 560, 1987.
88. Philippe, J., Drucker, D. J., and Habener, J. F., Glucagon gene transcription in an
islet cell line is regulated via a protein kinase C-activated pathway, J. Bioi. Chern., 262,
1823, 1987.
89. Philippe, J., Drucker, D. J., Knepel, W., Jepeal, L., Misulovin, Z., and Habener,
J. F., Alpha-cell-specific expression of the glucagon gene is conferred to the glucagon
promoter element by the interactions of DNA-binding proteins, Mol. Cell. Bioi., 8,
4877, 1988.
90. Efrat, S., Teitelman, G., Anwar, M., Ruggiero, D., and Hanahan, D., Glucagon
gene regulatory region directs oncoprotein expression to neurons and pancreatic A cells,
Neuron, I, 605, 1988.
91. Drucker, D. J. and Brubaker, P. J., Proglucagon gene expression is regulated by a
cyclic AMP-dependent pathway in rat intestine, Proc. Nat!. Acad. Sci. U.S.A., 86, 3953,
1989.
92. Rindi, G., Grant, S. G. N., Yiangou, Y., Ghatei, M. A., Bloom, S. R., Bautch,
V. L., Solcia, E, and Polak, J, M., Development of neuroendocrine tumors in the
gastrointestinal tract of transgenic mice, Am. J. Pathol., 136, 1349, 1990.
Chapter 5
CALCIUM HOMEOSTASIS, ENDOPLASMIC

RETICULAR FUNCTION, AND THE REGULATION OF
mRNA TRANSLATION IN MAMMALIAN CELLS
Margaret A. Brostrom and Charles 0. Brostrom
TABLE OF CONTENTS
I. Overview ........................................................ 118
II. The Central Role of the Endoplasmic Reticulum in Ca2 +

Homeostasis and Protein Synthesis .............................. 118
III. Regulation of Translation by Ca 2 + ...•.......•........•.......... 121

A. Effect of Ca 2 + on Translational Initiation ................ 121
B. Requirement for Sequestered Ca 2 + ...................•.•• 123
C. Effect of Ca 2 + on the Phosphorylation/
Dephosphorylation of Initiation Factors .................. 127
D. Conditions That Dissociate Initiation from
Regulation by Ca 2 + . . . • . . . . . . . . . • . . . . . . . . . . . . . . . . . . . . • . . . 127
E. Effect of Ca 2 + on Peptide Chain Elongation ............. 128
IV. Ca 2 + and the Induction of GRP78/BiP ........................... 131
V. A Ca 2 + Requirement for Secretory Protein Processing

and Export ....................................................... 133
VI. Summary ........................................................ 135
References ............................................................... 135
0·8493-6961-4/93/$0 ()()+$.50
© 1993 by CRC Pre», Inc. 117
I. OVERVIEW
Cells have an extraordinary ability to impose order upon complex anays

of substituents and reactions. It is fascinating to reflect upon potential mech-
anisms whereby intracellular processes are coordinated and prioritized in the
maintenance of this organization. Ordered relationships must exist among
housekeeping processes, such as energy production, protein synthesis, and
membrane transport systems, that maintain the health of the cell, and cell-
specific functions involving stimulus-response coupling. For example, pro-
vision of the conect types and amounts of cellular enzymes and structural
proteins would appear, a priori, to mandate that protein synthesis be rigorously
coupled to and subordinated to the overall nutritional and functional status of
the cell. It is inconceivable that this energy-intensive process would smoothly
service the cell unless it were interdigitated with other cellular processes
through coordinating regulation. Such regulation would presumably involve
not only longer term adaptive alterations in the relative amounts of each protein
to be synthesized requiring hours to days but also more immediate adjustments
in the overall rate of peptide chain assembly. This brief review is concerned
with highlighting a developing literature relating the control of protein syn-
thesis at translation to the Ca2 + signal transduction system basic to the reg-
ulation of a wide variety of specialized processes. It is designed to update
and extend an earlier review pertaining to Ca2 + -dependent regulation of pro-
tein synthesis in intact mammalian cells. 1
II. THE CENTRAL ROLE OF THE ENDOPLASMIC

RETICULUM IN Ca2 + HOMEOSTASIS AND
PROTEIN SYNTHESIS
The endoplasmic reticulum (ER), as commonly described, is comprised
of a convoluted, bilayer membrane sheathing a continuous luminal or cisternal
space that occupies up to 10% or more of the total cell volume. Conventionally
the ER is viewed as consisting primarily of the "rough" or ribosomally
decorated region and the "smooth" or ribosomal-free region. Morphologi-
cally, the rough ER penetrates to the deepest recesses of the cell in being
continuous with the outer membrane of the nucleus. The rough ER tends to
occur in plate-life folds extending throughout much of the cytoplasm and
possessing an intraluminal space of 20 to 30 nm. At various loci these folds
taper into elements of the smooth ER, a dynamic network of anastomosing
tubules 30 to 60 nm in diameter. 2 In addition to these traditional divisions,
various subregions of the ER have been described possessing cisternae as-
sociated with cytoplasmic glycogen, mitochondria, cytoskeletal components,
and the plasmalemma. Vesicles that transport protein for processing to the
Golgi are derived from and represent another ER compartment. These gen-
eralities, however, fail to communicate a sense of the remarkable diversity
Brostrom and Brostrom 119
in ER content and structure existing from one cell type to another. While
both the rough and smooth ER possess considerable overlap in their protein
content, a number of proteins concerned with ribosomal docking and protein
processing are enriched in the roughER. 3.4 Prominent among lumenal resident
proteins, or reticuloplasmins, are endoplasmin or GRP94, GRP78 or BiP, and
protein disulfide isomerase. An oxidizing environment is thought to exist
within the lumen supporting protein disulfide bond formation and protein
hydroxylation. Functions commonly associated with the ER include early
protein processing, phospholipid biosynthesis, oxidative metabolism of hy-
drophobic molecules including many drugs, and storage of Ca2 + releasable
in response to extracellular stimuli. It is apparent, therefore, that the ER is
a highly complex organelle that infiltrates the cytoplasmic space of the cell
and interfaces with and produces much of the structural material comprising
other organelles. Overall, the ER possesses the structural and functional prop-
t:rties that would be expected for an organelle functioning in the integration
and coordination of major cellular processes.
A wealth of literature supports the central role of Ca2 + in intracellular
signaling related to stimulus-response coupling. 4 - 12 Prominent processes reg-
ulated by the cation include secretion, membrane transport and permeability,
glycogen metabolism, and muscle contraction. As intracellular-free Ca 2 + in
response to a stimulus rises severalfold from resting values near 0.1 f.LM, the
cation binds to Ca2 + receptor proteins, such as calmodulin and troponin C,
which then become capable of activating various enzymatic processes_ Con-
tributions of Ca 2 + to the free pool are derived from the extracellular fluid
and/or from intracellular sites of storage; release of cation from either source
is driven by concentration gradients on the order of 104 • Ca 2 + entry across
the plasmalemma involves voltage and/or ligand gated Ca2 + channels; 6 •8 Ca 2 +
efflux is generaHy achieved by a Na+ /Ca 2 + anti port and an active Ca2 + pump.
Despite much investigative effort, controversy continues regarding the relative
contribution of intracellular sequestered Ca 2 + from one cell type to another
in Ca2 + signaling, the localization and characteristics of Ca2 + stores, and
their function in maintaining Ca2 + homeostasis. s- 11 At least two stores of Ca2 +
are frequently proposed to exist, one of which is sensitive to release by inositol
I ,4,5-trisphosphate (IP 3 ) generated in response to hormonal action and another
which may support spontaneous oscillations in intracellular free Ca2 +. 7 - 12
With the discrediting of the mitochondrion as a prominent storage site for
Ca2 + in normal cells5 and the increasing appreciation of IP3 as a second
messenger triggering release of intracellular sequestered Ca2 + from the mi-
crosomal fraction, 7 the ER gained acceptance as a Ca2 + storing organelle.
Unfortunately the ER is sufficiently heterogeneous in terms of its biochemical,
functional, and morphological properties, that subsequent characterization of
its role in Ca2 + homeostasis proved to be elusive and controversial. 9 - 11 Cellular
contents of ER and IP3 -releasable Ca2 + varied widely from one cell type to
another. Discrepancies appeared to exist in the distribution of various enzyme

markers for the ER and Ca2 + uptake, IP 3 binding and IP 3 -induced Ca2 + release
for some tissues, and the distribution of various membrane-associated Ca 2 +
binding proteins. Ambiguities of this nature ultimately led to the controversial
proposal that IP 3 -sensitive Ca 2 + storage occurred in a separate organelle,
termed the "calciosome". 13 After much additional study and considerable
debate, the earlier conclusion that elements of the ER support Ca 2 + storage
subject to release by IP 3 has been largely reaffirmed and strengthened. 9 - 11 • 14
Oxalate augments Ca2 + uptake into the IP 3 sensitive pool, a property of the
sarcoplasmic reticulum (SR) that would also be expected of the ER. Electron
probe microanalysis of saponin permeabilized, oxalate loaded liver, and pho-
toreceptor cells have provided strong evidence for an IP3 -sensitive Ca 2 + stor-
age site associated with the ER. 9 •14 The structure of the ER and its ability to
store and release Ca2 + have been particularly well characterized in the ventral
nerve photoreceptor of Limulus. 9 Accumulation of Ca 2 + by the ER occurs in
both the photosensitive rhabdomeral lobe and the insensitive arhabdomeral
lobe. Only the rhabdomeral lobe releases Ca2 + in response to the injection
of IP 3 • Thapsigargin, an inhibitor of active accumulation of Ca2 +, that acts
preferentially, if not specifically, on the ER influences an IP 3 -sensitive pool
in saponin permeabilized cells. 15 Antibodies against purified IP 3 receptor have
revealed immunoreactivity in Purkinje cells associated with both the rough
and smooth ER as well as with the cis cisternae of the Golgi and subplas-
malemmal cisternae but not the plasmalemma or mitochondria. More quan-
titative analysis of gold-immunolabeled ultrathin cryosections revealed that
the IP3 receptor was much more concentrated on smooth cisternae than on
the perinuclear membrane or rough cisternae. 16 As described later, sequestered
Ca2 + is required for early protein processing, which is established to occur
in the ER. 3 .4 Depletion of cellular Ca2 + accelerates protein degradation in the
ER 17 and induces the synthesis of GRP78, an ER resident protein. 18 Calre-
ticulin, a high capacity, low affinity Ca2 + binding protein, is reported to exist
in both the SR of muscle cells and the ER of a number of cell types. 19
The emerging picture is that the ER has the ability to accumulate Ca2 +
throughout most of its structure including the outer nuclear membrane. In
contrast, the IP 3 receptor may localize in particularly high densities at cisternae
of the smooth ER, especially those in close approximation to the plasma-
lemma. The distribution of calreticulin appears to largely parallel that of the
IP 3 receptor. Assuming that Ca2 + has the ability to migrate once inside the
ER, it would naturally accumulate in regions containing the highest concen-
trations of Ca 2 + binding sites. Should such migration prevail, the ER would
function to recover Ca 2 + from the deeper areas of the cell for subsequent
transport to cisternae located near the cell surface. While this function at this
point is entirely hypothetical, Ca 2 + uptake by the SR, a closely related struc-
ture, reportedly is separated from the Ca2 + release sites of the junctional
region of the terminal cisternae. A number of extremely interesting questions

remain unresolved at this point. Are the interior components of the ER or-
ganized into a matrix that is influenced by Ca 2 +? What is the role of Ca2 +
in early protein processing? If the calciosome exists in isolatable form, is it
a distinct organeiJe or a specialized region of the ER that may detach within
some celJs but not others?
III. REGULATION OF TRANSLATION BY Ca2+
A. EFFECT OF Ca 2 + ON TRANSLATIONAL INITIATION

Although the primary regulation of specific gene expression in mam-
malian ceiJs occurs at the transcriptional step of protein synthesis, mRNA
translation is currently thought to represent a second important site of control.
The translational process has traditionaiJy been divided into the stages of
initiation, elongation, and termination, each of which involves complex in-
teractions of proteins and RNA. Cytoplasmic proteins are often perceived,
perhaps questionably, as being synthesized on "free" ribosomes, whereas
proteins destined for insertion into membranes or for secretion are synthesized
on membrane-bound ribosomes and inserted into the ER. Insertion may occur
cotranslationaiiy or foiiowing completion of polypeptide. Translational control
in mammalian ceiis is viewed as occurring largely, although not exclusively,
at the level of initiation. zo-n Translational initiation is thought to be influenced
by the intrinsic translatability of a given mRNA, by the ease of movement
of mRNAs from mRNPs into polysomes, and by the phosphorylation states
of initiation factors such as eiF-2 and eiF-4F. The following discussion is
concerned with a growing body of evidence supporting the proposal that
intracellular Ca 2 + functions in the maintenance of high rates of translational
initiation in intact ceiJs. As discussed in later sections of this review and
outlined in Table 1, the Ca 2 + requirement is indirect; the rate of translational
initiation appears to be closely controlled by the functional status of the ER
within which Ca 2 + serves to support glycoprotein processing.
Cells that are exposed to EGTA buffered medium normaiJy exhibit an
extremely rapid fall (seconds) of cytosolic-free Ca2 + concentration as mon-
itored with Ca 2 + fluorescent dyes. 23 With continuing exposure to EGTA,
intracellular-sequestered Ca2 + is graduaiJy depleted by spontaneous release
to the cytosolic pool. Some cell types, such as GH:, pituitary cells, release
their Ca 2 + stores relatively rapidly and become largely depleted within 15 to
30 min, whereas other cell types such as HeLa and CHO cells are relatively
resistant to depletion by this procedure. Amino acid incorporation into nearly
all protein populations in normal rat hepatocytes/4 C6 glial tumor cells,Z5 and
GH 3 pituitary cells 26 - 29 was found to be inhibited 80 to 90% upon depletion
of intracellular Ca2 + pools with EGTA buffered media within 30 min. Addition
of 1 mM Ca2 + in excess of chelator restored the rate of protein synthesis
TABLE 1
The Functional Status of the Endoplasmic Reticulum and Regulation
of Translational Initiation
Metabolic/nutritional Functional status Translational Cytosolic free

alteration of the ER initiation Ca2+
None Ca2 + Sequestration unimpaired Optimal Resting level

Oxidizing environment present (~0.1 fJ.M)
Protein processing optimal
Ca2 +-Mobilizing Ca2+ -Depleted Suppressed Transiently elevated
agonist (minutes) Glycoprotein processing
retarded
Reducing environment Hypoxic Suppressed At or near resting
(minutes) Disulfide bond formation level
impaired
Ca'+ -Mobilizing GRP78 Content increased Partly restored At or near resting
agonist or reducing Malprocessed proteins level
environment (hours) removed
Protein translocation facilitated
Note: Relationships observed to exist among ER sequestered Ca 2 + and GRP78 concentrations,

protein processing, and mechanisms of translational control are highlighted above. Inhibition
of protein processing by the ER is seen to interdict translational initiation, such that the
synthesis of almost all proteins, including those not subject to processing, is curtailed.
The ER is proposed to coordinate mRNA translation with protein processing in a manner
coupled to its functions: (I) as reservoir for hormonally releasable Ca' + and (2) in providing
an environment suited for protein disulfide bonding. In such an arrangement, rates of
protein synthesis would be subordinated to the overall metabolic or nutritional status of
the cell via monitoring of sequestered Ca2 + or changes in oxidation/reduction status. The
accommodation of translation observed following extended incubations in low Ca2 + medium
or under reducing conditions is consistent with a cellular adaption to assure sufficient
rates of protein synthesis for survival during chronic stress. Acquisition of translational
tolerance is viewed to arise as the consequence of increased concentrations of newly
synthesized GRP78 within the ER lumen. This adaptive response is observed to function
prominently against a variety of chronic nutritional, metabolic, physical, and chemical
stresses in partially restoring protein synthesis. Although translational elongation is thought
to be inhibited when cytosolic free Ca2+ is elevated, translational initiation appears to
be unaffected by the transient changes in cytosolic free Ca2 + that may occur in response
to Ca2 +-mobilizing agonists.
within 7 to 10 min to that of nondepleted control preparations. Ca2+ specif-

ically among physiologically occurring cations restored amino acid incor-
poration over a broad range of Mg 2 +, Na+, and K+ concentrations, pH, and
osmolarity in both minimal and enriched media either with or without sera.
The effects of Ca2 + depletion were not traceable to changes in amino acid
uptake, aminoacylation of transfer RNA, RNA synthesis, protein catabolism,
removal of cells from growth surfaces, or changes in viability as measured
by dye exclusion, replating, and determinations of ATP and GTP contents.
Ca 2 + depletion with 1 mM EGTA resulted in the disappearance of polysomes

and an accumulation of monosomes and ribosomal subunits typical of slowed
rates of translational initiation, as well as a large decrease in 43S preinitiation
complex (40S·eiF-2·Met-tRNAr"GTP).30 Reintroduction of Ca2 + rapidly (min)
restored cellular contents of 43S ribosomal preinitiation complex and poly-
somes with corresponding decreases in monosomal and ribosomal subunits.
Comparable polysomal profiles were found for Ca 2 +-depleted and restored
cells exposed to cycloheximide which slowed polypeptide chain elongation
to rate-limiting values in the overall translation process. Average ribosomal
transit times for both Ca2 +-depleted and restored cells were identical and were
extended in parallel as a function of increasing cycloheximide concentration,
indicating that neither elongation nor termination is directly affected by de-
pletion/repletion of the cation. Lysates of GH 3 cells exhibited incorporation
that was proportional to the polysomal contents derived from the original
intact cell preparations, lacked the ability to initiate new peptide synthesis,
and were not directly affected by Ca 2 + or EGTA. Lysates from either Ca 2 +-
depleted or restored cells exposed to cycloheximide provided lysates with
identical elongation activity and polysomal contents to lysates prepared from
Ca2 +-restored controls. Ca2+ was therefore proposed to support translation
by maintaining the rate of initiation rather than polypeptide chain elongation
or termination.
B. REQUIREMENT FOR SEQUESTERED Ca2+

Hormones such as angiotensin II, vasopressin, and a-adrenergic agonists
act on hepatocytes to generate IP 3 which in tum mobilizes approximately 50%
of sequestered Ca2 + stores from the ER to the cytoplasm over a few min. 31
Most of this Ca2 + is immediately pumped into the extracellular fluid in a
manner retarded by high concentrations of the cation. Over several min of
exposure, these hormones were found to inhibit protein synthesis in isolated
hepatocytes incubated with Ca 2 + concentrations within the physiologic range
and at physiologic pH. 24 This inhitition was largely reversible by the addition
of supraphysiologic concentrations of Ca 2 + to the extracellular medium. Cor-
responding reductions were observed in the polysomal contents of excised
portions of perfused rat liver in response to those hormones and manipulations
of the Ca2 + content of the perfusing medium. 32 Since protein synthesis was
inhibited both by EGTA, which extracts cellular Ca2 +, and by hormones,
which mobilize sequestered Ca2 + with some elevation of cytosolic-free
Ca2 +, it was proposed that Ca2 + sequestered by the ER, rather than cytosolic-
free Ca 2 +, was required for maintenance of adequate rates of translational
initiation. The divalent cation ionophores, A23187 and ionomycin, are widely
used as Ca 2 + ionophores to perturb Ca2 + homeostasis. While conventionally
viewed as irreversible promoters of Ca2 + influx across the plasmalemma,
these agents also act to release internally sequestered Ca 2 + as analyzed by
Ca 2 + kinetics in fluorescent dye and 45 Ca loaded cells. At low concentrations

these ionophores have been found to promote nearly quantitative efflux of
Ca 2 + from GH 3 and P3X63Ag8 myeloma cells, whereas influx of the cation
occurred at higher concentrations. 33 The net direction of Ca2 + movement for
a given cell type and ionophore concentration appears to be determined by
whether ionophore fosters entry from the extracellular fluid at rates that
overwhelm the plasmalemma! transport systems pumping the cation from the
cell. Low concentrations of Ca 2 + ionophores inhibited translation comparably
to EGTA and Ca 2 + -mobilizing hormones. 34 Amino acid incorporation was
reduced in a graded fashion in CaH -restored cells upon addition of nM to
J..LM concentrations of ionophore. Inhibitions were rapid and extensive, with
90 to 95% reduction in protein synthesis observed after 10 min exposure to
1 J..LM drug. Incorporation in cells depleted of Ca2 + with EGTA, however,
was largely unaffected by ionophore. Rapid disruption of polysomes and
reduced formation of 43S preinitiation complexes were observed upon treat-
ment of CaH -restored preparations with A23187. Average ribosomal transit
times were unchanged. Ca 2 + ionophores also provided a means of revealing
the strong underlying Ca2 + dependence of initiation in cell types, such as
HeLa and CHO, that possess only a modest overt Ca2 + -dependent component
in the presence of EGTA. The agents were proposed to function as initiation
inhibitors by provoking mobilization of sequestered intracellular Ca2 +. The
inhibitory effects of ionomycin and A23187 on amino acid incorporation and
on CaH accumulation have been recently found to be reversible upon exposure
of the cells to fatty acid-free bovine serum albumin. 33
Free arachidonic acid, which has been associated with the actions of
various secretagogues 35 and with ischemic stress, 36 directly releases seques-
tered Ca 2 + from cells 35 •37 and isolated organelles. 38 The potential role of
arachidonate and related fatty acids in the regulation of protein synthesis via
Ca 2 + mobilization was therefore examined. 39 Unsaturated fatty acids at J..LM
concentrations inhibited protein synthesis in a variety of mammalian cell types
in a manner dependent on degree of unsaturation and cell number. Arachi-
donate was generally the most, and the fully saturated arachidic acid the least,
potent of the fatty acids tested. At 2 X 106 GH 3 cells/ml, amino acid incor-
poration into a broad spectrum of polypeptides was inhibited 80 to 90% by
10 to 20 J..LM fatty acid. Inhibition was maximal at 4 to 8 min and was
attenuated by 1 to 2 hand more pronounced at lower pH. Indomethacin did
not relieve the inhibition indicating that cyclooxygenase pathway metabolites
were not involved. Protein synthesis was maximally inhibited when arachi-
donate mobilized approximately 40% of cell-associated Ca 2 +. At lower con-
centrations (10 J..LM), arachidonate suppressed translational initiation with the
inhibition being reversed as extracellular Ca2 + concentrations were increased
to supraphysiologic values. These properties of arachidonate were consistent
with those expected for a physiologic regulator that inhibits protein synthesis
through mobilizing Ca2 + sequestered in the ER. At higher concentrations (20
J.LM), however, arachidonate inhibited peptide chain elongation in a Ca 2 +-

independent manner. While the mechanism of elongation blockade was not
determined, low concentrations (0.0007%) of the nonionic detergent Lubrol
PX were noted to confer comparable elongation block without significant
mobilization of Ca2+ or cell lysis. Both arachidonate and Lubrol PX blocked
elongation, but not initiation, in reticulocyte lysates. The effects of arachi-
donate in intact cells were reversible with time via its metabolism. Concen-
trations of arachidonate synthesized during ischemic stress appear adequate
for inhibition of translation by either mechanism. 36
Synthetic oligopeptides specific for inhibition of metalloendoprotease
activity suppress a variety of events involving membrane contact and fu-
sion,40.42 supporting the concept that proteolysis functions directly in such
events. These inhibitors also alter transmembranal Ca2+ movements43 and,
at concentrations that block the constitutive secretory pathway in HepG2
hepatoma cells, inhibit amino acid incorporation without affecting ATP con-
centration or viability. 41 The possibility that these agents inhibit protein
synthesis by perturbing sequestered Ca2+ stores was therefore explored. 44
Cbz-Gly-Phe-NH2 provided rapid inhibition of amino acid incorporation into
a broad spectrum of proteins in various cultured cells possessing ER but not
in reticulocytes. Polysome accumulation and incorporation in GH 3 cells were
reduced concurrently, indicating that the dipeptide acted to slow translational
initiation. Inhibitions were largest at low extracellular Ca2+, were reversed
by increasing extracellular Ca2+, were comparable to those achieved in the
presence of EGTA or Ca2+ ionophores, and were observed with assorted
metalloendoprotease antagonists but not with leupeptin. At concentrations
inhibitory to protein synthesis, Cbz-Gly-Phe-NH 2 mobilized cell-associated
45
Ca, lowered cytosolic-free Ca 2 +, and did not generate inositol phosphates.
It was concluded that metalloendoprotease antagonists suppress translational
initiation as a consequence of their capacity to mobilize sequestered Ca2 +
stores. Two hypotheses were advanced to explain the effects of the antago-
nists on Ca2 + homeostasis. The first assumes that vesicular fusion events,
requiring the action of metalloendoprotease(s), function in maintenance of
Ca2+ sequestration by the ER. The return from Golgi to ER of transport
vesicles carrying the Ca 2 + -binding proteins of the ER lumen has been pro-
posed to function in the restoration of the Ca2+ lost initially during movement
of such transport vesicles from ER to Golgi. 45 This process for recovery of
ER luminal components, referred to as the "salvage" pathway, 1" may be
inactivated in the presence of membrane fusion inhibitors, such as Cbz-Gly-
Phe-NH2, resulting in the draining of ER and cytosolic cation. Alternatively
Cbz-Gly-Phe-NH2 may directly activate a Ca2+extrusion system that lowers
both cytosolic-free and sequestered cation.
Several conclusions emerge from the preceding body of information.
Amino acid incorporation in a number of cell types is suppressed by extraction
of cellular Ca2+ with EGTA or by Ca2 + mobilizing agents including hormones
that generate IP 3 , Ca2 + ionophores, arachidonate, and various dipeptides. All

of these agents are reversed to various degrees by high extracellular Ca2 +. It
is also clear that thapsigargin, an agent that blocks Ca2 + accumulation by the
ER, inhibits amino acid incorporation. 46 Not all reports, however, are har-
monious with the preceding body of data supporting a role for sequestered
Ca2 + in initiation. Based on results with lysates, Kumar et al. 47 have concluded
that a Ca 2 + requirement for protein synthesis in Ehrlich ascites tumor cells
relates to direct effects of the free cation on the translational apparatus. The
Ca2 +-selective chelator BAPTA at concentrations of 1 to 2 mM depressed
amino acid incorporation by lysates obtained from these cells; the inhibition
was reversed by Ca2 + but not by several other cations. The BAPTA-treated
preparations exhibited decreased levels of 43S preinitiation complexes and
decreased binding of factors eiF-2 and eiF-3 to 40S subunits. Similar effects
of high concentrations of chelator have been seen, however, in reticulocyte
lysates, 25 a system that does not appear to be supported by Ca2 +. 48
In further studies of Ehrlich cells, 49 calmodulin antagonists were found
to inhibit protein synthesis in both intact and cell-free preparations. Initiation
was preferentially inhibited as indicated by an increase in 80S monomers and
43S complexes accompanied by disaggregation of polysomes. It was sug-
gested that Ca2 + maintains translation by activation of a mechanism involving
calmodulin or a related Ca2 +-binding protein. The relatively high concentra-
tions of calmodulin antagonists required to suppress translation, however,
were significantly in excess of those required for inhibition of most calmo-
dulin-dependent enzymes50 and were well within the range reported to block
Ca2 + entry through Ca2 + channels. 28 A supporting role for calmodulin at
initiation would not fit well with the extensive body of information reporting
inhibition of peptide chain elongation via a calmodulin dependent mechanism
(see below).
The behavior of intact reticulocytes with respect to Ca2 + and A23187
was found to be at variance with that of nucleated cells, perhaps because
reticulocytes lack a well-defined, functional ER. 51 Rates of amino acid in-
corporation in intact rabbit reticulocytes were unaffected by depletion of Ca2 +
with EGTA. The Ca2 + ionophore A23187 strongly inhibited incorporation in
reticulocytes incubated with 1 mM Ca2 + but not with EGTA. Polysomal
profiles and the extension of average ribosomal transit times of cells treated
with Ca 2 + ionophore at 1 mM Ca 2 + were characteristic of translational elon-
gation block. Reticulocyte lysates are a popular system for investigating the
properties of translational initiation because the activity survives cell lysis
and will support the synthesis of protein from exogenously added mRNA.
This property is remarkable since initiation in most mammalian cells is nearly
eliminated by cell damage or disruption. It is interesting to speculate that
reticulocyte lysates retain high degrees of initiation because the process has
been uncoupled from control by vesicular membranes prevalent in most other
cell types.
C. EFFECT OF Ca2 + ON THE PHOSPHORYLATION/

DEPHOSPHORYLATION OF INITIATION FACTORS
The activities of a number of initiation factors are potentially subject to
regulation via protein phosphorylation. 20 - 22 In particular the phosphorylation
of the o. subunit and accompanying inhibition of eiF-2 has been extensively
studied. This factor catalyzes the formation of the ternary complex eiF-2·-
GTP·Met-tRNA, which functions in the binding of initiator Met-tRNA to the
40S ribosomal subunit to form the 43S preinitiation complex essential to the
initiation of all mRNAs. As noted earlier, this complex is depressed in
Ca 2 +-depleted cells. Reactivation of eiF-2 after utilization of the ternary
complex depends on an exchange of GDP bound to eiF-2 for GTP, catalyzed
by the factor eiF-2B. Phosphorylation of eiF-2o. by any one of a number of
kinases results in an increased affinity of eiF-2B such that eiF-2B is seques-
tered into an inactive eiF-2B·-eiF-2o.(P) complex. This complex is incapable
of supporting G nucleotide exchange in the recycling of eiF-2 and translational
initiation therefore becomes suppressed. Kimball and Jefferson 52 observed
that the activity of e1F-2B in extracts of perfused livers was reduced to 53%
of the control value in response to vasopressin, a Ca 2 + mobilizing hormone.
An increase in the proportion of the o. subunit of eiF-2 in the phosphorylated
form was also observed. These authors proposed that inhibition of translational
initiation in vasopressin-treated livers resulted from eiF-2o. phosphorylation.
Alternatively, phosphorylation of eiF-2o. could result as a consequence of
translational initiation block.
Another frequently advanced mechanism for translational regulation in-
volves phosphorylation of factor e1F-4E, a subunit of e1F-4F which serves to
recognize the 5' terminal mRNA cap in early initiation. Phosphorylation of
eiF-4E correlates with enhanced rates of protein synthesis and growth in
mammalian cells. 20 The phosphorylation state of the subunit in Ca2 +-depleted
GH 3 cells, however, did not differ from that in Ca2 + -restored preparations. 53
Phosphorylation of ribosomal protein S6, which is observed to correlate with
increased formation of ribosome initiation complexes, with ribosomal entry
into elongation, and with increased growth, 54 similarly was not altered as a
function of cellular Ca2 + depletion.
D. CONDITIONS THAT DISSOCIATE INITIATION FROM

REGULATION BY Ca2+
Certain types of major stress, including thermal or chemical stress, result
in an initial shutdown of normal protein synthesis in eukaryotic cells accom-
panied by loss of polysomal contents. Over several hours, a series of stress
proteins are induced and overall amino acid incorporation is largely restored,
but with minimal Ca2 + dependence. For example, after thermal stress at 46°or
exposure to sodium arsenite, amino acid incorporation into Ca2 +-replete GH 3
cells was reduced to that of unstressed, Ca 2 +-depleted cells, 55 and preparations
originally stressed in the Ca2 +-depleted condition were subsequently unable
to accumulate poly somes in response to Ca2 +. Ca2 + did not influence either
the recovery of amino acid incorporation or the induction of the heat shock
proteins (hsp) that occurred during the next several hours. Restoration of the
Ca 2 + requirement occurred by 24 h. More modest degrees of thermal stress
(41 °) resulted in the induction of hsp 68 without loss of Ca2 +-stimulated
initiation. Under this condition hsp 68 was synthesized in a Ca 2 +-dependent
manner.
Viral infections interrupt host cell protein synthesis by various mecha-
nisms while reprogramming the translational apparatus to synthesize viral
coded proteins. HEp-2 cells infected with HSV-1 were employed to examine
the Ca 2 + requirement for translation of viral mRNAs. 56 Early in viral infec-
tion, the synthesis of nearly all proteins, including viral (a) proteins, was
sensitive to inhibition by Ca2 + ionophore or EGTA. By 4 to 6 h after
infection, overall polypeptide synthesis in infected cells had become resistant
to depletion of Ca2 + stores. Specific viral mRNAs were readily detected in
polysomes, and the synthesis of viral polypeptides of early (13) and late (-y)
kinetic classes was found to be insensitive to the effects of Ca2 + depletion.
Production of mature forms of viral glycoproteins, however, was reduced by
treatment with EGTA and ionophore. Thus, productive infection of cells by
HS V-1, like heat shock or chemical stress, results in a modification that
overcomes the requirement of translational initiation for stored Ca2 +. Pro-
duction of most viral proteins may then proceed in a Ca 2 +-independent
fashion. A second, Ca2 +-sensitive step affecting the maturation (glycosyla-
tion and/or transport) of viral glycoproteins was also apparent which was not
bypassed during HSV -1 infection.
E. EFFECT OF Ca 2 + ON PEPTIDE CHAIN ELONGATION

Peptide chain elongation in healthy cells is readily seen to be rate-
limiting over initiation in that polysomal content accumulates to degrees that
are not increased by cycloheximide, an inhibitor that reduces incorporation
by slowing elongation and extending average ribosomal transit times. An
inhibitory regulatory input at this site would be expected to slow protein
synth;!sis quite rapidly. 22 In recent years, evidence has accrued for the exist-
ence of a nonselective and transitory mechanism for elongational slowing
coupled to increased concentrations of intracellular free Ca2 +. The physio-
logic relevance of this mechanism, which involves the rapid and transient
phosphorylation of elongation factor 2 (eEF-2) has not as yet been clearly
established. Elongation of polypeptide chains in mammalian cells is cata-
lyzed by two factors. n Elongation factor 1 ( eEF-1) is responsible for the
codon-dependent binding of aminoacyl tRNA to the acceptor (A) site on the
ribosome, whereas eEF-2 promotes translocation of the charged tRNA from
the A site to the P (peptide bond formation) site. The activity of eEF-2 is
associated with a monomeric 100 kDa, GTP-binding protein found in all
eukaryotic cells. Immediately prior to translocation, the anticodon end of the
peptidyl-tRNA is bound in the A site and the peptidyl end is located in the
P site. Although details of the mechanism have not been elucidated, it is
likely that binding of the eEF-2·GTP complex promotes movement of the
mRNA and the anticodon portion of the peptidyl-tRNA into the ribosomal P
site accompanied by the hydrolysis of the nucleotide to GDP.
Cytoplasmic extracts of mammalian cells are routinely found to incor-
porate phosphate from ATP into a major 100-kDa species. 57 In 1983 Palfrey5 8
reported that such incorporation was enhanced by Ca 2 + and the ubiquitous
Ca2 + mediator protein calmodulin. Independently, Palfrey and Nairn 58 .s9 and
Guroff and colleagues60 •61 isolated the I 00-kDa substrate and employed it in
procedures for partial purification of the protein kinase responsible for the in
vitro phosphorylation. The kinase, which possessed a molecular mass of 130
to 140 kDa, was identified in a broad variety of tissues, required Ca2 + and
calmodulin for optimal activity, and was distinct in substrate specificity from
previously characterized forms of Ca 2 +/calmodulin-dependent protein ki-
nase. 59 •62 The enzyme, named Ca 2 +/calmodulin-dependent protein kinase III
by Nairn eta!. , 59 was unable to phosphorylate a variety of substrates utilized
by other protein kinases. Its 100-kDa substrate was not phosphorylated by
other protein kinases.
Identification of the 100-kDa substrate as eEF-2 was reported in early
1987 by Ryazanov 63 and later by Nairn and Palfrey. 64 Since the factor con-
stituted the only known substrate for the enzyme, Ryazanov proposed the
enzyme be named eEF-2 kinase. 65 Partially purified eEF-2 kinase was ob-
served to phosphorylate three threonine residues of eEF-2 from reticulocytes 66
with threonine 56 advanced as the primary phosphorylation site. 67 The com-
parable factor from chick embryo exhibited additional phosphorylation sites
on tyrosine and serine residues. 68 The threonine phosphorylation sites of
reticulocyte eEF-2 were localized to a region near the N terminus that may
constitute the ribosome binding site. 66 Phosphorylation was found to stabilize
the factor against the action of trypsin. 69 Dephosphorylation of phosphory Ia ted
factor in cell-free systems could be accomplished by incubation with various
phosphatases, with type 2A phosphatase being the most effective. 70. 72 Recent
functional studies support the hypothesis that phosphorylated eEF-2 cannot
sustain elongation. Okadaic acid, a tumor promoter that potently inhibits
protein phosphatases, was found by Redpath and Proud 71 to increase the net
phosphorylation of eEF-2 in reticulocyte lysates while provoking inhibition
of peptide chain elongation. In contrast the stimulation of elongation rate
observed when reticulocyte Iysates are incubated with high concentrations of
cAMP correlated with dephosphorylation of eEF-2. 70 The dephosphorylated
form of eEF-2 was found to support poly(U)-directed polyphenylalanine syn-
thesis in a reconstituted elongation system when combined with eEF-1 , whereas
phosphorylated eEF-2 was ineffective. 64 •73 Phosphorylated eEF-2 is reported
to bind to 80S ribosomes but to be unable to catalyze translocation of peptidyl
tRNA from the A site to the P site for ribosomes carrying poly(U) and
phenylalanyl tRNA. 74
Transient phosphorylation of eEF-2 has been reported for a wide variety

of cell types following exposure to Ca 2 + -mobilizing substances. 5 7 For ex-
ample, treatment of fibroblasts with serum, bradykinin, vasopressin, epider-
mal growth factor (EGF), or Ca2 + ionophores, each of which provoked tran-
sient increases in [Ca2+ L resulted in two- to tenfold increases in eEF-2
phosphorylation. 75 Phosphorylation of the factor was maximal at 0.5 to 1 min
and attenuated at 5 min. Thrombin and histamine, which elevate [Ca 2 +]; in
umiblical vein endothelial cells, also provided a rapid and transient phos-
phorylation of eEF-2, whereas phorbol esters or cAMP-elevating agents were
ineffective. 76 Phosphorylation of eEF-2 in cells exposed to bradykinin, serum,
vasopressin, histamine, or thrombin was not reduced when conducted in Ca 2 +-
depleted medium, whereas phosphorylation involving EGF or ionophore re-
quired the presence of extracellular cation. Phosphorylation of the factor
following depolarization of PC-12 pheochromocytoma cells was observed by
Nairn and colleagues 77 to be overturned by nerve growth factor (NGF) or
cAMP-elevating agents. Significant decreases in eEF-2 kinase activity were
noted within 1 h of exposure of the cells to NGF or cAMP-elevating agent.
Data supporting the hypothesis that downregulation of eEF-2 kinase is at-
tributable to the action of cAMP-dependent protein kinase have recently been
reported. 78 Of additional interest is recent evidence from the laboratory of
Nygard et al. 79 that the phosphorylated form of eEF-2 kinase is the active
species.
Given the transitory nature of eEF-2 phosphorylation in most cells, the
consequent inhibition of translation would be expected to be comparably short-
lived and therefore difficult to measure experimentally by standard techniques.
It is not clear that eEF-2 becomes sufficiently phosphorylated in intact non-
erythroid cells to inhibit the rate of peptide bond formation significantly. If,
however, phosphorylated eEF-2 were to bind to ribosomes with higher affinity
than nonphosphorylated factor, then only a fraction of the pool of factor would
need to be phosphorylated for elongation to be inhibited. The physiologic
role played by a brief slowing of translational elongation is also not imme-
diately obvious. Conceivably inhibition through this mechanism would rapidly
divert energy toward supporting hormonally activated cell-specific functions.
This inhibition would attenuate with depletion of sequestered stores of Ca 2 +
and the concommitant development of inhibition of translational initiation.
Such model building would be precluded by an alternative proposal by
Ryazanov 57 that eEF-2 kinase may be active only at mitosis. Evidence cited
in support of this proposal includes a report by Celis et al. 80 that the proportion
of phosphorylated eEF-2 variants present in amnion cells increased consid-
erably during mitosis, a period during which the rate of translation is believed
to decrease and [Ca2 +]; to rise briefly. Additionally Severinov et al. 81 found
that eEF-2 kinase activity in Xenopus oocytes decreased substantially during
the final stages of oogenesis and was absent in fully grown oocytes. The
mechanism was hypothesized to enhance the degradation of short lived re-

pressors of genes, such as proto-oncogenes, that code for proliferative factors,
thereby recruiting new mRNAs to translation. Activation of such gene expres-
sion has been reported following exposure to mitogenes and nonspecific
protein synthesis inhibitors such as cycloheximide. 82 •83 In summary, several
major questions relating to the effects of Ca2 + on translation remain to be
clarified. What is the overall mechanism through which mobilization of se-
questered Ca2 + from the ER inhibits translational initiation? What events
occur during its reversal? What is the physiologic function of either this
inhibition or the inhibition of translational elongation directed by cytosolic
Ca2 +?
IV. Ca2+ AND THE INDUCTION OF GRP78/BiP
Deprivation of sequestered Ca 2 + is currently viewed as a form of meta-

bolic stress to which mammalian cells respond with the transcription-depen-
dent induction of various resident ER proteins. 84 The most prominently in-
duced of these is a 78-kDa protein that has been localized to the lumen.
Commonly termed GRP78 or BiP, 18 this reticuloplasmin is induced by agents
or conditions that alter metabolic or nutritional status, such as inhibitors of
protein glycosylation, glucose deprivation severe enough to deplete glycosyl
units, sulfhydryl reducing agents, amino acid analogs, viral infection, and
overexpression of secretory proteins. 84 - 88 GRP78 has been variously hypoth-
esized to function in the correct folding of proteins during early protein
processing, 89 in the retention of improperly folded proteins that accumulate
within the ER lumen when processing is distressed, 18 or in the translocation
of proteins from the cytosol to the ER for processing. 90 ·91 Transcription of
the GRP78 gene is thought to be activated consequent to altered rates of early
protein processing via mechanisms not as yet defined. The GRP78 gene,
which possess a highly conserved promoter region that confers Ca2 + ionophore
inducibility and binds trans-acting transcription factors, 92 ·93 is required for
the continued viability of yeast cells. 94 The following discussion describes
additional findings that implicate GRP78 in a mechanism whereby mammalian
cells rapidly develop translational tolerance to depletion of sequestered Ca2 +
and other metabolic stresses such as a reducing environment.
GH 3 cells depleted of sequestered Ca 2 + maintain their viability for several
hours with stable, albeit reduced, rates of protein synthesis. When exposed
for 2 to 3 h to a phorbol ester and/or a cAMP-elevating agent, however, such
preparations exhibited increased incorporation of amino acids into a broad
spectrum of polypeptides. 29 •95 Phorbol esters and cAMP fostered the accu-
mulation of small polysomes without affecting average ribosomal transit times,
indicative of increased rates of translational initiation. In EGTA-buffered or
ionophore-containing medium, a six- to eightfold stimulation of protein syn-
thesis occurred over 3 h, as compared to relatively small increments in Ca 2 +-

repleted controls. In contrast to the behavior of GH, cells, Ca 2 + -depleted C6
glial tumor cells displayed spontaneous increases in incorporation rates with
time that were not affected by cAMP or phorbol ester.
Cells first accommodated to low Ca 2 + medium and then restored with
optimal Ca 2 + exhibited rates of incorporation identical to those of non treated
controls, but remained resistant to inhibition on subsequent challenge with
EGTA or ionophore. Acquisition of translational tolerance to these agents did
not involve alterations of cellular capacity or affinity for Ca2 +, but was
invariably preceded specifically by transcriptional-dependent induction of
GRP78.Y 5 In GH 3 and immunoglobulin-secreting myeloma cells deprived of
Ca 2 +, the synthesis of GRP78 was promoted by phorbol ester and cAMP.
Interestingly, the development of translational tolerance to ionophore corre-
lated directly with new synthesis of GRP78 as measured by pulse labeling
of the protein with [35 S]-methionine, rather than with total GRP78 measurable
by staining. Cells treated with the reducing agent dithiothreitol (OTT), an
efficacious inducer of GRP78 that does not mobilize sequestered Ca2 +, ex-
hibited acute shutdown of translational initiation followed by the subsequent
development of translational tolerance not only to DTT but also to Ca2 +
ionophore and EGTA. Chronic tolerance to these agents was found for NS-1
myeloma cells, which are defective in immunoglobulin processing and con-
stitutively express high contents of GRP78 and its corresponding mRNA. 96
Antisense oligonucleotides directed against GRP78 mRNA reduced amino
acid incorporation by tolerant GH 3 cells, but not in nontolerant controls. GH 3
cells treated several h with the metalloendoprotease antagonist Cbz-Giy-Phe-
NH2 also acquired translational tolerance occurring in tandem with GRP78
induction.
In investigation of the roles of cAMP and phorbol ester in the induction
of GRP78 and the coordinate development of translational tolerance to Ca 2 +
ionophore in GH 3 cells, 97 induction of the reticuloplasmin was found to depend
on two factors. First, cAMP and phorbol ester enhanced GRP78 gene tran-
scription. GRP78 mRNA was induced 3- to 6-fold with ionophore alone as
compared to 12- to 20-fold with ionophore plus cAMP-elevating agent and
phorbol ester. GRP78 gene transcription in nuclei isolated from ionophore-
treated cells was also increased by cAMP and phorbol ester. Induction was
not apparent in the absence of ionophore regardless of the presence of cAMP
and phorbol esters, and these activators did not affect GRP78 mRNA stability
despite the increased polysomal content of cells undergoing translational re-
covery. This cAMP/phorbol ester requirement for optimal GRP78 mRNA
induction was initially surprising in that induction in other cell types was not
known to require the activators. However, examination of the sequence of
the rat GRP78 promoter leads us to speculate that a cAMP response element
(CRE), which differs from the phorbol ester response sequence (TRE) by
only one nucleotide, 98 potentially exists in this region. The sequence from
-189 to -182 is 5'-TGACGTGA-3 and conforms to the CRE-consensus. 99
The human GRP78 gene has a similar sequence from -153 to -147. These
sequences correspond to protected domains and are footprinted on the non-
coding strand. 100 With development of promoter deletion mutants, it should
be possible to identify sequences that mediate regulation of GRP78 gene
expression by cAMP and phorbol ester. The second factor favoring GRP78
induction in stressed GH 3 preparations was the preferential translation of
GRP78 mRNA. This mRNA associated almost exclusively with polysomes,
whereas other mRNAs were subject to initiation block. 97 This efficient ini-
tiation of GRP78 may be explained by the recent observation 101 that the 5'
leader of GRP78 mRNA, unlike that of most cellular mRNAs, can confer
initiation by an internal ribosome binding mechanism.
On the basis of these and other findings, alterations in translational activity
following Ca 2 + deprivation or exposure to a reducing environment appeared
to emanate from perturbation of ER function. As noted earlier, the ER is
thought to function prominently in maintenance of Ca2 + homeostasis and
early protein processing reactions requiring an oxidizing environment. Agents
that deplete sequestered Ca2 + stores or that generate a hypoxic environment
acutely suppress translation at initiation. On chronic exposure to either class
of agent, transcriptionally dependent translational accommodation is seen to
develop that appears to depend on the rapid induction of GRP78/BiP, an ER
resident protein proposed to serve in protein translocation and folding. Major
open questions regarding GRP78 include: ( 1) How does the ER signal the
nucleus to transcribe new mRNA for the protein? (2) How does GRP78
promote the recovery of amino acid incorporation in Ca 2 + depleted cells?
Does GRP78 coordinate the relative rates of translational initiation and protein
processing? Evidence that both sequestered Ca 2 + and adequate oxidizing
eqivalents are needed to sustain early protein processing events is summarized
in the following section.
V. A Ca2+ REQUIREMENT FOR SECRETORY

PROTEIN PROCESSING AND EXPORT
Although cytosolic free Ca 2 +, rather than sequestered Ca 2 +, is conven-

tionally considered to support regulated metabolic processes, certain findings
favor the possibility that sequestered Ca 2 + maintains posttranslational pro-
cessing. For example, the activity in vitro of a 1,2 mannosidase, a vesicular
enzyme involved in processing of N-linked oligosaccharides, is reportedly
stimulated by Ca 2 +. 102 Depletion of cellular Ca 2 + stores has been proposed
to accelerate degradation of certain heterologous proteins in the ER. 17 Of
particular relevance are the findings that A23187 and ionomycin are effica-
cious inhibitors of the processing and secretion of a 1-antitrypsin, and, to a
lesser extent other proteins, by HepG2 cells. 103 Following ionophore treatment
secretory glycoproteins with a high mannose configuration accumulated in
vesicles with the density of the rough ER. It was unclear from these studies,
however, that Ca2 + mobilization mediated the observed secretory arrest and
that resumption of normal glycoprotein processing and export was possible
upon removal of the stress.
HepG2 human hepatoma cells were employed as a model system to further
investigate relationships between early protein processing and Ca 2 + storage
by the ER. 104 Three lines of direct evidence were provided that Ca2 + per se
is required for processing and secretion. First, secretion of a 1 -antitrypsin and
to a lesser degree other newly synthesized polypeptides was reduced following
treatments in EGTA-containing medium. Second, inhibition of glycoprotein
processing by Cbz-Gly-Phe-NH 2 was found to depend on Ca2 +. Inhibitions
of a 1-antitrypsin processing and of complement factor 3 (C3) secretion by
dipeptide were largest in Ca2 + -depleted medium and were either markedly
reduced or prevented at elevated extracellular Ca2 +. Third, it was demon-
strated that arrest of a 1 -antitrypsin processing and export in the presence of
ionomycin was fully reversed by Ca2 + following extraction of the drug with
BSA. In addition to establishing a Ca 2 + involvement in the mechanism of
ionophore inhibition, accomplishment of the reversal illustrated (1) that io-
nomycin-treated preparations retained the normal precursor form of a 1-anti-
trypsin rather than a malprocessed intermediate destined for degradation, and
(2) that both vesicular processing and protein transport were not permanently
disrupted by ionophore treatment. In contrast to agents depleting Ca 2 + stores,
exposure to DTT reduced albumin export while affecting a 1-antitrypsin export
minimally. Although the inhibition of albumin export was not accompanied
by significant intracellular accumulation of native polypeptide, considerable
amounts of several cell-associated, immunoreactive polypeptides of differing
molecular mass were invariably seen to develop in DTT-treated cells. These
peptides appeared to be aggregative/degradative forms of albumin collectively
more than doubling cell associated retention of the protein.
Ca2 + has been proposed to be required for the transport of secretory
proteins from ER to Golgi. 105 That albumin secretion is affected minimally
by ionomycin and not affected by EGTA argues against a prominent role for
Ca 2 + in transport of all proteins from the ER. Alternatively, Ca 2 + stored in
the ER may be required for accurate folding which, in turn, is necessary for
transport of secretory glycoproteins to other vesicular compartments. This
mechanism assumes more stringent folding requirements for transport of gly-
coproteins, such as a 1-antitrypsin and C3, than for nonglycosylated secretory
species such as albumin. An additional possibility consistent with available
data is that depletion of sequestered Ca2 + specifically restricts the enzymatic
trimming of high mannose forms of glycoproteins, resulting in the accumu-
lation of species not recognized by transport systems. By contrast, these data
did not support a reciprocal relationship between sequestered Ca 2 + and the

rate of degradation of secretory proteins. Furthermore, sequestered Ca2 + did
not appear to function critically in protein disulfide bond formation. Pro-
cessing and export of albumin, a protein with 17 intrachain disulfide link-
ages, 106 was largely unimpeded by EGTA or ionomycin.
VI. SUMMARY
Inhibition of mRNA translation consequent to interruption of ER protein
processing appears to be of general significance in that suppression of amino
acid incorporation into total cellular proteins of HepG2 cells accompanied
inhibitions of protein processing by agents depleting sequestered Ca2 + stores
or by DTT. 104 Tunicamycin, which interferes with transfer of the core oli-
gosaccharide structure to newly synthesized polypeptides and retards secretion
of the nonglycosylated species, was found to inhibit translational initiation
in GH 3 cells. 95 It is most appealing, therefore, to propose that the ER effec-
tively coordinates rates of mRNA translation with protein processing. While
translation obviously provides the substrates for the processing apparatus of
the ER and Golgi, there is no evidence of any other influence of translation
on processing. 107 In contrast it is quite clear that some inhibitors of protein
processing interdict translational initiation, such that the synthesis of almost
all proteins, including those not subject to processing, is curtailed. It will be
of interest to ascertain whether cells expressing translational tolerance sub-
sequent to induction of GRP78 are capable of efficient protein processing
when challenged by stresses that perturb ER function. Other questions of
interest include: Where is the Ca2 + requirement in glycoprotein processing
exerted? Does sequestered Ca 2 + regulate the rate of glycoprotein processing?
REFERENCES
I. Brostrom, C. 0. and Brostrom, M. A., Calcium dependent regulation of protein
synthesis in intact mammalian cells, Annu. Rev. Physiol., 52, 577, 1990.
2. Lee, C. and Chen, L. B., Dynamic behavior of endoplasmic reticulum in living cells,
Cell, 54, 37, 1988.
3. Green, M. and Mazzarella, R. A., Biosynthesis and sorting of proteins of the endo-
plasmic reticulum, in Protein Transfer and Organelle Biogenesis, Das, R. C. and Robbins,
P. W., Eds., Academic Press, Orlando, 1988, 243.
4. Koch, G. L. E., The endoplasmic reticulum and calcium storage, Bioessays, 12, 527,
1990.
5. Carafoli, E., Intracellular calcium homeostasis, Annu. Rev. Biochem., 56, 395, 1987.
6. Meldolsi, J. and Pozzan, T., Pathways of Ca2 + entry at the plasma membrane: voltage-,
receptor-, and second messenger-operated channels, Exp. Cell Res., 171, 271, 1987.
7. Berridge, M. J., Inositol phosphates and cell signalling, Nature (London), 341, 197,
1989.
8. Tsien, R. Y., Calcium channels, stores, and oscillations, Annu. Rev. Cell Bioi., 6, 715,
1990.
9. Walz, B. and Baumann, B., Calcium-sequestering cell organelles: in situ localization,
morphological and functional characterization, Prog. H istochem. Cytochem. , 20, I , 1989.
10. Meldolesi, J., Madeddu, L., and Pozzan, T., Intracellular Ca 2 + storage organelles in
non-muscle cells: heterogeneity and functional assignment, Biochim. Biophys. Acta, I 055,
130, 1990.
II. Rossier, M. F. and Putney, J. W., Jr., The identity of the calcium-storing, inositol
1,4,5-trisphosphate-sensitive organelle in non-muscle cells: calciosome, endoplasmic re-
ticulum ... or both?, Trends Neurosci., 14, 310, 1991.
12. Berridge, M. J., Cytoplasmic calcium oscillations: a two pool model, Cell Calcium,
12, 63, 1991.
13. Volpe, P., Krause, K.-H., Hashimoto, S., Zorzato, F., Pozzan, T., Meldolesi, J.,
and Lew, D. P., "Calciosome," a cytoplasmic organelle: The inositol I ,4,5-trisphos-
phate-sensitive Ca'+ store of nonmuscle cells?, Proc. Nat!. Acad. Sci. U.S.A., 85, 1091,
1989.
14. Baumann, 0., Walz, B., Somlyo, A. V., and Somlyo, A. P., Electron probe microan-
alysis of calcium release and magnesium uptake by endoplasmic reticulum in bee pho-
toreceptors, Proc. Natl. Acad. Sci. U.S.A., 88, 741, 1991.
15. Bian, J., Ghosh, T. K., Wang, J.-C., and Gill, D. L., Identification of intracellular
calcium pools. Selective modification by thapsigargin, J. Bioi. Chern., 266, 8801, 1991.
16. Satoh, T., Ross, C. A., Villa, A., Supattapone, S., Pozzan, T., Snyder, S. H., and
Meldolesi, J., The inositol! ,4,5-trisphosphate receptor in cerebellar purkinje cells: quan-
titative immunogold labeling reveals concentration in an ER subcompartment, J. Cell
Bioi., Ill, 615, 1990.
17. Wileman, T., Kane, L. P., Carson, G. R., and Terhorst, C., Depletion of cellular
calcium accelerates protein degradation in the endoplasmic reticulum, J. Bioi. Chern.,
266, 4500, 1991.
18. Pelham, H. R. B., Control of protein exit from the endoplasmic reticulum, Annu. Rev.
Cell Bioi., 5, I, 1989.
19. Milner, R. E., Baksh, S., Shemanko, C., Carpenter, M. R., Smillie, L., Vance,
J. E., Opas, M., and Michalak, M., Calreticulin, and not calsequestrin, is the major
calcium binding protein of smooth muscle sarcoplasmic reticulum and liver endoplasmic
reticulum, J. Bioi. Chern., 26, 7155, 1991.
20. Sonenberg, N., Cap-binding proteins of eukaryotic messenger RNA: functions in initi-
ation and control of translation, Prog. Nucleic Acid Res., 35, 173, 1988.
21. Hershey, J. W. B., Protein phosphorylation controls translation rates. J. Bioi. Chern.,
264, 20823, 1989.
22. Hershey, J. W. B., Translational control in mammalian cells, Annu. Rev. Biochem., 60,
717, 1991.
23. Albert, P. R. and Tashjian, A. H., Jr., Relationship of thyrotropin-releasing hormone-
induced spike and plateau phases in cytosolic free Ca 2 + concentrations to hormone se-
cretion, J. Bioi. Chern., 259, 15350, 1984.
24. Brostrom, C. 0., Bocckino, S. B., Brostrom, M.A., and Galuska, E. M., Regulation
of protein synthesis in isolated hepatocytes by calcium-mobilizing hormones, Mol. Phar-
macal., 29, 104, 1986.
25. Brostrom, C. 0., Bocckino, S. B., and Brostrom, M. A., Identification of a Ca2 +
requirement for protein synthesis in eukaryotic cells, J. Bioi. Chern., 258, 14390, 1983.
26. Brostrom, M. A., Brostrom, C. 0., Bocckino, S. B., and Green, S. S., Ca2 + and
hormones interact synergistically to stimulate rapidly both prolactin production and overall
protein synthesis in pituitary tumor cells, J. Cell. Physio/., 121, 291, 1984.
27. Wolfe, S. E., Brostrom, C. 0., and Brostrom, M. A., Mechanisms of action of
inhibitors of prolactin secretion in GH 3 pituitary cells. I. Ca 2 +-dependent inhibition of
amino acid incorporation, Mol. Pharmacal., 29, 411, 1986.
28. Wolfe, S. E. and Brostrom, M. A., Mechanisms of action of inhibitors of prolactin
secretion in GH 3 pituitary cells. II. Blockade of voltage dependent Ca 2 + channels, Mol.
Pharmacal., 29, 420, 1986.
29. Brostrom, M.A., Chin, K.-V., Cade, C., Gmitter, D., and Brostrom, C. 0., Stim-
ulation of protein synthesis in pituitary cells by phorbol esters and cyclic AMP, J. Bioi.
Chern., 262, 16515, 1987.
30. Chin, K.-V., Cade, C., Brostrom, M. A., Galuska, E. M., and Brostrom, M. A.,
Calcium-dependent regulation of protein synthesis at translational initiation in eukaryotic
cells, J. Bioi. Chern., 262, 16509, 1987.
31. Abdel-Latif, A. A., Calcium-mobilizing receptors, polyphosphoinositides, and the gen-
eration of second messengers, Pharmac. Rev., 38, 227, 1986.
32. Chin, K.-V., Cade, C., Brostrom, M.A., and Brostrom, C. 0., Regulation of protein
synthesis in intact rat liver by calcium mobilizing agents, Int. J. Biochem., 20, 1313,
1988.
33. Gmitter-Yellen, D., Brostrom, C. 0., Kuznetsov, G., and Brostrom, M. A., un-
published data, 1991.
34. Brostrom, C. 0., Chin, K.-V., Wong, W. L., Cade, C., and Brostrom, M. A.,
Inhibition of translational initiation in eukaryotic cells by calcium ionophore, J. Bioi.
Chern., 264, 1644, 1989.
35. Wolf, B. A., Turk, J., Sherman, W. R., and McDaniel, M. L., Intracellular Ca 2 +
mobilization by arachidonic acid, J. Bioi. Chern., 261, 3501, 1986.
36. Yasuda, H., Kishiro, K., Izumi, N., and Nakanishi, M., Biphasic liberation of ar-
achidonic and stearic acids during cerebral ischemia, J. Neurochem., 45, 168, 1985.
37. Kolesnick, R.N., Musacchio, I., Thaw, C., and Gershengorn, M. C., Arachidonic
acid mobilizes calcium and stimulates prolactin secretion from GH 3 cells, Am. J. Physiol.,
246, E458, 1984.
38. Chan, K.-M. and Turk, J., Mechanism of arachidonic acid-induced Ca 2 + mobilization
from rat liver microsomes, Biochim. Biophys. Acta, 928, 186, 1987.
39. Rotman, E. I., Brostrom, M.A., and Brostrom, C. 0., Inhibition of protein synthesis
in intact mammalian cells by arachidonic acid, Biochem. J., 282, 487, 1992.
40. Strittmatter, W. J., Couch, C. B., and Mundy, D. I., Role of metalloendoprotease
in the fusion of biological membranes, in Cell Fusion, Sowers, A. E., Ed., Plenum
Press, New York, 1987, 99.
41. Strous, G. J., van Kerkhof, P., Dekker, J., and Schwartz, A. L., Metalloendoprotease
inhibitors block protein synthesis, intracellular transport, and endocytosis in hepatoma
42. Hammerschlag, R., Bolen, F. A., and Stone, G. C., Metalloendoprotease inhibitors
block fast axonal transport, J. Neurochem. 52, 268, 1989.
43. Lelkes, P. I. and Pollard, H. B., Oligopeptide inhibitors of metalloendoprotease inhibit
catecholamine secretion from bovine adrenal chromaffin cells by modulating intracellular
calcium homeostasis, J. Bioi. Chern., 262, 15496, 1987.
44. Brostrom, M.A., Prostko, C. R., Gmitter-Yellen, D., Grandison, L. J., Kuznetsov,
G., Wong, W. L., and Brostrom, C. 0., Inhibition of translational initiation by me-
talloendoprotease antagonists, J. Bioi. Chern., 266, 7037, 1991.
45. Kelly, R. B., Tracking an elusive receptor, Nature (London), 345, 480, 1990.
46. Wong, W. L., Brostrom, M.A., and Brostrom, C. 0., unpublished data, 1991.
47. Kumar, R. V., Wolfman, A., Panniers, R., and Henshaw, E. C., Mechanism of
inhibition of polypeptide chain initiation in calcium-depleted Erhlich ascites tumor cells,
J. Cell Bioi., 108, 2107, 1989.
48. Wong, W. L., Brostrom, M.A., and Brostrom, C. 0., Effects of Ca'* and ionophore
A23187 on protein synthesis in intact rabbit reticulocytes, Int. J. Biochem., 23, 605,
1991.
49. Kumar, R. V., Panniers, R., Wolfman, A., and Henshaw, E. C., Inhibition of protein
synthesis by antagonists of calmodulin in Ehrlich ascites tumor cells, Eur. J. Biochem.,
195, 313, 1991.
50. Asano, M. and Hidaka, H., Biopharmacological properties of naphthalenesul fonamides
as potent calmodulin antagonists, in Calcium and Cell Function, Vol. 5, Cheung, W. Y.,
Ed., Academic Press, Orlando, 1984, 123.
51. West, J. B., Physiological Basis of Medical Practice, lith ed., Williams and Wilkins,
Baltimore, 1985, 346.
52. Kimball, S. R. and Jefferson, L. S., Mechanism of inhibition of protein synthesis by
vasopressin in rat liver, J. Bioi. Chern., 265, 16794, 1990.
53. Fawell, E. H., Boyer, I. J., Brostrom, M.A., and Brostrom, C. 0., A novel calcium-
dependent phosphorylation of a ribosome-associated protein, J. Bioi. Chern., 264, 1650,
1989.
54. Traugh, J. A. and Pendergast, A. M., Regulation of protein synthesis by phosphor-
ylation of ribosomal protein S6 and aminoacyl tRNA synthetases, Pro I?,. Nuclei Acid Res.
Mol. Bioi., 33, 195, 1986.
55. Brostrom, M. A., Lin, X., Cade, C., Gmitter, D., and Brostrom, C. 0., Loss of a
calcium requirement for protein synthesis in pituitary cells following thermal or chemical
stress, J. Bioi. Chern., 264, 1638, 1989.
56. Pancake, B., Prostko, C. R., Brostrom, M. A., and Brostrom, C. 0., unpublished
data, 1989.
57. Ryazanov, A. G. and Spirin, A. S., Phosphorylation of elongation factor 2: a key
mechanism in regulating gene expression in vertebrates, The New Biologist, 2, 843, 1990.
58. Palfrey, H. C., Presence in many mammalian tissues of an identical major substrate (M,
100,000) for calmodulin-dependent protein kinase. FEBS Lett.. 157, 183. 1983.
59. Nairn, A. C., Bhagat, B., and Palfrey, H. C., Identification of calmodulin-dependent
protein kinase III and its major M, 100.000 substrate in mammalian tissues, Proc. Nat/.
Acad. Sci. U.S.A .. 82, 7939, 1985.
60. End, D., Tolson, N., Hashimoto, S., and Guroff, G., Nerve growth factor-induced
decrease in the cell free phosphorylation of a soluble protein in PC12 cells, J. Bioi.
Chern .. 258, 6549, 1983.
61. Hama, T. and Guroff, G., Distribution of NsplOO and NsplOO kinase, a nerve growth
factor-sensitive phosphorylation system, in rat tissues, J. Neurochem .. 45. 1279. 1985.
62. Togari, A. and Guroff, G., Partial purification and characterization of a nerve growth
factor-sensitive kinase and its substrate from PCI2 cells. J. Bioi. Chern., 260. 3804,
1985.
63. Ryazanov, A. G., CaZ+ /calmodulin-dependent phosphorylation of elongation factor 2,
FEBS Lett., 214.331, 1987.
64. Nairn, A. C. and Palfrey, H. C., Identification of the major M, 100.000 substrate for
calmodulin-dependent protein kinase III in mammalian cells as elongation factor 2, J.
Bioi. Chem., 262, 17299. 1987.
65. Ryazanov, A. G., Natapov, P. G., Shestakova, E. A., Severin, F. F., and Spirin,
A. S., Phosphorylation of the elongation factor 2: the fifth Ca'+ /calmodulin-dependent
system of protein phosphorylation, Biochimie, 70, 619. 1988.
66. Ovchinnikov, L. P., Motuz, L. P., Natapov, P. G., Averbuch, L. J., Wettenhall,
R. E. H., Szyzka, R., Kramer, G., and Hardesty, B., Three phosphorylation sites in
elongation factor 2. FEBS Lett., 275. 209, 1990.
67. Price, N. T., Redpath, N. T., Severinov, K. V., Campbell, D.G., Russell, J. M.,
and Proud, C. G., Identification of the phosphorylation sites in elongation factor-2 from
rabbit reticulocytes, FEBS Lett.. 282. 253, 1991.
68. Kim, Y. W., Kim, C. W., Kang, K. R., Byun, S.M., and Kang, Y.-S., Elongation
factor-2 in chick embryo is phosphorylated on tyrosine residues as well as serine and
threonine, Biochern. Biophys. Res. Cornrnun., 175,400, 1991.
69. Nilsson, L. and Nygard, 0., Altered sensitivity of eukaryotic elongation factor 2 for
trypsin after phosphorylation and ribosomal binding, J. Bioi. Chern., 266, 10578, 1991.
70. Sitikov, A. S., Simonenko, P. N., Shestakova, E. A., Ryazanov, A. G., and
Ovchinnikov, L. P., cAMP-dependent activation of protein synthesis correlates with
dephosphorylation of elongation factor 2, FEES Lett., 228, 327, 1988.
71. Redpath, N. T. and Proud, C. G., The tumor promoter okadaic acid inhibits reticulocyte-
lysate protein synthesis by increasing the net phosphorylation of elongation factor 2,
Biochern. J., 262, 69, 1989.
72. Gschwendt, M., Kittstein, W., Mieskes, G., and Marks, F., A type 2A protein
phosphatase dephosphorylates the elongation factor 2 and is stimulated by the phorbol
ester TPA in mouse epidermis in vivo, FEES Lett., 257, 357, 1989.
73. Ryazanov, A. G., Shestakova, E. A., and Natapov, P. G., Phosphorylation of elon-
gation factor 2 by EF-2 kinase affects rate of translation, Nature (London), 334, 170,
1988.
74. Ryazanov, A. G. and Davydova, E. K., Mechanism of elongation factor 2 (EF-2)
inactivation upon phosphorylation, FEES Lett., 251, 187, 1989.
75. Palfrey, H. C., Nairn, A. C., Muldoon, L. L., and Villereal, M. L., Rapid activation
of calmodulin-dependent protein kinase III in mitogen-stimulated human fibroblasts, J.
Bioi. Chern., 262, 9875, 1987.
76. Mackie, K. P., Nairn, A. C., Hampel, G., Lam, G., and Jaffe, E. A., Thrombin
and histamine stimulate the phosphorylation of elongation factor 2 in human umbilical
vein endothelial cells, J. Bioi. Chern., 264, 1748, 1989.
77. Nairn, A. C., Nichols, R. A., Brady, M. J., and Palfrey, H. C., Nerve growth factor
treatment or cAMP elevation reduces Ca2 + /calmodulin-dependent protein kinase III ac-
tivity in PC12 cells, J. Bioi. Chern., 262, 14265, 1987.
78. Brady, M. J., Nairn, A. C., Wagner, J. A., and Palfrey, H. C., Nerve growth factor-
induced down-regulation of calmodulin-dependent protein kinase-III in PC-12 cells in-
volves cyclic AMP-dependent protein kinase, J. Neurochern., 54, 1034, 1990.
79. Nygard, 0., Nilsson, A., Carlberg, U., Nilsson, L., and Amons, R., Phosphorylation
regulates the activity of the eEF-2-specific Ca2 +- and calmodulin-specific protein kinase
III, J. Bioi. Chern., 266. 16425, 1991.
80. Celis, J. E., Madsen, P., and Ryazanov, A. G., Increased phosphorylation of elongation
factor 2 during mitosis in transformed human amnion cells correlates with a decreased
rate of protein synthesis, Proc. Nat/. Acad. Sci. U.S.A., 87, 4231, 1990.
81. Severinov, K. V., Melnikova, E. G., and Ryazanov, A. G., Downregulation of the
translation elongation factor 2 kinase in Xenopus laevis occytes at the final stages of
oogenesis, The New Biologist, 2, 887, 1990.
82. Reed, J. C., Alpers, J. D., Nowell, P. C., and Hoover, R. G., Sequential expression
of protooncogenes during lectin-stimulated mitogenesis of normal human lymphocytes,
83. Greenberg, M. E., Hermanowski, A. L., and Ziff, E. B., Effect of protein synthesis
inhibitors on growth factor activation of c-fos, c-myc, and actin gene transcription, Mol.
Cell. Bioi., 6, 1050, 1986.
84. Lee, A. S., Coordinated regulation of a set of genes by glucose and calcium-ionophore
in mammalian cells, Trends Biochern. Sci., 12, 20, 1987.
85. Kim, Y. K., Kim, K. S., and Lee, A. S., Regulation of the glucose-regulated genes
by [3-mercaptoethanol requires de novo protein synthesis and correlates with inhibition
of protein glycosylation, J. Cell. Physioi., 133, 553, 1987.
86. Watowich, S. S. and Morimoto, R. 1., Complex regulation of heat shock- and glucose-
responsive genes in human cells, Mol. Cell. Bioi., 8, 393, 1988.
87. Sarnow, P., Translation of glucose-regulated protein 78/immunoglobulin heavy-chain

binding protein rnRNA is increased in poliovirus-infected cells at a time when cap-
dependent translation of cellular mRNAs is inhibited, Proc. Nat/. Acad. Sci. U.S.A., 86,
5795, 1989.
88. Dorner, A. J., Wasley, L. C., and Kaufman, R. J., Increased synthesis of secreted
proteins induced expression of glucose-regulated proteins in butyrate-treated Chinese
hamster ovary cells, J. Bioi. Chern., 264, 10602, 1989.
89. Hendershot, L. M., Immunoglobulin heavy chain binding protein complexes are dis-
sociated in vivo by light chain addition, J. Cell Bioi., Ill, 829, 1990.
90. Vogel, J. P., Misra, L. M., and Rose, M. D., Loss of BiP/GRP78 function blocks
translocation of secretory proteins in yeast, J. Cell Bioi., 110, 1885, 1990.
91. Nguyen, T. H., Law, D. T. S., and Williams, D. B., Binding protein BiP is required
for translocation of secretory proteins into the endoplasmic reticulum in Saccharomyces
cerevisiae, Proc. Nat!. Acad. Sci. U.S.A., 88, 1565, 1991.
92. Ting, J, and Lee, A. S., Human gene encoding the 78,000-dalton glucose-regulated
protein and its pseudogene: structure, conservation and regulation, DNA, 7, 275, 1988.
93. Li, X. and Lee, A. S., Competitive inhibition of a set of endoplasmic reticulum protein
genes (GRP78, GRP94, and ERp72) retards cell growth and lowers viability after ion-
ophore treatment, Mol. Cell. Bioi., 11,3446, 1991.
94. Normington, K., Kohno, K., Kozutsumi, Y., Gething, M.-J., and Sambrook, J., S.
cerevisiae encodes an essential protein homologous in sequence and function to mam-
malian BiP, Cell, 57, 1223, 1989.
95. Brostrom, M. A., Cade, C., Prostko, C. R., Gmitter-Yellen, D., and Brostrom,
C. 0., Accommodation of protein synthesis to chronic deprivation of intracellular se-
questered calcium, J. Bioi. Chern., 265, 20539, 1990.
96. Nakaki, T., Deans, R. J., and Lee, A. S., Enhanced transcription of the 78,000-dalton
glucose-regulated protein (GRP78) gene and association of GRP78 with immunoglobulin
light chains in a nonsecreting B-cell myeloma line (NS-1), Mol. Cell. Bioi.. 9, 2233,
1989.
97. Prostko, C. R., Brostrom, M. A., Galuska-Malara, E. M., and Brostrom, C. 0.,
Stimulation of GRP78 gene transcription by phorbol ester and cAMP in GH 3 pituitary
98. Fink, J. S., Verhave, M., Walton, K., Mandel, G., and Goodman, R. H., Cyclic
AMP and phorbol ester-induced transcriptional activation are mediated by the same
enhancer element in the human vasoactive intestinal polypeptide gene, J. Bioi. Chern.,
266, 3882, 1991.
99. Montminy, M. R., Gonzalez, G. A., and Yamamoto, K. R., Regulation of cAMP-
inducible genes by CREB, Trends Neurosci., 13, 184, 1990.
100. Resendez, E., Wooden, S. K., and Lee, A. S., Identification of highly conserved
regulatory domains and protein-binding sites in the promoters of the rat and human genes
encoding the stress inducible 78-kilodalton glucose-regulated protein, Mol. Cell. Bioi.,
8, 4579, 1988.
101. Macejak, D. G. and Sarnow, P., Internal initiation of translation mediated by the 5'
leader of a cellular rnRNA. Nature (London), 353, 90, 1991.
102. Schutzbach, J. S. and Forsee, W. T., Calcium ion activation of rabbit liver a! ,2-
mannosidase, J. Bioi. Chern., 265, 2546, 1990.
103. Lodish, H. F. and Kong, N., Perturbation of cellular calcium blocks exit of secretory
proteins from the rough endoplasmic reticulum, J. Bioi. Chern., 265, 10893, 1990.
104. Kuznetsov, G., Brostrom, M. A., and Brostrom, C. 0., Identification of a calcium
requirement for secretory protein processing and export. Differential effects of calcium
and dithiothreitol, J. Bioi. Chern., 267, 3932, 1992.
105. Balch, W. E., Biochemistry of interorganelle transport, J. Bioi. Chem., 264. 16965,
1989.
106. White, A., Handler, P., Smith, E. L., Hill, R. L., and Lehman, I. R., Principles of
Biochemistry, McGraw-Hill, New York, 1978, 136.
107. Wieland, F. T., Gleason, M. L., Serafini, T. A., and Rothman, J. E., The rate of
bulk flow from the endoplasmic reticulum to the cell surface, Cell, 50, 289, 1987.
Chapter 6
TISSUE-SPECIFIC REGULATION OF GLUCOKINASE
Mark A. Magnuson and Thomas L. Jetton
TABLE OF CONTENTS
l. Introduction ...................................................... 144
II. Hexokinase Gene Family ........................................ 145
Ill. Dual Functional Roles of Glucokinase ........................... 147

A. Role in the ~ Cell ........................................ 147
B. Role in the Liver ......................................... 149
IV. Cloning of Glucokinase .......................................... 150

A. Hepatic Glucokinase cDNAs ............................. 150
B. Islet Glucokinase cDNAs ................................. 150
V. Alternate Promoters in the Glucokinase Gene .................... 153

A. Downstream Promoter Regulation ........................ 154
I. Effects of Hormones on Hepatic
Glucokinase Gene Transcription .................. 155
B. Upstream Promoter Regulation ........................... 156
VI. Multiple Glucokinase Isoforms ................................... 157

A. Activity Differences of Glucokinase Isoforms ............ 158
B. Aberrant Glucokinase Isoforms in Liver and Islet ........ 159
C. Multiple Human Liver Glucokinase lsoforms ............. 160
D. Aberrant Splicing of Glucokinase mRNA
in Pituitary ............................................... 161
VII. Conclusions ...................................................... 162
References ............................................................... 164
0-849 3 -696!-4193/$0. 00 + $.50

© !993 by CRC Pre». Inc. 143
I. INTRODUCTION
Euglycemia is characterized by blood glucose concentrations in the 4 to

7 mM range and must be maintained in the face of constantly changing dietary
intake and metabolic needs. Deviations in the blood glucose concentration
on either side of this narrow range can lead to a myriad of acute and chronic
pathological states. While several tissues play roles that are involved in main-
taining euglycemia, the liver and pancreatic 13 cell are probably the most
important. The liver is a major site of glucose uptake during periods of
hyperglycemia and removes glucose from the circulation for conversion into
glycogen or other energy forms. However, during periods of hypoglycemia,
the liver produces glucose and releases it into the bloodstream. The pancreatic
13 cell, on the other hand, contributes to glucose homeostasis by sensing
changes in the plasma glucose concentration and adjusting the rate at which
insulin is synthesized and released. While the physiologic roles of the he-
patocyte and pancreatic 13 cell in the maintenance of euglycemia are clearly
different, both of these cell types share the ability to alter their utilization of
glucose as blood glucose concentrations either rise or fall. In the liver, the
ability to utilize glucose is primarily dependent upon the relative amounts of
circulating insulin and glucagon with contributory effects from other hormones
(e.g., glucocorticoids and thyroid hormones). As insulin levels rise, the liver
responds by a coordinate induction of enzymes mediating glycolysis and
glycogenesis, while at the same time those enzymes that favor glycogenolysis
and gluconeogenesis are inhibited. 1 Conversely, when insulin levels fall and
glucagon level rise, a coordinate induction of enzymes mediating gluconeo-
genesis and glycogenolysis occurs, thus allowing the liver to produce glucose.
In contrast, glycolysis in the 13 cell is regulated by glucose and not by insulin.
A common feature of both the liver and 13 cell that contributes to their
respective roles in maintaining glucose homeostasis is the expression of hex-
okinase type IV, more commonly known as glucokinase (ATP: D-hexose
6-phosphotransferase; E.C. 2. 7 .1.1). 2
In this chapter we discuss the findings of recent molecular genetic studies
that have helped to clarify aspects of the tissue-specific structure and regulation
of glucokinase. The physiologic importance of this enzyme in glucose ho-
meostasis suggests the need for a variety of control points for modulating
glucokinase gene expression. Indeed, studies in this field have identified
several mechanisms by which the enzyme can be expressed in both the he-
patocyte and 13 cell, yet can also be regulated differently in these cells. These
observations have provided a basis for understanding the tissue-specific func-
tion of the enzyme glucokinase and have provided a basis for future studies
in ihis field. The tissue-specific expression of an active glucokinase molecule
is achieved by regulation occurring at several steps in the transcription and
processing of the glucokinase mRNA. These include the use of tissue-specific
promoter regions in the glucokinase gene, the modulation of the rate of
Magnuson and Jetton 145
TABLE 1
Structural and Functional Features of Mammalian Hexokinases
Major sites
Hexokinase type Mass (kDa) Km for glucose G-6-P inhibition of expression
~too 0.04' Yes Brain, kidney

II ~ 100 0.13' Yes Muscle, fat
III ~wo 0.02b Yes Liver, kidney
IV (Glucokinase) ~so 5.0' No Liver, islet
Note: The mass, Km for glucose, inhibition by glucose-6-P, and major tissue sites of expression
are shown for hexokinase types I through IV.
' Hyperbolic kinetics.

b Inhibited by excess glucose.
Sigmoidal kinetics.
transcription by hormones in the liver, and by tissue-specific patterns of RNA

splicing. We will discuss what is known about each of these control mech-
anisms and how, together, they determine the tissue-specific regulation of
this enzyme.
II. HEXOKINASE GENE FAMILY

The phosphorylation of glucose and other simple sugars at the sixth
position is the first step in their metabolism. Since carbohydrates are nearly
a universal energy source, hexokinases may also be ubiquitous in their phy-
logenetic distribution as most organisms depend on glycolysis as the primary
pathway for glucose catabolism. The most thoroughly studied hexokinases
are those in mammals and yeast. Among mammals, four distinct hexokinases
have been identified, termed types I through IV (also A through D), that have
been defined by their order of elution from ion-exchange resins. 3 These iso-
zymes share certain structural and functional features and, thus, are thought
to be evolutionarily related. The distribution, kinetic, and structural features
of each is shown in Table I.
Glucokinase (hexokinase type IV) differs in several ways from hexoki-
nases types I through III. It is about half the size (~so vs. ~ 100 kDa), has
a much higher Km for glucose (~10 mM vs. ~20 to 130 rnM), is not subject
to end product inhibition by physiologic concentrations of G-6-P, and has
sigmoidal kinetics. The enzyme has been thought to be more specific for
glucose than are the other hexokinases, thus leading it to be termed "glu-
cokinase"; however, this distinction may be incorrect from an enzymologic
viewpoint. 4 Physiologically, glucose may be the most important substrate for
the enzyme, so the unique name remains appropriate. Nevertheless, when
taken together, these various characteristics of glucokinase suggest that the
enzyme is more similar to yeast hexokinase than it is to the mammalian

hexokinase types I through III.
The similar sizes of glucokinase and the yeast hexokinases led to the
suggestion that the hexokinases types I through III evolved by a gene dupli-
cation and tandem ligation event of a common ancestral gene similar to that
of the yeast hexokinases or glucokinase. 3 •5 •6 Comparison of the sequence of
glucokinase, 7 yeast hexokinase, 8 •9 and mammalian hexokinases 10- 12 has al-
lowed this notion to gain further acceptance, although a comparison of the
intron/exon structures of the glucokinase gene with a hexokinase gene has
not yet been reported. All of these hexokinases have now been shown to share
a high degree of amino acid sequence homology. Glucokimse is 33% identical
with yeast hexokinase and 53% identical with the carboxy terminal domain
of rat brain hexokinase J.7 When conservative substitutions are considered,
these similarities are even more striking. Conservation of primary structure
is especially evident in the putative glucose- and ATP-binding domains of
these molecules. 7 Indeed, similarity of a region of these hexokinases with the
ATP-binding sites in protein kinases led to the identification of a similar
domain in glucokinase. 7 In addition, several residues that were identified as
playing a role in glucose binding, as originally determined by X-ray crystal-
lographic analysis of yeast hexokinase, 13 are also conserved among the dif-
ferent hexokinases. 7· 14 Mutagenesis of one of these residues in hepatic glu-
cokinase (aspartate-205) decreases activity of the enzyme to 11500th of that
of the wild type enzyme thus establishing its importance in catalysis. 14 Lastly,
conservation of the carboxy- and amino-terminal halves of mammalian hex-
okinase type I has further reinforced the notion of gene evolution by dupli-
cation and tandem ligation. When the two halves of the protein were com-
pared, they were found to share over 50% primary sequence homology. 10 • 11
The distribution of the different mammalian hexokinases is tissue-spe-
cific, although all four hexokinase isoforms are present in rat liver. 3 ·4 Glu-
cokinase is the most restricted in distribution and has been found only in the
liver and in the pancreatic islet. 15 Interestingly, the tissue-specific expression
of the four different hexokinases roughly mirrors that of the different glucose
transporters, although there can be significant overlapping of expression in
certain tissues. For instance, both hepatocytes and pancreatic ~-cells express
a high Km glucose transporter termed GLUT-2. 16 This glucose transporter has
a Km for glucose of ~30 rnM, thus allowing the cytoplasm of hepatocytes
and ~-cells to rapidly equilibrate to plasma concentrations of glucose. Brain
and kidney, which express hexokinase type I, also express GLUT I, 17 a
glucose transporter with a very low Km for glucose, whereas striated muscle
and adipose tissue coexpress hexokinase II and the insulin regulatable GLUT-
4, 18 • 19 a transporter with an intermediate Km for glucose. The coexpression
of certain hexokinases with specific glucose transporters suggests they are
functionally coupled, although direct interactions have not been demonstrated.
III. DUAL FUNCTIONAL ROLES OF GLUCOKINASE
While glucokinase catalyzes the rate-limiting step in glycolysis in both

hepatocytes and f3 cells, thus showing the biochemical similarity of these
different cells at this regulatory control point, the enzyme is regulated dif-
ferentially in the hepatocyte and f3 cell. The tissue-specific regulation of
glucokinase is thought to allow the hepatocyte and pancreatic f3 cell to perform
their different physiologic functons in order to maintain glucose homeostasis.
A. ROLE IN THE J3 CELL

The pancreatic f3 cell responds to changes in glucose concentration and
constantly adjusts the rate of insulin secretion. Physiologic glucose concen-
trations monitored by the f3 cell are in the mM range, while the concentration
of most ligands that interact with cell surface receptors are in the nM or pM
range. In the absence of a surface receptor for glucose, the f3 cell senses
changes in the circulating concentration of glucose by a mechanism that
involves changes in the rate of glucose metabolism, 20 the so-called "fuel
hypothesis" of glucose sensing. 21 •22 After glucose enters the pancreatic f3 cell
via GLUT 2, the high Km glucose transporter, 23 ·24 its metabolism is initiated
by phosphorylation at the sixth position, a reaction catalyzed by both glu-
cokinase and at least one other hexokinase whose isotype is not known. 25 29
At low glucose concentrations, the lower Km hexokinase isotype contributes
substantially to glycolysis. However, as glucose concentrations rise, glucok-
inase becomes the predominant glucose phosphorylating enzyme in the f3
cell. 30
The unique kinetic characteristics of glucokinase allows the rate of glucose
metabolism to vary in proportion to glucose concentration. By recognizing
and modulating the rate of glucose phosphorylation, glucokinase acts as the
rate-limiting determinant for glucose usage by the f3 celJ.3° The high Km and
the lack of feedback inhibition by glucose-6-P are important kinetic parameters
of glucokinase that enable f3 cell glucose responsiveness. Additionally, with
the half-maximal saturation of glucokinase near the middle of the physiologic
range of glucose fluctuations and its sigmoidal kinetics, the enzyme is ideally
suited for the role of controlling the metabolism of glucose by the f3 cell.
Since the Km of GLUT 2 is ~30 mM, glucose transport into the f3 cell is not
thought to be rate-limiting except, perhaps, in certain pathologic states. 31 The
role of glucokinase as the so-called glucose sensor for the f3 cell, first proposed
in 1968 by Matshinsky and Ellerman, 20 is now widely accepted.
The metabolism of glucose by the f3 cell is, in tum, tightly linked to the
rate of insulin secretion by a series of intermediate steps as shown in Figure
1. Glycolysis is thought to result in an increase in the ATP/ ADP ratio, thereby
inhibiting an ATP-sensitive K + channel which results in a decrease of K +
efflux, the effect of which is to depolarize the membrane of the f3 cell. 32 ·33
The change in membrane potential is then thought to activate one or more
00
Voltage-sens itive ATP-sensit ive
-""
ca+-+channel K+Channel
Glycogen Pyruvate ~ Insulin 9

(+)'[Ca++l
(+){
Glucokinase
\ /c High Km
Glucose-6 -P • Glucose ~Glucose Glucokinase
HighKm
\.. (-) Hexokinase 2 GUuT 2 (+)glucose
~ LowKm
HEPATOCYTE BETA CELL ~

~·(3•
FIGURE 1. Different regulation and functional roles of glucokinase in the hepatocyte and p cell. The different roles of hepatic and p cell glucokinase in ::s
::::,
maintaining glucose homeostasis are diagramatically illustrated. Both of these cell types express GLUT-2, a high Krn glucose transporter, and glucokinase, a high
K"' hexokinase. In the liver, glucokinase serves as an insulin-inducible determinant of glucose usage, whereas in the p cell, the enzyme functions as the so-called ~
"glucose sensor". Glucose metabolism in the P cell is thought to be linked to insulin secretion through several intermediate steps, as illustrated. C'l
">
::s
">
~
"';::i
~
"'(3•
"'::s
voltage-dependent Ca2 + channels leading to a sharp increase in the intracel-

lular Ca 2 + concentration and stimulating the mechanisms involved in the
exocytosis of insulin from storage granules. 34
Glucokinase levels in the 13 cell are inducible by glucose, thus permitting
glucose to serve as a positive mediator of this process. 29 However, for glu-
cokinase to determine the rate of glucose usage by the 13 cell, and act as the
so-called "glucose sensor", 30 it must not be affected by differing concentra-
tions of insulin, particularly at the source where the concentrations are highest.
Instead, the enzyme must function as a stable, insulin-independent determinate
for glucose usage by the 13 cell. Indeed, in cultured islets, insulin does not
affect glucokinase activity. 29 The mechanism by which glucose modulates
glucokinase activity in the 13 cell is not known; however, glucokinase activity
in cultured islets can vary as much as fivefold in 7-day cultured islets in either
3- or 30-rnM glucose. 29
B. ROLE IN THE LIVER

In liver, glucokinase is regulated differently than in the islet, thus re-
flecting the different physiologic role of the enzyme in the hepatocyte. The
entry and utilization of glucose by the liver, in so far as the most proximal
steps are concerned, is similar to that in the 13 cell as shown in Figure I.
Hepatocytes also express GLUT-2, the same high Km glucose transporter as
the 13 cell, so the entry of glucose into the hepatocyte is not rate-limiting.
During periods of hyperglycemia and rising insulin levels, hepatic glucokinase
activity is induced. 35 The distinct kinetic features of glucokinase permit the
phosphorylation of glucose to proceed at a rate proportional to the concen-
tration of glucose, as in the 13 cell. Hexokinases I through III, because they
are inhibited by glucose-6-P and have lower Michaelis' constants, do not
contribute to glucose phosphorylation at the higher (mM) concentrations en-
countered in the portal vein after a meal. Glucokinase in the liver, therefore,
serves as an insulin-sensitive determinant for glucose entry into the hepatocyte,
essentially trapping glucose and thus maintaining a gradient for its entry during
periods of hyperglycemia. Glucose transport can proceed in either direction
depending upon its concentration gradient, so GLUT 2 gene expression need
not be regulated. 1 During periods of fasting or starvation, plasma insulin
levels decline and glucagon secretion by pancreatic a cells increases. Con-
comitantly, hepatic glucokinase activity falls while glucose-6-phosphatase
levels rise, thus allowing glucose to diffuse out of the liver into the blood-
stream.1
Thus, in the liver, glucokinase acts as an insulin-dependent modulator of
glucose utilization, whereas in the 13 cell, glucokinase acts as a glucose-
dependent glucose sensor. Indeed, the differential regulation of glucokinase
activity in the liver and 13 cell by insulin and glucose, respectively, may permit
feedback regulation of glucokinase activity. 3638
IV. CLONING OF GLUCOKINASE
In order to understand the mechanisms underlying the tissue-specific

regulation of glucokinase, the cloning of a eDNA for the enzyme was required.
Since glucokinase is expressed at a higher level in liver than in the islet,
particularly after refeeding or insulin treatment, and liver was a more available
source of tissue, the cloning of hepatic glucokinase was accomplished first.
A. HEPATIC GLUCOKINASE cDNAs

Hepatic glucokinase cDNAs were cloned initially by two different groups.
Iynedjian et a!. 39 utilized an antibody to hepatic glucokinase to screen a rat
liver eDNA expression library made from 21S hepatic RNA and was the first
to report a eDNA for the enzyme. Andreone et al. 7 obtained a eDNA by a
route that involved purifying glucokinase to homogeneity, obtaining amino
acid sequence information, then developing synthetic oligonucleotides that
were used to screen a rat liver eDNA library. Both groups used the glucokinase
cDNAs obtained to examine the regulation hepatic glucokinase mRNA and
showed that it was markedly induced by both insulin administration to diabetic
animals and during the fasted-to-fed transition. Iynedjian eta!. 39 also examined
the developmental expression of glucokinase mRNA and found it to be induced
at weaning, thus showing that expression of glucokinase mRNA coincides
with that of hepatic glucokinase activity.
The sequence of hepatic glucokinase was deduced by Andreone et a!. 7
who showed that the enzyme was 465 amino acids in length and was ho-
mologous to both yeast hexokinases and rat hexokinase I. A map of a full-
length hepatic glucokinase eDNA (GK.Z2) is shown in Figure 2, and the
amino acid sequence of the enzyme is shown in Figure 3. Within the coding
sequence of the enzyme, the putative glucose- and ATP-binding domains were
identified based on their similarity to sequences in other hexokinases and
protein kinases. While the identity of the eDNA cloned by Andreone et al. 7
as coding for glucokinase was established by finding the sequence of multiple
peptides within the reading frame deduced from the eDNA, additional proof
was obtained by the expression of a hepatic eDNA in bacteria and measure-
ment of the kinetic properties of the product that positively identified the
eDNA as, in fact, coding for glucokinase. 40
B. ISLET GLUCOKINASE cDNAs

The cloning of hepatic glucokinase permitted studies of islet glucokinase
gene expression to be initiated. Previous biochemical and immunological
studies that compared hepatic and islet glucokinase found them to be indis-
tinguishable by several criteria, e.g., kinetic characteristics, mass, charge,
and immunologic cross-reactivity. Thus, it was thought that the islet isoform
was very similar, if not identical, to the hepatic isoform. 15 Further support
for a single form of glucokinase was provided by Southern blot analysis of
GK.Z2 - - -:::J -:::J -..r::. -..r::. - X ~
:::J:: (ij l:l
>UJ > > a. a. Cl ( /J
(Hepatic) a.. (XJ OQ
a.. a.. OJ a.. en en (XJ
- ;:s
:;::
2347 bp IL I I _ _L I I
§""
l:l
;:s
~
<.....,
~
~ : - :::J
X
GK.Z9 :::J:: :::J ..r::. ..r::. ~
>UJ (ij > > a. a. Cl Ci) ;:s
(XJ a.. a.. en (XJ (XJ
(lnsulinoma) a.. a.. \l en
2216 bp
2500 2000 1500 1000 500 Obp
Putative ATP Putative glucose Islet-specific Uver-specific Sequences affected

II binding domain ~ binding domain D sequence ~ sequence II by alternate RNA splicing
FIGURE 2. Comparison of glucokinase cDNAs from rat liver and insulinoma tissues. The structural differences of glucokinase cDNAs isolated from different
tissue sources is illustrated. The GK.Z2 eDNA clone was isolated from a rat liver eDNA library and the GK.Z9 eDNA clone was isolated from a rat insulinoma
eDNA library. Differences in the cDNAs at the 5' end and within the coding sequence of the enzyme are indicated. The differences coincide with the location
of splice junctions indicating that alternate RNA splicing is involved in generating glucokinase isoform diversity.
....
....Ul
Rat islet glucokinase:

465 amino acids
52,087 kDa
pl5.08
MLDDRARMEATKKEKVEQILAEFQLQEEDLKKVMSRMQKEMDRGLRLETH
EEASVKMLPTYVRSTPEGSEVGDFLSLDLGGTNFRVMLVKVGEGEAGQWS
VKTKHQMYSIPEDAMTGTAEMLFDYISECISDLLDKHQMKHKKLPLGFTF
SFPVRHEDLDKGILLNWTKGFKASGAEGNNIVGLLRDAIKRRGDFEMDVV
AMVNDTVATMISCYYEDRQCEVGMIVGTGCNACYMEEMQNVELVEGDEGR
MCVNTEWGAFGDSGELDEFLLEYDRMVDESSANPGQQLYEKIIGGKYMGE
LVRLVLLKLVDENLLFHGEASEQLRTRGAFETRFVSQVESDSGDRKQIHN
ILSTLGLRPSVTDCDIVRRACESVSTRAAHMCSAGLAGVINRMRESRSED
VMRITVGVDGSVYKLHPSFKERFHASVRRLTPNCEITFIESEEGSGRGAA
LVSAVACKKACMLAQ
Rat hepatic glucokinase:

465 amino acids
51, 924 kDa
pl5.01
MAMDTTRCGAOLLTL ...
FIGURE 3. Cell type-specific sequences of the islet and liver isoforrns of glucokinase. The
different amino acids sequences of the islet (top) and hepatic (bottom) glucokinase isoforrns are
shown (underlined).
rat genomic DNA that showed there was only a single gene for glucokinase
in the rat. 35 .4 1 However, surprisingly, when the liver-derived eDNA was used
as a probe in RNA blot experiments of islet RNA, both Magnuson and
Shelton41 and Iynedjian et a!. 35 found that the pancreatic islet form of rat
glucokinase mRNA was longer by 200 to 400 bases than that in liver. More-
over, Iynedjian et a!. 35 found that the effects of a fasting/refeeding regimen
on hepatic and islet glucokinase mRNA and protein levels were also different.
While the amount of hepatic mRNA and immunoreactive protein varied with
fasting and refeeding, islet glucokinase mRNA and immunodetectable protein
did not change under these conditions. 35 To determine the structural basis for
the different sizes of the mRNAs, and to explore the basis for the differential
regulation, cDNAs for islet glucokinase mRNA were cloned by both groups.
Initial studies identifying structural differences in glucokinase isoforms
yielded intriguing results. The glucokinase cDNAs cloned by Magnuson and
Shelton41 were obtained from a rat insulinoma eDNA library, whereas the
cDNAs obtained by Hughes eta!. 42 were cloned from a rat islet eDNA library.
Both groups determined that glucokinase transcripts from islet tissue and
hepatocytes were not equivalent and that this difference was located at the
5 1 end of the mRNAs. The nature of these differences suggested that additional
mechanisms, such as alternate RNA splicing and/or the utilization of distinct
promoter regions, were involved in producing tissue-specific glucokinase
mRNAs. Indeed, when the insulinoma and hepatic cDNAs for glucokinase
were compared by Magnuson and Shelton, 41 two differences were found that
were due to both the use of alternate promoter regions and to alternate RNA
processing, as shown in Figure 2. The sequences at the 5 1 ends of the cDNAs
were distinct and the site at which this difference occurred coincided with
the splice junction between the first and second exons, thus indicating that
the hepatic and islet glucokinase mRNAs utilized different first exon se-
quences in the liver and islet, respectively. This finding implicated the use
of different promoter regions in the production of glucokinase mRNAs in the
different cell types. Moreover, since the translation initiation codons for liver
and islet glucokinase were located in different first exons, the amino acid
sequence at the amino terminus of the enzyme from liver and pancreatic islets
was affected, as previously shown in Figure 3. In addition to sequence dif-
ferences related to the use of alternate promoter regions, there were also
sequence differences within the reading frame of the enzyme in the GK.Z9
eDNA clone, thus implicating alternate RNA processing in second region.
Together, the differences in the hepatic and islet glucokinase cDNAs indicated
that variant glucokinase isoforms were being produced as a result of alternate
RNA splicing of the glucokinase gene product.
V. ALTERNATE PROMOTERS IN THE

GLUCOKINASE GENE
The transcription unit that produces the hepatic glucokinase mRNA was
characterized by Magnuson et al. 43 and found to contain 10 exons that span
about 15 kb. The location of the f3 cell first exon, however, was not present
in the genomic DNA fragments characterized in the cloning of the hepatic
gene despite having 7.5 kb of 5 1 flanking DNA sequence in the lambda phage
clone studied (A.GK5). To determine the location of the genomic DNA se-
quences that gave rise to the 5 1 noncoding sequence in the islet glucokinase
mRNA, another genomic DNA fragment was isolated. 41 This new DNA
fragment (A.GK7) did not overlap with the 5 1 end of A.GK5, the clone con-
taining the liver-specific first exon and promoter region, thus indicating that
the upstream promoter region had to be at least 12 kb upstream from the
downstream promoter region. Although the exact distance separating the dif-
ferent promoter regions has not yet been determined, an upper limit for this
distance has been suggested by the cloning of cosmid DNAs for the gene. 44
Both promoter regions are contained in several cosmid clones that have an
insert size of approximately 40 kb, thus placing an upper limit on the distance.
However, since these clones have not been fully mapped, the actual distance
Beta Cell and

Pituitary mRNAs ~
1--1
t 1 kb
~
18
(>12kb)
:I'
•
1L
Ill I
2a 2 3
t
4
II
56
~
I I I.
789 10
= Beta cell and Pituitary-specific
• = Liver-specific
Hepatic mRNAs I =Common
FIGURE 4. Glucokinase gene structure. The intronlexon structure of the glucokinase gene is
diagramatically illustrated. The gene has 2 different promoter regions that are separated by at
least 12 kb and which are linked to cell type-specific first exon sequences. The upstream promoter
region is active in the pancreatic [3 cell and pituitary, whereas the downstream promoter region
is active in the liver. In addition, the location of a cassette exon (2A) is shown which is utilized
in the production of about 5% of the glucokinase mRNA present in liver 59 (From Magnuson,
M. A., J. Cell. Biochem., 48, 115, 1992. With permission.)
is not known and may, in fact, be considerably less. A map of the glucokinase
gene, as currently understood, is shown in Figure 4. The gene spans over 27
kb and contains 12 different exons that are used in a tissue-specific manner.
Among the glucokinase isoforrns that are functional, only 10 exons are utilized
with the complete sequence of exons 2 through 10 being necessary for catalytic
activity.
The tissue-specific expression of the upstream and downstream promoter
regions has been tested by RNA-PCR amplification experiments. In these
studies, the glucokinase mRNAs containing exon 1~ sequences have been
detected only in the islet and pituitary, while sequences containing the 1H
sequences have only been identified in the liver. 41 •45
A. DOWNSTREAM PROMOTER REGULATION

The location of the downstream (hepatic) promoter region was established
by using primer extension and S 1 nuclease assays to map the start sites of
transcription. In liver, transcription begins at a site 127 bases upstream from
the translation initiation codon 43 and occurs over a 4 to 5 bp region, perhaps
related to nonoptimal TATA homology (TATIT at - 29 to - 25 relative to
the + 1 transcription start site). Mapping of the start site of transcription of
the hepatic glucokinase mRNA has also been reported by Noguchi et al. 46
with placement of the strongest start site one base upstream of that reported
by Magnuson et al. 43 The sequence of approximately 1448 bp of DNA up-
stream from the transcription start site has been determined. An analysis of
this sequence for potential regulatory elements has revealed several sequences
TABLE 2
Differential Regulation of Hepatic and
13 Cell Glucokinase
Stimulatory effect Inhibitory effect
Hepatic glucokinase: Insulin' cAMP'

Biotin'
Triiodothyronine'
13 Cell glucokinase: Glucose Exercise training
' Regulation at the level of gene transcription has been estab-

lished.
From Magnuson, M. A., J. Cell. Biochem., 48, 115, 1992. With

permission.
that might be important in liver-specific transcriptional activity. 43 A possible

binding site for the liver-specific transcription factors HNF-1 is present at
- 171 to - 166. 43 In addition, there are 3 tandem repeats of the binding site
for liver factor A1 at - 69 to -53 and 2 single sites further upstream. 43
Whether these sites play a role in directing expression of glucokinase to the
liver has not been determined but awaits development of a system by which
the function of these elements can be tested.
1. Effects of Hormones on Hepatic Glucokinase

Gene Transcription
The effects of hormones and other agents on hepatic glucokinase appears
to be controlled primarily at the level of gene transcription. Insulin, 43 .47 tri-
iodothyronine,48.49 and biotin50 stimulate gene transcription while glucagon
(acting via cAMP) is inhibitory, 51 as shown in Table 2. In diabetic rats, insulin
increases transcription at least 20-fold within 45 min. 47 Similarly, an inductive
effect of insulin on glucokinase mRNA can be seen in freshly cultured primary
hepatocytes. Comparison of the potency of insulin, proinsulin, and insulin-
like growth factor I indicates that the induction occurs through the insulin
receptor. 5 1 Moreover, the rapid inductive effect of insulin on hepatic gluco-
kinase mRNA suggests that the effect is direct and involves preexisting cellular
proteins. In addition, glucose does not alter the effect of insulin on glucokinase
gene transcriptionY The addition of cAMP, even in the presence of insulin,
has an inhibitory action on transcription, completely eliminating RNA exten-
sion within 30 min of addition to cultured hepatocytes. 51 In addition, cAMP
appears to accelerate decay of glucokinase mRNA, decreasing the half-life
of the mRNA from -300 min to only 35 min. 5 1
Biotin, a water soluble vitamin, increases glucokinase gene transcription
but the effect is short lived. In starved rats, biotin causes a 20-fold increase
within 1 h, but this peak decays rapidly. 50 Triiodothyronine (T3) also stimulates
transcription of the gene in liver, but the effect is thought to be permissive

since it is only observed during refeeding and not during starvation. 48 The
effect of T3 is more pronounced in hepatocytes from neonatal rats where it
causes a dose-dependent increase of glucokinase mRNA of a magnitude sim-
ilar to that seen with insulin. 49
The cis-regulatory elements that mediate the hormonal effects on tran-
scription from the hepatic glucokinase promoter region have not been iden-
tified. The absence of hepatoma cell lines that express the gene has hampered
studies in this direction. However, even when fusion genes have been trans-
fected into primary hepatocytes, no hormonal effects have been observed.
For instance, fusion gene experiments using DNA fragments up to -5.5 kb
from the start site in liver, fused to the chloramphenicol acetyltransferase
gene, have failed to show evidence for regulation by insulin when transfected
into primary hepatocytes. 46 In addition, the signal obtained in the primary
hepatocytes was small, suggesting that these constructs also lack liver-specific
enhancer elements. 46 Thus it is not yet known where the hormone regulatory
elements reside in the glucokinase gene. However, since the gene is positively
regulated by insulin, T3 , and biotin, response elements for these agents should
be located somewhere within the regulatory region of this gene. In addition,
the mechanism by which cAMP inhibits gene transcription is unexplained but
presumably is also acting through a response element in the hepatic promoter
region of the gene.
B. UPSTREAM PROMOTER REGULATION

The upstream promoter region of the glucokinase gene is both spatially
separate and functionally distinct from the downstream promoter region. Var-
ious structural features also set the two promoter regions apart. First, the
pattern of transcription initiation from the upstream promoter region is distinct.
While transcription initiation from the downstream region occurs over a 4 to
5 bp region, transcription initiation from the upstream region is even more
diffuse. Using RNA isolated from solid rat insulinoma tumors in both RNAse
protection and primer extension assays, transcription initiation was mapped
to a 60 to 65 bp region that included multiple start sites. This wide region of
transcripton initiation may be explained by the absence of any TATA ho-
mology in the proximal promoter region. In this regard, the transcription
initiation pattern is similar to that of other genes that lack a TATA box.
Second, the sequences of the two different DNA regions are distinct. DNA
sequence information up to 2.3 kilobase pairs 5' of the transcription initiation
sites has been obtained. Inspection of this DNA sequence shows no detectable
homologies with the downstream promoter region. However, examination of
the upstream promoter sequence revealed a region that was similar at 7 of 8 bp
to an element that occurs in the promoters of the insulin genes. 41 Subsequent
studies have shown that the regulation of the glucokinase promoter is distinct
from that of the insulin genes. 52 Indeed, since the glucokinase gene is also
expressed in the pituitary, whereas the insulin gene is not, a different set of
regulatory factors must be utilized to direct this different pattern of tissue
expression. With the exception of the pituitary, where the upstream promoter
region is utilized for mRNA expression, no other tissue has been found to
express this gene.
Fusion gene experiments have revealed that a small fragment of the
upstream promoter region is sufficient for expression in cultured ~ cell lines.
Indeed, a 294-bp fragment of DNA containing this region appears to be all
that is necessary for specific expression in~ cells. 52 Within this DNA fragment
are several novel elements that appear to contribute to transcription in ~ cells. 52
Moreover, in the islets of euglycemic rats, glucokinase appears to be expressed
specifically in the ~ cell as studied by immunohistochemical analysis. 53 In-
terestingly, glucokinase immunoreactivity among different pancreatic ~cells
is not uniform. Some ~ cells display striking amounts of glucokinase im-
munoreactivity while others have much less. 53 Evidence is accumulating in
support of the functional heterogeneity of pancreatic ~ cells although the basis
for this heterogeneity is not known. 54 •55 However, the observation of variable
glucokinase immunoreactivity itself is important since different amounts of
glucokinase activity in different ~ cells may cause the glucose sensitivity of
individual cells to be different.
Hormonal effects on the activity of the upstream glucokinase promoter
region have not been observed; however, this has been more difficult to study
given the lower abundance of islet glucokinase and the general difficulties
encountered in isolating pancreatic islets. However, there is evidence that
exercise training has a pronounced effect on islet glucokinase mRNA levels.
Koranyi et al. 56 determined that rats which had undergone an exercise training
regimen have decreased expression of both proinsulin and glucokinase mRNAs.
Whether the change in the amount of glucokinase mRNA in response to
exercise is due to a change in the rate of gene transcription or due to an
alteration in the stability of the mRNA has not been determined.
Regulation of glucokinase expression in the islet may not necessarily be
occurring at the level of gene transcription. For instance, the different 5'
noncoding sequences in the hepatic and islet glucokinase mRNAs might allow
for regulation at the translational level. The length of the leader sequence is
different in the liver and islet, which might affect the translational efficiencies
of the mRNAs. Moreover, the islet mRNA has two ATGs upstream of the
actual initiator ATG. The upstream ATGs have poor Kozac consensus se-
quences but may act to diminish translational efficiency of the RNA. 41
VI. MULTIPLE GLUCOKINASE ISOFORMS
The existence of different glucokinase isoforms, established by the clon-

ing and comparison of hepatic and islet glucokinase cDNAs, raised questions
about whether the structural differences had any functional consequences.
The cell type-specific isoforms of glucokinase, produced by the different

translation initiation codons encoded within the different first exons in the
glucokinase gene, have different amino acid sequences at their amino terminal
ends. Interestingly, though, both the islet and hepatic isoforms of glucokinase
are still exactly the same length despite having a portion of their sequence
encoded by different first exons. Only 3 of the 15 amino terminal amino acids
are conserved in both the liver and islet glucokinase isoforms, as shown in
Figure 3. While it has not been established whether the amino acids at the
N terminus of the different glucokinase isoforms are of any functional im-
portance, such differences could conceivably affect either the kinetics of the
enzyme or its partitioning into distinct subcellular domains. 37•57 For instance,
the amino terminal sequences of hexokinase I enable it to bind to the outer
mitochondrial membrane protein by interactions with a voltage-dependent
anion channel called porin. 58 However, despite differences at the amino ter-
minus in the liver and islet glucokinase isoforms, direct immunolocalization
studies using different antibodies to the enzyme indicate that the enzyme is
located in the cytoplasm of both the hepatocyte and pancreatic 13 cell, 53 as
has been previously thought.
In addition to the tissue-specific glucokinase isoforms related to the al-
ternate first exons in the glucokinase gene, other isoforms are produced by
additional alternate RNA splicing events involving the glucokinase pre-mRNA.
One such example of this is the glucokinase eDNA cloned from insulinoma
tissue. 41 The eDNA isolated was missing 51 bp in the portion coded by exon
4 due to the use of an alternate splice acceptor site. The cognate RNA encodes
a glucokinase isoform that has an in-frame deletion of 17 amino acids. This
finding generated two questions that needed to be answered. First, does this
isoform have different kinetic features, and secondly, if it does, was this a
potential means of regulating the activity of glucokinase? The impact of the
alternate splicing on glucokinase activity became apparent in experiments that
expressed the different isoforms.
A. ACTIVITY DIFFERENCES OF GLUCOKINASE ISOFORMS

Starting with glucokinase cDNAs isolated from rat liver (encoding the
isoform termed Ll) and a eDNA isolated from insulinoma (encoding the
isoform termed B2) it was possible to make chimeric cDNAs that encoded
two additional isoforms (termed L2 and Bl) by digesting and ligating the
different fragments obtained via a Pst I site in the eDNA. The use of the four
different cDNAs made it possible to test the effect of the different amino
terminal sequences (Ll compared with B 1) and test the effect of the deletion
due to the use of an alternate splice acceptor site in exon 4 (e.g., L1 vs. L2
and Bl vs. B2). The coding regions of these glucokinase isoforms were
inserted into a eukaryotic expression vector that utilized the strong cytomeg-
alovirus intermediate promoter to drive expression of these cDNAs in NIH-
3T3 cells. The activity and kinetics of the expressed glucokinase activities
:
Iso form 17a.a.
Location Km Vmax
L1 j.c Liver 22.3 mM 3.9 mol/kg
~
DNA!h
L2 j.c Liver ND ND
81
82
1------1
100 a.a.
: •
~
Liver-specific
sequences II
Islet-specific
sequences
~c
j.c
o
Islet
Islet
Common
sequences
15.5
ND
1111
Sequences
10.5
ND
deleted
by atternate RNA
splicing
FIGURE 5. Kinetics analysis of four different rat glucokinase isoforms produced by alternate
RNA splicing. The coding nucleotide sequences for each of four different glucokinase isoforms
were cloned into the eukaryotic expression vector pCMV4 and transfected into NIH-3T3 cells.
The Km and Vmu for each isoform was determined measuring glucose phosphorylation at different
glucose concentrations. 45 An endogenous low Km hexokinase activity was present in the NIH-
3T3 cells but did not affect the activity determinations for glucokinase. ND = not detected.
were then determined. 45 The results obtained (shown with the corresponding
isoform in Figure 5) indicated that both the L1 and B 1 isoforms were en-
zymatically active while the L2 and B2 isoforms were inactive. The Kms of
the enzymatically active L1 and B 1 isoforms were comparable, but a more
substantial difference was observed in the vmaxs of the enzyme expressed in
this manner. However, the differences in the Vmax of the L1 and B 1 isoforms
could be due to the expression of different amounts of protein since the
mRNAs encoding these isoforms have different sequences surrounding their
translation initiation codons. The differences in the Vmax and Km of the L1
and B 1 isoforms were small compared to the effect of the alternate splice
acceptor site in the 4th exon of the gene. Selection of this splice site resulted
in a glucokinase isoform without apparent catalytic activity. This was an
important findng since the generation of aberrant glucokinase isoforms that
were not functional could diminish glucose phosphorylation and glycolysis
in the 13 cell and liver, thereby affecting glucose sensing and glucose usage,
respectively.
B. ABERRANT GLUCOKINASE ISOFORMS IN

LIVER AND ISLET
The identification of aberrant glucokinase isoforms due to the use of an
alternate splice acceptor site in the 4th exon of the gene was only the first
example of such events (see summary in Table 3). Subsequently, another
aberrant glucokinase isoform, produced by splicing from an alternate donor
site in the 2nd exon of the gene, has been identified. Hayzer and Iynedjian59
cloned a glucokinase eDNA from a rat liver library that inserts sequence from
TABLE 3
Alternate RNA Splicing Events Involved in the Production
of Glucokinase mRNA that Have a Deleterious
Effect on Enzyme Activity
Location protein Tissues No. nucleotides

in gene (a.a.) involved added or deleted Frameshift Size of product
Exon 2A Liver 1 + 151 Yes 498

Exon 2 Pituitary 2 -25 Yes 68
Exon 2 Pituitary -52 Yes 58
Liver 1
Exon 4 13 cells -51 No 448
Liver
Pituitary 2
Note: The origin of each splicing variant is discussed in the text.
1
or 2
indicates that these two alternative splice sites were utilized in a single mRNA.
From Magnuson, M.A., J. Cell. Biochem., 48, 115, 1992. With permission.
a cassette exon (termed 2A) immediately before the sequence provided by

exon 2. This exon is used in approximately 5% of the glucokinase mRNA
present in liver. The use of the cassette exon, which adds 151 bases, occurred
in conjunction with the use of an alternate splice donor site in exon 2 which
removes 52 bases from the mRNA. Together, the insertion/deletion events
add 87 new residues to the reading frame of the enzyme resulting in a new
isoform of 498 amino acids. When this isoform was expressed in bacteria it
did not phosphorylate glucose. 60
C. MULTIPLE HUMAN LIVER GLUCOKINASE ISOFORMS

Recently,Tanizawa et al. 61 have cloned two human liver glucokinase cDNAs
and found that a similar cassette exon is utilized to produce two different
hepatic glucokinase isoforms, both of which are enzymatically active. The
human liver glucokinase isoform designated hLGLK1 6 ' has 97% homology
with rat liver glucokinase, although the 14 N terminal residues are entirely
different due to the use of a different translation initiation codon in the cassette
exon sequences. Although the insertion of a cassette exon at the junction of
exons 1 and 2 is similar to the situation described in rat liver, 59 the size of
the exon is different, 124 bp, and its sequence is not homologous to the rat
exon 2A. The human liver isoform hLGLK2, which is produced from a mRNA
lacking the cassette exon, is 98% homologous to rat liver glucokinase 6 ' and
has sequences at the amino terminal end which are very similar to those in
rat. The relative abundance of the two different human glucokinase isoforms
has not been determined.
D. ABERRANT SPLICING OF GLUCOKINASE mRNA

IN PITUITARY
The functional significance of aberrant glucokinase mRNAs in the islet
and liver is unresolved. However, the finding of aberrant glucokinase mRNAs
in pituitary may offer an explanation why such forms are found in the liver
and islet. Historically, glucokinase activity has been identified only in the
liver and in the pancreatic islet. Therefore, it was a surprise when Newgard
and his colleagues found that AtT-20 cells, a mouse corticotroph-derived cell
line, were a rich source of glucokinase mRNA. 42 This observation raised the
issue of whether or not glucokinase mRNA and glucokinase activity were
present in pituitary. Studies reported by two groups have addressed somewhat
different aspects of this question.
Hughes et al. 42 analyzed anterior pituitary RNA for the presence of a
glucokinase mRNA by Northern blot analysis and found a transcript that was
of similar size to that in islet. They then amplified glucokinase cDNAs from
AtT-20 cells, a corticotroph-derived cell line, and pituitary and determined
the sequence of both. The glucokinase cDNAs obtained in their experiments
showed that alternate RNA splicing of the glucokinase mRNA was occurring
in the pituitary that utilized the alternate splice donor site in exon 2 which
was described first by Hayzer and Iynedjian. 59 Of three AtT-20 cell PCR-
derived clones studied, two utilized the alternate acceptor site while one did
not. However, of three pituitary clones examined, all utilized the alternate
acceptor site. The alternate splice donor site in exon 2, when used by itself,
shifts the reading frame of the enzyme and predicts a peptide of only 58
amino acids, far too small to be catalytically active.
Liang et al. 45 addressed the question of glucokinase mRNA and activity
in AtT-20 cells and pituitary both by measuring glucokinase activity and
amplifying glucokinase mRNA from pituitary. In their studies no glucokinase
activity could be detected in either AtT-20 cells or pituitary, even after partial
purification by a chromatographic purification step. Amplification of a pi-
tuitary glucokinase eDNA and determination of its sequence showed a dif-
ferent pattern of alternate splice sites as shown in Figure 6. Of 3 PCR-derived
eDNA clones that were sequenced, all showed the use of an alternate splice
donor site in exon 2 that removes 25 bases and shifts the reading frame. In
addition, the same clones all used the same alternate splice acceptor site in
exon 4 that was identified previously as occurring in rat insulinoma tissue.
The mRNA cognate to this eDNA clone, if translated, would produce only
a 68 amino acid peptide. The absence of glucokinase in pituitary has also
been addressed by immunohistochemical staining of pituitary for the protein
using antibodies that detect the enzyme in liver and islet. In these studies,
no immunoreactive glucokinase was detected. 62
While the actual alternate splice sites identified by the different groups
who have performed these studies differ, in no case has a glucokinase mRNA
encoding a normal glucokinase isoform been identified from pituitary tissue.
25 bp deletion due 51 bp deletion due

to alternate donor to alternate acceptor
in Exon 2 in Exon 4
2/3 3/4
llllnnHnmiiHNHIIINIIIIIHIIIIIIIIIIJII••·
~ Putative ATP-binding domain
68 amino acid ~ Putative glucose-binding domain

open reading frame
1111111111 Exon 1 B sequences
*Transcription initiates from upstream promoter
FIGURE 6. Pituitary glucokinase eDNA: alternate RNA splicing disrupts the reading frame
for glucokinase. Glucokinase cDNAs from rat pituitary tissue were obtained by RNA-PCR as
described previously. 45 Two different alternate splicing events were identified as illustrated.
Either alternate splicing event, by itself, would be sufficient to prevent the production of an
active glucokinase isoform.
However, the situation with the radiation-induced AtT-20 cells is more con-
fusing but may simply reflect the phenotypic divergence of this cell line into
different sublines.
VII. CONCLUSIONS
Biochemical studies over at least the last 25 years have established the
importance of glucokinase in the metabolism of glucose by the liver and 13
cell. Studies within the last 5 years that have examined the molecular control
points of this system have shown that glucokinase gene expression is under
multiple levels of regulation, as is shown diagramatically in Figure 7. Most
striking in this regard is the finding of alternate promoter regions in a single
glucokinase gene that impart differential cell type-specific regulation to the
enzyme. An upstream promoter region in the gene drives expression of glu-
cokinase mRNA in both the pancreatic 13 cell and the pituitary while the
downstream promoter region in the gene controls expression of glucokinase
mRNA in the liver. This gene organization allows a different set of regulatory
factors to be involved in both tissues, presumably allowing the tissue-specific
control of glucokinase gene expression.
In addition to the use of alternate promoter regions in the glucokinase
gene, alternate RNA processing contributes in an important way to the reg-
ulation of glucokinase activity. While glucokinase mRNAs are produced in
GLUCOKINASE GENE
Upstream Promoter Downstream Promoter
~
Beta cell PROMOTER SELECTION Hepatocyte
and pituitary
Insulin
~
Exercise? TRANSCRIPTION RATE Biotin
T3
cAMP
Variable isoform ~ Variable isoform

activities, especially RNA PROCESSING activities
pituitary
~
Glucose? TRANSLATION RATE
Exercise?
GLUCOKINASE
FIGURE 7. Multiple control points affecting glucokinase gene expression. A diagram showing
the flow of information from the glucokinase gene to protein. The sites at which regulation of
glucokinase gene expression has been identified are illustrated.
the liver, ~ cell, and pituitary, posttranscriptional regulation through the

selection of alternate splice donor and acceptor sites, in addition to the use
of a cassette exon, have the potential to alter the reading frame of the enzyme
and lead either to early termination events or to the production of catalytically
inactive isoforms. This effect is most pronounced in the pituitary where
glucokinase protein and activity is not produced despite generation of a mRNA.
The mechanisms underlying the tissue-specific expression and regulation
of the glucokinase gene expression are just beginning to be studied in depth.
Undoubtedly studies of these mechanisms will provide a fertile area for future
investigation. Since glucokinase is a key regulatory control point for the
response of the ~ cell to glucose and for the utilization of glucose by the
liver, an understanding of the mechanisms involved is important. As additional
knowledge is obtained, it is likely that additional complexities will be dis-
covered relating to the interplay of nutritional, neuroendocrine, and genetic
factors that act to determine glucokinase gene expression.
ACKNOWLEDGMENTS
These studies were supported by grants from the National Institutes of
Health and the Juvenile Diabetes Foundation. We thank our colleagues and
collaborators for their contributions to the studies which are summarized here.
REFERENCES
I. Granner, D. and Pilkis, S., The genes of hepatic glucose metabolism, J. Bioi. Chern.,
265, 10173, 1990.
2. Iynedjian, P. B. and Girard, J., Nomenclature for mammalian glucokinase, Biochem.
J., 275, 821, 1990.
3. Ureta, T., The comparative isozymology of vertebrate hexokinases, Camp. Bichem.
Physiol., 718, 549, 1982.
4. Cornish-Bowden, A. and Cardenas, M. L., Hexokinase and glucokinase in liver me-
tabolism, Trends Biochem. Sci., 16, 281, 1991.
5. Weinhouse, S., Regulation of glucokinase in liver, in Current Topics in Cellular Reg-
ulation, Vol. II, Horecker, B. L. and Stadtman, E. R., Eds., Academic Press, New
York, 1976, I.
6. Lawrence, G. M. and Trayer, I. P., Hexokinase isoenzymes: Antigenic cross-reactivities
and amino acid copositional relatedness, Camp. Biochem. Physiol., 798, 233, 1984.
7. Andreone, T. L., Printz, R. L., Pilkis, S. J,, Magnuson, M.A., and Granner, D. K.,
The amino acid sequence of rat liver glucokinase deduced from cloned eDNA, J. Bioi.
Chem., 264,363,1989.
8. Kopetzki, E., Entian, K.-D., and Mecke, D., Complete nucleotide sequence of the
hexokinase PI gene (HXKl) of Saccharomyces cervisiae, Gene, 39, 95, 1985.
9. Stachelek, C., Stachelek, J,, Swan, J,, Botstein, D., and Konigsberg, W., Identifi-
cation. cloning, and sequence determination of the genes specifying hexokinase A and
B from yeast, Nucleic Acids Res., 14, 1986.
10. Nishi, S., Seino, S., and Bell, G. I., Human hexokinase: sequences of amino- and
carboxyl-terminal halves are homologous, Biochem. Biophys. Res. Commun., 157, 937,
1988.
11. Schwab, D. A. and Wilson, J. E., Complete amino acid sequence of rat brain hexokinase,
deduced from the cloned eDNA, and proposed structure of a mammalian hexokinase,
12. Arora, K. K., Fanciulli, M., and Pedersen, P. L., Glucose phosphorylation in tumor
cells, J. Bioi. Chem., 265, 6481, 1990.
13. Bennett, W. S. and Steitz, T. A., Structure of a complex between yeast hexokinase A
and glucose, J. Mol. Bioi., 140,211, 1980.
14. Lange, A. L., Xu, L. Z., Poelwijk, F. V., Lin, K., Granner, D. K., and Pilkis,
S. J,, Expression and site-directed mutagenesis of hepatic glucokinase, Biochem. J.,
277, 159, 1991.
15. Iynedjian, P. B., Mobius, G., Seitz, H. J,, Wollheim, C. B., and Renold, A. E.,
Tissue-specific expression of glucokinase: Identification of the gene product in liver and
pancreatic islets, Proc. Nat!. Acad. Sci. U.S.A., 83, 1998, 1986.
16. Thorens, B., Sarkar, H. K., Kaback, H. R., and Lodish, H. F., Cloning and functional
expression in bacteria of a novel glucose transporter present in liver, intestine, kidney,
and beta-pancreatic islet cells, Cell, 55, 281, 1988.
17, Mueckler, M., Family of glucose-transporter genes. Implications for glucose homeostasis
and diabetes, Diabetes, 39, 6, 1990.
18. James, D. E., Brown, B., Navarro, J., and Pilch, P. F., Insulin-regulatable tissues
express a unique insulin-sensitive glucose transport protein, Nature (London), 330, 183,
1988.
19. Vilaro, S., Palacio, M., Pilch, P. F., Testar, X., and Zorzano, A., Expression of an
insulin regulatable glucose carrier in muscle and fat endothelial cells, Nature (London),
342, 792, 1989.
20. Matschinsky, F. M. and Ellerman, J. E., Metabolism of glucose in the islets of
Langerhans, J. Bioi. Chern., 243, 2730, 1968.
2L Malaisse, W. J., Sener, A., Herchuelz, A., and Hutton, J. C., Insulin release: the
fuel hypothesis, Metabolism, 28, 373, 1979.
22. Malaisse, W. J., Insulin release: the fuel concept, Diabete Metab., 9, 313, 1983.
23. Orci, L., Thorens, B., Ravazzola, M., and Lodish, H. F., Localization of the pancreatic
beta cell glucose transporter to specific plasma membrane domains, Science, 245, 295,
1989.
24. Chen, L., Alam, T., Johnson, J. H., Hughes, S., Newgard, C. B., and Unger, R. H.,
Regulation of 13-cell glucose transporter gene expression, Proc. Nat!. Acad. Sci. U.S.A.,
87, 4088, 1990.
25. Giroix, M.-H., Sener, A., Pipeleers, D. G., and Malaisse, W. J., Hexose metabolism
in pancreatic islets, Biochem. J., 223, 447, 1984.
26. Sener, A., Malaisse-Lagae, F., Giroix, M.-H., and Malaisse, W. J., Hexose metab-
olism in pancreatic islets: Compartmentation of hexokinase in islet cells, Arch. Biochem.
Biophys., 265,61,1986.
27. Shimizu, T., Knowles, B. B., and Matschinsky, F. M., Control of glucose phosphor-
ylation and glucose usage in clonal insulinoma cells, Diabetes, 37, 563, 1988.
28. Shimizu, T., Parker, J. C., Najafi, H., and Matschinsky, F. M., Control of glucose
metabolism in pancreatic 13-cells by glucokinase, hexokinase, and phosphofructokinase,
Diabetes, 37, 1524, 1988.
29. Liang, Y., Najafi, H., and Matschinsky, F. M., Glucose regulates glucokinase activity
in cultured islets from rat pancreas,], Bioi. Chern., 265, 16863, 1990.
30. Meglasson, M. D. and Matschinsky, F. M., New perspectives on pancreatic islet
glucokinase, Am. I Physiol., 13, E1, 1984.
3L Unger, R. H., Diabetic hyperglycemia: link to impaired glucose transport in pancreatic
13 cells, Science, 251, 1200, 199L
32. Ashcroft, F. M., Harrison, D. E., and Ashcroft, S. J. H., Glucose induces closure
of single potassium channels in isolated rat pancreatic 13-cells, Nature (London), 312,
446, 1984.
33. Cook, D. L., Satin, L. S., Ashford, M. L. J., and Hales, C. N., ATP-sensitive K+
channels in pancreatic 13-cells, Diabetes, 37, 495, 1988.
34. Boyd, A. E., Sulfonylurea receptors, ion channels, and fruit flies, Diabetes, 37, 847,
1988.
35. Iynedjian, P. B., Pilot, P.R., Nouspikel, T., Milburn, J. L., Quaade, C., Hughes,
S., Ucla, C., and Newgard, C. B., Differential expression and regulation of the glu-
cokinase gene in liver and islets ofLangerhans, Proc. Nat( Acad. Sci. U.S.A., 86, 7838,
1989.
36. Bedoya, F. J., Matschinsky, F. M., Shimizu, T., ONeil, J. J., and Appel, M. C.,
Differential regulation of glucokinase activity in pancreatic islets and liver of the rat, J.
Bioi. Chern., 261, 10760, 1986.
37, Magnuson, M. A., Glucokinase gene structure: functional implications of molecular
genetic studies, Diabetes, 39, 523, 1990.
38. Matschinsky, F. M., Glucokinase as glucose sensor and metabolic signal generator in
pancreatic beta-cells and hepatocytes, Diabetes, 39, 647, 1990.
39. lynedjian, P. B., Ucla, C., and Mach, B., Molecular cloning of glucokinase eDNA,
J. Bioi. Chern., 262, 6032, 1987.
40. Chien, C.-T., Tauter, A., Lange, A. J., Chan, K., Printz, R. L., El-Maghrabi,
M. R., Granner, D. K., and Pilkis, S. J., Expression of rat hepatic glucokinase in
Escherichia coli, Biochem. Biophys. Res. Commun., 165, 817, 1989.
41. Magnuson, M. A. and Shelton, K. D., An alternate promoter in the glucokinase gene
is active in the pancreatic beta cell, J. Bioi. Chern., 264, 15936, 1989.
42. Hughes, S. D., Quaade, C., Milburn, J. L., Cassidy, L., and Newgard, C. B.,
Expression of normal and novel glucokinase mRNAs in anterior pituitary and islet cells,
J. Bioi. Chern., 266, 4521, 1991.
43. Magnuson, M.A., Andreone, T. L., Printz, R. L., Koch, S., and Granner, D. K.,
Rat glucokinase gene: Structure and regulation by insulin, Proc. Nat!. Acad. Sci. U.S.A.,
86, 4838, 1989.
44. Niswinder, K. and Magnuson, M. A., unpublished data.
45. Liang, Y., Jetton, T. L., Zimmerman, E., Najafi, H., Matschinsky, F. M., and
Magnuson, M. A., Effects of alternate RNA splicing on glucokinase isoform activities
in the pancreatic islet, liver, and pituitary, J. Bioi. Chern., 266, 6999, 1991.
46. Noguchi, T., Takenaka, M., Yamada, K., Matsuda, T., Hashimoto, M., and Tanaka,
T., Characterization of the 5' flanking region of rat glucokinase gene, Biochem. Biophys.
Res. Commun., 164, 1247, 1989.
47. lynedjian, P., Gjinovci, A., and Renold, A. E., Stimulation by insulin of glucokinase
gene transcription in liver of diabetic rats, J. Bioi. Chern., 263, 740, 1988.
48. Hoppner, W. and Seitz, H.-J., Effect of thyroid hormones on glucokinase gene tran-
scription in rat liver, J. Bioi. Chern., 264, 20643, 1989.
49. Narkewicz, M. R., lynedjian, P. B., Ferre, P., and Girard, J., Insulin and tri-
iodothyronine induce glucokinase mRNA in primary culture of neonatal rat hepatocytes,
Biochem. J., 271, 1990.
50. Chauhan, J. and Dakshinamurti, K., Transcriptional regulation of the glucokinase gene
by biotin in starved rats. J. Bioi. Chern., 266, 10035, 1991.
51. lynedjian, P. B., Jotterand, D., Nouspikel, T., Asfari, M., and Pilot, P.-R., Tran-
scriptional induction of glucokinase gene by insulin in cultured liver cells and its repression
by the glucagon-cAMP system, J. Bioi. Chern., 264, 21824, 1989.
52. Shelton, K. D., Franklin, A. J., Khoor, A., Beechem, J., and Magnuson, M. A.,
Multiple elements in the upstream glukokinase promoter contribute to transcription in
insulinoma cells. Mol. Cell. Bioi., in press.
53. Jetton, T. L. and Magnuson, M. A., Heterogeneous expression of glucokinase among
different pancreatic [3-cells, Proc. Nat/. Acad. Sci. U.S.A., 89, 2619, 1992.
54. Schuit, F. C., lntVeld, P. A., and Pipeleers, D. G., Glucose stimulates proinsulin
biosynthesis by a dose-dependent recruitment of pancreatic beta cell, Proc. Nat!. Acad.
Sci. U.S.A., 85, 3865, 1988.
55. Solomon, D. and Meda, P., Heterogeneity and contact-dependent regulation of hormone
secretion by individual beta cells, Exp. Cell Res., 162, 507, 1986.
56. Koranyi, L. I., Bourey, R. E., Slentz, C. A., Holloszy, J, 0., and Permutt, M.A.,
Coordinate reduction of rat pancreatic islet glucokinase and proinsulin mRNA by exercise
training, Diabetes, 40,401, 1991.
57. Newgard, C. B., Quaade, C., Hughes, S.D., and Milburn, J. L., Glucokinase and
glucose transporter expression in liver and islets: implications for control of glucose
homeostasis, Biochem. Soc. Trans., 18, 851, 1990.
58. Nakashima, R. A., Mangan, P. S., Columbini, M., and Pedersen, P. L., Hexokinase
receptor complex in hepatoma mitochondria: Evidence from N ,N' -dicyclohexylcarbodi-
imide-labeling studies for the involvement of the pore-forming protein VDAC, Biochem-
istry, 25, 1986.
59. Hayzer, D. J, and Iynedjian, P. B., Alternative splicing of glucokinase mRNA in rat
liver, Biochem. J., 270, 261, 1990.
60. Quaade, C., Hughes, S. D., Coats, W. S., Sestak, A. L., lynedjian, P. B., and
Newgard, C. B., Analysis of the protein products encoded by variant glucokinase tran-
scripts via expression in bacteria, FEBS Lett., 280, 47, 1991.
61. Tanizawa, Y., Koranyi, L. 1., Welling, C. M., and Permutt, M. A., Human liver
glucokinase gene: cloning and sequence determination of two alternately spliced cDNAs,
62. Jetton, T. L. and Magnuson, M. A., unpublished data.
63. Magnuson, M. A., Tissue-specific regulation of glucokinase gene expression, J. Cell.
Biochem., 48,115,1992.
Chapter 7
DIETARY AND HORMONAL REGULATION OF

L-TYPE PYRUVATE KINASE GENE EXPRESSION
Tamio Noguchi and Takehiko Tanaka
TABLE OF CONTENTS
I. Overview ........................................................ I 70
II. Regulation of L-Type PK Gene Expression in Normal

and Diabetic Rats ................................................ I 70
III. Regulaton of L-Type PK Gene Expression in Primary

Cultures of Rat Hepatocytes ..................................... I 74
IV. Identification of Cis-Acting Regulatory Regions of the

L-Type PK Gene ................................................. I 79
V. Summary ........................................................ I 83
References ............................................................... 184
0-8493-6961-4i93t$0.00 +$.50
© !993 by CRC Press, Inc. 169
I. OVERVIEW
Pyruvate kinase (PK), a key glycolytic enzyme, catalyzes formation of

pyruvate and ATP from phosphoenol-pyruvate and ADP. Four isozymes of
this enzyme have been identified in mammals. 1 They are named L-, R-,
M 1-, and M2 -types, and differ in primary structure, kinetic properties, and
tissue distribution. 1 However, the L-and R-types, and the M 1- and M2 -types
are encoded by the same genes called the PKU and PKM 3 •4 gene, respec-
tively. Both genes consist of 12 exons and ll introns. The introns interrupt
the protein coding sequences of the four isozymes at identical positions. The
first (R) and second (L) exons of the PKL gene contain 5' noncoding and
amino terminal sequences specific to the R- and L-types, respectively, while
the remaining downstream exons are common to the two isozymes. By use
of two alternative promoters, which are located upstream of the type-specific
exons, the R-and L-type isozymes are expressed in a tissue-specific manner.
On the other hand, the 9th (M 1) and lOth (M 2 ) exons of the PKM gene
encode sequences specific to the M 1- and M2 -types, respectively, while the
remaining exons are common to the two isozymes. Thus, the M 1- and M2 -
type isozymes are produced from the same transcript of the M gene by
alternative RNA splicing, which occurs tissue specifically.
The L-type PK is largely expressed in the liver. 1 Gene expression of this
enzyme is regulated by hormones and diet 1 like the expression of genes of
other key enzymes involved in glycolysis and lipogenesis, including gluco-
kinase (GK), glucose-6-phosphate dehydrogenase, malic enzyme, acetyl CoA
carboxylase, and fatty acid synthase. 5 A high glucose diet or insulin induces
expression of these enzymes, while starvation, diabetes, or glucagon sup-
presses their expression. Dietary fructose, which is known to stimulate li-
pogenesis more than dietary glucose, 6 also induces all these enzymes except
glucokinase, 7 - 9 Its effect is observed even in diabetic liver in contrast to that
of dietary glucose. Since fructokinase, which catalyzes the first step offructose
metabolism, is not regulated by hormones or diet, 10 the L-type PK is also a
key enzyme of fructose metabolism as well as glucose metabolism. Thus,
this enzyme is very important in not only glycolysis but also lipogenesis in
the liver. In this chapter, we describe the molecular mechanism of induction
of the L-type PK by insulin and dietary fructose.
II. REGULATION OF L-TYPE PK GENE EXPRESSION

IN NORMAL AND DIABETIC RATS
The L-type PK mRNA level is markedly reduced in streptozotocin-

induced diabetes in the liver of rats fed a high dextrose (glucose) diet. 11 • 12
Insulin administration to diabetic rats results in an increase in the level of the
L-type PK mRNA in the liver. 11 • 12 As seen in Figure l, this response shows
Noguchi and Tanaka 171
Insulin Fructose ------,
LPK
•
GK
p-Actin
0 2 4 6 10 16 0 2 4 6 10 16
Treatment time (h)
FIGURE 1. Time courses of induction by insulin or dietary fructose of L-type pyruvate kinase
and glucokinase mRNAs in diabetic rat liver. LPK, L-type pyruvate kinase; GK, glucokinase.
a characteristic delay, the increase becoming detectable only after 6 h, 13

whereas induction of GK mRNA is nearly maximal 2 h after insulin injec-
tion. 14 • 18 These findings suggest that the mechanisms of induction of the two
enzymes are different. Similar time courses of induction of the two mRNAs
in normal rat liver have been observed after glucose feeding. 15 • 16 Dietary
fructose also increases the level of the L-type PK mRNA in diabetic liver, 17
but not that of GK mRNA (Figure 1). 18 The induction of the L-type PK
mRNA by fructose occurs faster than that by insulin. A similar response of
the L type PK mRNA to dietary fructose has been observed in normal rat
liver. 16 The effect of fructose on the L-type PK seems to be directly on the
liver because the possibility that hormones such as glucagon or adrenal or
thyroid hormones mediate the fructose effect is excluded. 13 Dietary glycerol
is also a potent inducer of the L-type PK in normal and diabetic rat liver
(Table 1). 7• 17• 19 Thus, we consider that the effect of fructose is attributable to
an intermediate common to the metabolism of both fructose and glycerol.
The L-type PK is expressed not only in the liver, but also in kidney and
small intestine. 1 The major sites for the metabolism of fructose are known to
be the liver, kidney, and small intestine, 6 whereas glycerol is metabolized
appreciably only in the liver and kidney. 20 Therefore, it was of interest to
examine the effects of various carbohydrates on the levels of the L-type PK
mRNA in the kidney and small intestine of normal rats. As shown in Table
1, dietary fructose induced an increase in the L-type PK mRMA level in these
TABLE 1
Effects of Various Carbohydrates on the Level
of L-Type Pyruvate Kinase mRNA in Liver,
Kidney, and Small Intestine
Amount of mRNA (fold change)

Carbohydrate Liver Kidney Small intestine
Control 1.0 ± 0.1 1.0 ± 0.6 1.0 ± 0.1

Fructose 5.1 ± 1.4 4.6 ± 1.9 3.0 ± 0.2
Glycerol 9.6 ± 6.0 3.2 ± 0.9 1.0 ± 0.2
Glucose 1.2 ± 0.2 1.0 ± 0.2 1.6 ± 0.1
Galactose 1.5 ± 0.5 0.7 ± 0.2 0.7 ± 0.1
Sorbitol 1.0 ± 0.5 0.6 ± 0.2 0.6 ± 0.2
Xylose 0.9 ± 0.2 0.5 ± 0.1 0.3 ± 0.1
Xylitol 0.9 ± 0.3 0.5 ± 0.3 0.5 ± 0.2
Note: Normal rats were used for determination of the mRNA

in the kidney and small intestine, and diabetic rats for
its determination in the liver. Rats were starved overnight
and then given the indicated diets for 6 h. Control rats
were killed at zero time. Data are normalized to the value
for control rats.
tissues. 19 The time courses of these inductions were very similar to that
observed in diabetic liver. Glycerol feeding also increased the mRNA level
in the kidney after 6 h, but did not affect the mRNA level in the small
intestine. Glucose induced a slight but significant (1.6-fold, p < 0.001)
increase in the level of the L-type PK mRNA in the small intestine, but not
in the kidney. No other carbohydrates, including galactose, sorbitol, xylose,
and xylitol, were effective inducers of the L-type PK mRNA in the kidney
or small intestine or in the liver of diabetic rats. Thus, the tissues in which
fructose and glycerol induce the L-type PK mRNA correspond to the tissues
in which these compounds are metabolized, supporting the metabolite hy-
pothesis mentioned above. The candidate molecule may not accumulate sig-
nificantly during metabolism of any other carbohydrate except glucose.
We examined the mechanism of induction of the L-type PK mRNA by
insulin and dietary carbohydrates in diabetic rat liver using a nuclear tran-
scription assay. No significant change in gene transcription was observed 3
h after insulin treatment. 13 As shown in Figure 2, however, insulin enhanced
the rate of transcription of this gene about 28-fold after 12 h. 13 • 18 This stim-
ulation was inhibited by cycloheximide, suggesting that insulin may stimulate
transcription of the L-type PK gene by stimulating the synthesis of some
unknown protein. On the other hand, only four- to sixfold stimulation of the
gene transcription was observed 4 h after fructose or glycerol feeding. 18 • 19 On
the contrary, Munnich et al. reported that dietary fructose stimulated tran-
scription of the L-type PK gene in the liver of normal rats. 16 We tested the
,2' ,2' ~ ,2'

::; ~ ~ ~
'-
o- l>
vo"'fll
"'q, o-
s·S' oc;-
~
.:;.fll
i- ,.:i
._o
.:f
J
,\(.i
....c:-"' () c.,"' 0 «" c]'
••
M2PK
LPK
pUC18
118 . 4 4. 3 2. 4 4.9 14.9 8. 5 Relative rate

of transcription
(ppm)
FIGURE 2. Effects of insulin, fructose, and glycerol on transcription of pyruvate kinase genes.
LPK, L-type pyruvate kinase; M2PK, M2-type pyruvate kinase.
(A)
Transcription rate
(Fold chan ge)
r- Normal ---, . - - - - - Diabetic
~
~
~
~ ..g4' ~
()
§ .5'
(j ~ . §
(j
• •
M PK
2
LPK
pUC18
M PK 1.0 2.3 1.0 1.6 0.8 1.7

2
LPK 1.0 26 . 7 1.0 2.4 11. 3 0.7
(B)
mRNA level
(Fold change)
LPK 1.0 4.4 1. 0 6. 2 6. 6 1.3

± 0.2 ± 1. 2 ± 0.7 ± 2. 2 ± 0. 3 ± 0.6
FIGURE 3. Effects of fructose on transcription of pyruvate kinase genes in normal and diabetic
liver. Data are normalized to the value for control rats. LPK, L-type pyruvate kinase; M2 PK,
M2-type pyruvate kinase.
possibility that this difference was due to a difference in the condition of the
animals used, that is, diabetic or normal. 19 Feeding fructose for 4 h resulted
in only 2.4-fold stimulation of the transcription of the L-type PK gene in
diabetic rats, but about 27-fold increase in its transcription in normal rats
(Figure 3). In contrast, the extent of increase in the L-type PK mRNA levels
was similar in these two conditions. To determine whether this difference in
the responses of gene transcription to fructose was due to insulin, we injected

insulin 30 min after the start of fructose or glucose feeding. Insulin treatment
resulted in 4. 7-fold further increase in transcription in fructose-fed diabetic
rats after 4 h, but no change in the transcription in glucose-fed diabetic rats,
consistent with results described above. On the contrary, insulin did not
increase the mRNA level further in fructose-fed rats. These results indicate
that the mechanism of induction of the L-type PK in the liver by fructose is
dependent on the plasma insulin level. Dietary fructose acts mainly at the
transcriptional level in normal liver, whereas it acts mainly at the posttran-
scriptional level in the liver in the insulin-deficient state. In the latter case,
we suggest that fructose stabilizes nuclear RNA species of the L-type PK
mRNA since increases in the levels of nuclear RNA precursors of this enzyme
after fructose feeding precede changes in the levels of cytosolic L-type PK
mRNA. 17
Another group reported that glucocorticoids and thyroid hormones were
involved in dietary induction of the L-type PK mRNA at a posttranscriptional
level. 21 We did not, however, observe any significant role of these compounds
in the induction of the mRNA by insulin and dietary fructose. 13 Glucagon or
cyclic AMP inhibited transcriptional stimulation of the L-type PK gene by
insulin or dietary fructose and also the posttranscriptional effect of fructose
in diabetic liver. 13 •21 These effects of glucagon may be related to inhibition
of carbohydrate metabolism in the liver because its metabolism is involved
in not only the effects of fructose but also the effect of insulin as discussed
later.
Next, we examined the rates of transcription of the L-type PK gene in
the kidneys of normal and diabetic rats fed a high fructose diet for 4 h (Figure
4). 19 Dietary fructose stimulated transcription only 1.8-fold and 2.2-fold in
normal and diabetic kidney, respectively. However, the magnitude of these
increases was much lower than that of the mRNA. Thus, in both normal and
diabetic rats, dietary fructose acts mainly at the posttranscriptional level in
the kidney, as in diabetic liver. These results are understandable since dietary
glucose and insulin have no effect on the expression of the L-type PK gene
in the kidney. 19
III. REGULATION OF L-TYPE PK GENE EXPRESSION

IN PRIMARY CULTURES OF RAT HEPATOCYTES
Based upon in vivo observations, we investigated the mechanisms of

action of insulin and carbohydrates further using cultured hepatocytes. 22 In-
sulin caused time- and dose-dependent increases in the amount of the L-type
PK mRNA in cultured hepatocytes maintained in a medium containing glucose
(28 mM). The maximal response to 10 nM insulin was observed after 24 to
36 h (Figure 5). The effect of insulin was dependent on the nature and
concentration of the carbohydrate used (Figure 6). The addition of glucose
(A)
Transc ription rate
(Fold c hange)
r Normal ---, Diabetic
()-
.$
.,"' ()- .,"' "'
.$"' ~
()-
0<:-
~
.:§'
~
0<:- 8
..._;:; .§' ,k
(j l<.' (j () l<.' (;'
M PK
1
LPK
pUC18
MlK 1. 0 1.1 1.0 1.0 0. 9 1. 1

LPK 1.0 1.8 1. 0 1.7 2. 2 3. 3
(B)
mRNA level
(Fold change)
LPK 1. 0 5. 5 1.0 0.8 4. 4 12 . 1
± 0. 1 ± 0. 4 ± 0. 1 ± 0. 1 ± 0. 4 ~ 1. 1
FIGURE4. Effects of fructose and glycerol on transcription of pyruvate kinase genes in normal
and diabetic kidney. Data are normalized to the control values. LPK, L-type pyruvate kinase;
M,PK, M 2-type pyruvate kinase.
so
Ql 40
01
r::
nl
.c. 30
u
"C
0
u. 20
10
0
Time after insulin addition (h)
FIGURE 5. Time courses of induction by insulin of L-type pyruvate kinase mRNA in the
presence of various carbohydrates. Hepatocytes were cultured with 0.1 fLM insulin in the presence
of 28 mM glucose (e), 5.6 mM fructose (O), or 5.4 mM glycerol (.") for the indicated times.
Data are expressed relative to the value at zero time.
Pyruvate
0
5
4
3
2
.
Ql
01
c: 5
.!:
u 4
"'0
3
0
u. 2
0
40
30
20
10
0
Concentration (mM)
FIGURE 6. Effects of carbohydrates at various concentrations on the level of L-type pyruvate

kinase mRNA in cultured hepatocytes. Hepatocytes were incubated for 48 h with the indicated
concentrations of carbohydrates in the presence (•) or absence (O) of 0.1 fLM insulin. Data are
expressed relative to the value at zero time.
to the medium caused a marked, dose-dependent increase in the induction of

the L-type PK mRNA by insulin. The additions of fructose and glycerol also
enhanced the induction by insulin, but the maximal increases, observed at
about 5.5 mM in both cases, were less than 20% of that with 28 mM glucose.
No induction of the L-type PK mRNA by insulin was observed in hepatocytes
cultured in the presence of pyruvate at any concentration examined. On the
other hand, none of these carbohydrates induced the L-type PK mRNA in
the absence of insulin. This was unexpected because fructose and glycerol
are potent inducers of the mRNA even in diabetic liver.
The time course of induction of the L-type PK mRNA by insulin in the
presence of fructose, but not of glycerol, was similar to that in the presence
of glucose (Figure 5). In the presence of glycerol, the mRNA increased in a
diphasic manner: the first increase reached a maximum after 3 h, whereas
the second increase corresponded to the increase in the presence of glucose.
The early effect of insulin/glycerol was inhibited by actinomycin D. Thus,
this effect can be regarded as a partial reproduction of the fructose effect in
normal liver, which occurred much faster than the insulin effect in diabetic
LPK GK p-Actin
• • Control
• Glucose
• • Glucose + insulin
• • 2-Deoxyglucose
• 2- Deoxyglucose + insulin
FIGURE 7. Effects of 2-deoxyglucose on the induction by insulin of L-type pyruvate kinase

and glucokinase mRNAs in cultured hepatocytes. Hepatocytes were incubated with glucose (28
mM) or 2-deoxyglucose (28 mM) in the presence or absence of insulin (0.1 1-lM) for 36 h. Cells
at zero time were used as controls. LPK, L-type pyruvate kinase; GK, glucokinase.
liver as described above. The reason why the effects of fructose in normal
and diabetic liver were not reproduced in cultured hepatocytes is not known .
Quantitative and/or qualitative reproduction of the in vivo metabolism of
fructose and glycerol in a culture system may be difficult because the effect
of fructose in vivo is attributable to a common metabolite of fructose and
glycerol. In any case, these results suggest that the metabolism of glucose is
required for the induction of the L-type PK mRNA by insulin . If so, the effect
of insulin should not be observed in the presence of 2-deoxyglucose, a non-
metabolizable glucose analog. As shown in Figure 7, thi s was, in fact, the
case: the effect of insulin was not observed in hepatocytes cultured in a
medium containing 2-deoxyglucose. This was not due to a nonspecific effect
of 2-deoxyglucose because the induction of GK mRNA by insulin was not
impaired by this glucose analog. In fact, the effect of insulin on GK mRNA
did not require the presence of glucose itself. Thus , we suggest that some
metabolite accumulating during the metabolism of glucose is necessary for
the induction of the L-type PK mRNA by insulin and that this metabolite can
be produced from fructose and glycerol to some extent, but not from pyruvate,
in cultured hepatocytes. However, the role of insulin in the induction remains
to be determined. Insulin may simply increase the concentration of this me-
tabolite by stimulating glucose metabolism, or, in addition to this effect,
another signal from insulin may also be necessary.
Consistent with in vivo studies, cycloheximide inhibited the induction of
the L-type PK mRNA by insulin , s uggesting that ongoing protein synthesis
is required for the effect of insulin on the L-type PK. The addition of H-7,
an inhibitor of protein kinase C, caused dose-dependent inhibition of the
induction of L-type PK mRNA by insulin . However, phorbol 12-myristate
13-acetate did not increase the level of the mRNA . Thus, the effect of H-7
may not involve protein kinase C, although the possibility cannot be ruled
TABLE 2
Effects of Various Hormones on the Level
of L-Type Pyruvate Kinase mRNA in
Cultured Hepatocytes
Amount of mRNA
(fold change)
Treatment Pyruvate kinase 13-Actin
Control 1.0 1.0

Dexamethasone (I f.LM) 0.1 1.3
Insulin (0.1 fJ..M) 26.8 1.4
8-CPT-cAMP (0.1 mM) 0.8 2.8
T 3 (10 f.LM) 1.8 1.7
Dexamethasone + Insulin 76.9 1.5
Dexamethasone + Insulin 0.2 1.3
+ 8-CPT-cAMP
Dexamethasone + T3 0.3 1.2
Dexamethasone + T3 77.1 1.0
+ Insulin
Note: Hepatocytes were incubated with various hormones at

the concentrations indicated in parentheses for 36 h. The
medium contained 28 mM glucose. Values are expressed
relative to the control value at zero time.
out that kinase C may be involved in the action of insulin but not sufficiently
so to induce the L-type PK mRNA by itself.
As mentioned above, another group reported that glucocorticoids and
thyroid hormones are involved in dietary induction of the L-type PK mRNA
by acting at a posttranscriptional level. 21 If this is the case in cultured he-
patocytes, the additions of these hormones alone to cultured hepatocytes
should cause some increase in the L-type PK mRNA level. As shown in Table
2, dexamethasone alone slightly decreased the mRNA level. This is consistent
with the in vivo findings that adrenalectomy resulted in a slight elevation of
the L-type PK mRNA level in the liver. 13 Interestingly, however, dexameth-
asone greatly enhanced the induction of the mRNA by insulin when added
with insulin. As glucocorticoids are known to be necessary to maintain the
differentiated function of hepatocytes in vitro, 23 this hormone may act to
maintain the levels of hepatocyte-specific transcription factors in vitro. Con-
sistent with in vivo findings, T 3 had no effect on the L-type PK mRNA level
either alone or with dexamethasone or on the induction of the mRNA by
insulin/dexamethasone. However, this hormone had a permissive effect on
the induction of GK mRNA by insulin. 8-CPT-cAMP, a stable analog of
cAMP, inhibited the induction of the mRNA by insulin, confirming the in
vivo results.
Rat Pyruvate Kinase L gene 234567 8

exonR exonl
- 3000
D t---+1 • 37
CAT
1. Liver
2. Kidney
3. Small intestine
4. Brain
5. Muscle
6. Spleen
7. Testis
8. Erythrocytes
FIGURE 8. CAT activities in various tissues of transgenic mice.
IV. IDENTIFICATION OF CIS-ACTING REGULATORY

REGIONS OF THE L-TYPE PK GENE
The in vivo and in vitro studies described above suggest that the mech-
anism of transcriptional stimulation of the L-type PK gene by insulin or dietary
fructose is very complicated. In both cases, metabolism of carbohydrates is
involved, and so a common metabolite could be responsible for the induction
of the L-type PK. If this is the case, the same cis- and trans-acting factors
are involved in both cases. To test this possibility, we tried to identify a cis-
acting element responsible for transcriptional stimulation of the L-type PK
gene by insulin or dietary fructose. First, we produced transgenic mice car-
rying the 5' flanking region of the L-type PK gene from about nucleotide
-3000 to + 37 linked to the chloramphenicol acetyltransferase (CAT) struc-
tural gene and determined the CAT activity in various tissues. 24 This activity
was detected in the liver, kidney, and small intestine, but not in other tissues
(Figure 8). Thus the tissues expressing CAT activity were the same as those
expressing the endogenous L-type PK. Dietary glucose or insulin induced
similar increases in the levels of CAT and L-type PK mRNAs in the liver of
transgenic mice (Figure 9). Dietary fructose also caused a 15-fold increase
in the level of CAT mRNA, but this increase extent was significantly less
than that of the L-type PK mRNA (41-fold increase), suggesting that stabi-
lization of the transcripts of the L-type PK gene plays a significant role in
the induction by fructose even in normal liver. Thus, the sequence of about
3 kb upstream of the L-type PK gene contains all the cis-acting elements
responsible for tissue-specific expression of this enzyme and its transcriptional
stimulation by dietary carbohydrates and insulin. Dietary fructose, but not
glucose, also induced a marked increase in the L-type PK mRNA level in
the kidney, but induced only a marginal increase in the level of CAT mRNA.
These results are consistent with our finding mentioned above that posttran-
Normal
Control Glucose Fructose
CAT
LPK
p - Actin
FIGURE 9. Northern blot analysis of liver RNA of transgenic mice after administration of
glucose or fructose diet or insulin for 16 h. LPK, L-type pyruvate kinase.
scriptional regulation is a major factor in the induction of the L-type PK in

the kidney by fructose.
For further analysis of the cis-acting element, we introduced plasmid
DNA containing 3.2 kb of the 5' flanking region of the L-type PK gene
linked to the CAT gene into adult rat hepatocytes by electroporation and
measured transient CAT expression. 25 Although hepatocytes transfected with
this plasmid showed high CAT activity, the addition of insulin did not affect
expression of the CAT gene. 26 This is in contrast with the results obtained in
transgenic mice for some unknown reason. The results, however, indicated
that strong promoter activity is present in the 3.2 kb upstream region. There-
fore, we carried out functional analyses of a series of 5' and internal deletion
constructs of the CAT fusion gene. In this way, we identified three positive
regulatory regions required for expression of the L-type PK in hepatocytes.
These regions, designated as PKL-1, PKL-11, and PKL-III, were located
between nucleotides -76 and -94, -126 and -149, and -150 and -170,
respectively (Figure 10). Then, we determined the individual activities of
these elements. As shown in Figure 11, PKL-1 showed enhancer-like activity
alone, whereas PKL-11 and PKL-III did not have any independent effects.
Combinations of L-1 + L-11 and L-11 + L-III, but not of L-1 + L-III, showed
synergistic enhancer activities when oriented in the same direction. The in-
clusion of all three elements oriented in the same direction had the maximum
Transcription
CGCCACGGGGCACTCCCGTGGTTCCTGGACTCTGGCCCCCAGTGT
(l-IIIF) (L-IIF)
FIGURE 10. Schematic representation of the regulatory regions of the rat pyruvate kinase
gene.
-e2 Relative
L-111 L-11 L-1 ~ activity(%)
-
_.,.. 3. 3 ± 0. 7
-- -
2. 7 ± 0.2
_.,.. 0. 11 ± 0. 01
0. 10 ± 0. 10
0. 15 ± 0.05
0.60 ± 0.26
2.0 ± 0.8
-- -
1.3 ± 0.6
_.,.. 21. 2 ± 6. 7
--- -- -
5.0 ± 1.2
_.,.. _.,.. 2. 8 ± 1.2
_.,.. 4.1 ± 1.6
-- -
66.3 ± 26.8
_.,..
---
25. 7 ± 12.8
7. 5 ± 3. 2
27. 1 ± 17.6
plcat62 0.24 ± 0.23
plcat189 100
FIGURE ll. Interaction of three cis-acting elements of the L·type pyruvate kinase gene. A
single copy of each double-stranded oligonucleotide was ligated to pLcat62 in the orientation
shown by arrows on the left, which is the 5' to 3' direction. The relative activities of the
constructs are expressed as percentages of that of pLeat 189, which includes the region up to
-189.
Relative
-==:J.
~
()
activity (%)
A)
-63 -185 ~- 100
II 1700
8)
-63 -185
IT=riill. 100
~- II 2400
•1-
r
C) I IC A Tl <0.5
-63 -185
s I 100
FIGURE 12. Interaction of the regulatory regions with heterologous promoter. A fragment
containing the three elements was inserted upstream (A) or downstream (B) of the SV40 early
promoter, or upstream of the pyruvate kinase M gene promoter (C) in the orientation shown by
the arrows, which are the 5' to 3' direction.
synergistic effect, indicating that the three elements function as a unit. To

examine whether this unit exerts its effect on heterologous promoters, a
fragment containing all three elements was inserted upstream or downstream
of the SV40 early promoter or the PKM gene promoter. In all cases, inclusion
of the unit markedly stimulated expression of the CAT gene (Figure 12). Thus
this unit enhanced expression from heterologous as well as homologous pro-
moters in a manner that was independent of its orientation and position relative
to the cap site. The activity of the unit was not detected in HeLa cells or
K562 erythroleukemia cells, suggesting that this unit has cell-type specificity.
Judging from this finding together with the results in transgenic mice, this
unit seems very likely also to be responsible for expression of the L-type PK
in the kidney and small intestine.
PKL-1 contains a sequence homologous to the LF-BP7 (also called HNFP8
or HNFla 29 ) binding site. PKL-11 and PKL-III include sequences homologous
to the binding sites of LF-Al 30 and the adenovirus major late transcription
factor (MLTF), 31 respectively. We carried out gel retardation assays to ex-
amine whether these nuclear proteins in fact bind to these three elements.
The results indicated that the different trans-acting factors interacted with the
three elements and that the trans-acting protein bound to PKL-1 was in fact
LF-Bl. However, the trans-acting factors bound to PKL-11 and PKL-III were
different from LF-Al and MLTF, respectively. We tentatively named these

factors L-IIF and L-lliF, respectively (Figure 10). Since the three elements
had synergistic effects on the promoter activity, these proteins, bound to the
elements, must interact with each other. Very recently, another trans-acting
protein that binds to the LF-B 1 binding site was identified and characterized.
This protein called LF-B3,3 2 vHNF1/ 3 or HNFla 29 is a transcriptional factor
that binds to DNA as a dimer like LF-B1 and forms a heterodimer with LF-
Bl. The tissue distribution of LF-B3 resembles that of LF-Bl in adult rats:
the two proteins are expressed in the liver, kidney, and small intestine, al-
though their relative levels are different. Thus, LF-B3 also plays an important
role in expression of the L-type PK gene in these tissues.
An important question is whether these cis- and trans-acting factors are
involved in transcriptional stimulation of the L-type PK gene by insulin and
dietary fructose. This problem could not be solved in our transfection system
using cultured hepatocytes, as mentioned above. Therefore, we produced
another type of transgenic mouse carrying a L-type PK/CAT fusion gene with
a 5' deletion up to nucleotide -610. 34 Expression of the CAT gene showed
similar tissue-specificity to that of the endogenous L-type PK gene, strongly
supporting the conclusion that the three cis-acting elements are important in
tissue-specific expression of the L-type PK gene. However, preliminary ex-
periments showed that dietary glucose and fructose caused only a slight change
in the CAT mRNA level although they greatly increased the L-type PK mRNA
level. Thus, the three elements may not alone be sufficient to confer insulin-
and fructose-responsiveness on the CAT reporter gene. On the contrary,
Thompson and Towle reported very recently that the 101-bp fragment from
nucleotide - 197 to -96 of the L-type PK gene conferred insulin/glucose-
responsiveness on the CAT reporter gene with the thymidine kinase promoter
when transfected into primary cultured hepatocytes using synthetic lipo-
somes. 35 This region contained PKL-11 and PKL-III. This is in contrast with
our finding that insulin had no effect on expression of the CAT gene when
the CAT fusion gene containing the upstream region of the L-type PK gene
up to - 3000 was introduced into hepatocytes by electroporation, as mentioned
above. Another group also could not identify the regulatory region responsive
to insulin by transfection of DNA into cultured hepatocytes by the calcium
phosphate method. 36 These different results may be due to the different trans-
fection procedures used. Further studies both in vitro and in vivo are necessary
to clarify this discrepancy and elucidate the molecular mechanism underlying
induction of the L-type PK by insulin and carbohydrates.
V. SUMMARY
Work to date in our laboratory as well as in others has demonstrated that

the expression of genes coded for pyruvate kinase is influenced by hormonal
and dietary factors. The four isozymes of this enzyme differ in their location
in the body. In turn, the genes for these isozymes are different in their
responsiveness to the kind of dietary carbohydrate and the hormones which
influence its expression. The identification of the cis-acting regulatory region
of the liver type pyruvate kinase gene has yet to be resolved.
REFERENCES
I. Imamura, K., Noguchi, T., and Tanaka, T., Regulation of isozyme patterns of pyruvate
kinase in normal and neoplastic tissues, in Markers of Human Neuroectodermnl Tumors,
Staal, G. E. J. and van Veelen, C. W. M., Eds., CRC Press, Boca Raton, FL, 1986,
chap. 13.
2. Noguchi, T., Yamada, K., Inoue, H., Matsuda, T., and Tanaka, T., The L-type and
R-type isozymes of rat pyruvate kinase are produced from a single gene by use of different
promoters, J. Bioi. Chern., 262, 14366, 1987.
3. Noguchi, T., Inoue, H., and Tanaka, T., TheM,- and M,-type isozymes of rat pyruvate
kinase are produced from the same gene by alternative RNA splicing, J. Bioi. Chern.,
261, 13807, 1986.
4. Takenaka, M., Noguchi, T., Inoue, H., Yamada, K., Matsuda, T., and Tanaka, T.,
Rat pyruvate kinase M gene. Its complete structure and characterization of the 5 '-flanking
region, J. Bioi. Chern., 264, 2363, 1989.
5. Goodridge, A. G., Dietary regulation of gene expression: enzymes involved in carbo-
hydrate and lipid metabolism, Annu. Rev. Nutr., 7, 157, 1987.
6. Van Den Berghe, G., Metabolic effects of fructose in the liver, in Current Topics in
Cellular Regulation, Vol. 13, Horecker, B. L. and Stadtman, E. R., Eds., Academic
Press, New York, 1978, 97.
7. Gunn, J, M. and Taylor, C. B., Relationships between concentration of hepatic inter-
mediary metabolites and induction of the key glycolytic enzymes in vivo, Biochem. J.,
136, 455, 1973.
8. Takeda, Y., Inoue, H., Honjo, K., Tanioka, H., and Daikuhara, Y., Dietary response
of various key enzymes related to glucose metabolism in normal and diabetic rat liver,
Biochim. Biophys. Acta, 136, 214, 1967.
9. Fukuda, H., Iritani, N., and Tanaka, T., Effects of high-fructose diet on lipogenic
enzymes and their substrate and effector levels in diabetic rats, J. Nutr. Sci. Vitaminol.,
29, 691, 1983.
10. Adelman, R. C., Spotter, P. D., and Weinhouse, S., Dietary and hormonal regulation
of enzymes of fructose metabolism in rat liver, J. Bioi. Chern., 241, 5467, 1966.
11. Noguchi, T., Inoue, H., and Tanaka, T., Regulation of rat liver L-type pyruvate kinase
mRNA by insulin and by fructose, Eur. J. Biochem., 128, 583, 1982.
12. Noguchi, T., Inoue, H., Chen, H., Matsubara, K., and Tanaka, T., Molecular cloning
of DNA complementary to rat L-type pyruvate kinase mRNA. Nutritional and hormonal
regulation of L-type pyruvate kinase mRNA concentration, J. Bioi. Chern., 258, 15220,
1983.
13. Noguchi, T., Inoue, H., and Tanaka, T., Transcriptional and post-transcriptional reg-
ulation of L-type pyruvate kinase in diabetic rat liver by insulin and dietary fructose, J.
Bioi. Chern., 260, 14393, 1985.
14. Iynedjian, P. B., Gjinovci, A., and Renold, A. E., Stimulation by insulin of glucokinase
gene transcription in liver of diabetic rats, J. Bioi. Chern., 263, 740, 1988.
15. Iynedjian, P. B., Ucla, C., and Mach, B., Molecular cloning of glucokinase eDNA.
Developmental and dietary regulation of glucokinase mRNA in rat liver, J. Bioi. Chern.,
262, 6032, 1987.
16. Munnich, A., Lyonnet, S., Chauvet, D., Van Schaftingen, E., and Kahn, A., Dif-
ferential effects of glucose and fructose on liver L-type pyruvate kinase gene expression
in vivo, J. Bioi. Chern., 262, 17065, 1987.
17. Inoue, H., Noguchi, T., and Tanaka, T., Rapid regulation of L-type pyruvate kinase
mRNA by fructose in diabetic rat liver, J. Biochern. (Tokyo), 96, 1457, 1984.
18. Noguchi, T. and Tanaka, T., unpublished data, 1989.
19. Matsuda, T., Noguchi, T., Takenaka, M., Yamada, K., and Tanaka, T., Regulation
of L-type pyruvate kinase gene expression by dietary fructose in normal and diabetic rats,
J. Biochern. (Tokyo), 197, 655, 1990.
20. Vernon, R. G. and Walker, D. G., Glycerol metabolism in the neonatal rat, Biochern.
J., 118,531, 1970.
21. Vaulont, S., Munnich, A., Decaux, J.-F., and Kahan, A., Transcriptional and post-
transcriptional regulation of L-type pyruvate kinase gene expression in rat liver, J. Bioi.
Chern., 261, 7621, 1986.
22. Matsuda, T., Noguchi, T., Yamada, K., Takenaka, M., and Tanaka, T., Regulation
of the gene expression of glucokinase and L-type pyruvate kinase in primary cultures of
rat hepatocytes by hormones and carbohydrates, J. Biochern. (Tokyo), 108, 778, 1990.
23. Feliu, J. E., Coloma, J., Gomez-Lechon, M.-J., Garcia, M.D., and Baguena, J.,
Effect of dexamethasone on the isozyme pattern of adult rat liver parenchymal cells in
primary cultures, Mol. Cell. Biochern., 45, 73, 1982.
24. Yamada, K., Noguchi, J., Miyazaki, J., Matsuda, T., Takenaka, M., Yamamura,
K., and Tanaka, T., Tissue-specific expression of rat pyruvate kinase Llchloramphenicol
acetyltransferase fusion gene in transgenic mice and its regulation by diet and insulin,
Biochern. Biophys. Res. Cornrnun., 171, 243, 1990.
25. Yamada, K., Noguchi, T., Matsuda, T., Takenaka, M., Monaci, P., Nicosia, A.,
and Tanaka, T., Identification and characterization of hepatocyte-specific regulatory
regions of the rat pyruvate kinase L gene. The synergistic effects of multiple elements,
J. Bioi. Chern., 265, 19885, 1990.
26. Noguchi, T., Yamada, K., and Tanaka, T., unpublished data, 1989.
27. Frain, M., Swart, G., Monaci, P., Nicosia, A., Stampfli, S., Frank, R., and Cortese,
R., The liver-specific transcription factor LF-Bl contains a highly diverged homeobox
DNA binding protein, Cell, 59, 145, 1989.
28. Chouard, T., Blumenfeld, M., Bach, 1., Vandekerhove, J., Cereghini, S., and Yaniv,
M., A distal dimerization domain is essential for DNA binding by the atypical HNFl
homeodomain, Nucleic Acids Res., 18, 5853, 1990.
29. Mendel, D. B., Hansen, L. P., Graves, M. K., Conley, P. B., and Crabtree, G. R.,
HNF-lo: and HNFl J3 (vHNF-1) share dimerization and homeodomains, but not activation
domains, and form heterodimers in vitro, Genes Dev., 5, 1042, 1991.
30. Hard on, E. M., Frain, M., Paonessa, G., and Cortese, R., Two distinct factors interact
with the promoter regions of several liver-specific genes, EMBO J., 7, 1711, 1988.
31. Chodosh, L. A., Carthew, R. W., Morgan, j. G., Crabtree, G. R., and Sharp,
P. A., The adenovirus major late transcription factor activates the rat -y-fibrinogen pro-
moter, Science, 238, 684, 1987.
32. De Simone, V., De Magistris, L., Lazzaro, D., Gerstner, J., Monaci, P., Nicosia,
A., and Cortes, R., LFB3, a heterodimer-forming homeoprotein of the LFBl family,
is expressed in specialized epithelia, EMBO J., 10, 1435, 1991.
33. Rey-Campos, J., Chouard, T., Yaniv, M., and Cereghini, S., vHNFl is a borneo-
protein that activates transcription and forms heterodimers with HNFl, EMBO J., 10,
1445, 1991.
34. Noguchi, T., Miyazaki, J., Yamamura, K., and Tanaka, T., unpublished data, 1991.
35. Thomson, K. S. and Towle, H. C., Localization of the carbohydrate response element
of the rat L-type pyruvate kinase gene, J. Bioi. Chern., 266, 8679, 1991.
36. Cognet, M., Bergot, M.-0., and Kahn, A., cis-Acting elements regulating expression
of the liver pyruvate kinase gene in hepatocytes and hepatoma cells. Evidence for tissue-
specific activators and extinguisher, J. Bioi. Chern., 266, 7368, 1991.
Chapter 8
GLUCOSE-6-PHOSPHATE DEHYDROGENASE: DIET

AND HORMONAL INFLUENCES ON DE NOVO
ENZYME SYNTHESIS
Rolf F. Kletzien and Carolyn D. Berdanier
TABLE OF CONTENTS
I. Introduction ...................................................... 188
II. Diet Effects on G6PD Activity .................................. 188
III. Signals That Increase G6PD Activity ............................ 190
IV. Use of Cellular Models and Molecular Probes to

Define the Signals Responsible for Regulation of
Glucose-6-Phosphate Dehydrogenase ............................ 191
V. Hormonal Regulation of G6PD .................................. 192
VI. Nutrient Regulation of G6PD .................................... 194
VII. Other Factors Capable of Regulating G6PD ..................... 195
VIII. Recapitulation of Work on Hepatocytes .......................... 196
IX. Recent Insights into G6PD Gene Structure: Implications

for Regulation of Expression ..................................... 197
X. Sequence and Structure of the Rat G6PD Gene

and Pseudo gene .................................................. 197
XI. Conclusions ...................................................... 202
References ............................................................... 203
0·8493·6961·4193/$0.00+$.50
© 1993 by CRC Pre». Inc. 187
I. INTRODUCTION
Among the many nutrient-gene interactions described in this volume,

none have been described as often or as fully as that which occurs when
animals are starved and realimented. Tepperman and Tepperman 1 described
the response to intermittent starvation as adaptive hyperlipogenesis. Periodic
famine even today afflicts human populations in third world countries and in
countries facing catastrophic events such as war, earthquakes, crop failures,
and economic failures. Those that can adapt and increase their metabolic
efficiency survive; those that cannot die. Those that live pass on their genetic
material to subsequent generations. Those that die do not. Truly, survival
under adverse nutritional circumstances is the ultimate result of nutrient-gene
interactions.
II. DIET EFFECTS ON G6PD ACTIVITY

Although survival and reproduction is the result of numerous nutrient-
gene interactions, this chapter will discuss only one: the influence of nutri-
tional state on the enzyme glucose-6-phosphate dehydrogenase G6PD (E.C.
1.1.1.49, D-glucose-6-phosphate: NADP+ oxidoreductase). This enzyme cat-
alyzes the first rate-limiting reaction in the pentose phosphate shunt. Together
with the other pentose shunt dehydrogenase (6-phosphogluconate dehydro-
genase E.C. 1.1.1.44), this reaction sequence provides approximately 50%
of the reducing equivalents needed to support fatty acid synthesis. The re-
maining reducing equivalents are provided by the malic enzyme reaction
( ~40%) and the isocitrate dehydrogenase reaction ( ~ 10%). Glock and McLean 2
were among the first to report that diet restriction would result in a decrease
in the activity of G6PD. This report was followed by others3 ·4 that described
the increase in enzyme activity in rats that were switched from a carbohydrate-
poor diet to one rich in simple sugars. Even more enzyme activity was
observed if the animals were first starved and then fed a sugar-rich diet. 1•5 - 15
There is a high correlation between the activity of G6PD and de novo fatty
acid synthesis. 13 As shown in Table l the percent increment in enzyme activity
depends on a variety of dietary and nondietary factors. Dietary factors include
the type and amount of carbohydrate, the kind and amount of fat, and the
level of protein. In Table 1 are presented some of the observations from a
variety of studies using rats starved for 48 hand refed for 48 h. The literature
is vast with respect to the responses of rats, chickens, cows, mice, pigs, fish,
and humans. Only data from rats are presented in Table I. Rats that are
starved and refed diets rich in sugars such as sucrose generally have a higher
G6PD response than rats refed an equivalent amount of starch or fructose. 13 •14
Those refed fructose have a lower G6PD response than those refed glucose. 11
Those refed a glucose-menhaden oil diet have a lower response than those
whose diet contained com oil or coconut oil or beef tallow. 15 • 16 Note the rather
Kletzien and Berdanier 189
TABLE 1
Influence of Diet Composition on the Percent Increment
in Hepatic G6PD Activity Due to Starvation-Refeeding
Diet % Increment• Reference
65% Glucose-5% com oil 91-514 II, 13, 15, 16

65% Starch-5% com oil 262-278 13, 14
65% Sucrose-5% com oil 190-233 13-15
65% Fructose-5% com oil 72 11
65% Glucose-5% menhaden oil 15, 16, 101-122
65% Glucose-5% beef tallow 270 16
65% Glucose-5% coconut oil 218 16
65% Starch-5% coconut oil 291 14
65% Sucrose-5% coconut oil 319 14
65% Protein-20% glucose-5% com oil !54 13
40% Fat, 30% glucose, 20% protein 274 13
Note: SR, starved 48 h; refed 48 h.
, % increment = G6PD activity in S-R rats - G6PD activity in AL rats

G6PD activity in AL rats
broad range in values for some of the diets. In part, this is due to the variation
within strains and in part to the variation in response between strains. Some
rats, notably those of the Wistar strain, are relatively unresponsive to star-
vation-refeeding, whereas Sprague-Dawley rats are more responsive to this
feeding paradigm. BHE rats are more responsive than Sprague-Dawley rats. 15
Not only are there differences due to the strain in enzyme activity, there are
also differences attributable to the diet fed the rat prior to its use in a starvation-
refeeding experiment, and, too, its sex and its age. If rats were fed a high
fat diet prior to use, the increment in enzyme activity was twice as high as
in rats fed a standard stock diet prior to use. 17 Middle-aged rats ( -12 months
of age) were similar to young rats 18 • 19 in their response to starvation-refeeding,
while aged rats (22 months of age) were less responsive to this treatment than
their younger (2 months old) cohorts. 2° Female rats are less responsive to
starvation-refeeding than are male rats. 21 This gender difference in response
is probably due to the fact that female rats have a higher G6PD activity in
the basal ad libitum fed state. In part, this is due to the cyclical presence of
estrogen which imposes a cyclical pattern of reductions and increases in food
intake. Thus, postpubescent female rats experience a 5-day food intake de-
crease/increase cycle which, in effect, is a modified starve-refeed cycle.
Prepubescent female rats do not evidence an estrus cycle and also do not have
cyclical variations in food intake. Estrogen has also been shown to be essential
to the induction of G6PD activity in the uterus. 22
Variation in food supply, as well, can affect the G6PD response to star-
vation-refeeding. Rats that are deprived of food during shipping have a larger
response to starvation-refeeding than rats obtained from a local source where

the stress of shipping is minimized. 23 •25 A larger response has also been
observed in rats that have been starved and refed more than once. ' 0 •25 •26 The
length of starvation also affects the responseY In rats starved for 3 to 6 days
the enzyme response was greater than in rats starved for only 2 days prior to
refeeding. All of these observations suggest that the increase in G6PD activity
in animals refed following starvation is due to signals generated by the star-
vation or food restriction that stimulates G6PD.
III. SIGNALS THAT INCREASE G6PD ACTIVITY
At first, it was thought that the increase in G6PD activity in starved-

refed rats was due to the fact that these rats overeat during the initial 2 days
of refeeding. 28 However, Gimenez and Johnson showed that the G6PD en-
zyme response still occurred in rats not allowed to overeat. 29 A number of
papers reported that the enzyme response could be suppressed if the antibiotic
8-azaguanine were administered during the refeeding period. 9 · 10 • 18 •25 - 27 8-
Azaguanine is a compound that is incorporated into messenger RNA as it is
being synthesized. However, because it is a substituted guanine, the messen-
ger RNA cannot be translated. Rats that are starved and refed and given this
drug will have enzyme activities similar to those found in ad libitum fed rats
but will not evidence the large increase in activity typical of the response to
starvation-refeeding. Thus, one can infer that the large increase in enyzme
activity is due to de novo synthesis of mRNA coded for the G6PD enzyme.
Although the mechanism for the increase in enzyme protein synthesis was
clarified by the use of 8-azaguanine, the signal or signals which stimulated
the mRNA synthesis were unknown. Some suggested that insulin might serve
as that signal since a high glucose diet was needed in order to demonstrate
the enzyme response. Glucose is a potent insulin secretogogue and insulin
does stimulate protein synthesis. However, a variety of experiments using
intact animals failed to demonstrate unequivocally that insulin was the primary
signal for the enzyme overshoot. 4 •10 •27•28 •30 •31
The fact that starvation was an absolute requirement for the overshoot
suggested that the stress response was involved in the generation of the enzyme
overshoot. This was tested by removing the adrenals (ADX) from rats and
then subjecting them to starvation-refeeding. 32 When refed, the ADX rats had
enzyme activity levels similar to those of the ad libitum fed rats but no
overshoot (Figure 1). When the glucocorticoids were replaced, the large
increase in G6PD activity in the starved-refed rat was restored. 32- 34 Using 8-
azaguanine, it was shown that the glucocorticoid effect was that of stimulating
G6PD mRNA synthesis. 31 This effect of glucocorticoid, although dependent
to some extent on other hormones (insulin and thyroxine), 33 •34 was indeed a
direct effect of the hormone glucocorticoid on the synthesis of mRNA coding
for G6PD.
Kletzien and Bn·danier 191
G6PD ACTIVITY IN STARVED · REFED RATS
50
" 0 Intact
~ ADX
. 40 Intact + 8AZ
....
.<:
"'
·;; ADX + 8 AZ
5:
>
-o 30
0
CD
(!)
0
~ 20
.~
c:
=>
10
AD LIB Starved Starved - Starved - Refed

Refed plus GC
FIGURE 1. Hepatic G6PD activity in ad libitum fed and starved-refed Sprague-Dawley rats.
Rats were fed or refed a 65% glucose diet. The starvation period and the refeeding period were
48 h each; 8 azaguanine was administered daily. Abbreviations: GC, glucocorticoid; 8 AZ, 8-
azaguanine. Data are from References 33 and 34.
IV. USE OF CELLULAR MODELS AND MOLECULAR

PROBES TO DEFINE THE SIGNALS RESPONSIBLE
FOR REGULATION OF GLUCOSE-6-PHOSPHATE
DEHYDROGENASE
The studies carried out with intact animals have demonstrated that a
variety of nutrients and hormones are either directly or indirectly involved
with the regulation of hepatic G6PD. However, it is often difficult to distin-
guish direct vs. indirect involvement of hormones or nutrients on gene expres-
sion in intact animals. For example, in attempting to determine whether insulin
or glucose is a primary signal for induction of G6PD, experimentation in
animals is compromised since blood glucose is maintained at 5 mM, and
elevating it by feeding glucose causes insulin release. Eliminating insulin by
making the animal diabetic results in high levels of glucagon which cause an
elevation of hepatic cAMP and gluconeogenesis, conditions that might repress
G6PD. The well-known induction of G6PD following injection of insulin into
the animal could be the result of a positive influence of insulin on the G6PD
gene or it could be result of derepression of the influence of glucagon. Thus,
direct assignment of the responsibility for control of G6PD expression requires
the use of less complex systems than the intact animal.
In order to sort out the various factors involved in the regulation of hepatic
G6PD, we have employed primary cultures of liver parenchymal cells main-
tained in chemically defined, serum-free medium. In this system, cells retain
the terminally differentiated phenotype for a discrete period of time (up to
several weeks or months, depending on culture conditions), 37 - 39 in contrast
to hepatoma cell lines which seldom display the full realm of differentiated
functions expected of a hepatocyte. Thus, metabolic responses of the primary
cultures to hormones or nutrients can be expected to mimic that observed in
vivo. With this system it is then possible to add hormones or nutrients,
individually or in combination, to define precisely the involvement of these
factors in the control of G6PD expression. Several groups have employed the
hepatocyte primary culture system in studies on G6PD regulation. 40 -45 While
culture conditions were not identical, some common themes have emerged
and some areas of disagreement are apparent. This review will attempt to
analyze and summarize results on nutrient and hormonal control of G6PD
from the six laboratories 40-45 employing this culture system.
An additional development that has greatly aided analysis of G6PD expres-
sion is the availability of molecular probes for the mRNA encoding rat liver
G6PD. 46 The complete sequence of the protein 47 and mRNA 48 has been es-
tablished and genomic fragments have been isolated. 49 These sequences along
with high titer antibodies against G6PD provide all of the necessary tools to
analyze expression in either animals or in cells in culture. Thus, it is now
possible to dissect in a molecular fashion the effects of the hormones or
nutrients on the control of expression. This is particularly important in those
situations where a hormone or nutrient might increase transcription of the
G6PD gene, but a different hormone might be required for the optimum rate
of translation of that mRNA species. If one simply measures enzyme activity
or enzyme protein by immunochemical techniques, this sort of interaction
would not be observed. Therefore, combining these molecular tools with the
primary culture model should provide precise information on the control of
G6PD expression.
V. HORMONAL REGULATION OF G6PD
Insulin has long been considered the primary signal for induction of G6PD
in addition to the other lipogenic enzymes in animals. Kurts and Wells 45 and
Nakamura et al. 41 were the first to demonstrate that insulin could induce
G6PD in primary cultures of hepatocytes. Subsequently, others extended this
observation to show that insulin increased the relative rate of G6PD synthesis 5°
and mRNA encoding G6PD. 50 ·51 The ability of insulin to induce G6PD is not
rapid in onset, 50 taking 36 to 48 h to reach its maximum effect. However,
induction of G6PD in vivo following feeding a high carbohydrate diet is also
a relatively slow process, taking 48 to 72 h to reach maximal levels of
activity. 49 ·52 Thus, although the absolute change in enzyme activity observed
in tissue culture following insulin treatment is not as great as that observed
in the intact animal following carbohydrate refeeding, the time course is quite
similar. This lends strength to the argument that insulin is the primary signal
involved but suggests that other factors may be involved in amplification of
the insulin response. The six laboratories 40 -45 that have utilized primary cul-
tures of hepatocytes are in agreement with the primary role of insulin in
control of G6PD expression.
The precise level at which insulin regulates G6PD expression has not
been established. The fact that insulin elevates G6PD mRNA in cultured
hepatocytes suggests that transcriptional regulation of the G6PD gene may
be a critical site of regulation. In addition, two studies employing transcription
"run on" assays from rat liver nuclei isolated from fasted-refed animals have
demonstrated that there is a transient activation of G6PD gene transcrip-
tion.49·53 However, it is also possible that insulin stabilizes the G6PD mRNA
and/or causes a selective translation of it. Changes in G6PD mRNA stability
following refeeding a high carbohydrate diet have been observed in rats. 49
The study of how insulin accomplishes regulation of G6PD and other insulin
regulated genes remains as a major area of interest for molecular endocri-
nologists.
The glucocorticoids have a well established role in lipogenesis (for review
see Reference 54), and, thus, the interaction of insulin and the glucocorticoids
in regulation of G6PD has been explored in the hepatocyte primary culture
model. The results from these studies have shown that some differences are
apparent which are perhaps the result of culture conditions. Kurtz and Wells 45
first demonstrated that a synthetic glucocorticoid, dexamethasone (Dex), in-
duced G6PD activity by itself and its effect was additive with insulin. In
addition, the study 45 suggested that the glucocorticoid played a critical ''per-
missive" role for insulin induction of G6PD since insulin failed to increase
G6PD in hepatocytes from adrenalectomized rats unless Dex was added to
the cultures. Kletzien and co-workers50 ·5l.5 5 concluded that while the gluco-
corticoids enhanced the insulin induction of G6PD activity, it exerted its main
effect on mRNA levels encoding G6PD. Using molecular probes and a high
titer, monospecific antibody against G6PD, they demonstrated that Dex ap-
pears to increase the mRNA encoding G6PD but translation of the active
protein does not occur unless insulin is also present. 50 ·51 Holten and co-workers
have also demonstrated that maximum induction of G6PD in the primary
culture model requires the presence of a glucocorticoid. 42 Thus, it would seem
that this class of hormones is involved in at least a permissive way, in reg-
ulating G6PD expression.
In contrast to the above results, two groups 41 .43 have reported that the
glucocorticoids are not involved in regulation of G6PD. However, Ichihara
and colleagues 41 maintained their cultures in the continuous presence of glu-
cocorticoids and serum. Thus, it is not surprising that addition of glucocor-

ticoids failed to cause any further increase in enzyme activity. 41 It is more
difficult to understand why Salati et al. 43 did not observe enhancement of
insulin induction of G6PD activity, but it may be related to the use of serum
in the tissue culture medium during the early phase of attachment of cells to
the culture dish. Serum may c:mtain sufficient glucocorticoids or other factors
that exhibit long range effects on subsequent insulin stimulation of gene
expression. Alternatively, many of these experiments 43 were conducted with
vitamin E present in the culture medium which would be expected to decrease
oxidant stress, a potential stimulant ofG6PD. The glucocorticoids may exhibit
different effects depending on the redox state of the cell.
Experiments conducted in animals some time ago have suggested that
triiodothyronine (T3) may be involved in regulation ofG6PD. 56 Little evidence
from the primary culture model has been forthcoming to support a role for
thyroxine. Fritz and Kletzien 57 demonstrated that T3 increased G6PD expres-
sion in the hypothyroid rat through a mechanism that did not involve an
increase in G6PD mRNA. However, Kletzien and co-workers have not been
able to demonstrate any direct effect of T3 on G6PD induction in the hepa-
tocyte primary cultures, a result that is in concert with the other three groups
that have directly examined this question. 42 .43 •45 Thus, it would appear that
T3 has no immediate, direct role in regulation of hepatic G6PD expression.
Yoshimoto et al. 58 have shown that epidermal growth factor (EGF) is
capable of inducing G6PD in primary cultures of hepatocytes and, in the
presence of insulin, additivity in the response was noted. This effect appeared
to be disconnected from lipogenesis since EGF repressed induction of malic
enzyme and the insulin stimulated incorporation of acetate into triglycerides.
EGF was shown to increase both the relative rate of G6PD synthesis and
mRNA encoding the enzyme. 5 8 An EGF growth stimulus may be responsible
for the increase in G6PD. Nevertheless, since it is sometimes recommended
that EGF be included in the medium for hepatocyte primary cultures, 38 the
influence of EGF or other growth or serum factors must be taken into account
when attempting to ascertain direct vs. indirect effects of these agents on
expression. Clearly, it would be ideal to always use a chemical defined,
serum-free medium although optimum conditions may not yet be defined.
VI. NUTRIENT REGULATION OF G6PD
Investigation of the role of carbohydrate in the regulation of G6PD in rat

hepatocytes in primary culture has led to the conclusion that carbohydrate
does40 •43 or does not42 .45 influence induction of the enzyme by itself or in
conjunction with hormones. In attempting to resolve this difference, it is clear
that culture conditions can influence the response of cells to carbohydrate.
Recent work from Holten's laboratory has shown that fructose, when present
at 5 mM in the culture medium, increases expression in the presence of Dex
and insulin, while at 10 rnM, expression is decreased. 42 In the same study,

these investigators found that the presence of D-glucose by itself or in com-
bination with hormones did not influence expression of G6PD. 42 Kurtz and
Wells45 also showed that glucose did not enhance basal or hormone-induced
expression of G6PD. In an equally clear fashion, two studies have demon-
strated that glucose enhances basal and hormone-regulated expression40 •43 with
the maximum observed effect of glucose being to increase basal expression
fivefold. 43 In attempting to resolve this difference, one could consider the
potential role of glucokinase expression. Both of the studies40 •43 that observed
a glucose effect used a Waymouth medium while the other studies 42 ,4 5 did
not. Since induction of glucokinase in primary cultures of hepatocytes is
highly dependent on culture medium conditions, 59 it seems entirely possible
that cells which lack glucokinase would not respond to glucose. The common
observation that high density primary cultures of hepatocytes maintained in
high glucose Waymouth medium produce excessive quantities of lactic acid
clearly demonstrates the cells' ability to use glucose. Presumably, a metabolite
of glucose is responsible for mediating the effect on gene expression. Thus,
it is likely that both fructose and glucose are working through the same
mechanism to produce the increased expression. No information is available
as to the type of regulation that glucose or fructose are exerting.
Nutrient suppression of hepatic G6PD has also been reported. Results
from studies in animals showed that a high level of fatty acid in the diet
suppressed the basal level of G6PD activity and that which could be induced
by a high carbohydrate diet. 5 2 Using primary cultures of hepatocytes, Salati
et al. 43 have demonstrated that linoleate suppressed insulin induction of G6PD
but not that elicited by glucose. Arachidonate was found to be a more effective
inhibitor than linoleate, and both fatty acids caused a right shift in the insulin
dose response curve. The insulin induction of two other lipogenic enzymes
(acetyl CoA carboxylase and malic enzyme) was similarly affected, 43 sug-
gesting that nutrient control of gene expression in this case may be linked to
a mechanism involving desensitization of hepatocytes to the lipogenic action
of insulin. Conversely, glucose enhancement of G6PD induction may rep-
resent sensitization of hepatocytes to insulin. Evidence has been presented to
suggest that fatty acid suppression of G6PD occurs at the level of gene
transcription. 53
VII. OTHER FACTORS CAPABLE OF

REGULATING G6PD
A variety of nonnutrient, nonhormonal agents are known to induce hepatic

G6PD. The effect of ethanol, alone and in combination with hormones and
glucose, has been extensively characterized. 40 •55 Ethanol increases both the
relative rate of G6PD synthesis and mRNA encoding the enzyme. 55 Partic-
ularly noteworthy is the ability of the drug to increase the glucocorticoid
enhancement of G6PD induction. Recent work 60 in this area suggests that a

metabolite of ethanol, acetaldehyde, may be the active species. Other agents,
such as carbon tetrachloride and the drug Primaquine, can also induce the
enzyme. 61 A common property of all these agents is that they are capable of
producing oxidant stress in the hepatocyte. The participation of G6PD in
controlling oxidant stress in a variety of tissues is well known. 61 . 63 The G6PD
gene may be a sentinel for cellular oxidant stress and capable of rapidly
responding to the need for cellular NADPH. If this notion is correct, it may
also explain some of the differences observed in the primary culture model.
For example, Manos et al. 42 reported that G6PD could be induced 12-fold in
hepatocytes maintained in medium with only amino acids as an energy source.
The lack of a readily metabolizable carbohydrate to run through the pentose
phosphate pathway could result in considerable oxidant stress and induction
of G6PD. Incubation of G6PD deficient erythrocytes in carbohydrate and
pyruvate-free medium is known to accelerate lysis. 63
VIII. RECAPITULATION OF WORK

ON HEPATOCYTES
Use of primary cultures of hepatocytes to define precisely the hormonal

and nutrient signals responsible for regulation of G6PD has demonstrated the
central role that insulin occupies. Most investigators have also demonstrated
glucocorticoid involvement, although this class of hormones seems to enhance
insulin's effects rather than directly increasing expression. Participation of
carbohydrate in the control of G6PD expression is more controversial with
suggestions that glucose is involved, 40 .43 is not involved, 42 •45 or that fructose
is involved. 42 Use of different culture conditions and culture media appear to
be responsible for these discrepancies. These differences could also contribute
to changes in the redox state of the cells which may influence G6PD activity.
Fatty acid has been demonstrated to be a negative regulator of expression, 43 •53
while hormones or conditions associated with fasting appear to not be involved
in directly repressing G6PD expression. 41
The level at which hormones and nutrients control G6PD expression has
not yet been established. It is evident that mRNA encoding G6PD is elevated
by insulin treatment42 •50 •51 and the glucocorticoids enhance the level by an
additional increment. 50 •5 1 Thus, there is the suggestion that expression is
controlled at the transcriptional level. This is supported by two recent studies49 •53
in which transcription "run off" assays were used to examine the effect of
carbohydrate refeeding on hepatic G6PD in rats. Both studies demonstrated
a transient increase in transcription within hours of feeding. However, one
study also demonstrated that the G6PD mRNA half-life increased substantially
(1.5 to 2 h for fasted animals vs. 9 to 11 h for refed) following refeeding of
the high sucrose diet. 49 Therefore, there is the expectation that control of
G6PD expression will be exerted at the transcriptional level, at the level of
mRNA stability, and perhaps, as suggested earlier,S 0 ·51 at the translational

level. The availability of molecular probes and appropriate genomic fragments
should permit conclusive experiments to be carried out in primary cultures
of hepatocytes.
IX. RECENT INSIGHTS INTO G6PD GENE

STRUCTURE: IMPLICATIONS FOR REGULATION
OF EXPRESSION
G6PD is an X chromosome-linked gene in mammals, and the enzyme is

subject to enormous genetic variability with over 300 variant forms described
to date in man. 64 ·65 An unusual feature of the deficiency syndrome is that the
loss of enzyme activity is manifest primarily in the red blood cell. 65 Since
classical genetic linkage studies have shown that G6PD is encoded by a single
allele, it is difficult to understand why only the red blood cell would exhibit
the decreased enzyme activity. 65 However, recent studies on L- and R-type
pyruvate kinase provide a possible framework for considering how this might
occur. 66 The L- and R-type isozymes are encoded by the same gene but
transcription is controlled by different promoters. Thus, the first exon in each
mRNA is unique to the species resulting in amino terminal heterogeneity in
the protein expressed in liver L-type vs. red blood cells (R-type).
Recently, partial G6PD protein sequence data for the rat liver protein has
been published. 47 Comparison of this sequence with that of the human red
blood cell protein67 shows a very high degree of sequence homology from
human and drosophilia sources and has shown a very high degree of conser-
vation except at the 5' end of the cDNA. 68 •69 Thus, it appears that the majority
of the G6PD sequence has been conserved in evolution from drosophilia to
man except the amino terminal sequence. Heterogeneity in the amino terminal
sequence may reflect different tissue-specific requirements in the regulation
of G6PD activity and/or control of gene expression as has been observed in
the key glycolytic enzyme, pyruvate kinase. 66
X. SEQUENCE AND STRUCTURE OF THE RAT G6PD

GENE AND PSEUDOGENE
The consensus sequence for rat G6PD derived from eDNA and genomic
clones is presented in Figure 2. We have previously shown that fragments of
this sequence will hybrid select mRNA encoding G6PD46 • Recently, partial
protein sequence for rat liver G6PD has been published47 allowing direct
comparison with our nucleic acid sequence. For the 71 amino acid amino
terminal sequence published by Jeffery et a!. ,47 there exists 100% identity
with our deduced amino acid sequence from the 5' coding region, unequiv-
ocally establishing this consensus sequence for rat G6PD. In addition, com-
parison of our consensus sequence with one of the published sequences for
1 2 3
9.4-
6.7-
4.4-
2. 3-
2 . 0-
0.56-
FIGURE 2. Southern blot of rat liver chromosomal DNA. Rat liver chromosomal DNA was
digested with Pstl, Eco Rl, and Bam HI (Lanes I, 2, and 3). The blot was probed with a
nicknamed eDNA from pGDY-38 46 which is a 3' untranslated region probe. The 9.4 and 9 kb
(lanes 2 and 3) fragment is the functional gene and the 6.7 and 6.2 kb (lanes 2 and 3) fragment
is the pseudogene.
human G6PD 70 revealed an overall sequence identity of 76.6%. If the indi-

vidual regions of the two sequences are compared, there is 87.4% identity in
the coding region while the 5' and 3' untranslated regions demonstrated a
higher degree of divergence at 42.8% and 64.7% homology, respectively.
The high degree of sequence conservation between the rat and human protein
is demonstrated when the deduced protein sequences are compared (93.2%
sequence identity), indicating that many of the nucleotide changes in the
coding region are in the wobble position. While the 3' untranslated flank was
less homologous than the coding sequence, there were, nevertheless, areas
within this region of near 100% homology. The sequences immediately fol-
lowing translation termination, in and around the unusual polyadenylation
TABLE 2
Comparison of Rat/Human Exon/Intron Boundaries
EXON ACTAAATTCA GT IVS (>110 bp) AG1 GAAAACATCA

CGA GCGCAG AC GCGTCAT
EXON 2 GGGTGCATCG GT IVS 2 (>500 bp) AG GGTGACCTGG
EXON 3 CTACCATCTG GT IVS 3 (110 bp) AG GTGGCTGTTC
c
EXON 4 CTTCTTTAAA GT IVS 4 (>250 bp) AG GTCACTCCAG
c G cc
EXON 5 TGAGTAGAC GT IVS 5 (>270 bp) AG AGGCTGGAAC
c T
EXON 6 TGGTGCTGAG GT IVS 6 (228 bp) AG ATTTGCCAAC
EXON 7 GGATCATCAG GT IVS 7 (306 bp) AG GGATGTCATG
c
EXON 8 TGATGAGAAG GT IVS 8 (128 bp) AG GTCAAAGTGT
G
EXON 9 CGGTGGGATG GT IVS 9 (270 bp) AG GAGTACCCTT
A GG
EXON 10 CAGATACAAG GT IVS 10 (119 bp) AG AATGTGAAGC
c
EXON 11 TTGTCCGTAG GT IVS 11 (96 bp) AG TGATGAACTC
c G c cc G
EXON 12 TCTATGGCAG GT IVS 12 (101 bp) AG CCGAGGTCCC
T c
EXON 13
Note: The rat sequence is shown on the top line, and in those positions where the human sequence
is different, the base present in the human sequence is on the bottom line.
signal, and in two other areas within this region are highly conserved between
rat and human.
While the human genome apparently does not contain a G6PD pseudo-
gene,71 a processed pseudogene for G6PD is present (Figure 2, 6.7 kb) in
the rat genome and is more highly represented than the single copy of the
functional gene (Figure 2, 9.4 kb). A genomic clone was isolated which
contained a 6.7 Eco Rl fragment within which was a 3.2 kb sequence having
the characteristics of a processed pseudo gene for G6PD. Sequence analysis
(Figure 3) revealed the absence of introns and the presence of a poly A tail,
indicative of a pressed pseudogene. As in the case of other processed pseu-
dogenes, a short direct repeat (GAAAAGACAA) which probably represents
the site of insertion was located flanking the sequence. It occurs immediately
3' of the poly A tail and 65 base pairs 5' of the "translation start" site.
Numerous base substitutions were found throughout the coding, 5' and 3'
untranslated regions. The overall sequence identity between the pseudogene
and consensus sequence is 95.8%.
•I•
COliS 6T&MI:6T6nT6&U61:&&tAACTAAAmASAMAUTIIA6AIIU56TS&tTTT6A6CC66ACCCA6&T6T6T6&6ATC 41
I I I
P6 ~C666CTAT6CAA!ttA&AltliT6AACiliTTT66CA6CA&tAACTAAATTtASAAAACAT A6A6tA&ATi&CTTTiA&tTS&AtCCA&iT616T666ATC
I I
CONS CT~&IM6AITTiTACtA666T&AT6tCmUCW&CT&ATACAUC.ITATTTATtATI:ATi&ST6CATmiT6ACt166ttAA6AA&AA&ATTTATmAmTCT66T6SCT6 161
I I II II
Pi CTM6&1M6Ain&TACCAT&ST&ATSCCmCioCAAS&n&ATACACA--TATTTATCATCAT666T6tATCS66T&ACCT66tCAAiAASAA&ATTTATCCTACCATCTS6T66CTS
TTI:t&&&AT66ttTTCIACctiAASoiCACtTTI:ATl6TA66tTAT6ttC6CTCAC&ICTCACA6166AI6ACATCC6CAAACA6A6~6AitcrTTCTTTAAAI61CACTCCA6AASAAASA 1111
I I I I I
Pi TTm&&AT&6CtTTIIACUAAA&ACACC1TtATTSTA66CTATim&nCAT&ACTCACA6T&SA16ACATCC6CAACCAiA616A&CCCTTCTTTAAAiltACTCCASAASAAA6A
COliS CttAAOI:TA6ASSA6TTCTTT6CCC&TAACTCCTATiTA6t166tCA6TATSATiATCCAiCmCTACAAiCACCTCAACA6tC-ACATSAA16tCCTSCACCAi66AATSCAS6ctAA 101

II I I I I I I I
P6 CCCAA61:1166Ai6A6mTTT&ttT&TAACACCTAT6TA&CTS&CCA6TATSAT6ATCCA&CCmTACAA6CAmCAATA&mAI:ATiAAT6ttCISI:ACCA6666Al&CA66tTAA
I
COliS mm&TTtTACCT&SCCTT6CCCmACTmTAT&AA&CA6TI:ACCAA6AACATTCAA6A6ATCI6tAT6AmAiACA&SCTi6AACmATtATA6Ti6AiAASCCCTTC666AS 111
I I I I I
P6 CCATt1AmTATt16CCCTT&cemcmmTAT&AAiCA6TCACtAA6AACATTCAAiAiATtT6TAT&A6TCA6ACA6iCT&SAACC&CATCATAi16iASAAStcCTTts6GAG
COliS A6Att16CA6A6CTCCAATCAACT6Tt6AACCACATtltt!CTtT6TTIC616A66ACCA6ATtTACC6CATT&ACCAC1AtCT&66CAAA6ASATSSTCCAiAACCTCATSSTSCTGAS Ml
I 111111111111
Pi A6ACCmA6A6CTCCAATCAACT&n6AoiCUCATCTmmT&TTTC&T6A66ACCA6ATtTACC6CATTSACCACTAttT66------------TCCA6AAmW&STSCTSA6
I
CONS ATTTSCI:AACA66ATCTT!&SACCCATCTi&AATC~ACAACATT6tTTST-----iliATCCTIACATTIAAA6A6CmTT6iTACISASSSTCSTSSS66tTATTTT&AT&AA 161
I 11111111 I I
P6 ATTI6tCAACA66ACCTTT&SACCCATCI&&AATC6A6ACAACATTSCTTSTmmTT&T&AmTTACATTTAAAiAStcCTT!iATACI&Ai66TTST&S&S&nATTTT6ATSAA
I I
l:llNS TTT666ATI:ATCA666AT6TI:AT&CA6AACCAUTCCTSCA6AT6TT6TmTA&T&SCCAT66AAAA6CCTSCmTACA6ATTI:A&ATiATSTCC6TSAT6A6AA661CAAA616TTA 1181
I II I I I I
Pi TTTA66ATCATA66&SAT6TTATACA&MI:CACmcmA&AT&TT6TiTCTA6166ttAT66AAAA6CCT6tCTtTACA6ATTCAAAT6ATiTCCA16Al&A6AAi61CAAAST6TT A
COliS AAAT6TATtTI:A&A&iT&6AAACT&ACAAC6T6&mTT66CCASTATiT66&iAACCCI:Ai!iiA6AAS~AASCTACCAAT666TACTTA6AliACCCCACA&TACCCCAT66&1CT 1001

I I II
P6 AAAT6TATCTCA&AiGT&iAAATT&ACAAC&T&iTCCTTi6tCA6TAT6T&i66AAttttA6T&iA&AA66A&AA&iT ACtAATi66TACTTAiATiACtcCACA6T ACCCCAT&STCCT
I
COliS ACCACT6tTACCITTSCA6CA&CTiTCCTtTAT6T&&A&AAT&AAC&&T&i&AT&iA&TACCCTTI:ATm&eSCTiliSCAM6CTt16AATiA6t&CAAA6t16AA61&AiACTTCA6 1111
I • II I I I
16 ACCACT&CTACCTTA61:A6CA&CT&TCCTCTAT6T&iA&AAT&AACAAT6&iATi~TACI:tTTCATCCT&c.ICT6T66CAAA6CTCT&AAT&A6T&CAAA&tT6AA&T6AAACTTCA6
TTWC&AT6T&itA6&T&ACATCTTCI:ACI:A&CA616tAASCiTAAC&A&CT66TCATCC6TiTii:A6tcrAAT6A6&t:i6TATACACI:AA6AT&AT6ACCAA&AASI:tT6&CAT&TTC 1141
I I I I IIIII I I
P& TTCTIC&AT&T&itA6&T&ACATCTTCCACCA&CA6T61:AAST6TAoiCAA6CT66TCATCTiT6T6CA--AT&A66C6&TATACACCAA&AT&AT&ACCAA&AACCCT66tCT6TTC
I
TTCAACttT6A66A&Tti&A61:T&iAmAACCTAT66CAAI:A&ATACAA&AAT6T&AA&CTI:ttT6ATSCCTAT6AAC6CCTCATCCT6&AT6TCTTCT61666A6tCAAAT&CACTTT 1365
11111111111111111111111111111111111111111111111111111111111111111111111
P& TTCAAI:ttT6A66A&TCT&A6CT&ilmAACCTAT&iCAACA6A-----------------------------------------------------CTTT
CIIIIS ITCCiTA6~&oiT&oiACTCA666M6CCT66CiTATCTTI:ACACCATTSCT&c.ICAA&ATT&ATC&A~CCI:ASCCI:ATCCC6TAT6TCTAT6&t:Jc&A66TCCCACA6A6&CA IIIII

I I 1111 I I I I I
PI 61CCATMT&oiT6AACTC66oii:ACitCT6&CAIATCTTCACAACAIT6CT&c.ICAA6ATT66TC~A6AASCCCCA6CCCATTCCTTAI6TCTAT&iCA6CC&AiiTCCCACA6A&&CA
C1IIS &oiT&A&CT&AT&oiA6A&oi61666CTTI:tA6TAT&A666TACCTACM6T66616AACCCTI:ACAASCTliT&6AAACTTAI:ACI:ATCT&I:ACTCT6tCTCTTtT6&CCACCCTT 1601

1111111111 I
P& 6AT&A&CT&oiT6AA&oi6AIT&66CTTI:tA6TAT&A666T-----&6616AACCCT!:ACAA61:T m&&AAACTTACACCATCT&CATTCT&ttTCTTCTmmmr
CONS TCI&CATCT&I:tCTICTI:ACCATCTAACCCTCIATTA&iACTITT&ACCTCATATT66AAA6ACTTT666ACCATA66CCTTA&CTACACA TICTA6TCm666CTT AOittACCATTC 1711

I I
P1 TCTICATCTIIUTlCTCAa:ATCTAACCCTATATTA&iACT~TTIIACtTCAIATT-.cTTTIISUCCATA66CCITA6CTACACA TIC! A61CTCT666CTTA66CCACCITTC
C1IIS TITI:tTATICT&CT&CI:ACT&CCACTACCACTAAICCUSCTACAmCTI:A&ATACI:A&6!:ATTI:AAAAC&CATT6CAAT&CTTTCA66ACCAttACT6TCCCTATCT&A&CCAmAr lEI:

I I
PI IITI:tTATICI&CTiti:ACT&CCACTACCACTAA&CCCA6CTACATICCTCAAATICCA&iCATTCAAAACACATT&CAAT6CTTTCAO&ACCACCACT&TCCCT A1CT6A6CCACCCA1
CIIIIS CTTTCCACAA6ACCT6AATI:ACCTCCTCCCCTCAATCtcCT&CA6AAA6AAC6CCTATCA&ltT&TmT&&ACTmTAA6ATAi6A6TTA66AACAATT666A66A6CCTT6&6CC 1 TIIi I

I I I I
I& CTTTCI:ACAA6ACCT&AATI:ACmCTCCCCTCCAACTm&CA&AAT6AACSCCTATCA&Tt16TCCt166ACTCCTTAA6ATAi&A6TTA66AACAATT&66A6&A&CCTT666CCTT
CIIIIS &iA666ACAAT&ACCAAACCACACITCCCT6A6ACT616i&CAA&CTCCTCAAAACTTAAA6T&ATCAA6&ACACCCATCT6A6A66AtCT6CCCATA6CCACACTA6CCTTA&T&ClAC 108l

II I I
P& &iA&&iACAAT&ACCAMCI:ACACTTtcCT&A&ACT6T&i&I:AAATTCCTCAAAACITAAA6T&AT6AA66ACACCCATTT&A&Ai6ACCT6CCCATAmACACTA6CCTTA6l&CT.C
em TTSACAna:rccrcACCA&CT&iAA&Ao\CTCTCATmmrmAATATTTT&&&&imTASATArcrmAA!tAAnmrAsrccArA&TCA&ccrcArccAACCCAT&&&eA& 1101
II I I
P& TT6ACATTCCTCCICACCA61:166AA&AAACCTCAT6tT6CCTA&C6ATATTTT66666tCATA6ATATCTCCTAAACAATTCCATA&TCCATA6TCAAttTCATCCAACCCAT666CA6
CONS CCTCCTTACCAM6&1AiBTAA6A&I:A&CA&CTA6AATTTTCCTACCCCAAmT&CCCCTCAAAAAAAAAAAAAAAAAA 11!1

I I Ill I
P& CCII:ATTACCAM66AA66TAA6A&l:A&AOCTA6AATTITCCIACcttAAmm T A CITCAAAAAA&TTAAAAAAAAiAAA~
The coding region and intron/exon boundaries were sequences from the
genomic fragments containing the functional G6PD gene from rat liver (Table
2 and Figure 3). The structure of the rat liver gene was found to be remarkably
similar to that of the human gene as reported by Martini et a!. 71 The number
of exons (13), the position of the introns in the coding sequence, and the size
of the corresponding exons are identical in both genes. The intron boundaries
(indicated by arrows in Figure 3) are highly conserved although intron size
varies considerably (Table 2). IVS I does not intersect the coding region, and
the exon/intron junction at this site in the 5' untranslated region is not con-
served. Exon 13 contains all of the 3' untranslated region, the termination
codon, and 88 base pairs of coding sequence.
Comparison of the rat consensus G6PD sequence with that of the human
demonstrates a high degree of homology except at the sequences encoding
the amino terminal end of the protein. However, two different sequences
have been reported for the human amino terminal sequence of G6PD. One
was determined by sequence analysis of the protein from red blood cells 67
and the other deduced from cDNAs produced from a variety of human cell
types and transformed celllines. 69 •70 The sequence reported by Jeffery et al. 47
for the rat liver protein corresponds to that deduced from the human cDNAs.
In order to determine the location of this difference in the sequence infor-
mation, we compared eDNA and protein sequence data from rat and human
with that of the G6PD gene. As shown in Figure 4, the point of divergence
is localized to amino acid position #35 in the G6PD protein sequence. At
the nucleic acid level, the site of sequence divergence is immediately 5' of
the IVS 2 region of the gene. IVS 2 is extremely large (> 11 kb in the human
gene) and is the first intron to intersect the coding sequence. In the case of
the rat liver sequence, only 120 bases that encode protein are upstream of
this intron. That this structural feature has been preserved in evolution is
indicative of a possible important structural role in gene function. Either
alternative splicing or use of different promoters as in the case of R-type vs.
L-type pyruvate kinase could give rise to amino terminal heterogeneity for
G6PD.
FIGURE 3. Comparison of the consensus coding sequence for rat G6PDH with a processed
pseudogene. Nucleotide residues are numbered from the start of translation. The consensus
sequence for G6PDH (upper, CONS) is from complete, overlapping eDNA clones from rat liver
and brain libraries and from the coding region of the functional gene. Each clone was sequenced
twice in both directions to verify the authenticity of the sequence. The pseudogene sequence
(lower, PG) starts with the direct repeat (underlined) and ends with the same sequence immediately
3' of the poly A taiL Bases that are different between the two sequences are indicated with a
bar. The polyadenylation signal, the start, and termination of translation are denoted by the box.
The positions of the 12 introns are denoted by arrows.
Human~7:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
eDNA
Rat
cDNA~~_Q~~~~~~~~~~~~~~~~~~~~~~~~~~~
protein
HUMAN
protein
IVS2
FIGURE 4. Amino terminal protein and 5' coding sequence relationships between rat and
human G6PDH. The rat protein sequence (I) and the human protein (19) and eDNA sequence
(17) are from published reports. Identities in the sequences are posed. The position of the second
intron of the G6PDH gene is noted by the arrow. The rat amino acid positions from 29 to 44
are shown and for the human, positions 46 to 61.
XI. CONCLUSIONS
In conclusion, genetic studies have established that G6PD is encoded by

a single copy X-linked gene in mammals, and the expectation has been that
the protein would be the same in all tissues. Comparison of the nucleic acid
sequence encoding G6PD from drosophilia to man has revealed a high degree
of sequence conservation except for that encoding the extreme amino terminal
end of the protein. The conservation of the coding sequence highlights the
importance of the protein and the reaction carried out by it for metabolism
in complex biological systems. The amino terminal heterogeneity may be
important in tissue-specific expression and control of G6PD activity. In this
regard, G6PD occupies a delicate place at the interface of carbohydrate and
lipid metabolism. While it is primarily a housekeeping enzyme in most tissues,
expression is clearly regulatable in others. Liver is most notable since, as
discussed above, a variety of hormones, nutrients, and other agents capable
of producing oxidant stress regulate the activity of the enzyme. Thus, the
cellular machinery responsible for controlling expression would be expected
to be complex in order that regulation be accomplished in those tissues. Here
it is inducible while maintaining a constitutive "housekeeping" level in other
tissues. The current problems associated with isolating and characterizing the
promoter for the G6PD gene may be a reflection of the complex nature of
the putative control region.
REFERENCES
1. Tepperman, H. M. and Tepperman, J., The hexose monophosphate shunt and adaptive
hyperlipogenesis, Diabetes, 7, 478, 1958.
2. Glock, G. E. and McLean, P., A preliminary investigation of the hormonal control of
the hexose monophosphate oxidative pathway, Biochem. J., 61, 390, 1955.
3. Fitch, W. M. and Chaikoff, J. L., Extent and patterns of adaptation of enzyme activities
in liver of normal rats fed diet high in glucose and fructose, J. Bioi. Chern., 235, 554,
1960.
4. Rudack, D., Chisholm, E. M., and Holten, D., Rat liver glucose 6 phosphate dehy-
drogenase. Regulation by carbohydrate diet and insulin, J. Bioi. Chern., 246, 1249, 1971.
5. Potter, V. R. and Ono, T., Enzyme patterns in rat liver and Morris hepatoma 5123
during metabolic transitions, Cold Spring Harbor Symp. Quant. Bioi., 26, 355, 1961.
6. Tepperman, J. and Tepperman, H. M., Effects of antecedent food intake pattern on
hepatic lipogenesis, Am. J. Physiol., 193, 55, 1958.
7. Tepperman, H. M. and Tepperman, J., On the response of hepatic glucose 6 phosphate
dehydrogenase activity to changes in diet composition and food intake pattern, Adv.
Enzyme Reg., !, 121, 1963.
8. Sassoon, H. F., Dror, Y., Watson, J. J., and Johnson, B. C., Dietary regulation of
liver glucose 6 phosphate dehydrogenase in the rat: starvation and dietary carbohydrate
induction, J. Nutr., 103, 321, 1973.
9. Szepesi, B. and f'reedland, R. A., Differential requirement for de novo RNA synthesis
in the starved refed rat; inhibition of the overshoot by 8-azaquanine after refeeding, J.
Nutr., 99, 449, 1969.
10. Szepesi, B. and Berdanier, C. D., Time course of the starve-refeed response in rats;
the possible role of insulin, J. Nutr., 101, 1563, 1971.
11. Szepesi, B. and Michaelis, 0. E., Comparison of the fructose effect and the starve-
refeed response of rat liver enzymes, Life Sci., II, 113, 1972.
12. Michaelis, 0. E. and Szepesi, B., Effect of various sugars on hepatic glucose 6 phosphate
dehydrogenase, malic enzyme and total lipid and the rat, J. Nutr., 103, 697, 1973.
13. Williams, B. H. and Berdanier, C. D., Effects of diet composition and adrenalectomy
on the lipogenic responses of rats to starvation-refeeding, J. Nutr., 112, 534, 1982.
14. Baltzell, J. K. and Berdanier, C. D., Effect of the interaction of dietary carbohydrate
and fat on the responses of rats to starvation-refeeding, J. Nutr., 115, 104, 1985.
15. Berdanier, C. D., Johnson, B., and Buchanan, M., Interacting effects of menhaden
oil and sucrose on the lipogenic response to starvation-refeeding, Nutr. Res., 9, 1167,
1989.
16. Johnson, B. J. and Berdanier, C. D., Effect of menhaden oil on the responses of rats
to starvation-refeeding, Nutr. Rep. Int., 36, 809, 1987.
17. Bue, J. and Berdanier, C. D., Effect of antecedent diet on the responses of BHE rats
to starvation-refeeding, Nutr. Rep. Int., 35, 791, 1987.
18. Szepesi, B., Berdanier, C. D., and Michaelis, 0. E., Effect of age on the responses
to starvation-refeeding in the rat, Proc. Soc. Exp. Bioi. Med., 147, 195, 1974.
19. Kim, M.-J., Pan, J.-S., and Berdanier, C. D., Age and thyroid hormone as factors
in the responses of BHE rats to starvation-refeeding, Proc. Soc. Exp. Bioi. Med., 183,
273, 1986.
20. Bois-Joyeux, B., Chanez, M., Aranda-Haro, F., and Peret, J., Age dependent hepatic
lipogenic enzyme activities in starved-refed rats, Diabetes Metab., 16, 290, 1990.
21. Berdanier, C. D., Effects of estrogen on the responses of male and female rats to
starvation refeeding, J. Nutr., 111, 1425, 1981.
22. Smith, E. R. and Barker, K. L., Effects of estradiol and NADP on the rate of synthesis
of uterine glucose 6 phosphate dehydrogenase, J. Bioi. Chern., 249, 6541, 1974.
23. Szepesi, B. and Michaelis, 0. E., Problems in the demonstration of the disaccharide
effect due to stress during shipment of rats, Nutr. Rep. Int., 12, 299, 1975.
24. Michaelis, 0. E., Scholfield, D. J., Nace, C. S., and Reiser, S., Demonstration of
the disaccharide effect in nutritionally stressed rats, J. Nutr., 108, 919, 1978.
25. Szepesi, B., Metabolic memory: effect of antecedent dietary manipulations on subsequent
diet induced responses of rats. I. Effects on body weight, food intakes, glucose 6 phosphate
dehydrogenase and malic enzyme, Can. J. Biochem., 51, 1604, 1973.
26. Szepesi, B., Effect of previous caloric restriction on two NADP-linked dehydrogenases
of rat liver, Nutr. Rep. Int., 10, 127, 1974.
27. Szepesi, B., Berdanier, C. D., Diachenko, S. K., and Moser, P. B., Effect of length
of starvation, refeeding and 8-azaquanine on serum insulin and NADP-linked dehydro-
genases of rat liver, J. Nutr., 101, 1147, 1971.
28. Homan, H. D. and Berdanier, C. D., Effect of adrenalectomy on the enzyme responses
of starved-refed rats, Nutr. Rep. Int., 22, 213, 1980.
29. Gimenez, R. S. and Johnson, B. C., Pair feeding in the dietary control of G6PD, J.
Nutr., Ill, 260, 1981.
30. Freedland, R. A., Cunliffe, T. L., and Zinkl, J, G., Effect of insulin on enzyme
adaptations to diets and hormones, J. Bioi. Chem., 241, 5448, 1966.
31. Szepesi, B., Vegors, R., and DeMosey, J, M., On the possible role of insulin in the
starve-refeed response, Nutr. Rep. Int., 5, 281, 1972.
32. Berdanier, C. D., Wurdeman, R., and Tobin, R. B., Further studies on the role of
the adrenal hormones in the responses ofrats to meal feeding, J. Nutr., 106, 1791, 1976.
33. Wurdeman, R., Berdanier, C. D., and Tobin, R. B., Enzyme overshoot in starved-
refed rats: role of adrenal glucocorticoid, J. Nutr., 108, 1457, 1978.
34. Bouillon, D. J, and Berdanier, C. D., Role of glucocorticoid in adaptive hyperlipo-
genesis in the rat, J. Nutr., 110, 286, 1980.
35. Berdanier, C. D. and Shubeck, D., Interaction of glucocorticoid and insulin in the
responses of rats to starvation-refeeding, J. Nutr., 109, 1766, 1979.
36. Nessmith, S., Baltzell, J., and Berdanier, C. D., Interaction of glucocorticoid and
thyroxine in the responses of rats to starvation-refeeding, J. Nutr., 113, 2260, 1983.
37. Kletzien, R. F., Pasiza, M. W., Becker, J, E., Potter, V. R., and Butcher, F. R.,
Induction of amino acid transport in primary cultures of adult rat liver parenchymal cells
by insulin, J. Bioi. Chem., 251, 3014, 1976.
38. Inoue, C., Yamamoto, H., Nakamura, T., Ichihara, A., and Okamoto, H., Nico-
tinamide prolongs survival of primary cultured hepatocytes without involving loss of
hepatocyte-specific functions, J. Bioi. Chem., 264, 4747, 1989.
39. Dunn, J, C. Y., Yarmush, M. L., Koebe, H. G., and Tompkins, R. G., Hepatocyte
function and extracellar matrix geometry: long-term culture in a sandwich configuration,
Fed. Proc., 3, 174, 1989.
40. Kelley, D. S. and Kletzien, R. F., Ethanol modulation of the hormonal and nutritional
regulation of glucose-6-phosphate dehydrogenase activity in primary cultures of rat he-
patocytes, Biochem. J., 217, 543, 1984.
41. Nakamura, T., Yoshimoto, K., Aoyama, K., and Ichihara, A., Hormonal regulations
of glucose-6-phosphate dehydrogenase and lipogenesis in primary cultures of rat hepa-
tocytes, J. Biochem., 91, 681, 1982.
42. Manos, P., Nakayama, R., and Holten, D., Regulation of glucose-6-phosphate de-
hydrogenase synthesis and mRNA abundance in cultured rat hepatocytes, Biochem. J.,
276, 245, 1991.
43. Salati, L. M., Adkins-Finke, B., and Clarke, S. D., Free fatty acid inhibition of the
insulin induction of glucose-6-phosphate dehydrogenase in rat hepatocyte monolayers,
Lipids, 23, 36, 1988.
44. Kelley, D. S., Nelson, G. J., and Hunt, J. E., Effect of prior nutritional status on the
activity of lipogenic enzymes in primary monolayer cultures of rat hepatocytes, Biochem.
J., 235, 87, 1986.
45. Kurtz, J. W. and Wells, W. W., Induction of glucose-6-phosphate dehydrogenase in

primary cultures of adult rat hepatocytes: requirement for insulin and dexamethasone, J.
Bioi. Chern .. 256, 10870, 1981.
46. Kletzien, R. F., Prostko, C. R., Stumpo, D. j., McClung, J. K., and Dreher, K. L.,
Molecular cloning of DNA sequences complementary to rat liver glucose-6-phosphate
dehydrogenase mRNA: nutritional regulation of mRNA levels, J. Bioi. Chern., 260,
5621, 1985.
47. Jeffery, J., Soderling-Barros, j., Murray, L.A., Hansen, R. j., Szepesi, B., and
Jornvall, H., Molecular diversity of glucose-6-phosphate dehydrogenase: rat enzyme
structure identifies NH 2-terminal segment, Proc. Nat/. Acad. Sci. U.S.A., 85, 7840,
1988.
48. Ho, Y.-S., Howard, A. J., and Crapo, j. D., Cloning and sequence ofacDNAencoding
rat glucose-6-phosphate dehydrogenase, Nucleic Acids Res., 16, 7746, 1988.
49. Prostko, C. R., Fritz, R. S., and Kletzien, R. F., Nutritional regulation of hepatic
glucose-6-phosphate dehydrogenase: transient activation of transcription, Biochern. J.,
258, 295, 1989.
SO. Stumpo, D. J. and Kletzien, R. F., Regulation of glucose-6-phosphate dehydrogenase
mRNA by insulin and the glucocorticoids in primary cultures of rat hepatocytes, Eur. J.
Biochern., 144, 497, 1984.
51. Fritz, R. S., Stumpo, D. S., and Kletzien, R. F., Glucose-6-phosphate dehydrogenase
mRNA sequence abundance in primary cultures of rat hepatocytes: effect of insulin and
dexamethasone, Biochern. J .• 237, 637, 1986.
52. Kelley, D. S., Nelson, G. J., Serrato, C. M., and Schmidt, P. C., Nutritional regulation
of hepatic lipogenesis in the rat, Nutr. Res., 7, 509, 1987.
53. Katsurada, A., Iritani, N., Fukuda, H., Matsumura, Y., Noguchi, T., and Tanaka,
T., Effects of nutrients and insulin on transcriptional and post-transcriptional regulation
of glucose-6-phosphate dehydrogenase synthesis in rat liver, Biochirn. Biophys. Acta,
1006, 104, 1989.
54. Berdanier, C. D., Role of glucocorticoids in the regulation of lipogenesis, Fed. Proc.,
3, 2179, 1989.
55. Stumpo, D. j. and Kletzien, R. F., The effect of ethanol, alone and in combination
with the glucocorticoids and insulin, on glucose-6-phosphate dehydrogenase synthesis
and mRNA in primary cultures ofhepatocytes, Biochern. J., 226, 123, 1985.
56. Mariash, C. N., Kaiser, F. E., Schwartz, H. L., Towle, H. C., and Oppenheimer,
J. H., Synergism of thyroid hormone and high carbohydrate diet in the induction of
lipogenic enzymes in the rat, J. Clin. Invest., 65, 1126, 1980.
57. Fritz, R. S. and Kletzien, R. F., Regulation of glucose-6-phosphate dehydrogenase by
diet and thyroid hormone, Mol. Cell Endo., 51, 13, 1987.
58. Yoshimoto, K., Nakamura, T., and Ichihara, A., Reciprocal effects of epidermal
growth factor on key lipogenic enzymes in primary cultures of adult rat hepatocytes, J.
Bioi. Chern., 258, 12355, 1983.
59. lynedjian, P. B., Jotterand, D., Nouspikel, T., Asfari, M., and Pilot, P., Transcrip-
tional induction of glucokinase gene by insulin in cultured liver cells and its repression
by the glucagon-cAMP system, J. Bioi. Chern., 264, 21824, 1989.
60. Teet, j. F., Kletzien, R. F., Ginsberg, L. C., Stevens, G. J., and Stapleton, S. R.,
Acetaldehyde increases mRNA levels of glucose-6-phosphate dehydrogenase in rat he-
patocytes in culture, Fed. Proc., 5, 310 (Abstr.), 1991.
61. Tarlou, A. R., Brewer, G. j., Carson, P. E., and Alving, A. S., Primaquine sensitivity,
Arch. Intern. Med., 109, 137, 1962.
62. Shapiro, B. M., The control of oxidant stress at fertilization, Science, 252, 533, 1991.
63. Beutler, E., Robson, M., and Buttenwies, E., The mechanism of glutathione destruction
and protection in drug-sensitive and non-sensitive erythrocytes, J. Clin. Invest., 36,
617, 1957.
64. Kirkman, H. N. and Hendrickson, E. M., Sex-linked electrophoretic differences in

glucose-6-phosphate dehydrogenase, Am. J. Human Genet., 15, 241, 1963.
65. Luzzato, L., Glucose-6-phosphate dehydrogenase: genetic and haematological aspects,
Cell. Biochem. Function, 5, 101, 1987.
66. Noguchi, T., Yamada, K., Inoue, H., Matsuda, T., and Tanaka, T., The L-and R-
type isozymes of rat pyruvate kinase are produced from a single gene by use of different
promoters, J. Bioi. Chern., 262, 14366, 1987.
67. Jakizawa, T., Huang, T.-Y., Ikuta, T., and Yoshida, A., Human glucose-6-phosphate
dehydrogenase: primary structure and eDNA cloning, Proc. Nat!. Acad. Sci. U.S.A., 83,
4157, 1986.
68. Fouts, D., Ganguly, R., Gutierrez, A. G., Lucchesi, J. C., and Manning, j. E.,
Nucleotide sequence of the Drosophilia G6PD gene (ZW +) and comparison with the
homologous human gene, Gene, 63, 261, 1988.
69. Vulliamy, T. J., D-Urso, M., Battistuzzi, G., Estrada, M., Foulkes, N. S., Martini,
G., Calabro, V., Poggi, V., Giordano, R., Town, M., Luzzatto, L., and Persico,
M. G., Diverse point mutations in the human glucose-6-phosphate dehydrogenase gene
cause enzyme deficiency and mild or serve hemolytic anemia, Proc. Nat/. Acad. Sci.
U.S.A., 85, 5171, 1988.
70. Persico, M. G., Vigletto, G., Martini, G., Toniolo, D., Paonessa, G., Moscatelli,
C., Dono, R., Vulliamy, T., Luzzatto, L., and D-Urso, M., Isolation of human glucose-
6-phosphate dehydrogenase eDNA clones: primary structure of the protein and unusual
5' non-coding region, Nucleic Acids Res., 14, 2511, 1986.
71. Martini, G., Toniolo, D., Vulliamy, T., Luzzatto, L., Dono, R., Viglietto, G., D'Urso,
M., and Persico, M. G., Structural analysis of the X-linked gene encoding human
g1ucose-6-phosphate dehydrogenase, EMBO J., 5, 1849, 1986.
Chapter 9
NUTRITIONAL AND HORMONAL REGULATION

OF GENES ENCODING ENZYMES INVOLVED
IN FAT SYNTHESIS
Hei Sook Sui, Naima Moustaid, Kenji Sakamoto, Cynthia Smas,

Nick Gekakis, and Ann Jerkins
TABLE OF CONTENTS
I. Introduction ...................................................... 208
II. Cloning of Nutritionally Regulated eDNA Sequences ............ 209
III. In Vivo Nutritional and Hormonal Regulation of

Gene Expression ................................................. 211
IV. Regulation of Transcription ...................................... 214
V. Differentiation-Dependent Expression in 3T3-Ll Cells .......... 216
VI. Hormonal Regulation of Gene Expression in 3T3-Ll

Adipocytes ....................................................... 219
VII. Conclusions ...................................................... 221
References ............................................................... 224
0-8493-6961-41931$0.00 +$.50
© 1993 by CRC Pre>s. Inc. 207
I. INTRODUCTION
Fatty acid and triacylglycerol synthesis for energy storage is known to

be regulated in response to the nutritional and hormonal state of the animal.
Subjecting rats to a few days of fasting results in the inhibition of lipogenesis;
when fasted animals are subsequently fed a diet high in carbohydrate and low
in fat, there is a prompt rise in the production of fatty acids and triacylglycerol
to levels above those seen in normally fed rats. '- 2 Under these conditions,
excess glucose in the cell is first converted to pyruvate via glycolysis. Pyruvate
is converted to acetyl CoA which is used for the synthesis of long chain fatty
acids, primarily palmitate. NADPH required for fatty acid synthesis is pro-
duced by the malic enzyme or via the pentose phosphate pathway. The fatty
acids produced are then used for esterification onto glycerol-3-phosphate to
generate triacylglycerol. Many key enzymes involved in these metabolic path-
ways are induced during the refeeding period. 1 The increase in the concen-
tration of enzymes, such as of fatty acid synthase, has been shown to be due
to the increased synthesis of enzyme protein. 3
It may be that expression of genes encoding some of these inducible
enzymes is regulated by nutritional and hormonal stimuli via common mech-
anisms. Changes in nutrient intake bring changes in circulating glucose and
amino acid concentrations, which, in tum, signal secretion of hormones. It
is generally accepted that insulin in the circulation is elevated during feeding
and has positive effects, whereas the glucagon level is decreased and has
negative effects on the expression of enzymes participating in fatty acid and
triacylglycerol synthesis. 4 Circulating T3 is also known to decrease during
fasting and increase during feeding. In addition, a high concentration of
glucose is known to be required for the induction of these enzymes. The
actual molecular mechanisms underlying this adaptive metabolic process,
however, are, as yet, poorly understood. It is, in general, accepted that the
thyroid hormone-receptor complex itself interacts with a thyroid hormone
response element to activate gene transcription. 5 In cAMP mediated tran-
scriptional activation, a nuclear protein is phosphorylated by the cAMP-
dependent protein kinase and interacts with the cAMP response element. This
protein is called the cAMP response element binding protein (CREB). 6 How-
ever, transcriptional regulation by the major anabolic hormone insulin is not
well understood. Moreover, how a combination of hormones, which regulate
specific genes coding for enzymes participating in fat synthesis, brings about
an appropriate response to dietary stimuli is also not known.
This review will discuss the regulation of expression of genes for three
enzymes participating in fatty acid and fat synthesis, liver phosphofructokinase
(E.C. 2. 7 .1.11), fatty acid synthase (E.C. 2.3.1.85), and p90 (E.C. 2.3.1.15).
These three proteins have been selected for investigation in our laboratory to
compare their regulation at the gene level by nutritional and hormonal stimuli.
Sui et al. 209
Each catalyzes an important step in different pathways through which met-

abolic flux increases during fat synthesis for energy storage. Phosphofruc-
tokinase is a key regulatory enzyme in glycolysis and exists in three isozymes
specific to the tissues in which each are found: liver, muscle, and brain. It
has been shown that the concentration of the major liver isozyme, a tetramer
of the liver type subunit, was sensitive to nutritional and hormonal changes
in that it was decreased by starvation or by insulin deprivation.9·1° Fatty acid
synthase, a dimer of 250 kDa subunit, carries out all the steps for synthesis
of fatty acids from acetyl CoA and malonyl CoA. The fatty acid synthase
activity is regulated by changes in enzyme concentration. Allosteric effectors
or covalent modifications of this enzyme are not known. 2 p90 has been
tentatively identified as glycerol-3-phosphate acyltransferase, an enzyme which
catalyzes the esterification of glycerol-3-phosphate to 1-monoacylglycerol-3-
phosphate (lysophosphatidate), the committed step in triacylglycerol biosyn-
thesis. Glycerol-3-phosphate acyltransferase activity is found in both micro-
somal and mitochondrial membranes, and it is thought that the mitochondrial
form in particular is decreased by starvation and increased by insulinY
II. CLONING OF NUTRITIONALLY REGULATED

eDNA SEQUENCES
To study the regulatory mechanisms of gene expression, we first isolated

eDNA sequences coding for these enzymes using the following strategies.
We isolated liver poly( A+) RNA from mice under the condition of high fat
synthesis by first fasting and then feeding a diet high in carbohydrate but
devoid of fat. Since fatty acid synthase has a large subunit, we enriched
poly( A+) RNA for higher size messages by sucrose density gradient centri-
fugation before eDNA library construction and screening. 11 Replica filters of
this library were hybridized to the total eDNA synthesized by reverse tran-
scribing mRNAs prepared from fasted or refed mouse liver; clones that hy-
bridized preferentially to the refed liver eDNA were picked and characterized.
The cloned eDNA sequence for murine fatty acid synthase hybridized to a
single 8.2 kb mRNA. This is in contrast to avian and rat tissues where two
species of mRNA for fatty acid synthase are detected. 12 · 13 In addition to the
fatty acid synthase eDNA clone, we have also isolated eDNA sequences to
three other mRNAs of unknown identity. The open reading frame of one such
mRNA with a size of 6.8 kb (p90 mRNA) showed sequence homology to E.
coli glycerol-3-phosphate acyltransferase and was tentatively identified as the
murine enzyme. 14 The observation that glycerol-3-phosphate acyltransferase
activity in rat liver, especially in the mitochondrial fraction, was increased
by feeding a high carbohydrate diet supports this tentative identification. For
isolation of the eDNA sequence for liver phosphofructokinase, an enzyme
known to be present in very low concentration, the polysome immunopre-
L-PFK FAS p90

3 .0 kb 8.2 kb 6.8 kb
fasted refed fasted refed fasted refed
FIGURE 1. Relative levels of liver (B-type) phosphofructokinase, fatty acid synthase, and
p90 rnRNAs during fasting and refeeding of a high carbohydrate, fat-free diet. Poly (A+) RNA
{4 fLg per lane) prepared from livers of mice starved 48 h or previously starved mice refed for
16 h were subjected to 0. 7% agarose gel electrophoresis, transferred to nitrocellulose, and probed
with 32 P-labeled eDNA inserts.
cipitation method was employed for enrichment of its message before prep-
aration of a eDNA library in A. gtll and screening with liver phosphofruc-
tokinase isozyme-specific antibodies. 15
Northern blot analysis using these eDNA sequences showed that the fatty
acid synthase and p90 mRNAs were present in high concentration in adipose
tissue and liver. 11 ·14 Liver phosphofructokinase mRNA, on the other hand
was detected in all tissues examined although liver showed the highest con-
centration.15 Fatty acid synthase and p90 mRNAs were not detectable in fasted
mouse liver and increased by two orders of magnitude when mice were refed
a high carbohydrate, fat-free diet. 11 In contrast, the liver phosphofructokinase
mRNA was easily detectable in fasted mice liver and increased fivefold during
refeeding (Figure l ). Dunaway and co-workers previously reported a decrease
Sul et al. 211
in liver phosphofructokinase concentration in starved or insulin-deprived an-

imals that was primarily due to an enhanced rate of enzyme degradation. 10
Our results indicate that liver phosphofructokinase synthesis is regulated by
changes in mRNA level. However, it is clear that the glycolytic enzyme liver
phosphofructokinase, although regulated in the same fashion as the lipogenic
enzymes, 4 • 11 • 16 • 17 is altered to a much smaller extent. It is worth noting that
in the liver, glycolysis is expected to occur even during gluconeogenic periods.
Moreover, since phosphofructokinase, unlike fatty acid synthase, is highly
regulated by metabolites, which act as allosteric effectors, control at the gene
level may not play a primary role in regulation of phosphofructokinase activity.
This is in contrast to the regulation of a well-characterized glycolytic enzyme,
pyruvate kinase; the mRNA level for pyruvate kinase seems to be altered
during fasting/refeeding to a similar order of magnitude as those for fatty acid
synthase and p90. 18
III. IN VIVO NUTRITIONAL AND HORMONAL

REGULATION OF GENE EXPRESSION
The roles for specific hormones in controlling expression of fatty acid

synthase during fasting/refeeding have been suggested in various studies. Such
nutritional manipulation causes an increase in circulating insulin and a de-
crease in glucagon levels. In addition, the increase in synthesis of fatty acid
synthase in rat liver caused by refeeding is not observed in diabetic animals.
Insulin specifically restores this increase in fatty acid synthase synthesis. 3
The lipolytic hormone glucagon decreases the concentrations of lipogenic
enzymes and also suppresses glycolysis. Evidence for the role of glucagon
was also demonstrated by an inhibition of the effects of refeeding on enzyme
activities. 3
With the cloned eDNA sequences, we were able to examine expression
of these nutritionally regulated genes at the pretranslational level. In normal
animals, there was a steady increase in both fatty acid synthase and p90
mRNAs for 16 h of refeeding after a lag period of a few hours. The lag was
probably due to the digestion and absorption processes (Figure 2). However,
in streptozotocin-diabetic animals, the mRNA levels approximated those of
fasted animals when measured 8 h after refeeding. This indicated that insulin
is required for the fasting/refeeding induction of the mRNAs for fatty acid
synthase and p90. Dibutyryl cAMP, administered at the time of refeeding to
normal fasted animals, inhibited the induction of the two mRNAs by 90%
(Figure 2). Because glucagon stimulates production of cAMP in liver, these
data suggest that the barely detectable levels of these mRNAs during fasting
may be partly due to the increased plasma glucagon level.
The effect of insulin was further examined by measuring steady-state
mRNA levels after insulin administration to streptozotocin-diabetic animals
fed a high carbohydrate, fat free diet (Figure 3). The induction of fatty acid
synthase and p90 mRNAs by insulin is rapid and marked. The mRNA levels
212 Nutrition a nd Gene Expression
70
~
w 10
~~
< 50
z
a:
E ~
~
w
en JO
<
w
u
a: 20
~
0 10
~
~
0
0 4
• 10 12 14 11
TIME !hl
- 8 .2 kb
- 6.8 kb
. ..
F RF F RF RF + Bt ~ cAMP
DiabetiC Normal
FIGURE 2. Effects of streptozotocin-induced diabetes and cAMP on fatty acid synthase and
p90 mRNA induction by fasting/refeeding. Poly (A + ) RNA was prepared from livers of diabetic
and normal mice fasted (F) or refed (RF) for 8 h. Dibutyryl cAMP and theophyll ine were injected
at the time of refeeding in normal, fasted mice. Northern analyses were carried out as in Figure
I. 2.1 kb mRNA represents a-actin .
Sul et al. 213
6.8 kb-
8 .2 kb -
0 2 4 6 8 8C
TIME (h )
20
1-
z
w
1-
5(.)
15
<(
z
c:
E 10
~
w
C)
~
I
(.)
a
c5u.
TIME(h)
FIGURE 3. Effect of insulin on mRNA levels for p90 and fatty acid synthase in streptozotocin-
diabetic mice. Poly (A+) RNA was prepared from livers of streptozotocin-diabetic mice killed
at the indicated times after administration of combined dose of regular insulin (3 U/100 g)
intraperitoneally and Lente Insulin (30 U/100 g) subcutaneously. Saline-injected mice were also
used at 8 h as a control (8C). Northern analyses were done as in Figures I and 2.
for fatty acid synthase and p90 increased four- and twofold 1 h after intra-
peritoneal insulin injection, respectively, and achieved maximal increases of
about 20-fold after 6 h. However, it is known that glucagon levels are de-
creased when insulin is administered to diabetic animals and this may be a
factor. In cultured avian embryo hepatocytes, Goodridge and co-workers
reported that insulin only amplified the effect of T3 on fatty acid synthase
gene expression. 19 ·20 It would be surprising if there is a difference in regulation
of avian and murine fatty acid synthase by insulin.
IV. REGULATION OF TRANSCRIPTION
mRNA concentrations can be, in theory, controlled by the rate of specific

transcription, processing, or turnover. The first approach would be to ask if
the altered transcription can quantitatively account for the observed effects
of fasting/refeeding, insulin, and cAMP on mRNA levels for fatty acid syn-
thase and p90. Therefore, transcription run-on analyses were carried out in
isolated liver nuclei from normal and diabetic animals treated as described
above. 14 •20 The transcription rates of the fatty acid synthase and p90 genes
were increased when previously fasted mice were refed a high carbohydrate
diet (Figure 4). The maximum increase in transcription rate of 39-fold for
fatty acid synthase was attained at 6 h after refeeding and was maintained up
to 16 h, whereas the transcriptional increase for the p90 gene was substantially
slower in that it was 2.5-fold at 6 h, 7-fold at 9 h, and reached 22-fold after
16 h of refeeding. However, there was no detectable transcription of fatty
acid synthase and p90 genes in fasted or fasted-refed streptozotocin-diabetic
animal liver, indicating a requirement of insulin for transcriptional induction
by fasting/refeeding. Furthermore, there was a rapid and marked increase in
transcription rates of fatty acid synthase and p90 genes when insulin was
given to diabetic mice; transcription rates of both of these genes increased
approximately fourfold after 30 min and attained a maximum increase of
seven- to eightfold at 2 h (Figure 5). The results demonstrate that these genes
are highly regulated at the transcriptional level by nutritional and hormonal
stimuli. We then asked whether the insulin effect on specific transcription
requires ongoing protein synthesis and/or a regulatory protein with a short
half-life. In the presence of cycloheximide, the increases in mRNA levels
and transcriptional rates for fatty acid synthase and p90 caused by insulin in
diabetic mice were not observed. Thus, it appears insulin is required for
ongoing protein synthesis and that a similar mechanism may be involved in
the insulin regulation of these two genes.
The effect of cAMP on fatty acid synthase and p90 gene transcription
was also examined. When dibutyryl cAMP was given at the start of refeeding
of normal animals, there was no increase in transcription of either fatty acid
synthase or p90 genes by fasting/refeeding, indicating a negative transcrip-
Sul et al . 215
NORMAL DIABETIC
--
pBR322
p13
pFAS-1
pAM 91
.. --
0 6 69 16 0 6
Bt 2 cAMP
TIME AFTER REFEEDING (h)
Q)
n;
a:
c: 30
·a0
"5
U)
c:
~ 20
I-
.E
Ql
Ol
~ to
.r.
(.)
"0
&
5 tO 15
Time After Reteeding (h)
FIGURE 4. Effects of fasting/refeeding, cAMP, and streptozotocin-diabetes on fatty acid

synthase and p90 gene transcription. Normal or streptozotocin-diabetic mice were treated as
described in Figure I. Nuclei were isolated from livers and run-on tran scription analyses were
carried out. Plasmids containing eDNA inserts coding for fatty acid synthase (pFAS-1), p90
(pl3), and actin (pAM91) were used. pBR322 is a negative control.
tiona! control of these genes by cAMP (Figure 4) . Similar effects of cAMP

were observed in liver phosphofructokinase gene expression. The increase in
mRNA level for liver phosphofructokinase in refed mice was reduced by SO%
by cAMP administration. Likewise, the four- to fivefold increase in tran-
scription rate of the phosphofructokinase gene observed during refeeding was
totally abolished by cAMP administration.
pBR322 • •
p13
pFAS-1
pAM91 •
0 0.5 1.0 2.0 6.0
TIME (h)
Q)
~
a:
c:
0
~
a.
·;;::
(.)
(/)
c:
~
1-
c:
Q)
0>
c:
cu
.s::
()
"0
0
lL
0.0 1.0 2.0 3 .0 4.0 5.0 6.0
Time (h)
FIGURE 5. Effect of insulin on transcriptional rates of p90 and fatty acid synthase genes in
streptozotocin-diabetic mice liver. Nuclei were isolated from diabetic mice treated with insulin
as in Figure 3 and run-on assays were carried out. (•) pl3; (eJ) pFAS-1; (+l pAM91.
V. DIFFERENTIATION-DEPENDENT EXPRESSION
IN 3T3-Ll CELLS
The murine 3T3-Ll cells, upon differentiation, acquire adipocyte mor-

phology, accumulate large amounts oftriacylglycerol, and become responsive
Sul et al. 217
pBR322
FAS pBiuescript
mANA
pFAS1
pSCD1
dO d5 dO d5
FIGURE 6. Comparisons of mRNA levels and transcription rates for fatty acid synthase gene
in 3T3-Ll preadipocytes and adipocytes. RNA (5 JLg per lane) prepared from confluent 3T3-Ll
cells (dO) and from cells differentiated by treatment of methylisobutylxanthine and dexamethasone
(d5) were used for Northern blot. Nuclei isolated from preadipocytes and adipocytes were used
for run-on transcription and hybridized with 10 JLg each of fatty acid synthase (pFAS-1), stearoyl
CoA desaturase (pSCDl), and control plasmids fixed on nitrocellulose.
to lipolytic and lipogenic hormones. The concentrations of several lipogenic

enzymes, along with other differentiation-dependent proteins, increase con-
comitantly with differentiation. These cells have been used as a model system
for studying the biochemical and molecular mechanisms underlying hormonal
regulation of lipogenic enzyme genes and adipose tissue development. 22 - 24
As with other lipogenic enzymes, fatty acid synthase activity, mass, and
rate of synthesis increase by an order of magnitude during conversion of 3T3-
Ll preadipocytes to adipocytes. 25 - 27 We recently showed that mRNA levels
for both fatty acid synthase and p90 increased about 20-fold during differ-
entiation. The increase in fatty acid synthase mRNA level can account for
the increase in fatty acid synthase synthesis observed, demonstrating regu-
lation at a pretranslationallevel. Interestingly, however, run-on assays showed
that fatty acid synthase gene transcription was significant in preadipocytes
and the transcription increased only 1.5-fold during adipocyte differentiation
(Figure 6). Increase in the transcription rate for the p90 gene was three- to
fourfold. In contrast, transcription of the stearoyl-CoA desaturase gene in-
creased markedly during adipose conversion as reported by others. 28 ·29 The
discrepancy between the induced level of fatty acid synthase and p90 mRNAs
and transcription rates strongly suggests that not only an increase in tran-
scription but altered RNA stability and/or processing may contribute to the
increase in mRNA levels for fatty acid synthase and p90 during adipocyte
conversion. To test whether changes in fatty acid synthase mRNA turnover
rates during differentiation could account for the induction, fatty acid synthase
preadipocytes adipocytes
, 234 4c 12344 c
Cl)
>
.!!
<
z
a:
E
0 10 20
time (hour)
FIGURE 7. Comparison of mRNA stability in 3T3-Ll preadipocytes a nd adipocytes. Pre-

adipocytes (day 0) and adipocytes (day 5) were treated with actinomycin D (10 f.l.g/ml). RNA
were then prepared from control cells (J ane I) or from cells 3 h (lane 2), 9 h (Jane 3), or 20 h
(lane 4) after actinomycin treatment. Lane 4c represents control cells maintained for 20 h with
no actinomycin D treatment. Northern blot shown is for the fatty acid synthase mRNA . Quan-
tification by densitometric scanning for relati ve decrease in fatty acid synthase and actin mRNAs
in preadipocytes (o. D) or adipocytes <•. •> were shown.
mRNA abundance was measured as a function of time after the addition of

a transcription inhibitor, actinomycin, in preadipocytes and adipocytes (Figure
7) . The decrease in fatty acid synthase mRNA occurred more rapidly in
preadipocytes compared to adipocytes; after 3 h of actinomycin treatment,
fatty acid synthase mRNA levels decreased by 64% in preadipocytes while
there was no change in adipocytes. Apparent half-lives of fatty acid synthase
mRNAs were estimated to be 2 .5 h in preadipocytes vs. 20 h in adipocytes.
The eightfold increase in fatty acid synthase mRNA half-life in adipocytes ,
along with the 1.5-fold increase in the transcription rate of the gene, accounts
for the 10- to 15-fold increase in fatty acid synthase mRNA observed during
differentiation. Therefore, the increase in fatty acid synthase mRNA during
Sul et al. 219
adipose conversion is due to both transcriptional activation of the gene and

posttranscriptional stabilization of the mRNA. The increase observed in most
of the developmentally regulated mRNAs during adipocyte differentiation,
such as glycerol-3-phosphate dehydrogenase, has been attributed to the tran-
scriptional activation of their gene. 30 •31 In the case of adipsin, mRNA stability
was also suggested to contribute to the increase in RNA abundance during
the differentiation of 3T3-F442A adipocytes. However, the transcription rates
or the mRNA stability for adipsin mRNA could not be compared due to its
low detectability. 30
It has been previously reported that, in addition to the activities of li-
pogenic enzymes, the activities of several glycolytic enzymes are increased
during adipocyte differentiation. 32 •33 Therefore, we have examined the rate
of synthesis of liver phosphofructokinase during 3T3-Ll differentiation. We
observed that the rate of synthesis of liver phosphofructokinase increased
fivefold during adipose conversion as measured by pulse-chase experiments.
In addition, the mRNA level for this enzyme was also increased an order of
magnitude during differentiation, indicating pretranslational control. We have
also found that 3T3-Ll cells expressed high level of muscle phosphofructo-
kinase mRNA during the fibroblastic stage but that during adipocyte differ-
entiation, the muscle isoform is downregulated as the liver phosphofructo-
kinase mRNA level increases (unpublished results). It is clear that isozyme
switching from muscle to liver phosphofructokinase occurs during adipocyte
differentiation. It may be that all three isoforms are expressed in embryonic
tissue and, as cells differentiate to become specialized, phosphofructokinase
isozyme(s) with regulatory properties suited for the cell function is expressed.
We also have previously shown that myoblasts express all three isozymes but
that during differentiation to myotubes, expression of the muscle isozyme
predominates. 34
VI. HORMONAL REGULATION OF GENE

EXPRESSION IN 3T3-Ll ADIPOCYTES
Mature 3T3-Ll adipocytes, which acquire responsiveness to lipolytic and

lipogenic hormones during the differentiation process, were utilized to de-
termine independent hormone effects on gene expression. When mature 3T3-
Ll adipocytes were treated with insulin for 16 h in a serum-free medium,
both fatty acid synthase and p90 mRNA levels increased approximately three-
fold (Figure 8). However, the increase in expression of fatty acid synthase
and p90 by insulin in streptozotocin-diabetic mice was much higher than the
increase observed in cultured adipocytes. Treatment of 3T3 adipocytes with
dibutyryl cAMP caused a 60 and 80% decrease in fatty acid synthase and
p90 mRNA levels, respectively. Overall, the effects of insulin and cAMP on
these two mRNAs in cultured adipocytes are in agreement with those obtained
200
...c
0
-
(_)
0
# 100
Control Bt 2 cAMP Insulin
FIGURE 8. Effect of insulin and cAMP on mRNA levels for fatty acid synthase p90 in 3T3-
Ll adipocytes. 3T3-Ll cells were differentiated, maintained in serum-free medium for 24 h,
and then treated either with I J.Lg/ml of insulin or 1 mM dibutyryl cAMP plus 1 mM theophylline
for 16 h. Poly (A+) RNA prepared were used in Northern blot analyses for fatty acid synthase
(D) and p90 (~).
in mouse liver described in the previous section, indicating independent effects

of these agents.
The role of thyroid hormone in regulation of fatty acid synthase activity
is well documented in studies which utilized animals of different thyroid
status. Liver fatty acid synthase activity was reduced in hypothyroidism, 35
and thyroid hormone injection for 7 days increased fatty acid synthase activity
two- to threefold in liver. 36 ·37 Possible synergistic effects of thyroid hormones
and a high carbohydrate diet have also been suggested in the regulation of
fatty acid synthase activity. 38 In cultured primary rat hepatocytes, treatment
with T3 for 24 h has been shown to produce a two- to threefold increase in
lipogenic enzyme activities, including fatty acid synthase. 39 Therefore, we
examined the effect ofT3 on fatty acid synthase in mature 3T3-Ll adipocytes.
The fatty acid synthase mRNA level increased 2.4-fold at 12 h of 10 nM T3
treatment and further increased to 4.5-fold at 24 h. A similar increase in p90
mRNA level was observed after T3 treatment. The increase in fatty acid
synthase mRNA level was correlated with the increase in synthesis of fatty
acid synthase. It was next examined if the increase in fatty acid synthase
mRNA was due to an increase in the transcription rate of this gene. Treatment
with 10 nM T3 for 6 h increased fatty acid synthase gene transcription to a
Sul et al. 221
(A)
FAS mANA-
0 6 12 24
Time (Hours)
(B)
p13
pFAS1
pBR322
0 6 12 24
Time (Hours)
FIGURE 9. Effect of T3 on mRNA level and transcription rate for fatty acid synthase. Dif-
ferentiated 3T3-Ll adipocytes were treated with T3 for indicated times. RNA and nuclei were
prepared and used for Northern blot (A) and transcription run-on assays (B), respectively.
maximum of fourfold (Figure 9). The increase in transcription preceded the

increase in mRNA levels. In contrast to the fatty acid synthase gene, tran-
scription of the p90 gene was not appreciably affected by T3, indicating
possible regulation at the posttranscriptional level. These two genes seem to
be regulated by thyroid hormone by different mechanisms.
VII. CONCLUSIONS
The in vivo and in vitro experiments described above led us to conclude

that mammalian fatty acid synthase and p90 genes are under transcriptional
control by insulin and cAMP, and that insulin upregulates and cAMP down-
regulates these genes. Liver phosphofructokinase is regulated in a similar
(a) (b)
0
2
15 30
3 2 3
0 15 30 45 0 0 15 30
0 15 30 45 0 0 15 30 45 0 0 15 30 45 0
- 530
r -10
A r
l
E - 510
L
-3 0
- 490
r
8
- - 50
l- -7 0 0
r
I L
c
L ...• •450
- 90
FIGURE 10. Binding of nuclear proteins to the mouse liver phosphofructokinase promoter
region. DNA fragments containing the mouse liver phosphofructokinase promoter regions. + 87
to -169 (a) and -351 to -699 (b), were labeled, incubated with indicated amounts (j.Lg) of
liver nuclear extracts, and digested with DNase I. The regions of protection are demarcated: (a)
I, no competing oligonucleotides; 2, 60-fold excess of oligonucleotides (- 65 to - 97); 3, 60-
fold excess of AP-I consensus sequence and (b) I, no competing oligonucleotides; 2, excess of
oligonucleotides (-445 to -475); 3, excess of oligonucleotides (-485 to -522).
fashion. Nevertheless, these genes may be regulated independently and dif-

ferent mechanisms may be involved.
At present, in order to identify hormone response elements and nuclear
factors which interact with these regulatory sequences, we have isolated gen-
omic clones coding for these enzymes and defined transcription units and 5'
and 3' flanking regions. We have also examined by DNase 1 footprinting
assays sequences at 5' flanking regions which may be potentially important
for transcriptional regulation. As shown in Figure 10, liver nuclear proteins
were found to interact with several regions of the liver phosphofructokinase
gene promoter in vitro. 40 Interestingly, multiple GC boxes which are usually
present in "housekeeping" genes and are probably Spl binding sites, showed
protection by liver nuclear proteins. Also protected was an AP-1-like se-
quence. Competition experiments showed that a distinct protein was involved
in binding. In addition, there were two unique sequences at the distal region
Sul et al. 223
of the liver phosphofructokinase gene which showed protection from DNase

digestion. The fatty acid synthase gene promoter contains multiple GC boxes
and several unique regions including possible hormone response elements
which show interaction with liver nuclear proteins (unpublished results). The
conserved 8-bp motif (CRE) present in a number of genes upregulated by
cAMP is not present in either liver phosphofructokinase or fatty acid synthase
gene promoter regions.
To determine the functional significance of these sequences in the hor-
monal regulation of these genes, we have constructed fusion genes linking
promoters to the chloramphenicol acetyltransferase reporter gene and tran-
siently transfected 3T3-Ll adipocytes. To date, we found that 2.1 kb and
0. 7 kb of the immediate 5' flanking sequences of the fatty acid synthase and
liver phosphofructokinase genes, respectively, contain cis-acting sequences
for the upregulation by insulin. The studies of these genes will enable us to
dissect nutritional and hormonal regulation of gene expression during energy
storage. Moreover, these studies will allow us to define cis-acting elements
and trans-acting factors which interact with these sequences for the transcrip-
tional activation by insulin. Unlike the transcriptional regulation by T3 and
cAMP, the mechanism of insulin regulation of gene transcription is not well
understood, although some progress has been made in characterizing se-
quences which may be important for insulin regulation in other genes, such
as phosphoenolpyruvate carboxykinase and glyceraldehyde-3-phosphate de-
hydrogenase.4'.42 Identification and studies of cis-trans components for insulin
regulation may also lead to understanding a signal transduction pathway for
insulin. In addition, by characterizing cis-trans components for various hor-
mone responsiveness, it will be possible to examine potential interactions of
trans-acting factors in achieving multihormonal regulation of these genes
during changes in the nutritional status of the animal.
ACKNOWLEDGMENTS
We thank former members of this laboratory, J. D. Paulauskis, S. C.

Gehnrich, and P. Rongnoparut whose efforts contributed to the studies re-
viewed here. The research was supported by National Institutes of Health
Grants DK-36264 and DK-40518 and the Juvenile Diabetes Foundation
(H. S. S.). N. M. is the recipient of Juvenile Diabetes Foundation Postdoc-
toral Research Fellowship.
REFERENCES
I. Gibson, D. M., Lyons, R. T., Scott, D. F., and Muto, Y., Synthesis and degradation
of the lipogenic enzymes of rat liver, Adv. Enzyme Regul., 10, 187, 1972.
2. Volpe, J. J. and Vagelos, P. R., Mechanisms and regulation of biosynthesis of saturated
fatty acids, Physiol. Rev., 56, 339, 1976.
3. Lakshmanan, M. R., Nepokroeff, C. M., and Porter, J, W., Control of the synthesis
of fatty acid synthetase in rat liver by insulin, glucagon, and adenosine 3':5'cyclic
monophosphate, Proc. Nat!. Acad. Sci. U.S.A., 69, 3516, 1972.
4. Goodridge, A. G., Dietary regulation of gene expression: Enzymes involved in carbo-
5. Brent, G. A., Moore, D. D., and Larsen, P. R., Thyroid hormone regulation of gene
expression, Annu. Rev. Physiol., 53, 17, 1991.
6. Gonzalez, G. A., Yamamoto, K. K., Fischer, W. H., Karr, D., Menzel, P., Biggs,
W., Vale, W. W., and Montiminy, M. R., A cluster of phosphorylation sites on the
cAMP-regulated nuclear factor CREB predicted by its sequence, Nature (London), 337,
749, 1989.
7. Bates, E. J, and Saggerson, E. D., A study of the glycerol phosphate acyltransferase
and dihydroxyacetone phosphate acyltransferase activities in rat liver mitochondrial and
microsomal fractions, Biochem. J., 182, 751, 1979.
8. Saggerson, E. D. and Carpenter, C. A., Effects of streptozotocin-diabetes and insulin
administration in vivo or in vitro on the activities of five enzymes in the adipose-tissue
triacylglycerol-synthesis pathway, Biochem. J., 243, 289, 1987.
9. Dunaway, G. A. and Weber, G., Effects of hormonal and nutritional changes on rates
of synthesis and degradation of hepatic phosphofructokinase isozymes, Arch. Biochem.
Biophys., 162, 629, 1974.
10. Dunaway, G. A., Leung, G. L.-Y., Thrasher, J, R., and Cooper, M.D., Turnover
of hepatic phosphofructokinase in normal and diabetic rats. Role of insulin and peptide
stabilizing factor, J. Bioi. Chern., 253, 7460, 1978.
I I. Paulauskis, J. D. and Sui, H. S., Cloning and expression of mouse fatty acid synthase
and other specific mRNAs. Development and hormonal regulation in 3T3-LI cells, J.
Bioi. Chern .. 263, 7049, 1988.
12. Yan, C., Wood, E. A., and Porter, J, W., Characterization of fatty acid synthetase
eDNA clone and its mRNA, Biochem. Biophys. Res. Commun., 126, 1235, 1985.
13. Back, D. W., Goldman, M. J,, Fisch, J. E., Ochs, R. S., and Goodridge, A. G.,
The fatty acid synthase gene in avian liver. Two mRNAs are expressed and regulated in
parallel by feeding, primarily at the level of transcription, J. Bioi. Chern., 261, 15179,
1986.
14. Shin, D. H., Paulauskis, J, D., Moustaid, N., and Sui, H. S., Transcriptional regulation
of p90 with sequence homology to E. coli glycerol-3-phosphate acyltransferase, J. Bioi.
Chern., 266, 23 834, 1991.
15. Gehnrich, S. C., Gekakis, N., and Sui, H. S., Liver (B-type) phosphofructokinase
mRNA. Cloning, structure, and expression, J. Bioi. Chern., 263, 11755, 1988.
16. Raymondjean, M., Vaulont, S., Cognet, M., Decaux, J. F., Puzenat, N., Bergot,
M. 0., and Kahn, A., Positive and negative regulation of gene expression by insulin
and glucagon: the model of L-type pyruvate kinase gene, Biochimie, 73, 41, 1991.
17. Sui, H. S., Wise, L. S., Brown, M. L., and Rubin, C. S., Cloning of eDNA sequences
for murine ATP-citrate lyase. Construction of recombinant plasmids using an immuno-
purified mRNA template and evidence for the nutritional regulation of ATP-citrate lyase
mRNA content in mouse liver, J. Bioi. Chern., 259, 1201, 1984.
18. Sui, H. S., Wise, L. S., Brown, M. L., and Rubin, C. S., Cloning of eDNA sequences
for murine malic enzyme and the identification of aberrantly large malic enzyme mRNA
in MOD-I null mice, J. Bioi. Chern., 259, 555, 1984.
Sul et al. 225
19. Wilson, S. B., Back, D. W., Morris, S.M., Jr., Swierczynski, J., and Goodridge,
A. G., Hormonal regulation of lipogenic enzymes in chick embryo hepatocytes in culture.
Expression of the fatty acid synthase gene is regulated at both translational and pretran-
slational steps, J. Bioi. Chern., 261, 15179, 1986.
20. Stapleton, S. R., Mitchell, D. A., Salati, L. M., and Goodridge, A. G., Triiodothy-
ronine stimulates transcription of the fatty acid synthase gene in chick embryo hepatocytes
in culture. Insulin and insulin-like growth factor amplify that effect, J. Bioi. Chern., 265,
18442, 1990.
21. Paulauskis, J. D. and Sui, H. S., Hormonal regulation of mouse fatty acid synthase
gene transcription in liver, J. Bioi. Chern., 264, 574, 1989.
22. Green, H. and Kehinde, 0., Sublines of mouse 3T3 cells that accumulate lipid, Cell,
I, 113, 1974.
23. Rubin, C. S., Hirsch, A., Fung, C., and Rosen, 0. M., Development of hormone
receptors and hormonal responsiveness in vitro, J. Bioi. Chern., 253, 7570, 1978.
24. Sui, H. S., Adipocyte differentiation and gene expression, Curr. Opinion Cell Bioi., 1,
1116, 1989.
25. Weiss, G. H., Rosen, 0. M., and Rubin, C. S., Regulation of fatty acid synthetase
concentration and activity during adipocyte differentiation. Studies on 3T3-LJ cells, J.
Bioi. Chern., 255, 4751, 1980.
26. Student, A. K., Hsu, R. Y., and Lane, M. D., Induction of fatty acid synthetase
synthesis in differentiating 3T3-LJ preadipocytes, J. Bioi. Chern., 255, 4745, 1980.
27. Wise, L. S., Sui, H. S., and Rubin, C. S., Coordinate regulation of the biosynthesis
of ATP-citrate lyase and malic enzyme during adipocyte differentiation. Studies on 3T3-
Ll cells, J. Bioi. Chern., 259, 4827, 1984.
28. Moustaid, N. and Sui, H. S., Regulation of expression of the fatty acid synthase gene
in 3T3-Ll cells by differentiation and triiodothyronine, J. Bioi. Chern., 266, 18550,
1991.
29. Bernlohr, D. A., BoJanowski, M. A., Kelly, T. J., and Lane, M. D., Evidence for
an increase in transcription of specific mRNAs during differentiation of 3T3-Ll prea-
dipocytes, J. Bioi. Chern., 260, 5563, 1985.
30. Cook, D. S., Hunt, C. R., and Spiegelman, B. M., Developmentally regulated mRNAs
in 3T3-adipocytes: analysis of transcriptional control, J. Cell. Bioi., 100, 514, 1985.
31. Djian, P., Phillips, M., and Green, H., The activation of specific gene transcription
in the adipose conversion of 3T3 cells, J. Cell. Physioi., 124, 554, 1985.
32. Mackall, J. C. and Lane, M.D., Role of pyruvate carboxylase in fatty acid synthesis:
alterations during preadipocyte differentiation, Biochern. Biophys. Res. Cornrnun., 79,
720, 1977.
33. Spiegelman, B. M. and Green, H., Control of specific protein biosynthesis during the
adipose conversion of 3T3 cells, J. Bioi. Chern., 255, 8811, 1980.
34. Gekakis, N., Gehnrich, S.C., and Sui, H. S., Phosphofructokinase isozyme expression
during myoblast differentiation, J. Bioi. Chern., 264, 3658, 1989.
35. Volpe, J. J. and Kishimoto, Y., Fatty acid synthetase of brain: development, influence
of nutritional and hormonal factors and comparison with liver enzyme, J. Neurochern.,
19, 737, 1972.
36. Mariash, C. N., Kaiser, F. E., and Oppenheimer, J. H., Comparison of the response
characteristics of four lipogenic enzymes to 3,5,3' -triiodothyronine administration: evi-
dence for variable degrees of amplification of the nuclear 3,5,3' -triiodothyronine signal,
Endocrinology, 106, 22, 1980.
37. Diamant, S., Gorin, E., and Shafrir, E., Enzyme activities related to fatty acid synthesis
in liver and adipose tissue of rats treated with triiodothyronine, Eur. J. Biochern., 26,
553, 1972.
38. Mariash, C. N., Kaiser, F. E., Schwartz, H. L., and Towle, H. C., Synergism of
thyroid hormone and high carbohydrate diet in the induction of lipogenic enzymes in the
rat. Mechanisms and implications, J. Clin. Invest., 65, 1126, 1980.
39. Spence, J, T. and Pilot, H. C., Induction of lipogenic enzymes in primary cultures of
rat hepatocytes. Relationship between lipogenesis and carbohydrate metabolism, Eur. J.
Biochem., 36, 15, 1982.
40. Rongnoparut, P., Verdon, C. P., Gehnrich, S. C., and Sui, H. S., Isolation and
characterization of the transcriptionally regulated mouse liver (B-type) phosphofructo-
kinase gene and its promoter, J. Bioi. Chern., 266, 8086, 1991.
41. O'Brien, R. M., Lucas, P. C., Forest, C. D., Magnuson, M. A., and Granner,
D. K., Identification of a sequence in the PEPCK gene that mediates a negative effect
of insulin on transcription, Science, 249, 533, 1990.
42. Nasrin, N., Ercolani, L., Denaro, M., Kong, X. F., Kang, I., and Alexander, M.,
An insulin response element in the glyceraldehyde-3-phosphate dehydrogenase gene binds
a nuclear protein induced by insulin in cultured cells and by nutritional manipulations in
vivo, Proc. Nat/. Acad. Sci. U.S.A., 87, 5273, 1990.
Chapter 10
REGULATION OF HEPATIC GENE EXPRESSION BY

DIETARY FATS: A UNIQUE ROLE FOR
POLYUNSATURATED FATTY ACIDS
Steven D. Clarke and Donald B. Jump
TABLE OF CONTENTS
I. Introduction ................................. ..................... 228
II. Posttranslational Regulation of Lipogenic Enzymes .............. 228
III. Polyunsaturated Fatty Acid Suppression of Fatty Acid

Synthase and S 14 mRNA Abundance ............................ 231
IV. Potency of Fatty Acid Inhibitors ................................. 233
V. Hepatic-Specific Inhibition ................................. ...... 234
VI. Structural Requirements and Fatty Acid Potency ................. 236
VII. Transcriptional Regulation by Polyunsaturated Fats .............. 238
VIII. Mechanisms of Transcriptional Control .......................... 240
IX. Summary and Speculation ................................. ...... 243
Acknowledgments ................................. ...................... 243
References ................................. .............................. 244
0-8493-6961-4/93/$0.00 +$.50
This chapter is dedicated to Dr. Sandy Abraham whose tireless

and frontiering efforts in the area of polyunsaturated fatty acid me-
tabolism and the control of gene expression served as the foundation
for the work presented here.
I. INTRODUCTION
Numerous proteins involved in carbohydrate and lipid metabolism (Table 1)

are regulated by dietary fat, but it has become increasingly clear over the last
decade that regulation of the expression of these proteins is differentially
affected by saturated and polyunsaturated fats. 1- 6 Using the fatty acid bio-
synthetic pathway as a model, Gibson and his colleagues discovered that only
polyunsaturated fatty acids inhibited hepatic fatty acid biosynthesis. 1 Dietary
saturated fats such as tallow were found to have little or no ability to suppress
hepatic lipogenesis. I-s This unique characteristic of dietary polyunsaturated
fats has led us to propose that at least part of the hypotriglyceridemic action
of dietary polyunsaturated fats is due to the inhibition of de novo fatty acid
synthesis which in tum reduces substrate availability for triglyceride synthesis,
and hence decreases the hepatic output of triglycerides. Consequently, the
amount of circulating plasma triglyceride is decreased.
The suppression of hepatic lipogenesis by dietary polyunsaturated fatty
acids appears to require that the fatty acid contain a minimum of 18-carbons
and two double bonds in the 9 and 12 positions, and at least one of these
bonds must be of the cis configuration (Figure 1). In addition, the inhibition
of fatty acid biosynthesis by polyunsaturated fats is: (1) specific for hepatic
lipogenesis, 47 (2) mediated by the fatty acid component of the dietary lipid, 4
(3) dependent upon the desaturation of the 18-carbon fatty acid by the delta-
6 desaturase, 8 •9 (4) not mediated by a prostaglandin or thromboxane meta-
bolite, 6 •9 and (5) related to the suppression of genes coding for proteins directly
involved fatty acid biosynthesis. 10- 12 The objective of this review is to trace
the evolution of our understanding of the mechanism by which polyunsaturated
fatty acids uniquely regulate the expression of lipogenic proteins, notably
fatty acid synthase, and to speculate as to the potential molecular mechanisms
by which polyunsaturated fatty acids regulate nuclear transcriptional events.
It should be noted that the data to be discussed are derived primarily from
studies with rodents, and that the unique effects of polyunsaturated fatty acids
may not include all species, e.g., polyunsaturated fats do not selectively inhibit
lipogenic enzymes in birds Y
II. POST-TRANSLATIONAL REGULATION OF

LIPOGENIC ENZYMES
The classic studies of Chaikoff and his colleagues 14 had indicated that fat
ingestion would suppress acetate incorporation into hepatic lipids. It was
C Iarke and Jump 229
TABLE 1
Dietary Fatty Acid Effects on Rat Liver Proteins
Enzyme Fat Amount mRNA Transcription Ref.
Glucokinase Saf oil -30% 5, 47, 49

Phosphofructokinase 18:2 nc 5
Pyruvate kinase 18:2 nc 5
Glucose-6-phosphate 16:0/18: I nc 4, 48
dehydrogenase 18:2/18:3 -60%
Saf oil -42% -52% 22
Malic enzyme 16:0/18:1 nc 4, 30
18:2/com -45% nc
Citrate lyase 18:2 -55% 47
Acetyi-CoA carbox- 16:0/Tallow nc 5
ylase 18:2/Saf -60% 5
Fatty acid synthase 16:0 nc nc nc II, 12
18:1 nc +30% II
18:2 60% -70% 6,11,20
Fish oil -80% -75% -94% II, 12
Delta-9 desaturase 18:2/18:3 -50% 49
Fish oil -85%
Delta-6 desaturase 18:2/18:3 +150% 49
Fish oil -20%
Delta-S desaturase 18:2/18:3 +350% 49
lipase Fish oil nc
Lard nc 50
Com oil +155%
Fish oil +105%
apo A I High PIS +22-50% 51
Sl4 16:0 nc nc II, 12
18:1 -17% nc
18:2 -80%
Fish oil -84% -80%
13-Actin 18:1 nc nc 12
PEPCK Fish oil nc
proposed that fatty acids or their CoA esters functioned as negative allosteric
modifiers of key glycolytic and lipogenic enzymes, e.g., phosphofructokinase
and acetyl-CoA carboxylase. 14 • 15 In our early studies, the putative fatty acid
allosteric binding sites of hepatic phosphofructokinase and/or acetyl-CoA
carboxylase were hypothesized to possess a higher affinity for polyunsaturated
fatty acyl-CoAs than for the saturated fatty acid CoA esters. Thus, polyun-
saturated fatty acids could selectively inhibit glucose flux to fatty acids by
selectively binding to regulatory enzymes and inhibiting their catalytic effi-
125~-----------------------------------------,
100
(])
(/)
c:
0
0. 75
(/)
(])
a:
(])
>
·.;:::; 50
<tl
(])
a:
25
FIGURE 1. Structural requirements for a fatty acid to suppress hepatic lipogenesis. Mice
(n = 4) were fed for 5 d a high glucose diet containing 5% (w/w) of the respective fatty acids
cited in the figure. In vivo fatty acid synthesis was determined by tritiated water incorporation
into fatty acids. Relative response is defined relative to mice fed a high glucose fat-free diet
(SEM < 15%).
ciency. 16 ·17 Recently, we reported that polyunsaturated fats were in fact better
acute inhibitors of hepatic fatty acid synthesis and malonyl-CoA accumulation
than were saturated fats. 16 However, the acute suppression of fatty acid syn-
thesis by the ingestion of a single meal containing polyunsaturated fat was
modest when compared to the 60 to 90% reduction in fatty acid synthesis
and lipogenic enzymatic activities brought on by the prolonged (3 to 4 d)
ingestion of polyunsaturated fat (Figure 4). 1-12 The acute suppression of lipo-
genesis by safflower oil was not the consequence of reduced catalytic effi-
ciency or capacity of pyruvate kinase or phosphofructokinase, i.e. glycolytic
activity was unaffected. 5 In addition, the quantity of catalytically functional
acetyl-CoA carboxylase (as determined by increased resistance to avidin ac-
tivation)18·19 was not reduced by the ingestion of saturated or polyunsaturated
fats (Figure 5). Subsequent analysis of the time course of inhibition of fatty
acid synthesis vs. acetyl-CoA carboxylase and fatty acid synthase activities
revealed that the polyunsaturated fatty acid suppression of hepatic fatty acid
biosynthesis paralleled the decline in the amount of acetyl-CoA carboxylase
and fatty acid synthase. 5 Thus, polyunsaturated fatty acids appeared to exert
their greatest impact on the fatty acid biosynthetic capacity via a coordinate
Clarke and Jump 231
suppression of the amount of lipogenic enzymes. 5 •20-22 In light of this con-

clusion, our attention focused on the events which regulate enzyme synthesis
and gene expression.
III. POLYUNSATURATED FATTY ACID SUPPRESSION

OF FATTY ACID SYNTHASE AND S14 mRNA
ABUNDANCE
Inhibition of enzyme protein synthesis is the major reason for the reduction
in hepatic activity of lipogenic enzymes (e.g., fatty acid synthase) caused by
the ingestion of polyunsaturated fatty acids. 6 •20 Two mechanisms may explain
this inhibition of enzyme synthesis: (1) interference with translation rate of
the mRNA and/or (2) reduction in the amount of mRNA. To address this
issue we have focused our efforts on two genes: fatty acid synthase and the
putative lipogenic protein, S 14. 10- 12
When young growing rats were fed for 6 d a diet containing 10% safflower
oil, the abundance of fatty acid synthase and S14 mRNA was decreased to
levels that were 30 and 50%, respectively, of the values observed with a fat-
free high glucose diet. 11 Comparable amounts of dietary monounsaturated or
saturated fat (e.g., triolein) had no suppressive effect. 11 Polyunsaturated fatty
acids not only impaired fatty acid synthase and Sl4 expression in the growing
and adult rodents, but also blocked the 20- to 30-fold rise in mRNA levels
associated with weaning rats onto a high carbohydrate diet. 11 The weanling
animal appeared to be more sensitive to the polyunsaturated fat inhibition
than was the growing or adult rat, e.g., 3% 18:2 (n-6) suppressed fatty acid
synthase expression 80 to 90% in the 30-d-old rat, whereas 10% 18:2 (n-6)
suppressed fatty acids synthase mRNA only 65% in the 3-month-old rat. 11
The regulation of gene expression by dietary polyunsaturated fats is not unique
to fatty acid synthase and Sl4, but also extends to the suppression of acetyl
CoA carboxylase 21 and glucose-6-phosphate dehydrogenase. 22 Thus, dietary
polyunsaturated fats appear to coordinately suppress the level of transcripts
coding for lipogenic enzymes, and in this way decrease the synthesis of these
enzymes which in tum results in a reduced lipogenic capacity. 4 •5 • 11 • 12 However,
each gene possesses its own regulatory sequences, and expression of each
gene coding for a protein in the fatty acid biosynthetic pathway is a summation
of integrated hormonal and nutrient signals. Consequently, individual proteins
may vary in the degree of response to dietary polyunsaturated fatty acids,
e.g., S14 mRNA is less dramatically reduced by polyunsaturated fats than is
fatty acid synthase mRNA. 11 Nevertheless, polyunsaturated fatty acids clearly
possess a selective ability to reduce the amount of protein and mRNA for
several enzymes of fatty acid biosynthesis.
While dietary fat rich in 18:2 (n-6) suppressed the expression of enzymes
involved in fatty acid biosynthesis, the level of mRNA and protein for the
liver fatty acid binding protein (1-FABP) is induced by dietary fatsY A
-- 150.--------------------------------------------,
+-'
Q)
0
+-'
If
0
z 100
'*'
Cl)
Cl)
Q)
~
+-'
c
>
Cl) 50
:'2
()
<0
>
+-'
+-'
If
Liver Gut Lung Brain Adipose
Tissue
FIGURE 2. Tissue-specific inhibition of lipogenesis by polyenoic fatty acids. Mice (n = 4)

were fed for 5 d a high glucose diet containing 10% hydrogenated cottonseed oil or triolein
(single slash bars), 10% com oil (solid bars), or 10% menhaden fish oil (double slash bars). In
vivo fatty acid synthesis was measured by tritiated water incorporation (SEM < 15%).
selective induction of 1-FABP due to polyunsaturated fats has not been firmly
established, but studies with rat liver cells in primary culture indicate 18:2
(n-6) directly enhanced 1-FABP synthesis, and resulted in higher levels of 1-
FABP protein than did 16:0. 24 Clearly, polyunsaturated fatty acids uniquely
regulate hepatic gene expression, and this effect may extend to genes coding
for proteins other than enzymes in fatty acid biosynthesis.
The polyunsaturated fatty acid control of lipogenic proteins is predomi-
nately a liver-specific phenomenon (Figures 2 and 3). 25 Fatty acid biosynthesis
and the activities of lipogenic enzymes within rodent lung, brain, and small
intestine were unaffected by a diet that contained 20 to 30% kcal as hydro-
genated cottonseed oil, com oil, and menhaden oil (Figure 2). 9 ·25 Comparable
levels of dietary fat reduced de novo fatty acid biosynthesis in mammary and
epididymal fat pads, but saturated and polyunsaturated fats were equipotent
(Figure 2). Interestingly, fatty acid synthase mRNA levels were unaffected
by either dietary saturated or polyunsaturated fat (Figure 3), 9 which is in sharp
contrast to the hepatic responses to these dietary lipids. 11 •23 Since dietary fats
suppress adipose fatty acid synthase activity and rates of fatty acid biosyn-
thesis, 15 the data suggest fatty acid synthase synthesis in adipose tissue may
be regulated at the level of translation as well as at the point of transcription.
Clarke and Jump 233
-Cl)
150
-
0
CD
u..
0
z
*'
<(
100
z
a:
E
Cl)
-
fll
CD
..r:
c: 60
>
Vl
"'0
"(j
<(
-
>
CD
u.. 0
Liver Mam Fat Epid Fat
Tissue
FIGURE 3. Tissue-specific suppression of fatty acid synthase by dietary polyunsaturated fats.

Mice (n = 4) were fed for 5 d a high glucose diet containing 10% hydrogenated cottonseed oil
(single slash bars), 10% com oil (solid bars), or 10% menhaden fish oil (double slashed bars).
Fatty acid synthase mRNA abundance was determined by DNNRNA hybridization. 11
Regardless of the mechanism of regulation, clearly the unique control of fatty

acid synthase expression by polyunsaturated fats noted for hepatocytes does
not extend to the adipocyte.
IV. POTENCY OF FATTY ACID INHIBITORS

As might be expected the inhibitory potency of dietary polyunsaturated
fatty acids varies depending upon the degree of unsaturation and chain length
(Figures 1 through 4). 2 - 5 • 11 In general, 18:3 (n-3) is more potent than 18:2 (n-
6), but, most importantly, products of the delta-6 desaturase are 2 to 4 times
more potent suppressors than are the respective fatty acid precursors (Figure
6). 8 ·9 ·22 Obviously, since dietary oils vary in their polyunsaturated fatty acid
content, they also differ in their capacity to inhibit lipogenesis and the expres-
sion of lipogenic enzymes. Thus, olive oil (rich in 18:1) has no suppressive
effect, safflower oil (75% 18:2 [n-6]) is more potent than com oil (50% 18:2
[n-6]), and menhaden fish oil is two- to threefold more potent than safflower
oil. The greater potency of fish oil is observed as a twofold greater maximal
suppression of fatty acid synthase expression, and as a threefold lower re-
120
-.....
Q)
i:5
100 ---- + - -~--t--
.....
'\,""'---J
'
co
LL. 80
z
0
~~~
-
'#.. 60 r -~~"
~
>
.-::
>
'_;j 40
--------
"i----
()
<(
en
20
·t-~---~-
~ ~--
0
0 3 6 10 15
Diet Fat Dose (% w/w)
FIGURE 4. Response of hepatic fatty acid synthase activity to varying doses and types of
dietary fat. Mice (n = 4) were fed a high glucose diet for 5 d containing the level of fat (w/w)
cited in the figure. Hydrogenated cottonseed oil, com oil, and menhaden fish oil are represented
by the open squares, open triangles, and crosses, respectively. Data are expressed as means ±
SEM for n = 4 mice per point.
quirement for fish oil to achieve 50% of maximum inhibition (Figure 4). 9 A
comparison of 18:3 (n-6) and 20:4 (n-6) with fish oil has shown that the key
difference in potency between fish oil and vegetable oil lies in the conversion
rate of 18:2 (n-6) to 18:3 (n-6), i.e., 20:4 (n-6) is equipotent to fish oil (Figure
6). Thus, the greater potency of fish oil as an inhibitor of fatty acid synthase
and S 14 gene expression is due primarily to its high proportion (> 30%) of
20- and 22-carbon polyenoic fatty acids.
V. HEPATIC-SPECIFIC INHIBITION
Dietary studies cannot determine if fatty acids directly inhibit hepatocyte

gene expression or indirectly function via peripheral release of a fatty acid
metabolite or hormonal factor. Results with rat hepatocytes in primary culture
indicate that polyunsaturated fatty acids do not require peripheral tissue me-
tabolism nor cause the release of a peripheral tissue factor or hormone. 26 •27
The hormonal induction of acetyl CoA carboxylase, malic enzyme, and glu-
cose-6-phosphate dehydrogenase activities as well as the induction of fatty
acid synthase mRNA is suppressed by the addition of albumin-bound poly-
C Iarke and 1 ump 235
co 80
en
.5!
>
X
0 ~--:::-
..0
..... / :;..e-
co 60 /
uI b/
<(
0
u()
<( /
....c: 40
/
/
....enco /
I"
/> /
~
"(j)
co
a: 20
c:
"U
<(
>
"#. 0
0 10 20 30 120
Minutes into Meal
FIGURE 5. The effect of dietary saturated and polyunsaturated fats on the activation of acetyl
CoA carboxylase during meal ingestion. Rats were trained to eat their entire daily food allotment
during a 3-h meal. The fraction of catalytically active acetyl CoA carboxylase present in hepatic
tissue during the ingestion of a single meal of 10% tripalmitin (solid circles), 10% safflower oil
(open squares), or no fat (solid squares) was determined using the avidin inactivation proce-
dure.'•·'9
enoic (n-6) fatty acids to the media (Figure 7). 26 •27 As occurs in vivo, 20:4
(n-6) was more potent than 18:2 (n-6). Approximately 100 IJ.M 20:4 (n-6)
was required to suppress fatty acid synthase mRNA induction 50%. 26 Ad-
mittedly, circulating levels of 20:4 (n-6) above 50 ~J.M are unlikely. However,
hepatocytes in culture rapidly utilize 20:4 (n-6) for triglyceride and phospho-
lipid synthesis. 28 Thus, the media concentration of 20:4 (n-6) must be main-
tained at a high level in order to provide an effective concentration sufficient
to exert an inhibitory action. As would be expected from the in vivo responses,
as much as 500 1J.M 20: 1 (n-9) did not inhibit fatty acid synthase expression,
and in fact, increased synthase mRNA levels twofold (Figure 7). 26 One pos-
sible explanation for the enhancement of fatty acid synthase gene expression
by 20:1 (n-9) may be that 20: 1 (n-9) displaced the inhibitor 20:4 (n-6).
Interestingly, as the media content of 20:4 (n-6) was increased from 50
to 500 IJ.M, the level of 13-actin mRNA rose 70 to 275%, but 20:1 (n-9) had
no effect on 13-actin expression (Figure 7). 26 13-Actin expression is used as
an indicator of hepatocyte differentiation status, i.e., high levels of 13-actin
mRNA may be indicative of a loss of adult differentiation traits. 29 It is curious
-CD
0
c
100
as
"C
c 75
::l
.c
as
,...
-
co
,...
50
0
-z
0~
<(
25
a:
E
CIJ
~ 0
Type of Diet Fat
FIGURE 6. Delta-6 desaturation of linoleate and o:-linolenate is required to suppress fatty acid
synthase gene expression. Mice were fed a high glucose diet for 5 days containing various fatty
acids with (solid bars) and without (slashed bars) the delta-6 desaturase inhibitor eicosatetraynoic
acid (ETYA). The level of fat supplementation to the diet was 5% for 18:2, 18:3, and 18:1, and
3% for 20:3 and 20:4. The 20:5/22:6 column represents 10% dietary menhaden fish oil which
provides 3% 20:5/22:6 (n-3).
as to what effects polyunsaturated fatty acids may exert on adult hepatic

function and differentiation characteristic. In any case, it appears that po-
lyunsaturated fatty acids specifically regulate gene expression events of the
hepatocyte and that these events may extend to hepatic genes other than those
coding for lipogenic proteins.
VI. STRUCTURAL REQUIREMENTS AND FATTY

ACID POTENCY
The basic requirement for a dietary fatty acid to inhibit expression of
lipogenic proteins is that it contain 18 carbons and possess at least 2 conjugated
double bonds in the 9,12-positions. 7 · 9 Thus, polyenoic fatty acids of the (n-
9) family as well as unusual fatty acids such as 5c, 13c-22:2 do not suppress
fatty acid biosynthesis (Figure 1). Similarly, loss of one of the double bonds
by hydroxylation renders the fatty acid inactive. 30 However, a fatty acid may
contain additional double bonds and still retain inhibitory potency. For ex-
ample, dietary columbinic acid (5t,9c, 12c-18:3) and alpha-linolenic acid
C Iarke and 1 ump 237
501!M
0 20:1 20:4 20:1 20:4 20:1 20:4
FAS ·
l'l · Actin·
FIGURE 7. Northern analysis demonstrating suppression of fatty acid synthase gene expression
in primary cultures of rat liver cells by albumin-bound arachidonate. Rat liver cells were main-
tained in primary culture 38 and incubated for 36 h with the levels of albumin-bound 20:4 (n-6)
or 20: I (n-9) cited in the figure.
(9c, 12c, 15c-18:3) both reduce the abundance of fatty acid synthase mRNA
and the rate of lipogenesis. 11 •20 With respect to double bond conformation,
one of the two double bonds may be in the trans configuration, but 9t,12t-
18:2 does not lower the level of hepatic fatty acid synthase rnRNA (Figure
7).7
While dietary 18:2 (n-6) and 18:3 (n-3) markedly reduce the abundance
of hepatic fatty acid synthase and S14 mRNAs, 11 the actual intracellular
modulator of gene expression is not 18:2 (n-6) or 18:3 (n-3) per se. 9 In order
for these fatty acids to suppress the expression of genes coding for lipogenic
enzymes, they must first undergo desaturation by the delta-6 desaturase. When
delta-6 desaturation was blocked by the fatty acid analog, eicosatetraynoic
acid (ETYA), the suppression of fatty acid synthase mRNA level by dietary
safflower oil, com oil, or 18:2 (n-6) was completely prevented (Figure 6),
but the hepatic concentration of 18:2 (n-6) dramatically increased. 8 •9 Adding
the product of the delta-6 desaturase (e.g., 6c,9c,12c-18.3) completely rein-
stated the inhibition of gene expression (Figure 6). Under these conditions,
18:3 (n-6) and 20:3 (n-6) content of hepatic lipids increased, but conversion
of 20:3 (n-6) to 20:4 (n-6) continued to be impaired because ETYA also
inhibited the delta-5 desaturase. Thus, while 20:4 (n-6) and 20:5 (n-3) are
potent suppressors of fatty acid synthase gene expression, the formation of
these fatty acids is not an a priori requirement for the polyenoic fatty acid
regulation of gene expression to occur.
The potent suppression of fatty acid synthase and S 14 mRNA levels by

fish oil rich in (n-3) fatty acids combined with the observation that formation
of 20:4 (n-6) was not required to observe the regulation of gene expression.
This suggests that prostanoids may not be involved in the polyunsaturated
fatty acid mechanism. Consistent with this conclusion is the failure of several
prostanoid inhibitors to prevent the suppression of fatty acid synthase gene
expression. 6 •8 •31 For example, the prostaglandin inhibitors indomethacin and
ibuprofin 6 •31 and the leukotriene inhibitors nordihydroquaiaretic acid, quer-
citin, and phenidone did not prevent dietary com oil or safflower oil from
suppressing hepatic fatty acid synthase and glucose-6-phosphate dehydroge-
nase activities. 9 Similarly, columbinic acid, which is not a substrate for pros-
tanoid synthesis, continued to suppress fatty acid synthase enzyme activity. 20
However, the inability to detect an effect does not mean an effect does not
exist. Selective regulation of gene transcription by a fatty acid may require
high affinity binding of a polyunsaturated fatty acid mediator to a nuclear
trans-acting protein. Such selectivity will likely require modification of the
fatty acid by mechanisms such as hydroxylation or oxidation. 32 Since nu-
merous possibilities exist for this hypothetical fatty acid metabolite, eluci-
dation of the nuclear proteins which regulate fatty acid synthase and S 14 gene
expression will greatly aid in identifying the fatty acid metabolite ligand
responsible for controlling gene expression.
VII. TRANSCRIPTIONAL REGULATION BY

POLYUNSATURATED FATS
The suppression of fatty acid synthase and S14 mRNA abundance in

hepatic tissue caused by the ingestion of polyunsaturated fatty acids could
result from: (1) an inhibition of gene transcription, (2) an interference with
the maturation of the nascent transcript, and/or (3) an acceleration of the rate
of mRNA degradation. Nuclear run-on assays with nuclei prepared from livers
of rats fed a diet containing 10% menhaden fish oil vs. tripalmitin or triolein
demonstrated that polyunsaturated fats determine the level of fatty acid syn-
thase and S14 mRNAs by controlling the rate of gene transcriptionY For
example, menhaden oil suppressed fatty acid synthase gene transcription
94% and reduced the level of its mRNA 80%. 12 Similarly, hepatic S14 gene
transcription and mRNA levels were both suppressed 60 to 70% by dietary
fish oil. 12 Since the transcription rates for phosphoenolpyruvate carboxy kinase
and 13-actin were unchanged, the effects of fish oil appear to be specific for
genes coding for lipogenic proteins. These results do not completely eliminate
the possibility that polyunsaturated fatty acids could subtly influence transcript
maturation processes and/or mRNA degradation rates, but certainly the pre-
dominate control mechanism by which polyunsaturated fatty acids regulate
the hepatic expression of lipogenic proteins appears to be via inhibition of
Clarke and Jump 239
gene transcnpt10n. Malic enzyme may be an exception to this conclu-

sion, 30 •33 •34 but this protein does not play a pivotal role as a determinant of
flux in the fatty acid biosynthetic pathway.
In light of the potent suppression of fatty acid synthase transcription by
polyunsaturated fatty acids, we have speculated that polyunsaturated fatty
acids or a metabolite regulate the activity of a specific transacting factor which
acts on a specific cis-acting element associated with the fatty acid synthase
and Sl4 genes. In order to address this question we have cloned from a rat
genomic DNA cosmid library a 39-kb insert that contains the entire fatty acid
synthase gene. Based upon a recently published sequence for the putative
start site of fatty acid synthase transcription, 35 our clone extends 2.5 kb
upstream from the start site of transcription. The region of - 2495 to + 1500
of the fatty acid synthase gene has been fully sequenced and contains several
nucleotide motifs which are characteristic of a promoter region. First, TATA
and inverted CAAT boxes were located within the first 100-bp 5' to the
putative transcription start site. The TATA and CAAT boxes were separated
by a 60-bp sequence that was comprised of more than 75% G-C nucleotides.
In addition, two G-C boxes with consensus sequences characteristic of a
binding site for transcription factor Spl (i.e., 5' -GGGCGG-3') were located
within 150 bp of the CAAT box. Also contained within the + I to - 2452
region were 2 sets of glucocorticoid-like response elements (GREs) which
were located at - 1534 and -2000, as well as a prospective thyroid response
element (TRE) located at -650.
Using the in vitro transcription initiation assay and transient transfection
of liver cells, we have found the region - 2195 to + 65 of the fatty acid
synthase gene does in fact contain a transcriptionally active promoter and
certain hormonal regulatory elements characteristic of the endogenous gene.
Nuclear protein extracts from hepatic tissue, but not spleen or kidney, have
been found to drive the in vitro transcription of an artificial G-free gene. 36
These responses required the 5' to 3' orientation. When the - 2195 to + 65
region of the fatty acid synthase gene was linked to the chloramphenicol
acetyltransferase reporter and rat liver cells in primary culture were transiently
transfected. CAT activity was expressed. Moreover, the CAT expression was
stimulated by glucocorticoids (Figure 8). The glucocorticoid-dependent CAT
expression is consistent with the glucocorticoid-dependent induction of the
endogenous fatty acid synthase gene (Figure 9) and consistent with the pres-
ence of the two GREs within this region. Neither T3 nor insulin enhanced
CAT expression (Figure 8). Since expression of fatty acid synthase is an
insulin-dependent process (Figure 9), 37 it would appear that the - 2195 region
does not contain all regulatory nucleotide sequences characteristic of the fatty
acid synthase gene. Thus, while the -2195 to +65 region contains an active
promoter for fatty acid synthase which is capable of initating transcription in
a tissue-specific manner, it appears that additional sequences are essential to
understanding in vivo expression in this gene. The successful cloning of the
250,------------------------------------------ ---.
Q)
c
0
E
0
2oo
I
-
0
z
150
0
'#.
c
0 100
CJ)
CJ)
CD
.....
a.
X
w 50
~
u
Hormonal Treatment
FIGURE 8. Stimulation of fatty acid synthase promoter activity by glucocorticoids. Rat liver
cells in primary culture 38 were transiently transfected by lipofectin with a pSV2 vector containing
- 2190 to + 65 of the fatty acid synthase gene, the chloramphenicol acetyltransferase reporter,
and the SV40 enhancer 5' to the promoter. The hormone concentrations were 0.1 f.LM and SEM
< 15%.
fatty acid synthase gene and its promoter, coupled with the extensive char-
acterization of the S14 gene, will permit us to identify potential cis-elements
and trans-acting nuclear proteins which may be under polyunsaturated fatty
acid control.
VIII. MECHANISMS OF TRANSCRIPTIONAL

CONTROL
Our early understanding of the mechanisms by which polyunsaturated

fats suppressed lipogenesis and of the activities of lipogenic enzymes sug-
gested that the process was a long-term adaptation that required peripheral
metabolism of the ingested polyunsaturated fatty acids to an active derivative
or required the release of a hormone which suppressed hepatic lipogenesis.
However, these early kinetic studies relied heavily upon changes in the ac-
tivities of lipogenic enzymes. 1' 5 Since the average half-life of a lipogenic
enzyme is 36 to 48 h, we could not evaluate the early changes which might
occur in gene transcription or mRNA abundance as a result of polyunsaturated
fat ingestion. Thus, the inhibitory effects of polyunsaturated fats would (by
Clarke and Jump 241
1 2 3 4 5 6 7 8
FAS-
28S-
tJ- Actin-
FIGURE 9. Hormonal induction of fatty acid synthase gene expression in primary cultures of
rat liver cells. Rat liver cells were maintained in primary culture" and treated with the hormones
cited in the figure. RNA was extracted and subjected to Northern analyses. The hormonal
concentrations were 0. I f.LM. Lane I, no hormone; Jane 2, insulin; Jane 3, insulin/dexamethasone;
lane 4, dexamethasone; lane 5, insulinldexamethasoneff3 ; Jane 6, T 3 ; Jane 7, T 3/insulin; and lane
8, T,!dexamethasone.
definition) appear to involve long-term events. Through the use of primary

rat hepatocytes in monolayer culture, 26 ·27•38 we have demonstrated that po-
lyunsaturated fatty acids directly regulate hepatocyte gene expression and do
not require their peripheral metabolism to an active metabolite (Figure 7).
However, one obvious and frequently argued adaptation is that ingestion
of oils rich in polyenoic fatty acids results in the enrichment of hepatic
membrane phospholipids with polyenoic fatty acids, and that the accompa-
nying rise in membrane fluidity causes an increase hepatocyte responsivity
to 13-agonists (e.g., glucagon) and/or decreased responsivity to lipogenic
agonists (e.g., insulin) . Such a shift in hormonal balance could lead to
increased cAMP-dependent protein kinase activity and subsequently suppres-
sion of fatty acid synthase and S 14 transcription. There is little question that
dietary polyunsaturated fats alter hormone binding and membrane signaling
mechanisms, 39 A 2 but the hormonal responsivity changes induced by polyun-
saturated fatty acids do not present a pattern consistent with suppression of
fatty acid synthase and S 14 gene expression. For example, although dietary
polyunsaturated fats increase glucagon-stimulated adenylate cylase activity
twofold, 39 they also significantly increase tissue sensitivity to insulin. 40 .4 2
Moreover, the dose of com oil required to elicit maximal increases in glu-
cagon sensitivity was only 2% of the diet, 39 while inhibition of fatty acid
synthase gene expression by com or safflower oil continues to be expressed
at doses up to 10% of the diet (Figure 4). 11 •37 The strongest argument against
the idea that fatty acid membrane compositional changes are responsible for
the control of gene expression lies in the kinetics of transcriptional inhibition
vs. membrane fatty acid changes. Removing fish oil from the diet for simply
one 3-h meal completely reverses the near 90% suppression of fatty acid
synthase gene expression. 11 Similarly, transcription of acetyl CoA carbox-
ylase was suppressed 50% within 6 h of adding com oil to the diet. 21 In our
opinion, 3 to 6 h is insufficient time for a dietary lipid to achieve significant
changes in membrane phospholipid fatty acid composition. Rather, the ki-
netics of transcriptional regulation are more consistent with direct control of
gene transcription by polyunsaturated fatty acids or their metabolites.
The final piece of evidence against a change in cAMP balance caused
by dietary polyunsaturated fats is derived from the effects of these fatty acids
on gene transcription. 12 If the polyunsaturated fatty acid suppression of fatty
acid synthase and Sl4 gene transcription were to involve a cAMP-dependent
process, then one would expect that the transcription rate of phosphoenol-
pyruvate carboxykinase (PEPCK) would be increased by dietary polyunsa-
turated fats. However, PEPCK transcription, which is exquisitely sensitive
to stimulation by cAMP and inhibition by insulin, 43 was not affected by dietary
polyunsaturated fat. 12 Thus, collectively the available data suggest that the
regulation of gene expression by polyunsaturated fats does not involve shifts
in the balance of hormones which modulate the cAMP second message system.
While enrichment of membrane lipids with polyunsaturated fats appears
not to be the mechanism by which polyunsaturated fatty acids regulate fatty
acid synthase and S 14 gene expression, it is feasible that long chain polyenoic
fatty acids interfere with nuclear events regulated by lipogenic hormones. For
example, fatty acids or nuclear fatty acid binding proteins might potentially
interact with T3 or glucocorticoid receptors in such a way as to interfere with
receptor/DNA interactions and thereby alter transcriptional events. While such
an idea is purely speculation, dietary polyunsaturated fats were found to
interfere with T3 induction of hepatic fatty acid synthase, malic enzyme, and
glucose-6-phosphate dehydrogenase. 44 The effect of a diet containing 10%
safflower oil was to shift the dose-response curve for T3 to the right, 44 but
not to reduce the maximum activity induced by high doses ofT 3 • Such a shift
Clarke and Jump 243
in T3 responsivity suggests dietary polyunsaturated fatty acids may function

as competitive inhibitors of the T3 induction of lipogenic enzymes. The mech-
anism of such a shift is unclear, but interference with T3 receptor binding
and/or receptor/DNA interactions may be potential sites of polyenoic fatty
acid action.
IX. SUMMARY AND SPECULATION
The tissue-specific response of the fatty acid synthase gene to polyun-

saturated fat inhibition, and the rapid transcriptional suppression of genes
coding for lipogenic enzymes, has led us to speculate that the nucleus of
hepatocytes contains a nuclear protein (e.g., nuclear fatty acid binding protein)
which selectively binds 18:3 (n-6) and 18:4 (n-4) polyenoic fatty acids and
their metabolites. Upon binding of the polyenoic fatty acid ligand, the putative
trans-acting protein would bind to a cis-acting element linked to the fatty
acid synthase (or Sl4) gene which would result in suppression of gene tran-
scription. Such a cis-element could be specific for the proposed nuclear fatty
acid binding protein or could involve interaction with a defined hormonal
response element, e.g., thyroid response element. Alternatively, transcrip-
tional regulation could also be regulated via fatty acid ligand binding to nuclear
fatty acid binding protein which would allow the protein to interact with other
nuclear hormonal receptors, such as the glucocorticoid or thyroid receptors,
and thereby alter the hormonal action of these signals. A single nuclear fatty
acid binding protein which bound to a specific nucleotide response sequence
that was common to all genes coding for lipogenic enzymes would explain
the apparent coordinate suppression of lipogenic enzymes by dietary polyun-
saturated fats. Obviously at this time our hypothesis is simply speculation,
but precedent for the concept is derived from the mechanism by which cho-
lesterol regulates HMG-CoA reductase and the LDL receptor, 45 as well as
from the recent demonstration of a nuclear binding protein for clofibrate and
its involvement in the induction of peroxisomes. 46
ACKNOWLEDGMENTS
The work presented within this review was supported in part by the Upjohn
Company, Kalamazoo, MI and the National Institutes of Health-DK 39302
(SOC) and GM 36851 (DBJ). The authors wish to thank Ormond MacDougal,
William Blake, Meera Kamdar, and Ana Mildner for the technical assistance
and thoughtful discussions. Dr. Clarke is the recipient of the Lillian Fountain
Smith Endowed Professorship of Human Nutrition at Colorado State Uni-
versity.
REFERENCES
I. AHmann, D. W. and Gibson, D. W., Fatty acid synthesis during early linoleic acid
deficiency in the mouse, J. Lipid Res., 6, 51, 1969.
2. Sabine, J. R., McGrath, H., and Abraham, S., Dietary fat and the inhibition of hepatic
lipogenesis in the mouse, J. Nutr., 98, 312, 1969.
3. Musch, K., Ojakian, M. S., and Williams, M. A., Comparison of o:-linolenate and
oleate in lowering activity of lipogenic enzymes in rat liver: evidence for a greater effect
of dietary linolenate independent of food and carbohydrate intake, Biochirn. Biophys.
Acta, 337, 343, 1974.
4. Clarke, S. D., Romsos, D., and Leveille, G. A., Differential effects of dietary methyl
esters of long-chain saturated and polyunsaturated fatty acids on rat liver and adipose
tissue lipogenesis, J. Nutr., 107, 1170, 1977.
5. Toussant, M. J,, Wilson, M. D., and Clarke, S. D., Coordinate suppression of liver
acetyl-CoA carboxylase and fatty acid synthase by polyunsaturated fat, J. Nutr., Ill,
146, 1981.
6. Flick, P. K., Chen, J,, and Vagelos, P.R., Effect of dietary linoleate on synthesis and
degradation of fatty acid synthetase from rat liver, J. Bioi. Chern., 252, 4242, 1977.
7. Emken, E. A., Abraham, S., and Lin, C. Y., Metabolism of cis-12-octadecenoic acid
and trans-9, trans-12 octadecadienoic acid and their influence on lipogenic enzyme ac-
tivities in mouse liver, Biochirn. Biophys. Acta, 919, Ill, 1987.
8. Clarke, B. A. and Clarke, S. D., Suppression of rat liver fatty acid synthesis by eicosa-
5,8,11, 14-tetraynoic acid without a reduction in lipogenic enzymes, J. Nutr., 112, 1212,
1982.
9. Clarke, S. D. and Abraham, S., Inhibition of fatty acid synthase gene expression by
polyunsaturated fat: a requirement for delta-6 desaturation, submitted.
10. Clarke, S. D., Armstrong, M. K., and Jump, D. B., Nutritional control of rat liver
fatty acid synthase and Sl4 mRNA abundance. J. Nutr., 120, 218, 1990.
II. Clarke, S. D., Armstrong, M. K., and Jump, D. B., Dietary polyunsaturated fats
uniquely suppress rat liver fatty acid synthase and S14 mRNA content, J. Nutr., 120,
225, 1990.
12. Blake, W. L. and Clarke, S. D., Suppression of hepatic fatty acid synthase and S14
gene transcription by dietary polyunsaturated fat, J. Nutr., 120, 1727, 1990.
13. Hillard, B. L., Lundin, P., and Clarke, S.D., Essentiality of dietary carbohydrate for
maintenance of liver lipogenesis in the chick,}. Nutr., 110, 1533, 1980.
14. Bortz, W. M., Abraham, S., and Chaikoff, I. L., Localization of the block in lipo-
genesis resulting from feeding fat, J. Bioi. Chern., 238, 1266, 1963.
15. Romsos, D. R. and Leveille, G. A., Effects of diet on activity of enzymes involved in
fatty acid and cholesterol synthesis, Adv. Lipid Res., 12, 97. 1974.
16. Wilson, M. D., Blake, W. L., Salati, L. M., and Clarke, S. D., Potency of polyun-
saturated and saturated fats as short term inhibitors of hepatic lipogenesis in rats, J. Nutr.,
120. 544, 1990.
17. Clarke, S. D. and Hillard, B. L., Suppression of hepatocyte fatty acid synthesis by
albumin-bound linoleate involves depolymerization of acetyl-CoA carboxylase filaments,
Lipids, 15, 207, 1981.
18. Ashcraft, B. A., Fillers, W. S., Augustine, S. L., and Clarke, S.D., Polymer-protomer
transition of acetyi-CoA carboxylase occurs in vivo and varies with nutritional conditions,
J. Bioi. Chern., 255, 10033, 1980.
19. Clarke, B. A. and Clarke, S.D., Polymer-protomer transition of acetyl-CoA carboxylase
as a regulator of lipogenesis in rat liver, Arch. Biochern. Biophys., 218, 92, 1982.
20. Schwartz, R. S. and Abraham, S., Effect of dietary polyunsaturated fatty acids on the
activity and content of fatty acid synthetase in mouse liver, Biochirn. Biophys. Acta, 711,
316, 1982.
Clarke and Jump 245
21. Katsurada, A., Iritani, N., Fukuda, H., Matsumura, Y., Nishimoto, N., Noguchi,
T., and Tanaka, T., Effects of nutrients and hormones on transcriptional and post-
transcriptional regulation of acetyl-CoA carboxylase in rat liver, Eur. }. Biochem., 190,
435, 1990.
22. Tomlinson, J, E., Nakayama, R., and Holten, D., Repression of pentose phosphate
pathway dehydrogenase synthesis and mRNA by dietary fat in rats, J. Nutr., 118,408,
1988.
23. Dempsey, M. E., Hargis, P. S., McGuire, D. M., McMahon, A., Olson, C. D.,
Salati, L. M., Clarke, S. D., and Towle, H. C., Role of sterol carrier protein in
cholesterol metabolism, Chern. Phys. Lipids, 38, 223, 1985.
24. Salati, L. M., Hargis, P. S., Olson, C. D., Clarke, S. D., and Dempsey, M. E.,
Rapid regulation of sterol carrier protein synthesis, secretion, and turnover in rat hepa-
tocytes, Fed. Proc., 45 (Abstr.), 1581, 1986.
25. Clarke, S. D., Wilson, M. D., and lbnoughazala, T., Resistance of lung fatty acid
synthesis to inhibition by dietary fat in the meal-fed rat, J. Nutr., 114, 598, 1984.
26. Armstrong, M. K., Blake, W. L., and Clarke, S. D., Arachidonic acid suppression
of fatty acid synthase gene expression in cultured rat hepatocytes, Biochem. Biophys.
Res. Commun., 177. 1056, 1991.
27. Salati, L. M., Adkins-Finke, B. A., and Clarke, S.D., Fatty acid regulation of glucose-
6-phosphate dehydrogenase in rat liver cell monolayers, Lipids, 23, 36, 1988.
28. Odin, R. S., Adkins-Finke, B. A., Blake, W. L., Phinney, S.D., and Clarke, S. D.,
Membrane phospholipid enrichment of cultured rat hepatocytes with omega-3 and omega-
6 fatty acids is associated with decreased triglyceride production and secretion, Biochim.
Biophys. Acta, 921, 378, 1987.
29. Reid, L. M., Narita, M., Fujita, M., and Rosenberg, L., Isolated and Cultured
H epatocytes, Guill ouzo, A. and Gugen-Guillouzo, C., Eds., Hohn Libbey, London, 1986,
225.
30. Schwartz, R. S. and Abraham, S., Effect of dietary fat on the activity, content, rates
of synthesis, and degradation and translation of messenger RNA coding for malic enzyme,
Arch. Biochem. Biophys., 221, 206, 1983.
31. Szepesi, B., Kamara, A. K., and Clarke, S.D., Lack of specificity of polyunsaturated
fats in the inhibition of rat liver glucose-6-phosphate dehydrogenase, J. Nutr., 119, 161,
1989.
32. Nebert, D. W., Proposed role of drug-metabolizing enzymes: regulation of steady state
levels of the ligands that effect growth, homeostasis, differentiation, and neuroendocrine
functions, Mol. Endocrinol., 5, 1203, 1991.
33. Dozin, B., Rail, J., and Nikodem, V. M., Tissue-specific control of rat malic enzyme
activity and messenger RNA levels by a high carbohydrate diet, Proc. Nat/. Acad. Sci.
U.S.A., 83, 4705, 1986.
34. Katsurada, A., Iritani, N., Fukuda, H., Noguchi, T., and Tanaka, T., Influence of
diet on the transcriptional and post -transcriptional regulation of malic enzyme induction
in the rat liver, Eur. J. Biochem., 168, 487, 1987.
35. Amy, C. M., Williams-Ahlf, B., Naggert, J,, and Smith, S., Molecular cloning of
the mammalian fatty acid synthase gene and identification of the promoter region, Biochem.
J., 271, 675, 1990.
36. MacDougal, 0., Clarke, S.D., and Jump, D. B., Unpublished observations.
37. Jump, D. B., Bell, A., Lepar, G., and Hu, D., Insulin rapidly induces rat liver Sl4
gene transcription, Mol. Endocrinol., 4, 1655, 1990.
38. Salati, L. M. and Clarke, S. D., Fatty acid inhibition of the hormonal induction of
acetyl-CoA carboxylase in hepatocyte monolayers, Arch. Biochem. Biophys., 246, 82,
1986.
39. Dax, E. M., Partilla, J. S., Pineyro, M.A., and Gregerman, R. 1., Altered glucagon-
and catecholamine hormone-sensitive adenylyl cyclase responsiveness in rat liver mem-
branes induced by manipulation of dietary fatty acid intake, Endocrinology, 127, 2236,
1990.
40. Field, C. J., Ryan, E. A., Thomson, A. B. R., and Clandinin, M. T., Diet fat
composition alters membrane phospholipid composition insulin binding, and glucose
metabolism in adipocytes from control and diabetic animals, J. Bioi. Chem., 265, 11143,
1990.
41. Blake, W. L. and Clarke, S. D., Type of dietary fat alters rat liver cell responsiveness
to vasopressin, J. Nutr. Biochem., 2, 87, 1991.
42. Clandinin, M. T., Cheema, S., Field, C. J., Garg, M. L., Vendatraman, J., and
Clandinin, T. R., Dietary fat: exogenous determination of membrane structure and cell
function, FASEB J., 5, 2761, 1991.
43. O'Brien, R. M., Lucas, P. C., Forest, C. D., Magnuson, M. A., and Granner,
D. K., Identification of a sequence in the PEPCK gene that mediates a negative effect
of insulin on transcription, Science, 249, 533, 1990.
44. Clarke, S. D. and Hembree, J., Dietary fat suppresses the triiodothyronine induction
of rat liver lipogenic enzymes, J. Nutr., 120, 625, 1990.
45. Dawson, P. A., Hofmann, S. L., Vander Westhuyzen, D. R., Sudhof, T. C., Brown,
M. S., and Goldstein, J. L., Sterol-dependent repression of low density lipoprotein
receptor promoter mediated by 16-base pair sequence adjacent to binding site for tran-
scription factor Spl, J. Bioi. Chem., 263, 3372, 1988.
46. Issemann, I. and Green, S., Activation of a member of the esteroid hormone receptor
superfamily by peroxisome proliferators, Nature (London), 347, 645, 1990.
47. Clarke, S. D., Romsos, D. R., and Leveille, G. A., Influence of dietary fatty acids
on liver and adipose tissue lipogenesis and on liver metabolites in meal-fed rats, J. Nutr ..
107, 1277, 1977.
48. Herzberg, G. R., The influence of dietary fatty acid composition on lipogenesis, Adv.
Nutr. Res., 5, 221, 1983.
49. Christiansen, E. N., Lund, J. S., Rortveit, T., and Rustan, A. C., Effect of dietary
n-3 and n-6 fatty acids on fatty acid desaturation, Biochim. Biophys. Acta. 1082, 57,
1991.
50. Baltzell, J. K., Wooten, J. T., and Otto, D. A., Lipoprotein lipase in rats fed fish oil:
apparent relationship to plasma insulin levels, Lipids, 26, 289.
51. Miller, J. C. E., Barth, R. K., Shaw, P. H., Elliot, R. W., and Hastie, N. D.,
Identification of a eDNA clone for mouse apoprotein A-1 (apo A 1) and its use in
characterization of apo A-1 mRNA expression in liver and small intestine, Proc. Nat/.
Acad. Sci. U.S.A., 80, 1511, 1983.
Chapter 11
DIETARY FAT, GENE EXPRESSION, AND

CARCINOGENESIS
Howard P. Glauert
TABLE OF CONTENTS
I. Introduction ...................................................... 248
II. Mechanisms of Carcinogenesis ................................... 248

A. Chemical Carcinogenesis ................................. 248
B. Multistage Carcinogenesis ................................ 249
C. Genes Important in Carcinogenesis ....................... 250
Ill. Epidemiological Studies ......................................... 250
IV. Experimental Carcinogenesis Studies ............................ 251

A. Nutritional Considerations ................................ 251
B. Skin Carcinogenesis ...................................... 252
C. Hepatocarcinogenesis ..................................... 252
D. Colon Carcinogenesis .................................... 253
E. Pancreatic Carcinogenesis ................................ 254
F. Mammary Carcinogenesis ................................ 255
G. Other Sites ............................................... 255
H. Effect of n-3 Fatty Acids ................................ 255
V. Mechanisms by which Dietary Fat May Influence

Carcinogenesis ................................................... 256
A. Membrane Fluidity ....................................... 256
B. Toxicity .................................................. 256
C. Eicosanoid Metabolism ................................... 256
D. Caloric Effects ........................................... 257
E. Effect on Initiation ....................................... 257
F. Induction of Specific Genes .............................. 258
VI. Summary and Conclusions ....................................... 258
References ............................................................... 259
-------
0-8493-6961-4/93/$0.00+$50
C0 1993 by CRC Press, 1nc. 247
I. INTRODUCTION
The amount and type of fat consumed in the diet may be important in
the development of human cancer. Several organizations have advocated
decreasing the fat content of the diet as a means of preventing the development
of cancer. For example, the American Cancer Society 1 advises people to "cut
down on total fat intake'', and the American Institute for Cancer Research 2
advises people to "reduce the intake of total dietary fat from the current
average of approximately 37% to a level of no more than 30% of total calories
and, in particular, reduce the intake of saturated fat."
In this chapter, the role of dietary fat on the development of human and
experimental cancer will be discussed. The role of gene expression as a
mechanism by which dietary fat may alter the development of cancer will be
examined. Cancer clearly is a disease of alterations both in genetic structure
and in genetic expression, both of which can be affected by dietary fat.
II. MECHANISMS OF CARCINOGENESIS
A. CHEMICAL CARCINOGENESIS
At present, epidemiologists believe that 60 to 90% of human cancers are
caused by environmental agents. 3 Worldwide, the incidence of cancer in
different tissues and organs varies widely from country to country. 4 When
people migrate from one country to another, their descendants tend to develop
the same types of cancers as the native population.
The main environmental agents which are thought to cause cancer are
viruses, radiation, and chemicals. 5 The induction of cancer by chemicals was
first documented in 1775 by Percival Pott, who found that men who worked
as chimney sweeps had a high incidence of scrotal cancer. 6 Many other human
carcinogens have since been identified, including azo dyes, benzene, vinyl
chloride, and cigarette smoke. 3 In animals, chemical carcinogenesis was first
observed in the early 1900s, when the application of coal tar to skin was
found to induce tumors. 7•8 Subsequently, chemicals isolated from coal tar,
the polycyclic aromatic hydrocarbons (PAH), were found to induce skin tu-
mors in mice. 9 • 10 Several other classes of chemicals have since been found
to induce tumors, including aromatic amines, dialkylnitrosamines, and al-
kylating agents. 3
In order to induce cancer, most chemical carcinogens have to be metab-
olized to some other form. The procarcinogens are generally metabolized by
mixed-function oxidases in the smooth endoplasmic reticulum to electrophilic
forms (the ultimate carcinogens). 3 These electrophiles then react with nu-
cleophilic sites in the cell, such as the nitrogen and oxygen atoms in DNA.
The binding of an ultimate carcinogen to DNA can lead to mispairing during
DNA replication and thus to the formation of mutations in DNA. Chemical
carcinogens or their reactive metabolites have been shown to be mutagenic
in bacterial and mammalian mutagenesis systems. 11 • 12
Glauert 249
B. MULTISTAGE CARCINOGENESIS
After the administration of a chemical carcinogen, there is a latency period
before the actual appearance of a neoplasm. The latency period may vary
depending on the carcinogen used, its dosage, and the tissue or organ being
investigated, but it may be considered to be a general feature of the patho-
genesis of neoplasia. 13 Concurrently with early investigations into the latency
phenomenon, Rous and Kidd 14 provided experimental evidence suggesting a
two-stage mechanism for carcinogenesis in skin. They found that coal tar
application could "initiate" tumor formation while wounding could "pro-
mote" this process. Subsequently, Berenblum and Shubik 15 clearly demon-
strated that skin carcinogenesis could be divided into an initiation stage fol-
lowed by a longer promotion stage. These authors and many others have
demonstrated that administering a low dose of an initiating agent (which may
be a chemical, UV light, or ionizing radiation) to mouse skin normally results
in few or no tumors being formed during the animal's lifetime. Subsequent
repeated application of a promoting agent (such as croton oil or its most potent
component, 12-0-tetradecanoylphorbol-13-acetate [TPA]) over several weeks
or months will, however, induce the formation of skin papillomas and car-
cinomas.
Subsequent studies have demonstrated that carcinogenesis in other tis-
sues-such as the liver, mammary gland, bladder, and pancreas-can also
be divided into the two stages of initiation and promotion. 5 Initiation and
promotion, regardless of the tissue in question, have certain common prop-
erties. In two-stage carcinogenesis, the administration of a single subcarcin-
ogenic dose of a carcinogen (i.e., initiation) is followed by the long-term
administration of a promoter. 16 Initiation is not reversible, whereas promotion
is. Neither the administration of an initiator alone, repetitive doses of the
promoter alone, nor the administration of the promoter before the initiator
will lead to tumor production: the application of the initiator must be followed
by repeated doses of the promoter. 16 Mechanistically, an initiator has been
defined as an agent which irreversibly alters the native molecular structure
of DNA (i.e., produces a mutation), whereas a promoter has been defined
as an agent that alters the expression of genetic information in the cell. 13
An additional model for multistage carcinogenesis has been proposed by
Moolgavkar and colleagues. 17 In this model, two mutations must occur in
order for a cell to become neoplastic. The rate at which a tumor appears can
be influenced by two factors: an increase in the mutation rate, which could
be brought about by a genotoxic carcinogen, or an increase in the rate of cell
proliferation, which would expand the population of normal and intermediate
cells (i.e., cells with one mutation), and increase the likelihood of a mutation
occurring. A tumor-promoting agent could bring about this second mecha-
msm.
C. GENES IMPORTANT IN CARCINOGENESIS

The identity of the genes which are mutated by chemical carcinogens or
radiation is of great importance in understanding the carcinogenic process.
Dietary fat likely influences carcinogenesis by bringing about an alteration
in the expression of one or more of these genes. Cellular oncogenes or tumor
suppressor genes are the most probable targets whose mutation would lead
to the development of tumors. Cellular oncogenes, or proto-oncogenes, are
the cellular analogs of viral oncogenes, which have the ability to transform
cells and produce tumors. Over 50 viral oncogenes have been discovered.
When mutations are produced in specific locations in cellular oncogenes, they
gain the ability to transform cells. Cellular oncogenes containing mutations
in specific codons have been demonstrated in several model systems. 18 In
mouse skin, papillomas and carcinomas induced by initiation/promotion pro-
tocols contain activated ras oncogenes. Activated ras oncogenes are also seen
in mouse liver tumors induced by several chemicals.
Another probable target is one or more of the tumor suppressor genes.
These genes are normally expressed in nontransformed cells and appear to
play a role in chromosome stability or in the regulation of cellular differen-
tiation or cellular proliferation. 19 Certain types of genetic human tumors, such
as retinoblastoma, involve the loss of specific genes. The loss of tumor
suppressor genes may involve point mutation in the specific gene or deletion
of part of the chromosome containing the specific gene. Mutations and allelic
losses of the p53 tumor suppressor gene have been observed in several human
tumors. 20- 22
III. EPIDEMIOLOGICAL STUDIES

Epidemiological studies do not provide definitive conclusions as to the
effect of dietary fat on human cancer. Prospective studies are few in number
and are not in agreement. Several epidemiological studies have examined the
relationship between the intake of dietary fat or meat and the development
of colon cancer. 23 ·24 Several correlational and case-control epidemiological
studies have found a positive association between dietary fat and colon cancer,
but others have not. Prospective epidemiological studies, however, have found
no effect in three studies, a positive association in one study, and an actual
protective effect of high fat intakes in two studies. 23 ·24 The interpretation of
these studies may be complicated by the total energy intake; 23 •25 •26 energy
intake has been correlated to colon cancer risk in several correlational and
case-control studies. 23 ·26
Numerous epidemiological studies have attempted to identify factors which
influence breast cancer risk in humans. Established breast cancer risk factors
include age of first pregnancy, body build, age at menarche or menopause,
and the amount of radiation received by the chest. 27 The effect of dietary fat
has been studied in correlational, case-control, and prospective epidemiological
Glauert 251
studies. Studies examining international correlations between dietary fat intake

and breast cancer risk and migrant studies have reported a positive association
between dietary fat intake and breast cancer risk. 27 Case-control studies gen-
erally have also supported a connection between total fat intake and breast
cancer risk. 28 Recently, a combined analysis of 12 case-control studies found
a highly significant positive association between breast cancer risk and sat-
urated fat intake. 29 Two prospective studies, however, did not find any link
between dietary fat intake and the development of breast cancer, 30 ·3 ' whereas
another found a positive correlation between fat intake and the development
of breast cancer. 32
In the pancreas, dietary fat consumption has been found to be associated
with an increased human cancer risk in the few correlational and case-control
epidemiological studies that have been published. 33 . 35 In correlational epi-
demiological studies, high dietary fat intakes are also associated with a higher
incidence of ovarian and prostatic cancer. 36
IV. EXPERIMENTAL CARCINOGENESIS STUDIES
A. NUTRITIONAL CONSIDERATIONS
In experiments examining the effect of dietary fat, fat has to be substituted
for carbohydrate isocalorically in order to ensure that fat is the only experi-
mental variable. In some studies, fat was added to a nutritionally complete
diet, such as a ground unrefined diet. Since animals consume approximately
the same number of calories, animals on such a high fat diet would consume
less of all other nutrients in the diet. Therefore, any changes seen in the high
fat group could not strictly be attributed to a higher consumption of fat. In
other studies, fat was substituted for a carbohydrate (such as corn starch or
dextrose) on a gram-for-gram basis. Since fat contains 9 kcal per gram and
carbohydrate contains only 4 kcal per gram/ 7 such a protocol would also
result in lower consumption of other nutrients in the diet. In order to substitute
fat for carbohydrate on a metabolizable energy basis, 4 g of fat have to be
substituted for 9 g of carbohydrate.
Another issue in studies examining dietary fat and carcinogenesis is whether
the enhancing effect of dietary fat, when it is seen. is due to dietary fat per
se (for example, by an effect on bile acid metabolism in colon carcinogenesis)
or rather to a more efficient utilization of calories. Even when the ratio of
calories to nutrients remains the same when the dietary fat content of the diet
is raised, animals frequently gain more weight than animals on a low fat diet.
When fat is substituted for carbohydrate isocalorically, the substitution is done
on the basis of metabolizable energy rather than net energy. Donato and
Hegsted/ 8 however, have demonstrated that fat is used more efficiently than
carbohydrate, so that equating their metabolizable energy values may be
inaccurate.
B. SKIN CARCINOGENESIS
Mouse skin is one of the oldest and most widely used systems for studying
chemical carcinogenesis, including multistage carcinogenesis. As discussed
earlier, two-stage carcinogenesis (initiation-promotion) was first observed in
mouse skin and involves initiation by a subcarcinogenic dose of radiation or
of a chemical, such as PAH, followed by the long-term administration of
croton oil or its active ingredient TPA.
Most studies examining dietary fat have studied complete carcinogenesis
by PAH or ultraviolet light. Early studies demonstrated that high fat diets
enhanced skin carcinogenesis induced by tar39 or PAH. 40 -46 In studies where
skin tumors were induced by ultraviolet light, Mathews-Roth and Krinsky47
found that high fat diets increased skin carcinogenesis, whereas Black et a!. 48
found that high fat diets did not increase skin carcinogenesis, but that feeding
a saturated fat inhibited tumorigenesis.
The effect of fatty acids on the initiation and promotion of skin carcin-
ogenesis has also been studied. Certain fatty acids-oleic acid and lauric
acid-have promoting activity when applied daily to mouse skin after a single
application of 7, 12-dimethylbenz(a)anthracene (DMBA); stearic acid and pal-
mitic acid do not have any effect. 49 When diets varying in their fat content
are fed during the initiation or promotion stages of DMBA-initiated, TPA-
promoted skin carcinogenesis, high fat diets enhance the promotion of skin
carcinogenesis and slightly inhibit initiation. 50 •51
C. HEPATOCARCINOGENESIS
Many early studies of dietary fat and cancer used the liver as the target
organ. In these studies, aromatic amines and azo dyes were frequently used
to induce hepatocellular carcinomas. In later studies, effects of dietary fat on
initiation and promotion in the liver were examined. In initiation-promotion
protocols, the administration of a single subcarcinogenic dose of a carcinogen
(such as diethylnitrosamine [DEN] or DMBA) along with a proliferative
stimulus (such as partial hepatectomy) followed by the long-term feeding of
chemicals such as phenobarbital, 2,3,7,8-tetrachlorodibenzo-p-dioxin, or po-
lyhalogenated biphenyls, leads to a high incidence of hepatocellular adenomas
and carcinomas. 13 •52 In addition, foci of putative preneoplastic hepatocytes
appear before the development of gross tumors. These foci, known as altered
hepatic foci or enzyme-altered foci, contain cells which exhibit qualitatively
altered enzyme activities or alterations in one or more cell functions. 53 The
enzymes most frequently studied include -y-glutamyltranspeptidase (GGT) and
placental glutathione-S-transferase, which are normally not present in adult
liver, but which are often present in foci, and ATPase and glucose-6-phos-
phatase, which are normally present but which are frequently missing from
foci. 54 •55 Altered hepatic foci can also be identified on hematoxylin and eosin-
stained tissue. 56 •57 The appearance of foci has been correlated with the later
development of malignant neoplasms. 58 •59
Glauert 253
The first studies examined the effect of dietary fat on the induction of
hepatocellular carcinomas by complete hepatocarcinogens. In the liver, in-
creasing the fat content of the diet enhances the development of 2-acetylam-
inofluorene (AAF)-, p-dimethylaminoazobenzene (DAB)-, and aflatoxin B 1
(AFB)-induced tumors and GGT-positive foci in rats. 60 ·63 Furthermore, he-
patocarcinogenesis by DAB is enhanced by feeding a diet which contains a
greater proportion of polyunsaturated fatty acids. 64 •65 In these studies, how-
ever, the diets were administered at the same time as the carcinogen injections,
so that the stage of carcinogenesis which was affected could not be determined.
More recent studies have examined whether this enhancement of hepa-
tocarcinogenesis is caused by an effect on the initiation of carcinogenesis,
the promotion of carcinogenesis, or both. Misslbeck et al. 66 found that in-
creasing the com oil content of the diet after the administration of I 0 doses
of aflatoxin increased the number and size of GGT -positive foci, but Baldwin
and Parker, 67 using a similar protocol, found no effect of dietary com oil.
Glauert and Pitot68 similarly found that increasing the safflower oil or palm
oil content of the diet did not promote DEN-induced GGT-positive foci or
greatly affect phenobarbital promotion of GGT-positive foci. The promotion
of GGT-positive foci by dietary tryptophan also is not affected by dietary
fat. 69 Newbeme et al. 70 found that increasing dietary com oil (but not beef
fat) during and after the administration of AFB increased the incidence of
hepatic tumors, but not when the diets were fed only after AFB administration.
Baldwin and Parke~ 7
also found that increasing the com oil content of the
diet before and during AFB administration increased the number and volume
of GGT-positive foci. Recently, it has been found that when rats are fed diets
high in polyunsaturated fatty acids (but not in saturated fatty acids) before
receiving the hepatocarinogen DEN, they develop more GGT-positive and
ATPase-negative foci than rats fed low fat diets. 71 The results of these studies
suggest that the enhancement of hepatocarcinogenesis by dietary fat is pri-
marily due to an effect on initiation, and that polyunsaturated fats have a
greater effect than do saturated fats.
D. COLON CARCINOGENESIS
Studies in experimental animals have produced differing results. Because
no animal models develop colon tumors spontaneously, a variety of chemi-
cals have been used to induce colon tumors, usually in rats or mice. These
include 1,2-dimethylhydrazine (DMH) and its metabolites azoxymethane
(AOM) and methylazoxymethanol (MAM); 3,2 '-dimethyl-4-aminobiphenyl
(DMAB); methylnitrosourea (MNU); and N-methyl-N' -nitro-N-nitrosoguan-
idine (MNNG). 72 "75 DMH and AOM have been used most frequently to study
nutritional effects. Both can induce colon tumors by single76. 80 or multi-
ple81.85 injections.
Animal studies examining the effect of dietary fat have used a variety of
protocols, and the results obtained often have been dependent on the inves-
tigator's protocol. In several studies, fat was either added to a chow diet or
was substituted for carbohydrate on a weight basis, so that the ratio of calories
to essential nutrients was altered. 83 ·84 •8691 These studies all found that in-
creasing dietary fat enhanced colon carcinogenesis, but the effect could have
been due to a lower consumption of essential nutrients rather than to an effect
of fat. In other studies, rats or mice were subjected to multiple doses of a
colon carcinogen, with the dietary fat content being varied (isocalorically)
during, and frequently before or after, the carcinogen injections. Some of
these studies found an enhancement when the dietary fat content of the diet
was increased, but most saw no effect and one saw an inhibition of tumor
development. 73 ·75 •81 •82 •92 -95 Feeding unsaturated rather than saturated fat in the
diet also was found to increase colon carcinogenesis induced by multiple
injections of AOM. 96 More recent studies have examined the effect of the
amount and type of dietary fat on the initiation and promotion of colon
carcinogenesis induced by one, two, or three injections of AOM. Reddy and
colleagues found that increasing the unsaturated fat, but not the saturated fat,
content of the diet during the promotional phase enhanced colon carcinogen-
esis; the initiation of colon carcinogenesis by AOM, however, was not affected
by the level of unsaturated fat, but was slightly enhanced by increased levels
of saturated fat. 97 -99 Other investigators, however, did not see an enhancement
of the promotional phase by dietary polyunsaturated fat. 100 •101
E. PANCREATIC CARCINOGENESIS
Dietary fat has been studied extensively in animal models, primarily rats
and hamsters. In rats, azaserine is the chemical carcinogen usually used. 102
Single or multiple doses of this chemical induce carcinomas derived from
acinar cells. Putative preneoplastic lesions-basophilic and acidophilic foci-
have also been described in the rat pancreas. 102 The carcinogen N-nitrosobis(2-
oxopropyl)arnine (BOP) has been used primarily to induce pancreatic tumors
in hamsters, although it also has been shown to be effective in rats. 102 In
hamsters, tumors are derived from ductal cells, as are the majority of human
pancreatic tumors. The hamster model, therefore, may be more relevant to
human pancreatic cancer.
Dietary fat has been found to influence tumorigenesis in both animal
models. In rats, feeding high-fat diets after, or during and after, the injection
of azaserine enhances the development of pancreatic tumors and putative
preneoplastic lesions. 103 -' 11 Pancreatic carcinogenesis induced by N-nitroso(2-
hydroxypropyl) (2-oxopropyl)amine in rats is also enhanced by feeding high
fat diets. 112 In several of these studies, the effect cannot be attributed un-
equivocally to dietary fat because fat was substituted for carbohydrate on a
weight basis. In hamsters, BOP-induced pancreatic carcinogenesis is also
increased by feeding high-fat diets. 108- 110 •113- 116 Roebuck and colleagues 104 •105 •111
found that polyunsaturated fat, but not saturated fat, enhanced pancreatic
carcinogenesis, and that a certain level of essential fatty acids are required
Glauert 255
for the enhancement of pancreatic carcinogenesis. Birt et al., 116 however,

found that feeding a saturated fat (beef tallow) enhanced pancreatic carcin-
ogenesis in hamsters greater than a polyunsaturated fat (com oil).
Finally, it has been observed in 2-year carcinogenesis studies in which
com oil gavage has been used as the vehicle for the carcinogen that a higher
incidence of pancreatic acinar cell adenomas is present in com oil gavage-
treated male Fischer-344 control rats than in untreated controls. 117• 118 This
association was not observed in female rats or in male or female B6C3F 1
mice. The mechanism of this effect is not known.
F. MAMMARY CARCINOGENESIS
The effect of dietary fat on mammary carcinogenesis in experimental
animals has been examined extensively: over 100 experiments have been
conducted. 119 Two animal models are primarily used: a rat model (usually the
Sprague-Dawley strain) in which mammary tumors are induced by DMBA
or MNU and a mouse model, in which tumors develop spontaneously. The
use of these models is advantageous because tumor latency, tumor size, and
tumor progression can easily be quantified by palpation of mammary tumors
as they appear. In both models, increasing the fat content of the diet clearly
enhances the development of mammary tumors. 119 In the rat model, a high
fat diet increases tumorigenesis both when it is fed during and after carcinogen
administration, and when it is fed only after carcinogen injection.
G. OTHER SITES
Dietary fat has also been studied for its effect on experimental carcino-
genesis in other organs. In the lung, dietary fat enhanced BP- or BOP-induced
carcinogenesis in hamsters, 114 • 120 whereas in mice, a high fat diet did not
affect spontaneous carcinogenesis in one study, but the feeding of egg extracts
enhanced it in another. 44 • 121 Spontaneous brain tumors in mice and prostate
tumors in rats were increased by the feeding of high fat diets, as was the
induction of BOP-induced renal tumors, but DMH- or MAM-induced renal
or ear duct tumors were not affected by dietary fat. 87• 114 • 122 • 123
H. EFFECT OF n-3 FATTY ACIDS

Feeding diets high in n-3 fatty acids generally inhibits chemical carci-
nogenesis; in most of these studies, oils high in n-3 fatty acids are substituted
for com oil. 124 In the pancreas, feeding diets high in menhaden oil decreases
the development of azaserine-induced preneoplastic lesions. 125 • 126 In the colon,
feeding menhaden oil in place of com oil, or eicosapentaenoic acid in place
of linoleic acid, decreases the development of DMH- or AOM-induced colon
tumors, but adding menhaden oil to a low-fat diet does not affect colon
carcinogenesis. 99 • 127. 129 In the mammary gland, feeding a diet high in men-
haden oil instead of a high com oil diet decreases the incidence of DMBA-
induced tumors in rats. 130 • 131 Locniskar et al., 132 however, found that substi-
tuting menhaden oil for com oil or coconut oil did not affect skin tumor
promotion by TPA.
V. MECHANISMS BY WHICH DIETARY FAT MAY

INFLUENCE CARCINOGENESIS
A. MEMBRANE FLUIDITY
An important function of dietary fatty acids is their presence in membrane
lipids. Altering the fatty acid content of the diet alters the composition of
membrane lipids, particularly in certain tissues; feeding diets high in n-6 or
n-3 fatty acids increases the concentrations of these fatty acids in membrane
lipids. 133 The activities of membrane-bound enzymes are increased in mem-
branes that are more fluid, i.e., that have a higher content of polyunsaturated
fatty acids. 133 Higher amounts of unsaturated fatty acids in cell membranes
are also associated with a higher rate of cell proliferation. 134 The alteration
by dietary fatty acids of the catalytic abilities of membrane-bound enzymes,
such as cytochrome P-450, may play an important role in carcinogenesis.
B. TOXICITY
One possible mechanism by which dietary fat may enhance carcinogenesis
is by the toxicity of fatty acids or of metabolites that increase after the feeding
of high fat diets. Such toxicity would bring about a proliferative response in
the tissue to replace lost cells. Cellular genes involved in cell proliferation,
including cellular oncogenes, would likely be increased.
In the colon, toxicity may play a role in the enhancement of carcinogenesis
by dietary fat. One hypothesis for the effect of dietary fat is that dietary fat
increases the concentration of metabolites with carcinogenic or promoting
activity in the fecal stream. Bile acids, particularly secondary bile acids, have
promoting activity in the colon; 135 - 137 their concentration in the feces has been
found to be increased by dietary fat in some but not all studies. 83 • 138- 141 Bile
acids function as detergents; therefore, high concentrations may be toxic to
epithelial cells in the colon. Increasing the concentration of bile acids as well
as fatty acids in the colon has been demonstrated to increase colon epithelial
cell proliferation. 142 - 146 It has been further demonstrated that the increased
expression of protein kinase C may play a role in bile acid-induced cell
proliferation. 147
C. EICOSANOID METABOLISM
Another mechanism by which dietary fat may influence carcinogenesis
is by altering the synthesis of eicosanoids. Fatty acids that are consumed in
the diet can be metabolized to a variety of other compounds, including longer
and more unsaturated fatty acids, prostaglandins, leukotrienes, thromboxanes,
hydroperoxyeicosatetraenoic acids, and hydroxyeicosatetraenoic acids. 148 Al-
tering the type of fatty acid in the diet changes the composition of the ei-
Glauert 257
cosanoids that are produced by the body. 149 For example, increasing the n-3
content of the diet will alter the type of eicosanoids that are produced as well
as their function. 150 Specific eicosanoids bind to receptors and cause specific
alterations in gene expression and cellular function, 151 some of which may
be related to carcinogenesis. It has been found that inhibition of eicosanoid
synthesis inhibits DNA synthesis and tumor promotion in both the skin and
liver. 152-154
D. CALORIC EFFECTS
The issue of whether the enhancing effect of fat in carcinogenesis IS
due to higher consumption of calories or more efficient utilization of energy
has been examined in several tissues. The earliest study was conducted by
Boutwell et a!. 46 using the mouse skin carcinogenesis system; they attributed
most of the enhancing effect of dietary fat to an increased consumption of
calories. Birt eta!., 51 however, found that the promotion of skin carcinogenesis
was enhanced even though the high-fat diets were pair-fed. The greater caloric
density of fat has also been proposed to play a role in colon tumorigenesis.
Caloric restriction inhibits chemically induced colon carcinogenesis, even if
the percentage of dietary fat in the diet is greatly increased. 155 • 156 In the
pancreas, the enhancement by dietary fat appears to be an effect of dietary
fat rather than of an increased consumption of calories, as pair-feeding does
not inhibit the enhancing effect of dietary fat in hamsters. 115 Several studies
have suggested that the enhancement of mammary carcinogenesis by dietary
fat may be caused, at least in part, by an alteration in the efficiency of energy
utilization. 157 - 159 Recently, Freedman et a!., 119 using a combined statistical
analysis of over 100 animal experiments, concluded that the enhancing effect
of high-fat diets was caused in part by a higher caloric intake and in part by
some other effect of dietary fat. Finally, caloric restriction has been found to
inhibit tumorigenesis in many tissues in experimental animals. 160
Several mechanisms have been proposed to explain how caloric con-
sumption affects carcinogenesis. These include effects on growth factors,
immunity, neuroendocrine function, and metabolic regulation.' 60 • 161 Any of
these effects could be brought about by alterations in gene expression.
E. EFFECT ON INITIATION
Dietary fat may also affect the initiation stage of carcinogenesis. Since
initiation involves the mutation of DNA, its alteration (by dietary fat or other
agents) would mainly affect the structure of genes rather than their expression.
In the liver, dietary fat appears to enhance carcinogenesis primarily by an
effect on initiation. In other tissues, many protocols have varied the levels
of dietary fat during the time of carcinogen injections; therefore dietary fat
may be affecting some aspect of initiation in these studies. Higher levels of
dietary fat may enhance initiation of carcinogenesis by several mechanisms,
including alterations in absorption of the carcinogen from the gut, transport
to the target organ, uptake by the target organ, metabolism by cytochrome

P-450 or other drug-metabolizing enzymes to a form which can react with
DNA, and DNA repair. Any of these processes could be altered by changes
in the expression of the appropriate gene. But dietary fat could affect these
processes by mechanisms independent of gene expression. Several of these
processes occur in membranes, whose lipid composition can be altered by
changing the amount or type of dietary fat. 162 • 163 Increasing the fat content
of the diet increases cytochrome P-450 and related activities. 164- 167 Western
blot analysis has indicated that higher amounts of enzyme protein are present
after feeding diets high in polyunsaturated fat. 168 Therefore dietary fat may
be affecting both gene expression and the surrounding matrix necessary for
optimum enzyme activity. The metabolism of several chemicals, including
hexobarbital, aniline, ethylmorphine, benzo[a]pyrene (BP), and dimethylni-
trosamine, is also enhanced by feeding diets high in polyunsaturated fatty
acids. 164.165.169-171
F. INDUCTION OF SPECIFIC GENES

Altering the level of dietary fat changes the expression of many genes.
Most of the genes studied, however, are related to carbohydrate or lipid
metabolism and are not likely to play a role in carcinogenesis. m-m As men-
tioned above, the expression of cytochrome P-450 genes is influenced by the
level of dietary fat. 164- 168
Very little work has been published on the effect of dietary fat on the
expression of genes which are likely to be important in carcinogenesis: on-
cogenes and tumor suppressor genes. In the colon, dietary fat did not affect
the expression of either c-myc or c-H-ras oncogenes in normal or tumor
tissue. 101 In the mammary gland, substituting 1-3 fatty acid-rich oils for com
oil decreased the expression of the H-ras oncogene in one study but had no
effect in the other. 176 • 177
Another approach has been to examine for genes whose expression is
altered greatly when the level of dietary fat is changed. Using subtractive
hybridization, Elliott et a!. 178 isolated a clone that is expressed fivefold higher
in the mammary gland of mice fed low fat diets. This gene was found to
have over 80% homology to a gene induced by UV irradiation (LFM 1). 179
The relationship of this gene to tumorigenesis is currently unknown.
VI. SUMMARY AND CONCLUSIONS
Clearly there is much variability in studies of dietary fat and cancer, both
in epidemiological and experimental studies. In epidemiological studies, a
relationship between dietary fat and colon, pancreatic, and mammary cancer
has been found in correlational and case-control studies. Few prospective
studies, however, have found an enhancement of these cancers by dietary fat,
and, for colon cancer, two studies have found that a high-fat diet is protective.
Glauert 259
In experimental studies, dietary fat generally enhances chemically induced

skin, liver, pancreatic, and mammary carcinogenesis, whereas conflicting
results have been seen in colon carcinogenesis. Dietary fat appears to act
primarily during the promotional stage of carcinogenesis in all of these models
except the liver, where the effect of dietary fat is primarily on initiation.
Because of the variability seen in studies of dietary fat and cancer (particularly
prospective epidemiological studies), it cannot be stated unequivocally at this
time that human cancer can be prevented by decreasing the fat content of the
diet.
The mechanisms by which high-fat diets enhance experimental carcino-
genesis are unclear, but probably involve both a higher intake of net energy
and other effects, some of which are organ-specific. Nearly all of the mech-
anisms by which dietary fat may influence carcinogenesis involve alterations
in genetic expression. Because of the variability that is seen in studies of
dietary fat and cancer, it is critical that basic scientific studies be conducted
to elucidate the mechanisms by which dietary fat influences carcinogenesis.
At present, few studies have been conducted which have examined the effect
of dietary fat on genes which are important in the carcinogenic process. The
determination of genes which are turned on or off by dietary fat will likely
provide answers as to the role of dietary fat in carcinogenesis.
REFERENCES
I. American Cancer Society, Cancer Facts and Figures 1991, American Cancer Society,
Atlanta, 1991.
2. American Institute for Cancer Research, Dietary Guidelines to Lower Cancer Risk, Amer-
ican Institute for Cancer Research, Washington, D.C., 1990.
3. Miller, E. C., Some current perspectives on chemical carcinogenesis in humans and
experimental animals, Cancer Res., 38, 1479, 1978.
4. Armstrong, B. and Doll, R., Environmental factors and cancer incidence and mortality
in different countries with special references to dietary practices, Br. J. Cancer, 15,
617, 1975.
5. Pitot, H. C., Fundamentals of Oncology, 3rd ed., Marcel Dekker, New York, 1986.
6. Pott, P., Chirurgical observations relative to cancer of the scrotum, Reprinted in Nat/.
Cancer lnst. Monogr., 10, 7, 1963.
7. Yamagiwa, K. and Ichikawa, K., Experimental study of the pathogenesis of carcinoma,
J. Cancer Res., 3, I, 1918.
8. Tsutsui, H., Ober das kunstlich erzeugte Cancroid bei der Maus, Gann, 12, 17, 1918.
9. Cook, J. W., Hewett, C. L., and Heiger, I., The isolation of cancer-producing hydro-
carbon from coal tar, J. Chem. Soc., 395.
10. Kennaway, E. L. and Hieger, I., Carcinogenic substances and their fluorescence spectra,
Br. Med. J., ii, 1044, 1930.
II. McCann, J., Choi, E., Yamasaki, E., and Ames, B. N., The detection of carcinogens
as mutagens in the Salmonella/microsome test: assay of 300 chemicals, Proc. Nat/. Acad.
Sci. U.S.A., 72, 5135, 1975.
12. DeMarini, D. M., Brockman, H. E., de Serres, F. J., Evans, H. H., Stankowski,
C. F., and Hsie, A. W., Specific-locus mutations induced in eukaryotes (especially
mammalian cells) by radiation and chemicals: a perspective, Murat. Res., 220, I I, 1989.
13. Pitot, H. C. and Sirica, A. E., The stages of initiation and promotion in hepatocarcin-
ogenesis, Biochem. Biophys. Acta, 605, 191, 1980.
14. Rous, P. and Kidd, J, G., Conditional neoplasms and subthreshold neoplastic states,
J. Exp. Med., 73, 365, 1941.
15. Berenblum, I. and Shubik, P., The role of croton oil applications, associated with a
single painting of a carcinogen, in tumour induction of the mouse's skin, Br. J. Cancer,
I, 379, 1947.
16. Diamond, L., O'Brien, T. G., and Baird, W. M., Tumor promoters and the mechanism
of tumor promotion, Adv. Cancer Res., 32, I, 1980.
17. Moolgavkar, S. H. and Knudson, A. G., Mutation and cancer: a model for human
carcinogenesis, J. Nat/. Cancer lnst., 66, 1037, 1981.
18. Stowers, S. J., Maronpot, R. R., Reynolds, S. H., and Anderson, M. W., The role
of oncogenes in chemical carcinogenesis, Environ. Health Perspect., 75, 81, 1987.
19. Sager, R., Tumor suppressor genes: the puzzle and the promise, Science, 246, 1406,
1989.
20. Smith, M. L., Yeleswarapu, L., Locker, J., and Lombardi, B., p53 mutation(s) in
diethylnitrosamine-induced foci of putative preneoplastic hepatocytes in male Fischer 344
rats, Proc. Am. Assoc. Cancer Res., 32, 136, 1991.
21. Hsu, I. C., Metcalf, R. A., Sun, T., Welsh, J, A., Wang, N.J., and Harris, C. C.,
Mutational hotspot in the p53 gene in human hepatocellular carcinomas, Nature (London),
350, 427, 1991.
22. Okamoto, A., Sameshima, Y., Yokoyama, S., Terashima, Y., Sugimura, T., Terada,
M., and Yokota, J,, Frequent allelic losses and mutations of the p53 gene in human
ovarian cancer, Cancer Res., 51, 5171, 1991.
23. Kolonel, L. N., Fat and colon cancer: how firm is the epidemiologic evidence?, Am. J.
Clin. Nutr., 45, 336, 1987.
24. Willett, W. C., Stampfer, M. J., Colditz, G. A., Rosner, B. A., and Speizer, F. E.,
Relation of meat, fat, and fiber intake to the risk of colon cancer in a prospective study
among women, New Engl. J. Med., 323, 1664, 1990.
25. Willett, W. C., Implications of total energy intake for epidemiologic studies of breast
and large-bowel cancer, Am. J. Clin. Nutr., 45, 354, 1987.
26. Lyon, J, L., Mahoney, A. W., West, D. W., Gardner, J, W., Smith, K. R., Sorenson,
A. W., and Stanish, W., Energy intake: its relationship to colon cancer risk, J. Nat/.
Cancer lnsl., 78, 853, 1987.
27. Kelsey, J, L. and Berkowitz, G. S., Breast cancer epidemiology, Cancer Res., 48,
5615, 1988.
28. Prentice, R. L., Pepe, M., and Self, S. G., Dietary fat and breast cancer: a quantitative
assessment of the epidemiological literature and a discussion of methodological issues,
Cancer Res., 49, 3147, 1989.
29. Howe, G. R., Hirohata, T., Hislop, T. G., Iscovich, J, M., Yuan, J. M., Katsouyanni,
K., Lubin, F., Marubini, E., Modan, B., Rohan, T., Toniolo, P., and Shunzhang,
Y., Dietary factors and risk of breast cancer: combined analysis of 12 case-control studies,
J. Nat/. Cancer lnsl., 82, 561, 1990.
30. Willett, W. C., Stampfer, M. J., Colditz, G. A., Rosner, B. A., Hennekens, C. H.,
and Speizer, F. E., Dietary fat and the risk of breast cancer, New Engl. J. Med., 316,
22, 1987.
31. Jones, D. Y., Schatzbin, A., Green, S. B., Block, G., Brinton, L.A., Ziegler, R. G.,
Hoover, R., and Taylor, P. R., Dietary fat and breast cancer in the National Health
and Nutrition Examination Survey I. Epidemiological followup study, J. Nat/. Cancer
lnst., 79, 465, 1987.
Glauert 261
32. Howe, G. R., Friedenreich, C. M., Jain, M., and Miller, A. B., A cohort study of
fat intake and risk of breast cancer, J. Nat/. Cancer Inst., 83, 336, 1991.
33. Wynder, E. L., An epidemiological evaluation of the causes of cancer of the pancreas,
Cancer Res., 35, 2228, 1975.
34. Durbec, J, P., Chevilotte, G., Bidart, J, M., Berthezene, P., and Sarles, H., Diet,
alcohol, tobacco and risk of cancer of the pancreas: a case-control study, Br. J. Cancer,
47, 463, 1983.
35. Norell, S. E., Ahlborn, A., Erwald, R., Jacobson, G., Lindberg-Navier, 1., Okin,
R., Tiirnberg, B., and Wiechel, K. L., Diet and pancreatic cancer: a case-control study,
Am. J. Epidemiol., 124, 894, 1986.
36. Carroll, K. K., Lipids and carcinogenesis, J. Environ. Patho/. Toxicol., 3, 253, 1980.
37. National Research Council, Recommended Dietary Allowances, lOth edition, National
Academy Press, Washington, D.C., 1989.
38. Donato, K. and Hegsted, D. M., Efficiency of utilization of various sources of energy
for growth, Proc. Nat/. Acad. Sci. U.S.A., 82, 4866, 1985.
39. Watson, A. F. and Mellanby, E., Tar cancer in mice. II. The condition of the skin
when modified by external treatment or diet, as a factor in influencing the cancerous
reaction, Br. J. Exp. Pathol., II, 311, 1930.
40. Baumann, C. A., Jacobi, H. P., and Rusch, H. P., The effect of diet on experimental
tumor production, Am. J. Hygiene, 30A, 1, 1939.
41. Jacobi, H. P. and Baumann, C. A., The effect of fat on tumor formation, Am. J.
Cancer, 39, 338, 1940.
42. Lavik, P. S. and Baumann, C. A., Dietary fat and tumor formation, Cancer Res., 1,
181, 1941.
43. Lavik, P. S. and Baumann, C. A., Further studies on the tumor-promoting action of
fat, Cancer Res., 3, 749, 1943.
44. Tannenbaum, A., The genesis and growth of tumors. III. Effects of a high fat diet,
Cancer Res., 2, 468, 1942.
45. Tannenbaum, A., The dependence of the genesis of induced skin tumors on the fat
content of the diet during different stages of carcinogenesis, Cancer Res., 4, 683, 1944.
46. Boutwell, R. K., Brush, M. K., and Rusch, H. P., The stimulating effect of dietary
fat on carcinogenesis, Cancer Res., 9, 741, 1949.
47. Mathews-Roth, M. M. and Krinsky, N. I., Effect of dietary fat level on UV-B induced
skin tumors, and anti-tumor action of beta-carotene, Photochem. Photobiol., 40, 671,
1984.
48. Black, H. S., Lenger, W., Phelps, A. W., and Thornby, J, 1., Influence of dietary
lipid upon ultraviolet-light carcinogenesis, Nutr. Cancer, 5, 59, 1983.
49. Holsti, P., Tumor promoting effects of some long chain fatty acids in experimental skin
carcinogenesis in the mouse, Acta Pathol. Microbiol. Scand., 46, 51, 1959.
50. Birt, D. F., Pelling, J. C., Tibbets, M. G., and Schweickert, L., Acceleration of
papilloma growth in mice fed high-fat diets during promotion of two-stage skin carcin-
ogenesis, Nutr. Cancer, 12, 161, 1989.
51. Birt, D. F., White, L. T., Choi, B., and Pelling, J, C., Dietary fat effects on the
initiation and promotion of two-stage skin tumorigenesis in the SEN CAR mouse, Cancer
Res., 49, 4170, 1989.
52. Silberhorn, E. M., Glauert, H. P., and Robertson, L. W., Carcinogenicity of poly-
halogenated biphenyls: PCBs and PBBs, Crit. Rev. Toxicol., 20, 439, 1990.
53. Beer, D. G. and Pitot, H. C., Biological markers characterizing the development of
preneoplastic and neoplastic lesions in rodent liver, Arch. Toxicol., 10 (Suppl.), 68, 1987.
54. Pitot, H. C., Glauert, H. P., and Hanigan, M., The significance of biochemical markers
in fhe characterization of putative initiated cell populations in rodent liver, Cancer Lett.,
29, I, 1985.
55. Hendrich, S., Campbell, H. A., and Pitot, H. C., Quantitative stereological evaluation
of four histochemical markers of altered foci in multistage hepatocarcinogenesis in the
rat, Carcinogenesis, 8, 1245, 1987.
56. Harada, T., Maronpot, R. R., Morris, R. W., Stitzel, K. A., and Boorman, G. A.,
Morphological and stereological characterization of hepatic foci of cellular alteration in
control Fischer 344 rats, Toxicol. Pathol., 17, 579, 1989.
57. Bannasch, P., Enzmann, H., Klimek, F., Weber, E., and Zerban, H., Significance
of sequential cellular changes inside and outside foci of altered hepatocytes during he-
patocarcinogenesis, Toxicol. Pathol., 17, 617, 1989.
58. Kunz, H. W., Tennekes, H. A., Port, R. E., Schwarz, M., Lorke, D., and Schaude,
G., Quantitative aspects of chemical carcinogenesis and tumor promotion in liver, En-
viron. Health Perspect., SO, 113, 1983.
59. Emmelot, P. and Scherer, E., The first relevant cell stage in rat liver carcinogenesis.
A quantitative approach, Biochim. Biophys. Acta, 605, 247, 1980.
60. Kline, B. E., Miller, J, A., Rusch, H. P., and Baumann, C. A., Certain effects of
dietary fats on the production of liver tumors in rats fed p-dimethylaminoazobenzene,
Cancer Res., 6, 5, 1946.
61. Sugai, M., Witting, L.A., Tsuchiyama, H., and Kummerow, F. A., The effect of
heated fat on the carcinogenic activity of 2-acetylaminofluorene, Cancer Res., 22, 510,
1962.
62. McCay, P. B., King, M., Rikans, L., and Pitha, J, V., Interactions between dietary
fats and antioxidants on DMBA-induced hyperplastic nodules and hepatomas, J. Environ.
Pathol. Toxicol., 3, 451, 1980.
63. Baldwin, S. and Parker, R. S., The effect of dietary fat and selenium on the development
of preneoplastic lesions in rat liver, Nutr. Cancer, 8, 273, 1986.
64. Miller, J, A., Kline, B. E., Rusch, H. P., and Baumann, C. A., The carcinogenicity
of p-dimethylaminoazobenzene in diets containing hydrogenated coconut oil, Cancer Res.,
4, 153, 1944.
65. Miller, J, A., Kline, B. E., Rusch, H. P., and Baumann, C. A., The effect of certain
lipids on the carcinogenicity of p-dimethylaminoazobenzene, Cancer Res., 4, 756, 1944.
66. Misslbeck, N. G., Campbell, T. C., and Roe, D. A., Effect of ethanol consumed in
combination with high or low fat diets on the post initiation phase of hepatocarcinogenesis,
J. Nutr., 114, 2311, 1984.
67. Baldwin, S. and Parker, R. S., Influence of dietary fat and selenium in initiation and
promotion of aflatoxin B,-induced preneoplastic foci in rat liver, Carcinogenesis, 8, 101,
1987.
68. Glauert, H. P. and Pitot, H. C., Effect of dietary fat on the promotion of diethylni-
trosamine-induced hepatocarcinogenesis in female rats, Proc. Soc. Exp. Bioi. Med., 181,
498, 1986.
69. Sidransky, H., Verney, E., and Wang, D., Effects of varying fat content of a high
tryptophan diet on induction of gamma-glutamyltranspeptidase positive foci in the livers
of rats treated with hepatocarcinogen, Cancer Lett., 31, 235, 1986.
70. Newberne, P.M., Weigert, J,, and Kula, N., Effects of dietary fat on hepatic mixed-
functions oxidases and hepatocellular carcinoma induced by aflatoxin B I in rats, Cancer
Res., 39, 3986, 1979.
71. Glauert, H. P., Lay, L. T., Kennan, W. S., and Pitot, H. C., Effect of dietary fat
on the initiation of hepatocarcinogenesis by diethylnitrosamine or 2-acetylaminofluorene
in rats, Carcinogenesis, 12, 991, 1991.
72. Druckrey, H., Production of colonic carcinomas by 1,2-dialkyhydrazines and azoalkanes,
in Carcinoma of the Colon and Antecedent Epithelium, Burdette, W. J., Ed., Springfield,
IL, Thomas, 1970, 267.
73. Reddy, B. S. and Ohmori, T., Effect of intestinal microflora and dietary fat on 3,2'-
dimethyl-4-aminobiphenyl-induced colon carcinogenesis in F344 rats, Cancer Res., 41,
1363, 1981.
Glauert 263
74. Rogers, A. E. and Nauss, K. M., Rodent models for carcinoma of the colon, Dig. Dis.
Sci., 30, 87S, 1985.
75. Nauss, K. M., Locniskar, M., Sondergaard, D., and Newberne, P. M., Lack of
effect of dietary fat on N-nitrosomethylurea (NMU)-induced colon tumorigenesis in rats,
Carcinogenesis, 5, 255, 1984.
76. Glauert, H. P. and Weeks, J. A., Dose- and time-response of colon carcinogenesis in
Fischer-344 rats after a single dose of I ,2-dimethylhydrazine, Toxicol. Letters, 48, 283,
1989.
77. Karkare, M. R., Clark, T. D., and Glauert, H. P., Effect of dietary calcium on colon
carcinogenesis induced by a single injection of I ,2-dimethylhydrazine in rats, J. Nutr.,
121, 568, 1991.
78. Schiller, C. M., Curley, W. H., and McConnell, E. E., Induction of colon tumors by
a single oral dose of 1,2-dimethylhydrazine, Cancer Lett., II, 75, 1980.
79. Decaens, C., Gautier, R., Bara, J., and Burtin, P., Induction of rat intestinal carcin-
ogenesis with single doses, low and high repeated doses of I ,2-dimethylhydrazine, Car-
cinogenesis, 10, 69, 1989.
80. Ward, J. M., Dose response to a single injection of azoxymethane in rats, Vet. Pathol.,
12, 165, 1975.
81. Glauert, H. P., Bennink, M. R., and Sander, C. H., Enhancement of 1,2-dimethyl-
hydrazine-induced colon carcinogenesis in mice by dietary agar, Food Cosmetics Toxicol.,
19, 281, 1981.
82. Nauss, K. M., Locniskar, M., and Newberne, P.M., Effect of alterations in the quality
and quantity of dietary fat and I ,2-dimethylhydrazine-induced colon carcinogenesis in
rats, Cancer Res., 43,4083, 1983.
83. Reddy, B.S., Weisburger, J. H., and Wynder, E. L., Effects of dietary fat level and
dimethylhydrazine on fecal bile acid and neutral sterol excretion and colon carcinogenesis
in rats, J. Nat!. Cancer lnst., 52, 507, 1974.
84. Bull, A. W., Soullier, B. K., Wilson, P. S., Hayden, M. T., and Nigro, N. D.,
Promotion of azoxymethane-induced intestinal cancer by high fat diet in rats, Cancer
Res., 39, 4956, 1979.
85. Sakaguchi, M., Minuora, T., Hiramatsu, T., et at., Effects of dietary saturated and
unsaturated fatty acids on bile acids and colon carcinogenesis induced by azoxymethane
in rats, Cancer Res., 46, 61, 1986.
86. Nigro, N. D., Singh, D. V., Campbell, R. L., and Pak, M. S., Effect of dietary fat
on intestinal tumor formation by azoxymethane in rats, J. Natl. Cancer lnst., 54, 439,
1975.
87. Reddy, B.S., Watanabe, K., and Weisburger, J. H., Effect of high-fat diet on colon
carcinogenesis in F344 rats treated with l ,2-dimethylhydrazine, methylazoxymethanol
acetate, or methylnitrosourea, Cancer Res., 37, 4156, 1977.
88. Bull, A. W., Bronstein, j. C., and Nigro, N. D., The essential fatty acid requirement
for azoxymethane-induced intestinal carcinogenesis in rats, Lipids, 24, 340, 1989.
89. Reddy, B. S., Narisawa, T., Vukusich, D., Weisburger, J. H., and Wynder, E. L.,
Effect of quality and quantity of dietary fat and dimethylhydrazine on colon carcinogenesis
in rats, Proc. Soc. Exp. Bioi. Med., 151, 237, 1976.
90. Bansal, B. R., Rhoads, J. E. J., and Bansal, S. C., Effects of diet on colon carcin-
ogenesis and the immune system in rats treated with I ,2-dimethylhydrazine, Cancer Res.,
38, 3293, 1978.
91. Schmaehl, D., Habs, M., and Habs, H., Influence of a non-synthetic diet with a high
fat content on the local occurrence of colonic carcinomas induced by N-nitroso-acetox-
ymethylmethylamine (AMMN) in Sprague-Dawley rats, Hepato-gastroenterol., 30, 30,
1983.
92. Locniskar, M., Nauss, K. M., Kaufmann, P., and Newberne, P. M., Interaction of
dietary fat and route of carcinogen administration on I ,2-dimethylhydrazine-induced colon
tumorigenesis in rats, Carcinogenesis, 6, 349, 1985.
93. Nauss, K. M., Bueche, D., and Newberne, P.M., Effect of beef fat on DMH-induced
colon tumorigenesis: influence of rat strain and nutrient composition, J. Nutr., 117, 739,
1987.
94. Pence, B. C. and Buddingh, F., Inhibition of dietary fat-promoted colon carcinogenesis
in rats by supplement calcium or vitamin 0 3 , Carcinogenesis, 9, 187, 1988.
95. Hardman, W. E., Heitman, D. W., and Cameron, I. L., Suppression of the progression
of I ,2 dimethylhydrazine (DMH) induced colon carcinogenesis by 20% dietary corn oil
in rats supplemented with dietary pectin, Proc. Am. Assoc. Cancer Res., 32, 131, 1991.
96. Sakaguchi, M., Hiramatsu, Y., Takada, H., Yamamura, M., Hioki, K., Saito, K.,
and Yamamoto, M., Effect of dietary unsaturated and saturated fats on azoxymethane-
induced colon carcinogenesis in rats, Cancer Res., 44, 1472, 1984.
97. Reddy, B.S. and Maeura, Y., Tumor promotion by dietary fat in azoxymethane-induced
colon carcinogenesis in female F344 rats: influences of amount and source of dietary fat,
J. Nat/. Cancer Inst., 72, 745, 1984.
98. Reddy, B. S. and Maruyama, H., Effect of different levels of dietary corn oil and lard
during the initiation phase of colon carcinogenesis in F344 rats, J. Nat/. Cancer lnst.,
77, 815, 1986.
99. Reddy, B. S. and Sugie, S., Effect of different levels of omega-3 and omega-6 fatty
acids on azoxymethane-induced colon carcinogenesis in F344 rats, Cancer Res., 48,
6642, 1988.
100. Wargovich, M. J., Allnutt, D., Palmer, C., Anaya, P., and Stephens, L. C., Inhibition
of the promotional phase of azoxymethane-induced colon carcinogenesis in the F344 rat
by calcium lactate: effect of simulating two human nutrient density levels, Cancer Lett.,
53, 17, 1990.
101. Guillem, J. G., Hsieh, L. L., O'Toole, K. M., Forde, K. A., LoGerfo, P., and
Weinstein, I. B., Changes in expression of oncogenes and endogenous retroviral-like
sequences during colon carcinogenesis, Cancer Res., 48, 3964, 1988.
102. Longnecker, D., Experimental pancreatic cancer: role of species, sex and diet, Bull.
Cancer, 77, 27, 1990.
103. Roebuck, B. D., Yager, J.D., Longnecker, D. S., and Wilpone, S. A., Promotion
by unsaturated fat of azaserine-induced pancreatic carcinogenesis in the rat, Cancer Res.,
41' 3961' 1981.
104. Roebuck, B. D., Longnecker, D. S., Baumgartner, K. J., and Thron, C. D., Car-
cinogen-induced lesions in the rat pancreas: effects of varying levels of essential fatty
acid, Cancer Res., 45, 5252, 1985.
105. Roebuck, B. D., Effects of high levels of dietary fats on the growth of azaserine-induced
foci in the rat pancreas, Lipids, 21, 281, 1986.
106. Roebuck, B. D., Kaplita, P. V., Edwards, B. R., and Praissman, M., Effects of
dietary fats and soybean protein on azaserine-induced pancreatic carcinogenesis and plasma
cholecystokinin in the rat, Cancer Res., 47, 1333, 1987.
107. O'Connor, T. P., Roebuck, B. D., and Campbell, T. C., Dietary intervention during
the postdosing phase of !-azaserine-induced preneoplastic lesions, J. Nat/. Cancer Inst.,
75, 955, 1985.
108. Woutersen, R. A. and van Garderen-Hoetmer, A., Inhibition of dietary fat promoted
development of (pre )neoplastic lesions in exocrine pancreas of rats and hamsters by
supplemental vitamins A, C and E, Cancer Lett., 41, 179, 1988.
109. Woutersen, R. A., van Garderen-Hoetmer, A., Bax, J., and Scherer, E., Modulation
of dietary fat-promoted pancreatic carcinogenesis in rats and hamsters by chronic ethanol
ingestion, Carcinogenesis, 10, 453, 1989.
110. Woutersen, R. A., van Garderen-Hoetmer, A., Bax, J., and Scherer, E., Modulation
of dietary fat -promoted pancreatic carcinogenesis in rats and hamsters by chronic coffee
ingestion, Carcinogenesis, 10, 311, 1989.
Glauert 265
Ill. Roebuck, B. D., Yager, J. D. J., and Longnecker, D. S., Dietary modulation of
azaserine-induced pancreatic carcinogenesis in the rat, Cancer Res., 41, 888, 1981.
112. Longnecker, D. S., Roebuck, B. D., and Kuhlmann, E. T., Enhancement of pancreatic
carcinogenesis by a dietary unsaturated fat in rats treated with saline or N-nitroso(2-
hydroxypropyl) (2-oxopropyl)amine, J. Natl. Cancer /nst., 74, 219, 1985.
113. Birt, D. F., Salmasi, S., and Pour, P.M., Enhancement of experimental pancreatic
cancer in Syrian golden hamsters by dietary fat, J. Natl. Cancer /nst., 67, 1327, 1981.
114. Birt, D. F. and Pour, P. M., Increased tumorigenesis induced by N-nitrosobis(2-oxo-
propyl)amine in Syrian golden hamsters fed high-fat diets, J. Nat[. Cancer /nst., 70,
1135, 1983.
115. Birt, D. F., Julius, A. D., White, L. T., and Pour, P.M., Enhancement of pancreatic
carcinogenesis in hamsters fed a high-fat diet ad libitum and at a controlled calorie intake,
Cancer Res., 49,5848,1989.
116. Birt, D. F., Julius, A. D., Dwork, E., Hanna, T., White, L. T., and Pour, P. M.,
Comparison of the effects of dietary beef tallow and com oil on pancreatic carcinogenesis
in the hamster model, Carcinogenesis, 11,745, 1990.
117. Eustis, S. L. and Boorman, G. A., Proliferative lesions of the exocrine pancreas:
relationship to com oil gavage in the National Toxicology Program, J. Nat/. Cancer /nst.,
75, 1067, 1985.
118. Haseman, J. K., Huff, J. E., Rao, G. N., Arnold, J. E., Boorman, G. A., and
McConnell, E. E., Neoplasms observed in untreated and com oil gavage control groups
of F344/N rats and (C57BL/6N X C3H/HeN)F 1 (B6C3F,) mice, J. Nat!. Cancer /nst.,
75, 975, 1985.
119. Freedman, L. S., Clifford, C., and Messina, M., Analysis of dietary fat, calories,
body weight, and the development of mammary tumors in rats and mice: a review, Cancer
Res., 50, 5710, 1990.
120. Beems, R. B. and van Beek, L., Modifying effect of dietary fat on benzo(a)pyrene-
induced respiratory tract tumors in hamsters, Carcinogenesis, 5, 413, 1984.
121. Szepsenwol, J., Carcinogenic effect of ether extract of whole egg, alcohol extract of
egg yolk, and powdered egg free of the ether extractable part in mice, Proc. Soc. Exp.
Bioi. Med., 116, 1136, 1964.
122. Szepsenwol, J., Brain nerve cell tumors in mice on diets supplemented with various
lipids, Pathol. Microbiol., 34, 1, 1969.
123. Pollard, M. and Lockert, P. H., Promotional effects of testosterone and dietary fat on
prostate carcinogenesis in genetically susceptible rats, Prostate, 6, I, 1985.
124. Cave, W. T., Dietary n-3 polyunsaturated fatty acid effects on animal tumorigenesis,
FASEB J., 5, 2160, 1991.
125. O'Connor, T. P., Roebuck, B. D., Peterson, F., and Campbell, T. C., Effect of
dietary intake of fish oil and fish protein on the development of L-azaserine-induced
preneoplastic lesions in the rat pancreas, J. Natl. Cancer /nst., 75, 959, 1985.
126. O'Connor, T. P., Roebuck, B. D., Peterson, F. J., Lokesh, B., Kinsella, J. E., and
Campbell, T. C., Effect of dietary omega-3 and omega-6 fatty acids on development of
azaserine-induced preneoplastic lesions in rat pancreas, J. Nat/. Cancer Inst., 81, 858,
1989.
127. Nelson, R. L., Tanure, J. C., Andrianopoulos, G., Souza, G., and Lands, W. E. M.,
A comparison of dietary fish oil and com oil in experimental colorectal carcinogenesis,
Nutr. Cancer, 11, 215, 1988.
128. Minoura, T., Takata, T., Sakaguchi, M., Takada, H., Yamamura, M., Hioki, K.,
and Yamamoto, M., Effect of dietary eicosapentaenoic acid on azoxymethane-induced
colon carcinogenesis in rats, Cancer Res., 48, 4790, 1988.
129. Reddy, B. S. and Maruyama, H., Effect of fish oil on azoxymethane-induced colon
carcinogenesis in male F344 rats, Cancer Res., 46, 3367, 1986.
130. Abou-El-Ela, S. H., Prasse, K. W., Carroll, R., Wade, A. E., Dharwadkar, S., and
Bunce, 0. R., Eicosanoid synthesis in 7, 12-dimethylbenz(a)anthracene-induced mam-
mary carcinomas in Sprague-Dawley rats fed primrose, menhaden or com oil diets, Lipids,
23, 948, 1988.
131. Abou-El-Ela, S. H., Prasse, K. W., Farrell, R. L., Carroll, R. W., Wade, A. E.,
and Bunce, 0. R., Effects of D.L-2-difluoromethylomithine and indomethacin on mam-
mary tumor promotion in rats fed high n-3 and/or n-6 fat diets, Cancer Res., 49, 1434,
1989.
132. Locniskar, M., Belury, M. A., Cumberland, A. G., Patrick, K. E., and Fischer,
S. M., Lack of a protective effect of menhaden oil on skin tumor promotion by 12-0-
tetradecanoylphorbol-13-acetate, Carcinogenesis, 11, 1641, 1990.
133. Murphy, M.G., Dietary fatty acids and membrane protein function, J. Nutr. Biochem.,
I, 68, 1990.
134. Welsch, C. W., Enhancement of mammary tumorigenesis by dietary fat: review of
potential mechanisms, Am. J. Clin. Nutr., 45, 192, 1987.
135. Narisawa, T., Magadia, N. E., Weisburger, J. H., and Wynder, E. L., Promoting
effect of bile acids on colon carcinogenesis after intrarectal instillation of N-methyl-N '-
nitro-N-nitrosoguanidine in rats, J. Nat/. Cancer Inst., 53, 1093, 1974.
136. Reddy, B.S., Narisawa, T., Weisburger, J. H., and Wynder, E. L., Promoting effect
of sodium deoxycholate on colon adenocarcinomas in germ-free rats, J. Nat/. Cancer
Inst., 56, 441, 1976.
137. Reddy, B. S., Watanabe, K., Weisburger, J. H., and Wynder, E. L., Promoting
effect of bile acids in colon carcinogenesis in germ-free and conventional F344 rats,
Cancer Res., 37, 3238, 1977.
138. Reddy, B. S., Mangat, S., Sheinfil, A., Weisburger, J. H., and Wynder, E. L.,
Colon carcinogenesis. Effect of type and amount of dietary fat and 1,2-dimethylhydrazine
on biliary bile acids and neutral sterols in rats, Cancer Res, 37, 2132, 1977.
139. Glauert, H. P. and Bennink, M. R., Influence of diet or intrarectal bile acid injections
on colon epithelial cell proliferation in rats previously injected with I ,2-dimethylhydra-
zine, J. Nutr., 113,475, 1983.
140. Reddy, B. S., Hanson, D., Mangat, S., Mathews, L., Sbaschnig, M., Sharma, C.,
and Simi, B., Effect of high-fat, high-beef diet and of mode of cooking of beef in the
diet on fecal bacterial enzymes and fecal bile acids and neutral sterols, J. Nutr., 110,
1880, 1980.
141. Gallaher, D. D. and Franz, P. M., Effects of com oil and wheat brans on bile acid
metabolism in rats, J. Nutr., 120, 1320, 1990.
142. Deschner, E. E., Cohen, B. 1., and Raicht, R. F., Acute and chronic effect of dietary
cholic acid on colonic epithelial cell proliferation, Digestion, 21, 290, 1981.
143. Cohen, B. 1., Raicht, R. F., Deschner, E. E., Takahashi, M., Sarwal, A. N., and
Fazzini, E., Effect of cholic acid feeding on N-methyl-N-nitrosourea-induced colon
tumors and cell kinetics in rats, J. Nat/. Cancer Inst., 64, 573, 1980.
144. Skraastad, 0. and Reichelt, K. L., An endogenous colon mitosis inhibitor and dietary
calcium inhibit increased colonic cell proliferation induced by cholic acid, Scand. J.
Gastroenterol., 23, 801, 1988.
145. Wargovich, M. J., Eng, V. W. S., Newmark, H. L., and Bruce, W. R., Calcium
ameliorates the toxic effect of deoxycholic acid on colonic epithelium, Carcinogenesis,
4, 1205, 1983.
146. Wargovich, M. J, Eng, V. W. S., and Newmark, H. L., Calcium inhibits the damaging
and compensatory proliferative effects of fatty acids on mouse colon epithelium, Cancer
Lett., 23, 253, 1984.
147. Craven, P. A., Pfanstiel, J., and DeRubertis, F. R., Role of activation of protein
kinase C in the stimulation of colonic epithelial proliferation and reactive oxygen formation
by bile acids, J. Clin. Invest., 79, 532, 1987.
Glauert 267
148. Rosenthal, M.D., Fatty acid metabolism of isolated mammalian cells, Pro g. Lipid Res ..
26, 87, 1987.
149. Lands, W. E. M., Biosynthesis of prostaglandins, Annu. Rev. Nutr., II, 41, 1991.
150. Dyerberg, J., Linolenate-derived polyunsaturated fatty acids and prevention of athero-
sclerosis, Nutr. Rev., 44, 125, 1986.
151. Nicosia, S. and Patrono, C., Eicosanoid biosynthesis and action: novel opportunities
for pharmacological intervention, FASEB J., 3, 1941, 1989.
152. Fischer, S. M., Baldwin, J, K., Jasheway, D. W., Patrick, K. E., and Cameron,
G. S., Phorbol ester induction of 8-lipoxygenase in inbred SENCAR (SSIN) but not
C57BLI6J mice correlated with hyperplasia, edema, and oxidant generation but not or-
nithine decarboxylase induction, Cancer Res., 48, 658, 1988.
153. Fischer, S. M., Fiirstenberger, G., Marks, F., and Slaga, T. J., Events associated
with mouse skin tumor promotion with respect to arachidonic acid metabolism: a com-
parison between SENCAR and NMRI mice, Cancer Res., 47, 3174, 1987.
154. Denda, A., Ura, H., Tsujiuchi, T., Tsutsumi, M., Eimoto, H., Takashima, Y.,
Kitazawa, S., Kinugasa, T., and Konishi, Y., Possible involvement of arachidonic acid
metabolism in phenobarbital promotion of hepatocarcinogenesis, Carcinogenesis, 10,
1929, 1989.
!55. Reddy, B.S., Wang, C. X., and Maruyama, H., Effect of restricted caloric intake on
azoxymethane-induced colon tumor incidence in male F344 rats, Cancer Res., 47, 1226,
1987.
156. Klurfeld, D. M., Weber, M. M., and Kritchevsky, D., Inhibition of chemically induced
mammary and colon tumor promotion by caloric restriction in rats fed increased dietary
fat, Cancer Res., 47, 2759, 1987.
157. Boissonneault, G. A., Elson, C. E., and Pariza, M. W., Net energy effects of dietary
fat on chemically induced mammary carcinogenesis in F344 rats, J. Natl. Cancer lnst.,
76, 335' 1986.
158. Kritchevsky, D., Weber, M. M., and Klurfeld, D. M., Dietary fat versus caloric
content in initiation and promotion of 7, 12-dimethylbenz(a)anthracene-induced mammary
tumorigenesis in rats, Cancer Res., 44, 3174, 1984.
159. Welsch, C. W., House, J. L., Herr, B. L., Eliasberg, S. J., and Welsch, M. A.,
Enhancement of mammary carcinogenesis by high levels of dietary fat: a phenomenon
dependent on ad libitum feeding, J. Natl. Cancer lnst., 82, 1615, 1990.
160. Boissonneault, G. A., Calories and carcinogenesis: modulation by growth factors, in
Nutrition, Toxicity, and Cancer, Rowland, I. R., Ed., CRC Press, Boca Raton, FL,
1991' 413.
161. Snyder, D. L., Dietary Restriction and Aging, Alan R. Liss, New York, 1989.
162. Baldwin, S. and Parker, R. S., Effects of dietary fat level and aflatoxin B, treatment
on rat hepatic lipid composition, Food Chern. Toxicol., 23, 1049, 1985.
163. Neelands, P. J, and Clandinin, M. T., Diet fat influences liver plasma-membrane lipid
composition and glucagon-stimulated adenylate cyclase activity, Biochem. J., 212, 573,
1983.
164. Wade, A. E., Norred, W. P., and Evans, J. S., Lipids in drug detoxification, in
Nutrition and Drug Interrelations, Hathcock, J. N. and Coon, J., Eds., Academic Press,
New York, 1978, 475.
165. Hammer, C. T. and Wills, E. D., Dependence of the rate of metabolism of benzo(a)pyrene
on the fatty acid composition of the liver endoplasmic reticulum and on dietary lipids,
Nutr. Cancer, 2, 113, 1980.
166. Rutten, A. A. J. j. L. and Flake, H. E., Influence of high dietary levels of fat on rat
hepatic phase I and II biotransformation enzyme activities, Nutr. Rep. Int., 36, 109,
1987.
167. Cassano!, P., Bonnamour, D., Grolier, P., Pelissier, M.A., Amelizad, Z., Albrecht,
R., and Narbonne, J, F., The effect of dietary imbalances on the activation of
benzo[a]pyrene by the metabolizing enzymes from rat liver, Mutat. Res., 191,67, 1987.
168. Kim, H. J., Choi, E. S., and Wade, A. E., Effect of dietary fat on the induction of
hepatic microsomal cytochrome P450 isozymes by phenobarbital, Biochem. Pharmacal.,
39, 1423, 1990.
169. Lam, T. C. L. and Wade, A. E., Influence of dietary lipid on the metabolism of
hexobarbital by the isolated, perfused rat liver, Pharmacology, 21, 64, 1980.
170. Lam, T. C. L. and Wade, A. E., Effect of dietary lipid on benzo(a)pyrene metabolism
by perfused rat liver, Drug Nutr. Interact., 1, 31, 1981.
171. Wade, A. E., Harley, W., and Bunce, 0. R., The effects of dietary com oil on the
metabolism and mutagenic activation of N-nitrosodimethylamine (DMN) by hepatic mi-
crosomes from male and female rats, Mutat. Res., 102, 113, 1982.
172. Clarke, S. D., Armstrong, M. K., and Jump, D. B., Dietary polyunsaturated fats
uniquely suppress rat liver fatty acid synthase and Sl4 mRNA content, J. Nutr., 120,
225, 1990.
173. Brannon, P.M., Adaptation of the exocrine pancreas to diet, Annu. Rev. Nutr., 10, 85,
1990.
174. Ribeiro, A., Mangeney, M., Cardot, P., Loriette, C., Chambaz, J., Rayssiguier,
Y., and Bereziat, G., Nutritional regulation of apolipoprotein genes: effect of dietary
carbohydrates and fatty acids, Adv. Exp. Med. Bioi., 285, 407, 1991.
176. Karmali, R. A., Chao, C. C., Basu, A., and Modak, M., II, Effect of n-3 and n-6
fatty acids on mammary H-ras expression and PGE-2 levels in DMBA-treated rats,
Anticancer Res., 9, 1169, 1989.
177. Ronai, Z., Lau, Y. Y., and Cohen, L.A., Dietary n-3 fatty acids do not affect induction
of Ha-ras mutations in mammary glands of NMU-treated rats, Mol. Carcinog., 4, 120,
1991.
178. Elliott, T. S., Swartz, D. A., Visek, W. J., and Kaput, J., Molecular analyses of
differences in gene expression due to changes in the level of dietary fat, FASEB J., 4,
A916, 1990.
179. Swartz, D. A., Visek, W. J., and Kaput, J., Genomic organization of a fat regulated
gene, FASEB J., 5, AI649, 1991.
Chapter 12
REGULATION OF ACETYL CoA CARBOXYLASE AND

GENE EXPRESSION
John B. Allred and Diana F. Bowers
TABLE OF CONTENTS
I. Overview ...................... ...................... ............ 270
II. Is Carbohydrate Converted to Fatty Acids in Humans? .......... 271
III. The Gene for Acetyl CoA Carboxylase ...................... .... 272
IV. Rat Acetyl CoA Carboxylase mRNA ...................... ...... 273
V. Evidence of Isozymic Forms of Acetyl CoA

Carboxylase ...................... ...................... .......... 278
VI. Possible Phsiological Relevance of Multiple Forms of

mRNA Coding for Acetyl CoA Carboxylase ..................... 279
VII. Regulation of Acetyl CoA Carboxylase: Relative

Importance of Transcription, Translation, and
Posttranslational Processing ...................... ................ 280
VIII. Subcellular Distribution of Acetyl CoA Carboxylase as a

Potential Posttranslational Regulatory Mechanism ............... 287
IX. Biotinylation ...................... ...................... ......... 290
X. Short-Term Regulatory Mechanisms ...................... ....... 290
XI. Conclusion ...................... ...................... ........... 291
References ...................... ...................... ................... 291
0-8493-6961-4/93/$0.00 +$.50
I. OVERVIEW
Acetyl CoA carboxylase is most often considered in its role in de novo

fatty acid biosynthesis. In that context, acetyl CoA carboxylase is usually
cited as the key enzyme which regulates fatty acid biosynthesis, 1 but it is
only one of several enzymes required for the process in vivo.2 As shown in
Figure 1, the "lipogenic" enzymes, in addition to acetyl CoA carboxylase,
include ATP-citrate lyase, fatty acid synthase, glucose-6-phosphate dehydro-
genase, and malic enzyme, all of which are active in cytosol where fatty acid
biosynthesis occurs. 1 When experimental animals are subjected to enhanced
lipogenic conditions (e.g., refeeding a high carbohydrate diet to previously
fasted rats), there is an increase in activity of acetyl CoA carboxylase in liver
with a concomitant, coordinated rise in the activities of the other lipogenic
enzymes. 2 At the same time the product of acetyl CoA carboxylase, malonyl
CoA, can inhibit camitine palmitoyltransferase, a key enzyme necessary for
the transport of long-chain fatty acyl CoA esters into mitochondria. This
inhibition is important in the regulation of 13-oxidation of long-chain fatty
acids, as discussed in Chapter 14. 3 In the simplest interpretation, this inhibition
by the product of acetyl CoA carboxylase may be a mechanism to minimize
the possibility that synthesis of long-chain fatty acids and their degradation
do not occur at the same time.
In addition to its role in de novo fatty acid biosynthesis, acetyl CoA
carboxylase is also required for elongation of long-chain fatty acids. While
the initial product of the fatty acid synthesis system is palmitate, part of it is
elongated to form stearate and oleate before exiting the liver as a triglyceride
in the form of VLDL. 4 The conversion of linoleic acid to arachidonic acid
by the microsomal elongation system requires malonyl CoA produced by
acetyl CoA carboxylase. 5 Such a conversion is the reason that linoleic, but
not arachidonic acid, is a dietary essential. Arachidonic acid is an integral
constituent of membranes and is necessary for the synthesis of an important
family of prostaglandins. 6
The fact that acetyl CoA carboxylase has a dual role in metabolism is
important in terms of potential dietary/hormonal regulation of the enzyme.
Early work showed that fatty acid biosynthesis in animal models increased
in response to a diet high in carbohydrate. 2 The two hormones most closely
associated with changes in dietary carbohydrate level, insulin and glucagon,
have opposing effects on lipogenesis and the activity of acetyl CoA carboxylase7
in isolated rat hepatocytes. On the other hand, even when carbohydrate con-
sumption is limited, the elongation system and consequently a basal level of
the activity of acetyl CoA carboxylase is still necessary.
Allred and Bowers 271
0 Glucoae-G-P Triglyceride
6-P-Giucon?.:tone 1 l
Palmitate
~ Fruiae-8-P
Trloae-P
Malonyl Co A
Pyruvate
/
+
(5) -Malate
! 0l
Acetyl CoA
Oxaloaceta~e
Acetyl CoA @
[M]oU©©i'ml@ au ©hr o©liTl
Oxaloacetate ~ Citrate --+--+ Citrate
FIGURE l. Fatty acid biosynthesis from glucose. 2 Acetyl CoA carboxylase catalyzes the
formation of malonyl CoA (Reaction I) from acetyl CoA produced in cytosol by the action of
ATP-citrate lyase (Reaction 2). Conversion of malonyl CoA to palmitate is catalyzed by fatty
acid synthase (Reaction 3), using NADPH produced by glucose-6-phosphate dehydrogenase
(Reaction 4) and malic enzyme (Reaction 5).
II. IS CARBOHYDRATE CONVERTED TO FATTY

ACIDS IN HUMANS?
In the context of human nutrition, the question of the importance of de
novo fatty acid biosynthesis must be considered. There is no doubt that
humans, in common with other higher animals, have acetyl CoA carboxylase
as well as the other necessary enzymes for the elongation of linoleic to
arachidonic acid. In fact, the gene for acetyl CoA carboxylase has been shown
to be present in human genetic material, reportedly located on chromosome
l7. 8 Further, an inborn error of metabolism in which acetyl Co A carboxylase
is missing is known to exist in humans. 9
While acetyl CoA carboxylase is clearly functional as a part of the long-
chain fatty acid elongation system in human tissue, how important is it and
the other lipogenic enzymes in de novo fatty acid synthesis from carbohydrate?
It is generally accepted that when de novo fatty acid synthesis happens in
humans, the process occurs in liver and not adipose tissue. 10 Still, there have
been a number of reports over the years which have questioned the importance
of the conversion of carbohydrate to fat in humans. Early work 11 measured
the respiratory quotients of subjects who were given an oral carbohydrate load
and found RQ values less than 1, whereas the synthesis of fatty acids from
carbohydrate should give values greater than 1. A subsequent studyl 2 con-
firmed that after a single, high-carbohydrate meal equivalent to 479 g of
starch, nonprotein-RQ values did not exceed 1 over the next 8 h. The authors
concluded that the ingested carbohydrate was not converted to fat under these
conditions.
The question was recently addressed more directly by administering ace-
tate labeled with the nonradioactive, heavy isotope of carbon, 13C, to human
volunteers who were given either an infused or oral glucose load. 4 While it
was demonstrated that the administered [ 13C]acetate did label the liver cy-
tosolic acetyl CoA pool, surprisingly little isotope was found in fatty acids
of triglycerides of blood VLDL. It was calculated that less than 2% of the
VLDL palmitate and less than 1% of the VLDL stearate came from de novo
fatty acid biosynthesis. Thus, the authors concluded that the glucose load
failed to stimulate de novo fatty acid biosynthesis.
If this conclusion is extrapolated to the general case, it raises a number
of questions. Not the least of these is, if excessive carbohydrate consumption
does not increase de novo fatty acid synthesis and deposition, what happens
to the calories? Further, since feeding diets high in carbohydrate for extended
periods is known to stimulate fatty acid synthesis in the tissue of a number
of animal models, including rats, 13 • 14 mice, 15 pigs, 16 and even ruminants, 17
the lack of an effect of dietary carbohydrate would imply that humans have
a markedly different lipogenic regulatory system than that found in other
higher animals. It seems much more likely that the apparent lack of a car-
bohydrate-stimulated increase in lipogenesis in humans in these three
reports 4 • 11 • 12 was due to the lack of subject adaptation to an increased car-
bohydrate intake. That is, in each of these human studies, subjects were
allowed to consume their "normal" diet (presumably containing substantially
more fat and less carbohydrate than the test diet) prior to the experiment and
measurements were made in the few hours following a single carbohydrate
load. Numerous studies have made it clear that while some of the regulatory
mechanisms that control lipogenesis in animals occur very rapidly, others
require substantially more time. One such time-dependent, adaptive mecha-
nism is the regulation of the quantity of the active cytosolic form of acetyl
CoA carboxylase. Much of the remainder of this chapter is devoted to a
discussion of the time-dependent adaptive mechanisms which regulate acetyl
CoA carboxylase activity and thus lipogenesis.
III. THE GENE FOR ACETYL CoA CARBOXYLASE

Insight into the molecular biology of acetyl CoA carboxylase has a rel-
atively short history. Kim's laboratory was the first to purify a functional
acetyl CoA carboxylase mRNA from mammalian tissue (rat mammary gland) 18
and to obtain acetyl CoA carboxylase e-DNA 19 which could be used as a
probe to study the molecular biology of the rat enzyme. Using these tools in
subsequent experiments, overlapping e-DNA clones were found which co-
vered the entire coding sequence of rat acetyl CoA carboxylase mRNA. 20
The sequence of the coding section, consisting of 7035 bases, was used to
predict the primary structure of rat acetyl CoA carboxylase. By this analysis,
rat acetyl CoA carboxylase contained 2345 amino acids and had a molecular
weight of 265,220. Although the entire predicted primary structure of rat
acetyl Co A carboxylase has been published, 2° Figure 2 shows only selected
segments of the enzyme to illustrate some features which may be important
in its synthesis and regulation.
A e-DNA probe for chicken acetyl CoA carboxylase has also been
prepared 24 and used25 to determine the sequence of bases in the coding section
of DNA for acetyl CoA carboxylase mRNA, from which the primary amino
acid sequence of the chicken enzyme was deduced. Chicken acetyl CoA
carboxylase was predicted to contain 21 fewer amino acid residues than the
rat enzyme and have a molecular weight of 262,706. Not surprisingly, acetyl
CoA carboxylase from the two sources have very similar primary structures.
It is somewhat more surprising that both the rat2° and chicken 25 liver enzyme
exhibit a high degree of homology with carbamoyl phosphate synthetase at
the amino terminal end of the protein.
IV. RAT ACETYL CoA CARBOXYLASE mRNA

The use of a e-DNA probe to study rat acetyl CoA carboxylase mRNA
has provided some very interesting observations and, at the same time, showed
that the transcriptional process and its regulation are far from simple. The
basic technique usually used to study mRNA is called Northern blotting. It
involves the electrophoretic separation by size of RNA extracted from a given
tissue and then transferring the RNA from the electrophoresis gel to a solid
support, usually a nylon membrane. The membrane is subsequently probed
with a e-DNA which has been made radioactive by the incorporation of 32P
(nick-translated). Any mRNA on the membrane whose base sequence is com-
plementary to that of the e-DNA will bind the radioactive e-DNA probe, in
proportion to the quantity of mRNA. When the membrane is placed in contact
with X-ray film, the film is exposed wherever radioactivity occurs. From the
developed film, the approximate size(s) of any mRNA which bound the e-
DNA probe can be estimated. The relative quantity of each mRNA species,
over a limited concentration range, can be estimated by scanning the film
with a densitometer.
Northern blot analysis of mRNA from rat liver 19 showed that the acetyl
CoA carboxylase e-DNA probe hybridized to a mRNA of varying sizes, but
two major bands were observed. The larger of the two contained approxi-
mately 10 kilobases (kb) while the smaller, more diffuse band was estimated
to contain about 3 kb. Since acetyl CoA carboxylase is such a large protein
Met·Asp·Glu·Pro·Ser·Pro-Leu·Ala-Lys·Thr-Leu·Glu·Leu·Asp-Gln·His·Ser·Arg·Phe·lle 20 N
-...I
1 1 1
21 lle·Gly-Ser·Val·Ser·Glu-Asp·Asn-Ser·Glu-Asp·Glu-lle·Ser-Asn·Leu-Val·Lys·Leu·Asp 40 ""
1 1
61 Leu·Gly·lle·Ser·Ala-Leu·Gln·Asp·Gly·Leu-Ala·Phe·His·Met·Arg·Ser·Ser-Met·Ser·Gly 80
1
81 Leu·His·Leu·Val·Lys·Gln·Gly·Arg·Asp-Arg·Lys-Lys·lle·Asp·Ser·Gln·Arg·Asp·Phe·Thr 100
Biotin
781 lle·Glu·Val-Met·Lys·Met·Val-Met·Thr·Leu·Thr·Ala·Val-Glu·Ser·Gly·Cys·lle·His·Tyr 800
1-----------Deletion------------1 1
1181 Phe·Met·Leu·Pro-Thr·Ser-His·Pro-Asn·Arg-Gly·Asn·lle-Pro·Thr·Leu·Asn-Arg-Met·Ser 1200
1
<:
::::
1201 Phe·Ala-Ser-Asn·Leu·Asn-His-Tyr-Gly·Met-Thr-His·Val·Ala·Ser·Val·Ser·Asp·Val-Leu 1220 3:.
g·
~
~
2340 CJ
2321 Met-Thr·Gln-His·lle-Ser·Pro-Thr-Gln·Arg-Ala·Glu-Val-Val·Arg-lle-Leu-Ser-Thr-Met
"'~
2341 Asp·Ser·Pro·Ser·Thr. 2345
~
"\:i
~
FIGURE 2. Selected segments of the primary structure of acetyl CoA carboxylase, based on the amino acid sequence predicted from analysis of e-DNA clones """"
of the enzymew The arrows ( t ) show the eight serine residues which can be phosphorylated. 21 The lysine residue in the indicated ( Biotin ) tetrapeptide is ~-
biotinylated in the holoenzyme'" 22 and the indicated octapeptide (----Deletion----) is deleted in a putative isozymic forrn of the enzyme."
Exona for s• Untranalated Region
r-----.
/ll\~
Genomic DNA
FL56 type m-RNA Coding Region
pAU type m-RNA Coding Region
FIGURE 3. Reported 27 exon-intron structure of genomic DNA coding for the 5' untranslated
region of acetyl CoA carboxylase, showing differential splicing of exons to produce heterogeneous
acetyl CoA carboxylase mRNA. The protein synthesis initiation codon (AUG) is located in exon
5. PI and P2 show the approximate locations of two putative promoters of transcription.
that mRNA coding for it must contain at least 7 kb, 20 it has been concluded
that the 10-kb band represented translatable acetyl CoA carboxylase mRNA
while the 3-kb band represented degradative products. 18
Although electrophoresis can separate species of mRNA which have sub-
stantial differences in molecular weight, detection of more subtle structural
differences require more sophisticated techniques. Using e-DNA probes com-
plementary to specific base sequences in acetyl CoA carboxylase mRNA for
Northern blot analysis and primer extension analysis, acetyl CoA carboxylase
mRNA has been examined in mammary gland from lactating rats 26 •27 and in
rat liver. 28 These analyses established that there was substantial heterogeneity
of the mRNA within the 10-kb band in both tissues. Five different species
of mRNA were found in rat mammary gland all of which had the same base
sequence in the coding region but differed in length and/or base sequence of
the 5' untranslated region. The coding region of genomic DNA for the 5'
untranslated portion of mRNA was found in five exons containing a total of
645 nucleotides, scattered over a 50-kb region. The authors concluded that
these heterogeneous species of mRNA occurred because of differential splic-
ing of the transcripts of these exons (Figure 3). For example, it was proposed
that the 5' untranslated region of the most abundant specie of mRNA in rat
mammary gland (FL56) is formed by splicing mRNA transcripts from exons

2, 4, and 5 without incorporation of the transcripts of exons 1 and 3. The 5'
untranslated region of the second most abundant form of mRNA in rat mam-
mary gland (FL63) is formed by splicing the transcripts of exons 2, 3, 4, and
5 but not that of exon 1.
Acetyl CoA carboxylase mRNA in rat liver was found to be even more
heterogeneous than mammary gland mRNA, again attributable to differential
splicing of the transcripts of the five exons coding for the 5' untranslated
region of mRNA. 28 The most abundant mRNA species in the liver of rats fed
laboratory chow was the same as the most abundant type (FL56) found in
mammary gland in lactating rats. However, when rats were fasted and then
refed a diet high in carbohydrate to stimulate hepatic lipogenesis, a type of
mRNA (pAU) not found in the mammary gland (regardless of diet) became
the most abundant species in liver although the FL56 type was still present.
As with various mRNA species in mammary gland, the pAU type of mRNA
did not differ in the coding region from other forms, but it did differ from
other species in the length and, in part, base sequence of the 5' untranslated
region. Analysis indicated that in the pAU type of mRNA, the 5' untranslated
region was made by splicing the transcripts of exons 1, 4, and 5 (Figure 3).
That is, the transcript of exon 2 found in the FL56 type was replaced in the
pAU type by the transcript of exon 1.
These interesting observations led to the hypothesis that transcription of
exons for the 5' untranslated region of acetyl CoA carboxylase mRNA is
regulated by two different gene promoters.27. 28 It was proposed that Promoter
I is located upstream from exon 1 while Promoter II is after exon 1 but ahead
of exon 2 (Figure 3). The presence of two promoters, differentially activated
depending upon physiological conditions, could account for the formation of
dissimilar mRNA species. For example, the 5' untranslated region of the
major species of mRNA (FL56 = ACC[2:4:5] mRNA) in liver of rats fed
laboratory chow and in mammary gland of lactating rats contained the tran-
script of exon 2, indicating that Promoter II (but not I) was active. In contrast,
when rats were fasted and then refed a diet high in carbohydrate, the 5'
untranslated region of the most abundant species of mRNA (pAU = ACC[l :4:5]
mRNA) contained the transcript of exon 1, indicating that Promoter I was
activated under these conditions. In adipose tissue, pAU is by far the major
species of acetyl CoA carboxylase mRNA, regardless of whether rats are fed
a chow diet or refed a diet high in carbohydrate.
A more detailed analysis of the physiological regulation of these two
promoters, using primer extension analysis, has appeared. 29 While this method
provides only an indirect measure of promoter activity, the results indicated
that when rats were fed laboratory chow ad libitum, only basal levels of
Promoter I in adipose tissue and Promoter II in liver were found. After rats
were fasted and then refed a diet high in carbohydrate, both promoters were
activated in liver but only Promoter I was activated in adipose tissue. The
activities of promoters in both liver (Promoter I and II) and adipose tissue
(Promoter I) were depressed in streptozotocin-diabetic rats. Administration
of insulin rapidly increased Promoter I activity in adipose tissue but, sur-
prisingly, had no effect on the activity of either promoter in liver of the same
rats even after a period of 24 h where insulin was administered every 6 h.
It has been proposed 29 that acetyl CoA carboxylase mRNA species can
be classified as either class 1 (represented by the pAU type of mRNA in
Figure 3) or class 2 (represented by the FL56 type of mRNA in Figure 3),
depending upon whether they were produced in response to activation of
Promoter I or II, respectively. Activation of Promoter I in liver occurs only
under increased lipogenic conditions while Promoter II is active in liver of
chow fed rats as well as in other tissue, presumably to maintain a basal level
of acetyl CoA carboxylase. In fact, it has been suggested that Promoter II is
a "house-keeping" promoter. 30 This scenario implies that the presence of
two different promoters may be related to the dual function of acetyl CoA
carboxylase as discussed in previous sections. That is, the class 1 mRNA
transcript produced following the activation of Promoter I may be involved
in increased de novo synthesis of fatty acids destined for stored triglyceride
while the type of mRNA (class 2) produced from activation of Promoter II
may be involved in long chain fatty acyl CoA elongation and/or synthesis of
fatty acids for membrane lipids.
While this scenario represents an interesting hypothesis, it may be an
oversimplification because the reported29 promoter activity of mammary gland
and adipose tissue does not precisely fit. That is, since initiation of lactation
is accompanied by a substantial increase in de novo fatty acid biosynthesis
in the mammary gland, it might be expected on the basis of the proposed
model that Promoter I would become activated during lactation, but Promoter
I was not found to be active in mammary gland under any condition. Instead,
Promoter II activity increased during lactation. In adipose tissue, only Pro-
moter I (and not Promoter II) activity has been found, even under nonlipogenic
conditions. Further, the expression of Promoter II has been found to require
enhancer elements. 30 Enhancer sequences, which activate transcription, are
not usually found in "house-keeping" genes. 3°Finally, additional information
on Promoter II regulation has been obtained by the study of insulin-dependent
differentiation of 30A5 preadipocytes into adipocytes. 31 Differentiation of
these cells, which contain only Class 2 acetyl CoA carboxylase mRNA species
whose production is under the control of Promoter II, is accompanied by an
increased activity/quantity of acetyl CoA carboxylase and the concentration
of acetyl CoA carboxylase mRNA. 32 Insulin apparently activates Promoter
II, but such acute hormonal regulation of "house-keeping genes" would not
be expected.
Surprisingly, insulin induction of acetyl CoA carboxylase and cell dif-
ferentiation required prior exposure of the 30A5 preadipocytes to cyclic-
AMP. 31 Specific regions of genomic DNA were shown to be required for this
cyclic-AMP effect. It is not clear that these results are relevant to the regulation
of acetyl CoA carboxylase mRNA production in cells of growing and adult
rats because hormonal effects during differentiation may be quite different
than their effects on differentiated cells. Semenkovich et al. 33 have shown
that the primary effect of insulin on lipoprotein lipase during the differentiation
of 3T3 preadipocytes into adipocytes was at the transcriptional level, but in
differentiated adipocytes, insulin had no effect on transcription. Rather, in
differentiated cells, exposure to insulin affected lipoprotein lipase production
only at the translational and posttranslational levels. Still, Swierczynski et
al. 34 showed that triiodothyroxine-stimulated accumulation of acetyl CoA
carboxylase mRNA in cultured chicken embryo hepatocytes, as well as in-
creased concentrations of mRNA of other lipogenic enzymes, was blocked
by protein kinase inhibitors. The authors concluded that ongoing protein
phosphorylation was specifically required for triiodothyroxine to stimulate
transcription of genes of lipogenic enzymes.
It is clear from these reports that regulation of acetyl CoA carboxylase
mRNA synthesis is not simple. Understanding of this complex process has
been made even more difficult by evidence that there are isozymic forms of
acetyl CoA carboxylase.
V. EVIDENCE OF ISOZYMIC FORMS OF ACETYL

CoA CARBOXYLASE
While the heterogeneous forms of acetyl CoA carboxylase mRNA found

in rat tissue discussed above differ in the noncoding region, Kong et al. 23
found rat tissue acetyl CoA carboxylase mRNA which does differ in base
composition in the coding region. Specifically, e-DNA clones were isolated
with and without 24 bases which code for 8 amino acids corresponding to
residues 1189 through 1196 (Figure 2). Further, it was shown that the quantity
of mRNA coding for the "long" (2345 amino acid residues) and "short"
(2337 amino acids) forms of acetyl CoA carboxylase varied among tissues
and physiological conditions. The "short" form of mRNA accounted for
about 10% of the total acetyl CoA carboxylase mRNA in liver whether rats
were fed laboratory chow or fasted rats were refed a diet high in carbohydrate.
In adipose tissue from rats fed a chow diet, the "short" form was 20% of
the total acetyl CoA carboxylase mRNA but increased to about 36% of the
total adipocyte acetyl CoA carboxylase mRNA in response to refeeding the
diet high in carbohydrate. In the mammary gland of lactating rats, the "short"
form of mRNA accounted for about 70% of total acetyl CoA carboxylase
mRNA. Assuming that the two types of acetyl CoA carboxylase mRNA are
translated with somewhat equal efficiency in vivo, these results indicate that
rat tissues contain isozymic forms of acetyl CoA carboxylase.
Work in our laboratory has shown that cytosolic rat liver acetyl CoA
carboxylase in purified35 •36 as well as crude 36 preparations occurs as a double
band after SDS-PAGE, indicating the presence of two forms of the enzyme
which differ in subunit molecular weight. That observation has been con-
firmed by Thampy 37 who concluded that the two forms were isozymes on the
basis of antibody reactivity and tissue distribution. More recently, Bianchi et
al. 38 also concluded that these two subunit molecular weight forms of acetyl
CoA carboxylase represent isozymes of the enzyme, based on differences in
binding by two different monoclonal antibody preparations. It was reported
that while liver contained both subunit forms, heart and skeletal muscle con-
tained only the larger subunit form and adipose tissue contained only the
smaller form.
While the reports by Thampy, 37 Kong et al., 23 and Bianchi et al. 38 all
concluded that there are isozymic forms of acetyl CoA carboxylase, a more
detailed analysis of the observations shows very little agreement on how the
putative isozymes differ from each other. First, our observations 36 indicated
that the two high molecular weight forms of acetyl CoA carboxylase in liver
cytosol differed in mass by about 13,000 daltons (13 kDa) while Bianchi et
al. 38 and Thampy 37 estimated a 15 kDa difference. In contrast, the mass of
isozymes with and without the 8 amino acids as shown in Figure 2 would
differ by less than 1 kDa, an amount too small to be detected by SDS-PAGE
in such a large protein. Second, tissue distribution of the isozymic forms were
markedly different. For example, while Bianchi et al. 38 found only one iso-
zymic form in rat adipose tissue, Kong et al. 23 reported that adipocytes con-
tained substantial amounts of mRNA for both isozymes. Finally, the mono-
clonal antibodies used by Bianchi et al. 38 to differentiate the two isozymic
forms of the enzyme were raised against synthetic peptides containing only
the 15 amino acids in theN terminal or C terminal fragments of acetyl CoA
carboxylase where the amino acid composition of the two isozymes reported
by Kong et al. 23 do not differ (Figure 2). In spite of the lack of agreement
on specifics, the available evidence does suggest that rat tissues contain
isozymic forms of acetyl CoA carboxylase which may have different tissue
distribution and regulatory properties.
VI. POSSIBLE PHYSIOLOGICAL RELEVANCE

OF MULTIPLE FORMS OF mRNA CODING FOR
ACETYL CoA CARBOXYLASE
Although it is clear that there are multiple forms of acetyl Co A carboxy lase
mRNA in tissue, much work remains to be done before the physiological
relevance of them can be understood. In the case of the species of acetyl Co A
carboxylase mRNA which differ only in the 5' untranslated region, translation
of each of them would produce enzymes with identical primary structure.
Since translation of mRNA occurs from the 5' to 3' direction, these 5' regions
are upstream from the coding section. In fact, the protein synthesis initiation
codon, AUG, is located within exon 5 (Figure 3). The function of the 5'
untranslated region of mRNA is not known but it is possible that the length
and base sequence of this section, while not affecting amino acid sequence
of the protein, may affect interaction with ribosomes. In the case of acetyl
CoA carboxylase mRNA species, differences in the 5' untranslated region
were found to affect translation in an in vitro rabbit reticulocyte protein
synthesis assay. 29 With that system, translation was about I 0 times better with
the pAU type of mRNA than with the FL56 type.
If this difference in translation efficiency occurs in vivo, the cell could
effectively control the rate of translation simply by regulating relative amounts
of different types of mRNA produced. Further, from an investigative view-
point, this would mean that experiments to determine the relative importance
of transcriptional vs. posttranscriptional processes in the control of enzyme
synthesis are difficult to interpret unless relative amounts of acetyl CoA
carboxylase mRNA differing in the 5' untranslated region are determined.
That is, a multifold increase in the quantity of class 2 type of acetyl CoA
carboxylase mRNA might produce no more enzyme protein per unit of time
than would be obtained from a much smaller increase in the quantity of class
l type of acetyl CoA carboxylase mRNA. Unfortunately, simple Northern
blot analysis which is most often used for determining dietary/hormonal effects
on acetyl CoA carboxylase mRNA levels does not differentiate between class
l and class 2 types of mRN A.
Similarly, simple Northern blot analysis does not differentiate between
the two acetyl CoA carboxylase mRNAs which code for isozymic forms of
the enzyme. Yet, the relative amounts of the two forms differ among tissues
and their relative concentrations vary under different physiological condi-
tions. 23 The ratio of the two isozymic forms of acetyl CoA carboxylase may
be very important because their regulatory properties differ. It has been shown
that deletion of the eight amino acids from the larger form of acetyl CoA
carboxylase allowed cyclic-AMP-dependent protein kinase to catalyze the
phosphorylation of the serine residue located at position 1200 which makes
the enzyme more dependent upon citrate for activation. 23
VII. REGULATION OF ACETYL CoA CARBOXYLASE:

RELATIVE IMPORTANCE OF TRANSCRIPTION,
TRANSLATION, AND POSTTRANSLATIONAL
PROCESSING
There is an abundance of evidence indicating that the quantity of the

active forms of acetyl CoA carboxylase increases in liver cytosol under in-
creased lipogenic conditions, e.g., when fasted rats are refed a diet high in
carbohydrate. Indeed, over two decades ago, Nakanishi and Numa 14 showed
that fasting of rats decreased acetyl CoA carboxylase quantity and its rate of
synthesis, as determined by the incorporation of injected [3 H]-leucine into
immunologically precipitated cytosolic acetyl CoA carboxylase, relative to
that of fed control rats. Refeeding a diet high in carbohydrate markedly

increased both the quantity and the apparent rate of synthesis of the cytosolic
enzyme.
Dietary/hormonal regulation of the synthesis of active acetyl CoA car-
boxylase could occur at any of several points in the multistep process. Reg-
ulation could occur either at the level of acetyl CoA carboxylase mRNA
synthesis (transcription), synthesis of the nascent protein (translation), or
during subsequent steps which converts the nascent protein to the active
holoenzyme. The simplest hypothesis to account for the increased quantity
of enzyme in liver of fasted/refed rats is that dietary/hormonal conditions
which increase lipogenesis result in a higher rate of transcription, leading to
a higher steady state concentration of acetyl CoA carboxylase mRNA. This
hypothesis assumes that the production of acetyl CoA carboxylase mRNA is
rate limiting and that translation and posttranslational processing are relatively
fast.
Data from reports where the activity/quantity of cytosolic acetyl CoA
carboxylase and mRNA concentration are measured at a point in time after
refeedings are consistent with the hypothesis that regulation of acetyl CoA
carboxylase can occur, at least in part, at the transcriptional level. Pape et
al. 39 confirmed previous reports that fasting (48 h) decreased and refeeding
a fat-free diet (72 h) markedly increased acetyl CoA carboxylase activity in
liver cytosol relative to that found in liver of fed controls. Northern blot
analysis of RNA from liver of the same rats, using a nick-translated e-DNA
probe for acetyl CoA carboxylase mRNA, showed very little of acetyl CoA
carboxylase mRNA in liver of chow-fed rats, no detectable acetyl CoA car-
boxylase mRNA in liver of fasted rats, and relatively large amounts in liver
of rats refed the fat-free diet. While the paper did not report quantitative
amounts of acetyl CoA carboxylase mRNA under any dietary conditions (e.g.,
from densitometer scanning of X-ray film from Northern blots), the authors
concluded that while regulation of acetyl CoA carboxylase synthesis may
occur in part by transcriptional control, posttranscriptional processing may
also be involved because acetyl CoA carboxylase activity was only three times
greater in liver of refed than fed rats while the increase in acetyl CoA car-
boxylase mRNA concentration was much larger.
An almost identical experiment was conducted by Batenburg and Whitsett')
using an acetyl CoA carboxylase e-DNA probe provided by Kim (i.e., the
same probe as used by Pape et al. 39 ), except that the relative concentration
of acetyl CoA carboxylase mRNA was determined by densitometer scanning
of the X-ray film from Northern blots. In this study, acetyl CoA carboxylase
mRNA was found to increase by only 2.5-fold in liver of rats refed a fat-free
diet (72 h) relative to that in liver of fed control rats, whereas Pape et al. 39
concluded that there was a much larger increase in acetyl CoA carboxylase
mRNA in response to refeeding a similar diet for the same time period.
Further, in contrast to the results reported by Pape et al. who found that acetyl
CoA carboxylase mRNA was not detectable in liver of fasted (48 h) rats, 39
Batenburg and Whitsett40 reported that fasting (48 h) resulted in only a small,
statistically insignificant decrease in rat liver acetyl CoA carboxylase mRNA
relative to that in liver from fed control animals. Unfortunately, Batenburg
and Whitsett40 did not report changes in acetyl CoA carboxylase quantity or
activity in response to fasting, but Pape et al. 39 reported that fasting rats for
48 h resulted in a 96% decrease in both acetyl CoA carboxylase quantity and
activity in liver relative to that in liver of fed control rats. Thus, it is highly
unlikely that the level of acetyl CoA carboxylase mRNA regulates enzyme
synthesis during fasting, but the data in these reports are compatible with the
hypothesis that refeeding increases transcription of the gene for acetyl CoA
carboxylase, although neither report indicated which of the multiple forms of
acetyl CoA carboxylase mRNA were affected.
Data in these reports 39 •40 showed that both acetyl CoA carboxylase mRNA
and enzyme activity/quantity were increased after rats were refed a diet high
in carbohydrate. In each case, however, acetyl CoA carboxylase mRNA was
measured only at a single point in time (72 h) after refeeding. Such an
experiment does not indicate whether the rise in acetyl CoA carboxylase
activity/quantity was due to increased acetyl CoA carboxylase mRNA. The
relative importance of each potential regulatory mechanism can better be
assessed by determining the time-dependent relationships of transcription rate,
acetyl CoA carboxylase mRNA levels, and quantity of active enzyme. When
Katsurada et al. 41 conducted such an experiment, the results (Figure 4) were
not consistent with the hypothesis of simple transcriptional control. They
found that the rate of acetyl CoA carboxylase mRNA synthesis increased
significantly (p< 0.05) within an hour after refeeding a fat-free, high car-
bohydrate diet and remained at a similarly high rate until 8 h after refeeding.
Meanwhile, acetyl CoA carboxylase mRNA concentration steadily increased
over the first 8 h and then remained relatively constant through 24 h of
refeeding. Over the next 24 and 48 h, acetyl CoA carboxylase mRNA con-
centration steadily dropped to reach a level at 72 h after refeeding that was
significantly (p< 0.05) below the 8- to 24-h plateau. In contrast, cytosolic
acetyl CoA carboxylase activity (and quantity as judged by immunoreactive
protein) increased at a much slower rate than the rate of accumulation of
mRNA and did not reach a plateau until 48 h after refeeding. It should be
noted that at a point in time (72 h) after refeeding, both acetyl CoA carboxylase
mRNA and acetyl CoA carboxylase activity/quantity were elevated relative
to that in liver of fasted rats, in agreement with the results of Pape et al. 39
and Batenburg and Whitsett, 40 but measurement of changes in these parameters
over time leads to a different conclusion. That is, since there was a time lag
of several hours between measurable increased acetyl CoA carboxylase mRNA
and the increased enzyme activity/quantity, the data are not consistent with
the hypothesis that formation of cytosolic enzyme is limited simply by acetyl
CoA carboxylase mRNA template availability. Rather, posttranscriptional
11 ~--------------------------~
. .
I 1.~. I.
'
+
lt . . . , ..······--···
fl
1 ..
T••">-,,
~ ',.l
..
I r: I
I
,.. ...~
..~
l
0 Ww~~~~~~~~~~~~~~
0
0 8 18
Refeecllng Time (Hrs)
FIGURE 4. Effect of refeeding a high carbohydrate diet on acetyl CoA carboxylase activity,
mRNA synthesis rate, and mRNA levels in rat liver, adapted from Katsurada et al. 41 Enzyme
activity was shown to be directly proportional to quantity of immunoprecipitated enzyme.
processes appeared to be of primary importance in determining the quantity

of cytosolic acetyl CoA carboxylase in liver of fasted-refed rats.
Determination of the temporal relationships of acetyl CoA carboxylase
mRNA synthesis rates, acetyl CoA carboxylase mRNA concentration, and
acetyl CoA carboxylase activity/quantity in liver of insulin-treated strepto-
zotocin-diabetic rats led to essentially the same result as refeeding a high-
carbohydrate diet. 41 That is, acetyl CoA carboxylase mRNA synthesis rate
rapidly increased after insulin injection which led to a steady increase in acetyl
CoA carboxylase mRNA concentration. Acetyl CoA carboxylase mRNA con-
centration reached a plateau in 8 to 16 h after insulin injection and remained
at this elevated concentration through the 72-h experiment. In contrast, acetyl
CoA carboxylase activity/quantity did not change for the first 24 h postin-
jection, after which time it rose linearly for the next 48 h. Again the time
lag between the rise in acetyl CoA carboxylase mRNA concentration and the
increase in enzyme quantity indicated that insulin-dependent formation of
cytosolic acetyl CoA carboxylase is not regulated simply by the availability
of the mRNA template but rather by posttranscriptional processes.
While the studies described in the preceding paragraphs dealt with intact
rats, a number of investigations on hormonal regulation of acetyl CoA car-
boxylase have been conducted with primary cultures of hepatocytes. Use of
such cell cultures has an advantage over studies with intact animals in that
the substrate and hormonal milieu of the cells can be controlled, but have the
disadvantage that it is not certain whether results from the use of these systems
accurately reflect in vivo regulatory mechanisms. With that caveat, it appears
that insulin stimulates the synthesis of acetyl CoA carboxylase in cultured rat
liver hepatocytes. Giffhorn and Katz42 found that the activity of acetyl CoA
carboxylase per mg DNA was increased 24 h after insulin and glucose were
added to cultured hepatocytes. Acetyl CoA carboxylase synthesis, as measured
by incorporation of [35 S]-methionine into immunoprecipitated enzyme, was
also stimulated by a combination of insulin (.01 J.1M) and glucose (20 mM).
Spence et a!Y conducted a more in-depth study of regulation of acetyl
CoA carboxylase in cultured rat hepatocytes in which they determined the
time-dependent changes in acetyl CoA carboxylase activity and synthesis
rates. Enzyme activity was measured44 by a [' 4 C]-bicarbonate fixation assay
in cytosol after preincubation with 20 mM citrate for 30 min at 37°. Protein
synthesis rates were determined at each time point by measuring the amount
of [3H]-leucine in 20 min into protein precipitated by polyclonal antiacetyl
CoA carboxylase antibody. Enzyme activity began to increase immediately
after insulin was added and reached a plateau after 6 h, but the rate of protein
synthesis increased much more slowly (Figure 5). Although the quantity of
acetyl CoA carboxylase was not measured in this study, preincubation of the
enzyme preparation with citrate has been found by others 39 to negate the
effects of any short-term regulatory mechanisms, in which case activity is
directly proportional to quantity. Therefore, it is likely that the increase in
acetyl CoA carboxylase activity well before synthesis of the enzyme was
increased indicates that the initial effects of insulin were on posttranslational
processes rather than transcription or translation.
Instead of measuring acetyl CoA carboxylase mRNA by Northern blot,
Spence et al. 43 determined time-dependent changes in acetyl CoA carboxylase
mRNA template activity. Template activity ("translatable acetyl CoA car-
boxylase mRNA' ') was measured at each time point by determining the
relative synthesis rates of immunoprecipitated acetyl CoA carboxylase vs.
synthesis rates of total soluble protein in a rabbit reticulocyte-lysate assay,
using total RNA extracted from hepatocytes as template. Following insulin
exposure, the acetyl CoA carboxylase mRNA activity increased at a sub-
stantially slower rate than enzyme activity but at a considerably faster rate
than enzyme synthesis (Figure 5). The temporal relationship between enzyme
synthesis rates and acetyl Co A carboxylase mRNA template activity was much
closer in cultured hepatocytes (Figure 5) than that found between enzyme
quantity and acetyl CoA carboxylase mRNA determined by Northern blot
analysis in liver of intact rats (Figure 4). Although this may be due, at least
5 5
- "':"
4 4
ic
.11
·--
. _. ··-
••
0
.I•.J
lJoo
:;:~-
.!! 3 3
·-·
.. 1,
.I II
:sc:c 0
2 2 ~~--
~ II II.~
:loo:loo II:
.i ~
Ill
1 1
E
I
0 0
0 2 4 8 8 10
lnaulln Expoaure (Hra)
FIGURE 5. Activity, synthesis, and mRNA template activity of acetyl CoA carboxylase over
time after exposure of cultured rat hepatocytes to insulin, adapted from Spence et at"' Activity
is very likely directly proportional to enzyme quantity since it was detem1ined by the bicarbonate
fixation assay after preincubation with 20 mM citrate. Protein synthesis rates were determined
at each time point by measuring the amount of [-'H)-leucine in 20 min into protein precipitated
by polyclonal antiacetyl CoA carboxylase antibody. Template activity of mRNA was measured
at each time point by determining the relative synthesis rates of immunoprecipitated acetyl CoA
carboxylase vs. synthesis rates of total soluble protein in a rabbit reticulocyte-lysate assay, using
total RNA extracted from hepatocytes as template.
in part, to the fact that these are two different systems, it could also be argued
that measurement of acetyl CoA carboxylase mRNA template activity gives
a more accurate assessment of the relative importance of transcription and
translation in regulating the synthesis of acetyl CoA carboxylase. Unlike
results from standard Northern blot analysis, measurement of acetyl CoA
carboxylase mRNA template activity would reflect both the quantity of mRNA
and variations in translation efficiency of the heterogeneous forms of acetyl
CoA carboxylase mRNA. Thus, the data in Figure 5 could be interpreted to
mean that acetyl CoA carboxylase synthesis over several hours may be reg-
ulated by an insulin effect, at least in part, at the transcriptional level, but
again there was a substantial time lag between the rise in acetyl CoA car-
boxylase mRNA and increased acetyl CoA carboxylase synthesis.
Using similar primary cultures of rat hepatocytes, Spence et al. 43 also
made the interesting observation that incubation of cultured cells with 25 rnM
fructose (but not glucose). in the absence of insulin, had an effect on enzyme
activity, rates of enzyme synthesis, and acetyl CoA carboxylase mRNA tem-
plate activity almost identical to the insulin response, except for an initial 2-
h time lag following fructose addition before any of the three parameters
changed. Combination experiments showed that insulin and fructose effects
on all three parameters were additive. The results of this study indicate that
effects of fructose, like insulin, can be observed in primary cultures of rat
hepatocytes on both transcription and posttranscriptional processes. Studies
with intact rats have similarly indicated that feeding diets high in fructose
(but not glucose) to streptozotocin-diabetic rats had almost the same effect
on acetyl CoA carboxylase activity/quantity, 4 I.43 acetyl CoA carboxylase mRNA
template activity, 43 acetyl CoA carboxylase mRNA concentration, 41 and acetyl
CoA carboxylase synthesis rates 41 in liver as insulin injection. The mechanism
by which dietary fructose affects these parameters has not been established.
In addition to fructose, the effects of other dietary components on reg-
ulation of acetyl CoA carboxylase have been studied. Katsurada et a!. 45 com-
pared the effects in liver on activity of lipogenic enzymes, including acetyl
CoA carboxylase, after fasted rats were refed a diet containing 85% sucrose
without protein vs. rats refed a diet composed of 67% sucrose/18% casein.
In each case, the remainder of the diet was composed of cellulose (9.9%),
vitamins, and minerals. Acetyl CoA carboxylase activity increased at the
same rate for the first 2 days of refeeding and remained at this high level
(about sixfold greater than that of fasted rats) until day 4 regardless of protein
content of the diet. In contrast, maximal increased activity of fatty acid
synthase, glucose-6-phosphate dehydrogenase and malic enzyme was ob-
served only when the diet contained protein. At the end of the 4 days, [3 H]-
leucine was injected 2 h before rats were killed, and radioactivity in immu-
noprecipitated cytosolic enzymes was determined as a measure of enzyme
synthesis rates. While the presence or absence of dietary protein had no effect
on acetyl CoA carboxylase synthesis rates, the rates of synthesis of both
glucose-6-phosphate dehydrogenase and malic enzyme were significantly lower
(p< 0.01) in liver of rats refed the protein-free diet. The levels of mRNA
template activity for glucose-6-phosphate dehydrogenase and malic enzyme45
were unaffected by dietary protein, indicating that dietary protein affected
the synthesis of these two enzymes at posttranscriptional level. Subsequent
experiments, 41 in which fasted rats were refed these identical diets, confirmed
that the presence or absence of dietary protein had no effect on either the
increase in acetyl CoA carboxylase activity/quantity, acetyl CoA carboxylase
mRNA synthesis rates, or acetyl CoA carboxylase mRNA concentration.
The effects of dietary triglycerides on acetyl CoA carboxylase depend
upon the chain length and degree of unsaturation of the fatty acids present.
For example, after rats were fed a diet in which 32% of total calories were
derived from medium chain triglycerides (predominant fatty acids = c8:0 and
C 10,0 ) for 21 d, acetyl CoA carboxylase activity in liver was not different than
that in liver of rats fed a high carbohydrate (corn starch), low fat (I% corn
oil) diet. 46 In contrast, when an equivalent amount of energy was derived
from com oil, liver acetyl CoA carboxylase activity was reduced by a factor
of 2.
Other studies have shown that dietary polyunsaturated fatty acids are
much more effective than saturated or monounsaturated fatty acids in de-
pressing lipogenesis and acetyl Co A carboxylase activity. For example, dietary
safflower oil had a greater inhibitory effect on rat liver acetyl CoA carboxylase
activity than tallow47 or palmitate. 47.4 8 Emken et al. 49 compared the effect of
feeding various long chain fatty acids to mice on the activity in liver of acetyl
CoA carboxylase and other lipogenic enzymes. They found that linoleate
dramatically decreased the activity of all of the lipogenic enzymes, including
acetyl CoA carboxylase, but neither palmitate nor stearate had any effect on
the activity of acetyl CoA carboxylase or any of the other lipogenic enzymes.
The mechanism by which polyunsaturated fatty acids decrease acetyl CoA
carboxylase activity remains unknown although Katsurada et al. 41 reported
that when rats were fed a diet containing 57% sucrose, 18% casein, and 10%
com oil, acetyl CoA carboxylase activity, acetyl CoA carboxylase mRNA
synthesis, and acetyl CoA carboxylase mRNA concentration in liver were all
significantly lower (p< 0.001) than when rats were fed a diet containing 67%
sucrose and 18% casein but no fat. While these data might suggest that dietary
corn oil affected transcription, time-dependent changes in these parameters
were not reported. The authors suggested that dietary polyunsaturated fatty
acids might influence insulin secretion or insulin binding to cellular receptors.
Salati et al., 50 who studied the effects of long chain fatty acids on induction
of acetyl CoA carboxylase and other lipogenic enzymes in rat hepatocyte
cultures, similarly concluded that fatty acids might affect insulin interaction
with the hepatocytes. It should be noted, however, that the effect of feeding
long chain fatty acids to rats is different than the effects observed when
primary cultures of hepatocytes are exposed to them. When Salati and Clarke51
determined the effects of adding either 0.2 mM palmitate, oleate, linoleate,
or arachidonate to the media of primary cultures of rat hepatocytes, they found
that each of them caused a small (30 to 35%) albeit statistically significant
reduction in acetyl CoA carboxylase activity. But, in contrast to results from
studies with intact rats, the effect did not vary with either chain length or
degree of unsaturation.
VIII. SUBCELLULAR DISTRIBUTION OF ACETYL

CoA CARBOXYLASE AS A POTENTIAL
POSTTRANSLATIONAL REGULATORY MECHANISM
Witters et a!. 52 published the first report in recent years which suggested
that there may be particulate forms of acetyl CoA carboxylase. They found
substantial acetyl CoA carboxylase activity was precipitated when rat liver
homogenates were subjected to high speed centrifugation and concluded that

acetyl CoA carboxylase was associated with microsomes, the extent of which
depended upon dietary conditions. Imesch et a!. 53 drew a similar conclusion
from the study of rat adipose tissue. Subsequently, Evans and Witters 54 sep-
arated proteins in rat liver preparations by SDS-PAGE and transferred them
to nitrocellulose (Western blot) which was then probed with polyclonal and
monoclonal antiacetyl CoA carboxylase antibody. They found that only 57%
of the enzyme was soluble. Nine percent of the total acetyl CoA carboxylase
was found in low-speed (20,000 X g) and 34% was found in high-speed
(105,000 X g) precipitates. It was concluded that much of the enzyme was
microsomal although no marker enzymes were measured to confirm that
conclusion. More recently, a report from the same laboratory55 concluded that
there were no particulate forms of acetyl CoA carboxylase in rat liver, based
upon the observation that neither mitochondrial nor microsomal forms of the
enzyme were detectable on Western blots of rat liver preparations using mono-
clonal antiacetyl CoA carboxylase antibodies raised against fragments of ace-
tyl CoA carboxylase. 38
In the meantime, it was discovered that radioactive avidin would bind to
biotinylated proteins and remain bound through SDS-PAGE. 36 That obser-
vation has been confirmed by others. 38 Radioactive bands could be found by
exposing the dried gel to X-ray film, and the quantity of each biotinyl enzyme
applied to the gel could be determined by measuring the amount of radio-
activity in each band. 56 Using this method, it was concluded57 that as much
as three fourths of the total acetyl CoA carboxylase found in liver of fed rats
was particulate and that it was associated with the outer mitochondrial mem-
brane. Only a trace amount of enzyme was found in the microsome fraction.
There were two mitochondrial forms of acetyl CoA carboxylase, both of
which were precipitated by polyclonal antiacetyl CoA carboxylase antibody.
The larger of the two forms had a molecular mass similar to that of cytosolic
acetyl CoA carboxylase while the smaller was about 30 kDa less. Both of
these mitochondrial biotinyl proteins were larger by at least 100 kDa than the
three known mitochondrial biotinyl enzymes, propionyl CoA carboxylase,
pyruvate carboxylase, and 3-methyl crotonyl CoA carboxylase.
The discovery that the amount of the active cytosolic forms of acetyl CoA
carboxylase vs. the amount of the relatively inactive mitochondrial forms
varied as a function of diet led to the hypothesis that the mitochondrial forms
represented a storage reservoir of acetyl CoA carboxylase which could be
mobilized/activated under increased lipogenic conditions. 57 When rats were
fasted (48 h), the quantity of active cytosolic acetyl CoA carboxylase de-
creased while the quantity of mitochondrial forms increased, with no change
in the total quantity of acetyl CoA carboxylase. When fasted rats were refed
a high-carbohydrate diet, the quantity of the active cytosolic forms increased,
beginning about 8 h after refeeding 58 while mitochondrial forms decreased,
again with no change in the total quantity of acetyl CoA carboxylase. The
study of acute (3 d) alloxan-diabetic rats 59 indicated that mobilization/acti-
vation of mitochondrial forms of acetyl CoA carboxylase was insulin depen-
dent. Examination of the subcellular distribution of acetyl CoA carboxylase
in the liver of Zucker rats 60 indicated that obese rats had no more enzyme
than that of lean litter mates, but rather more of the enzyme was in the active
cytosolic form.
Evidence of particulate forms of two other lipogenic enzymes which are
normally active in cytosol has been found. A form of ATP-citrate lyase has
been shown to be associated with mitochondria. 61 The distribution of this
enzyme between cytosol and mitochondria is dietary dependent. 62 More re-
cently, it was concluded63 that epidermal growth factor increased the activity
of glucose-6-phosphate dehydrogenase of rat renal cortical cells by causing
its release from a particulate form within the cell into the cytosol.
Although the mechanism by which acetyl CoA carboxylase might be
bound to the outer mitochondrial membrane has not been established, several
enzymes have now been shown to be attached to the cellular plasma membrane
via a glycosylated phosphatidyl inositol linkage. 64 The discovery that acetyl
CoA carboxylase is a glycoprotein65 ·66 raises the possibility that it may be
attached to the mitochondrial membrane by a similar mechanism. Analysis
of the carbohydrates present in purified acetyl CoA carboxylase by GC-MS
confirmed a previous report 67 that the enzyme contained inositol and estab-
lished that it also contained mannose and galactose, as well as trace amounts
of N-acetyl glucosamine, fucose, and 6-deoxy mannose. It was calculated
that the combined mass of the carbohydrates amounted to about 6 kDa. While
glycosylation is most often associated with exported proteins, 64 it has been
established that another enzyme intimately involved in lipid metabolism,
HMG CoA reductase, is a glycoprotein. 68 HMG CoA reductase remains within
the cell in which it is produced and, unlike exported glycoproteins, is not
processed in the Golgi apparatus. 69
Ifrat liver contains mitochondrial storage forms of acetyl CoA carboxylase
which can be mobilized under lipogenic conditions, it could provide a mech-
anism by which cells could more rapidly respond to dietary/hormonal changes
than de novo enzyme synthesis might allow. Such a mechanism could account
for the rapid rise in enzyme quantity implied by the data in Figure 5 well
before acetyl CoA carboxylase synthesis is increased after cultured rat he-
patocytes are exposed to insulin. In this scenario, insulin stimulation of acetyl
CoA carboxylase transcription and translation would still be necessary in
order to replace mobilized mitochondrial enzyme. While the existence of
storage forms of acetyl CoA carboxylase represents a plausible hypothesis,
definitive proof of the physiological significance of particulate forms of the
enzyme has not been obtained.
IX. BIOTINYLATION
Acetyl CoA carboxylase is synthesized as an apoenzyme which must be

biotinylated to form the holoenzyme. It is generally assumed that the bioti-
nylation occurs rapidly except in biotin-deficient animals and therefore does
not play a regulatory role. 70 It is known that in the active form of acetyl CoA
carboxylase, biotin is covalently bonded to the epsilon amino group of lysine
in the tetrapeptide [val-met-lys-met] located toward the amino terminal end
of the protein at position 783 to 786 (Figure 2). 21 The biotinylation site of
chicken liver acetyl CoA carboxylase contains the same tetrapeptide. 24 In
contrast, in all other biotinyl enzymes present in tissues of higher animals
(propionyl CoA carboxylase, pyruvate carboxylase, and 3-methyl crotonyl
CoA carboxylase), the tetrapeptide sequence contains alanine instead of valine71
and is located very near the carboxyl end of the protein. 72 While it has generally
been assumed that all four biotin enzymes are biotinylated by the same en-
zyme, 70 studies involving structural alterations of acetyl CoA carboxylase
suggest that this may not be the case. When site-directed mutagenesis was
used to change the coding sequence of e-DNA so that alanine was substituted
for the valine located at position 783, biotinylation of the protein expressed
in a bacterial system was abolished. 22 When the amino acids surrounding the
[ala-met-lys-met] site of acetyl CoA carboxylase were further altered to make
a protein more closely resembling the structure of the other biotinyl enzymes,
biotinylation of the lysine did occur. This suggests that acetyl CoA carboxylase
and the other biotinyl enzymes may be biotinylated by distinct enzymes,
differing in specificity.
X. SHORT-TERM REGULATORY MECHANISMS
A number of short-term mechansism which may regulate the activity of

acetyl CoA carboxylase have been found, including covalent modification,
citrate activation, and enzyme polymeric state. Most of the information about
these mechanisms lie outside the scope of this review with the exception that
there is evidence of interaction of these regulatory mechanisms with gene
expression.
Shortly after the initial report that acetyl CoA carboxylase activity could
be affected by its phosphorylation state/3 it was discovered that derivatives
of cyclic-AMP rapidly inhibited lipogenesis in raf4 and chicken75 liver slices
and that the addition of cyclic-AMP decreased the activity of acetyl CoA
carboxylase in rat liver homogenates. 76 These observations have been con-
firmed in numerous subsequent reports. It has now been shown that acetyl
CoA carboxylase contains at least eight serine residues (Figure 2) which can
be phosphorylated by various protein kinases. 2 ' Phosphorylation of the serine
residue at position 79 is catalyzed by AMP-dependent (but not cyclic-AMP-
dependent) protein kinase which inactivates the enzyme, 77 while phosphor-

ylation of the serine at position 1200, catalyzed by cyclic-AMP-dependent
protein kinase, does not inactivate acetyl CoA carboxylase directly but rather
affects its sensitivity to citrate activation. 23 Cyclic-AMP-dependent protein
kinase was found to catalyze the phosphorylation of serine-1200 only when
the smaller (2337 amino acids) isozymic form of acetyl CoA carboxylase, in
which 8 amino acids are deleted (Figure 2), served as the source of enzyme. 23
Thus, cyclic-AMP and citrate effects depend upon which isozyme is being
studied. In tum, the polymeric state of acetyl CoA carboxylase depends upon
phosphorylation state and citrate sensitivity. 78
XI. CONCLUSION
It is clear that dietary/hormonal regulation of acetyl CoA carboxylase
ultimately invovles gene expression. It is equally apparent, however, that
regulation of the enzyme is remarkably complex. The available data are not
compatible with the hypothesis that the quantity of active forms of the cytosolic
enzyme is solely dependent upon the rate of gene transcription. Rather, di-
etary/hormonal regulation of acetyl CoA carboxylase can occur at the tran-
scriptional, translational, and posttranslationallevels.
REFERENCES
I. Volpe, J. J. and Vagelos, P.R., Mechanisms and regulation of biosynthesis of saturated
fatty acids, Physiol. Rev.. 56, 339, 1976.
2. Gibson, D. M., Lyons, R. T., Scott, D. F., and Muto, Y., Synthesis and degradation
of the lipogenic enzymes of rat liver, Adv. Enzyme Regul., 10, 187, 1970.
3. Brady, L., Regulation of hepatic camitine palmitoyltransferase, Nutrition and Gene
Expression, CRC Press, Boca Raton, FL, 1992.
4. Hellerstein, M. K., Christiansen, M., Kaempfer, S., Kletke, C., Wu, K., Reid,
J, S., Mulligan, K., Hellerstein, N. S., and Shackleton, C. H. L., Measurement of
de novo hepatic lipogenesis in humans using stable isotopes, J. Clin. Invest., 87, 1841,
1991.
5. Mead, J. F., Alfin-Siater, R. B., Howton, D. R., and Popjak, G., Lipids: Chemistry,
Biochemistry and Nutrition, Plenum Press, New York, 1986, chaps. 8 and 9.
6. Needleman, P., Turk, J., Jakschik, B. A., Morrison, A. R., and Lefkowith, J. B.,
Arachidonic acid metabolism, Annu. Rev. Biochem., 55, 69, 1986.
7. Bijleveld, C., Vaartjes, W. J., and Geelen, M. J,, Time course of hormonal effects
on acetyl-CoA carboxylase as measured in digitonin-permeabilized rat hepatocytes, Horm.
Metab. Res., 21,602,1989.
8. Milatovich, A., Plattner, R., Heerema, N. A., Palmer, C. G., Lopez-Casillas, F.,
and Kim, K.-H., Localization of the gene for acetyl-CoA carboxylase to human chro-
mosome 17, Cytogenet. Cell. Genet., 48, 190, 1988.
9. Blom, W., DeMuinck-Keizer, S.M. P. F., and Scholte, H. R., Acetyl CoA c.arboxylase
deficiency: an inborn error of de novo fatty acid synthesis, New Engl. J. Med., 305, 465,
1981.
10. Shrago, E., Spennetta, T., and Gordon, E., Fatty acid synthesis in human adipose
tissue, J. Bioi. Chern., 244, 2761, 1969.
II. Passmore, R. and Swindells, Y. E., Observations on the respiratory quotients and weight
gain of man after eating large quantities of carbohydrate, Brit. J. Nutr., 17, 331, 1963.
12. Acheson, K. J,, Flatt, J, P., and Jequier, E., Glycogen synthesis versus lipogenesis
after a 500 gram carbohydrate meal in man, Metabolism, 31, 1234, 1982.
13. Tepperman, H. M. and Tepperman, J,, Adaptive hyperlipogenesis, Fed. Proc., 23,
73, 1964.
14. Nakanishi, S. and Numa, S., Purification of rat liver acetyl CoA carboxylase and
immunochemical studies on its synthesis and degradation, Eur. J. Biochem., 16, 161,
1970.
15. Buckley, M. G. and Rath, E. A., Regulation of fatty acid synthesis and malonyl-CoA
content in mouse brown adipose tissue in response to cold-exposure, starvation or re-
feeding, Biochem. J., 243, 437, 1987.
16. Mersmann, H. J, and Koong, L.-J., Effect of plane of nutrition on adipose tissue
metabolism in genetically obese and lean pigs, J. Nutr., 114, 862, 1986.
17. Smith, S. B., Prior, R. L., Ferrell, C. L., and Mersmann, H. J,, Interrelationships
among diet, age, fat deposition and lipid metabolism in growing steers, J. Nutr., 114,
153, 1984.
18. Lopez-Casillas, F., Pape, M. E., Bai, D. H., Kuhn, D. N., Dixon, J, E., and Kim,
K. H., Preparation of functional acetyl-CoA carboxylase mRNA from rat mammary
gland, Arch. Biochem. Biophys., 257, 63, 1987.
19. Bai, D. H., Pape, M. E., Lopez-Casillas, F., Luo, X. C., Dixon, J, E., and Kim,
K. H., Molecular cloning of eDNA for acetyl-coenzyme A carboxylase, J. Bioi. Chern.,
261, 12395, 1986.
20. Lopez-Casillas, F., Bai, D. H., Luo, X. C., Kong, I. S., Hermodson, M. A., and
Kim, K. H., Structure of the coding sequence and primary amino acid sequence of acetyl-
coenzyme A carboxylase, Proc. Nat/. Acad. Sci. U.S.A., 85, 5784, 1988.
21. Davies, S. P., Sim, A. T., and Hardie, D. G., Location and function of three sites
phosphorylated on rat acetyl-CoA carboxylase by the AMP-activated protein kinase, Eur.
J. Biochem., 187, 183, 1990.
22. Bai, D. H., Moon, T. W., Lopez-Casillas, F., Andrews, P. C., and Kim, K. H.,
Analysis of the biotin-binding site on acetyl-CoA carboxylase from rat, Eur. J. Biochem.,
182, 239, 1989.
23. Kong, I. S., Lopez-Casillas, F., and Kim, K. H., Acetyl-CoA carboxylase mRNA
species with or without inhibitory coding sequence for Ser-1200 phosphorylation, J. Bioi.
Chern., 265, 13695, 1990.
24. Takai, T., Wada, K., and Tanabe, T., Primary structure of the biotin-binding site of
chicken liver acetyl-CoA carboxylase, FEBS. Lett., 212, 98, 1987.
25. Takai, T., Yokoyama, C., Wada, K., and Tanabe, T., Primary structure of chicken
liver acetyl-CoA carboxylase deduced from eDNA sequence, J. Bioi. Chern., 263, 2651,
1988.
26. Lopez-Casillas, F., Luo, X. C., Kong, I. S., and Kim, K. H., Characterization of
different forms of rat mammary gland acetyl-coenzyme A carboxylase mRNA: analysis
of heterogeneity in the 5' end, Gene, 83,311, 1989.
27. Luo, X. C., Park, K., Lopez-Casillas, F., and Kim, K. H., Structural features of the
acetyl-CoA carboxylase gene: mechanisms for the generation of mRNAs with 5' end
heterogeneity, Proc. Nat/. Acad. Sci. U.S.A., 86, 4042, 1989.
28. Lopez-Casillas, F. and Kim, K. H., Heterogeneity at the 5' end of rat acetyl-coenzyme
A carboxylase mRNA. Lipogenic conditions enhance synthesis of a unique mRNA in
liver, J. Bioi. Chern., 264, 7176, 1989.
29. Lopez-Casillas, F., Ponce-Castaneda, M. V., and Kim, K. H., In vivo regulation of
the activity of the two promoters of the rat acetyl coenzyme-A carboxylase gene, En-
docrinology, 129, 1049, 1991.
30. Luo, X. C. and Kim, K. H., An enhancer element in the house-keeping promoter for
acetyl-CoA carboxylase gene. Nucleic Acids Res., 18, 3249, 1990.
31. Park, K. and Kim, K. H., Regulation of acetyl-CoA carboxylase gene expression:
Insulin induction of acetyl-CoA carboxylase and differentiation of 30A5 preadipocytes
require prior cAMP action on the gene, J. Bioi. Chern., 266. 12249, 1991.
32. Pape, M. E. and Kim, K. H., Transcriptional regulation of acetyl coenzyme A car-
boxylase gene expression by tumor necrosis factor in 30A-5 preadipocytes, Mol. Cell.
Bioi., 9, 974,1989.
33. Semenkovich, C. F., Wims, M., Noe, L., Etienne, J., and Chan, L., Insulin regulation
of lipoprotein lipase activity in 3T3-Ll adipocytes is mediated at the posttranscriptional
and posttranslationallevels, J. Bioi. Chern., 264. 9030, 1989.
34. Swierczynski, J., Mitchell, D. A., Reinhold, D. S., Salati, L. M., Stapleton, S. R.,
Klautky, S. A., Struve, A. E., and Goodridge, A. G., Triiodothyronine-induced ac-
cumulations of malic enzyme, fatty acid synthase, acetyl-coenzyme A carboxylase, and
their mRNAs are blocked by protein kinase inhibitors. Transcription is the affected step,
J. Bioi. Chern., 266, 17459, 1991.
35. Allred, J. B., Harris, G. J., and Goodson, J. Regulation of purified rat liver acetyl
CoA carboxylase by phosphorylation, J. Lipid Res., 24, 449, 1983.
36. Goodson, J., Pope, T. S., and Allred, J. B., Molecular weights of subunits of acetyl
CoA carboxylase in rat liver cytoplasm. Biochern. Biophys. Res. Cornrnun., 122, 694.
1984.
37. Thampy, K. G., Formation of malonyl coenzyme A in rat heart. Identification and
purification of an isozyme of acetyl CoA carboxylase from rat heart, J. Bioi. Chern.,
264, 17631, 1989.
38. Bianchi, A., Evans, J. L., Iverson, A. J., Nordlund, A. C., Watts, T. D., and
Witters, L. A., Identification of an isozymic form of acetyl-CoA carboxylase, J. Bioi.
Chern., 265, 1502, 1990.
39. Pape, M. E., Lopez-Casillas, F., and Kim, K. H., Physiological regulation of acetyl-
CoA carboxylase gene expression: effects of diet, diabetes, and lactation on acetyl-CoA
carboxylase mRNA, Arch. Biochern. Biophys., 267, 104, 1988.
40. Batenburg, J. J. and Whitsett, J. A., Levels of mRNAs coding for lipogenic enzymes
in rat lung upon fasting and refeeding and during perinatal development, Biochirn. Bio-
phys. Acta., 1006, 329, 1989.
41. Katsurada, A., Iritani, N., Fukuda, H., Matsumura, Y., Nishimoto, N., Noguchi,
T., and Tanaka, T., Effects of nutrients and hormones on transcriptional and post-
transcriptional regulation of acetyl-CoA carboxylase in rat liver, Eur. J. Biochern., 190,
435, 1990.
42. Giftlmrn, S. and Katz, N. R., Glucose-dependent induction of acetyl-CoA carboxylase
in rat hepatocyte cultures, Biochern. J., 221, 343, 1984.
43. Spence, J. T., Koudelka, A. P., and Tseng-Crank, J. C., Role of protein synthesis
in the carbohydrate-induced changes in the activities of acetyl CoA carboxylase and
hydroxymethylglutaryl CoA reductase in cultured rat hepatocytes, Biochern. J., 227, 939,
1985.
44. Craig, M. C., Dugn, R. E., Muesing, R. A., Slakey, L. L., and Porter, J. W.,
Comparative effects of dietary regimen on the levels of enzymes regulating the synthesis
of fatty acids and cholesterol in rat liver, Arch. Biochern. Biophys., 151, 128, 1972.
45. Katsurada, A., Iritani, N., Fukuda, H., Noguchi, T., and Tanaka, T., Effects of
dietary nutrients on lipogenic enzyme and mRNA activities in rat liver during induction,
Biochirn. Biophys. Acta., 877, 350, 1986.
46. Chanez, M., Bois-Joyeux, B., Arnaud, M. J., and Peret, J., Metabolic effects in rats
of a diet with a moderate level of medium-chain triglycerides, J. Nutr., 121, 585, 1991.
47. Wilson, M. D., Hays, R. D., and Clarke, S. D., Inhibition of liver lipogenesis by
dietary polyunsaturated fat in severely diabetic rats, J. Nutr., 116, 1511, 1986.
48. Clarke, S. D., Wilson, M. D., and Ibnoughazala, T., Resistance of lung fatty acid
synthesis to inhibition by dietary fat in the meal-fed rat, J. Nutr., 114, 598, 1984.
49. Emken, E. A., Abraham, S., and Lin, C. Y., Metabolism of cis-12-octadecenoic acid
and trans-9.trans-12-octadecadienoic acid and their influence on lipogenic enzyme activ-
ities in mouse liver. Biochim. Biophys. Acta., 919. Ill, 1987.
50. Salati, L. M., Adkins-Finke, B., and Clarke, S. D., Free fatty acid inhibition of the
insulin induction of glucose-6-phosphate dehydrogenase in rat hepatocyte monolayers,
Lipids, 23, 36. 1988.
51. Salati, L. M. and Clarke, S. D., Fatty acid inhibition of hormonal induction of acetyl-
coenzyme A carboxylase in hepatocyte monolayers. Arch. Biochem. Biophys.. 246, 82,
1986.
52. Witters, L. A., Friedman, S. A., and Bacon, G. W., Microsomal acetyl CoA car-
boxylase: evidence for association of enzyme polymer with liver microsomes. Proc. Nat/.
Acad. Sci. U.S.A., 78, 3639, 1981.
53. Imesch, E., Wolczunowicz, M., and Rous, S., Enzymatic activities of cytoplasmic and
of microsomal acetyl CoA carboxylase of rat epididymal adipose tissue; different regu-
latory effects of a short-term fast and palmitoyl CoA on these two enzymes, Int. J.
Biochem., 15, 977, 1983.
54. Evans, J. L. and Witters, L.A., Quantitation by immunoblotting of the in vivo induction
and subcellular distribution of hepatic acetyl-CoA carboxylase, Arch. Biochem. Biophys.,
264, 103, 1988.
55. Iverson, A. J., Bianchi, A., Nordlund, A. C., and Witters, L. A., Immunological
analysis of acetyl-CoA carboxylase mass, tissue distribution and subunit composition,
Biochem. J., 269, 365, 1990.
56. Roman-Lopez, C. R., Goodson, J., and Allred, J. B., Determination of the quantity
of acetyl CoA carboxylase by [14C]methyl avidin binding, J. Lipid Res., 28, 599, 1987.
57. Allred, J, B. and Roman-Lopez, C. R., Enzymatically inactive forms of acetyl-CoA
carboxylase in rat liver mitochondria, Biochem. J., 251, 881, 1988.
58. Roman-Lopez, C. R., Shriver, B. J., Joseph, C. R., and Allred, J. B., Mitochondrial
acetyl-CoA carboxylase: time course of mobilization/activation in liver of refed rats,
Biochem. J., 260, 927, 1989.
59. Roman-Lopez, C. R. and Allred, J. B., Acute alloxan diabetes alters the activity but
not the total quantity of acetyl CoA carboxylase in rat liver, J. Nutr., 117, 1976, 1987.
60. Allred, J. B., Roman-Lopez, C. R., Jurin, R. R., and McCune, S. A., Mitochondrial
storage forms of acetyl CoA carboxylase: mobilization/activation accounts for increased
activity of the enzyme in liver of genetically obese Zucker rats, J. Nutr., 119, 478, 1989.
61. Janski, A.M. and Cornell, N. W., Association of ATP citrate lysase with mitochondria,
Biochem. Biophys. Res. Commun., 92, 305, 1980.
62. Cornell, N. W., Janski, A.M., and Rendon, A., Compartmentation of enzymes: ATP
citrate lyase in hepatocytes from fed and fasted rats, Fed. Proc., 44, 2448, 1985.
63. Stanton, R. C., Seifters, J. L., Boxer, D. C., Zimmerman, E., and Cantley, L. C.,
Rapid release of bound glucose-6-phosphate dehydrogenase by growth factors. Correlation
with increased enzymatic activity, J. Bioi. Chern., 266, 12442, 1991.
64. Low, M. G., Biochemistry of the glycosyl-phosphatidylinositol membrane protein an-
chors, Biochem. J., 244, I, 1987.
65. Bowers, D. F. and Allred, J. B., Acetyl CoA carboxylase is a glycoprotein, FASEB
J., 5, A462, 1991.
66. Bowers, D. F., Identification and Characterization of Acetyl CoA Carboxylase as a
Glycoprotein, Ph.D. dissertation, The Ohio State University, Columbus, 1991.
67. Heger, H. W. and Peter, H. W., Phosphatidylinositol as an essential constituent of the
acetyl CoA carboxylase from rat liver, Int. J. Biochem., 8, 841, 1977.
68. Liscum, L., Cummings, R. D., Anderson, R. G. W., DeMartino, G. N., Goldstein,
J. L., and Brown, M.S., 3-Hydroxy-3-methylglutaryl CoA reductase: a transmembrane
glycoprotein of the endoplasmic reticulum with N-linked "high-mannose" oligosac-
charide, Proc. Nat/. Acad. Sci. U.S.A .. 80, 7165, 1983.
69. Haro, D., Marrero, P. F., Ayte, J,, and Hegardt, F. G., Identification of a cholesterol-
regulated 180-kDa microsomal protein in rat hepatocytes, Eur. 1. Biochern .. 188, 123,
1990.
70. Moss, J, and Lane, M. D., The biotin-dependent enzymes, Adv. Enzyrnol .. 35, 321,
1971.
71. Shenoy, B. C., Paranjape, S., Murtif, V. L., Kumar, G. K., Samols, D., and Wood,
H. G., Effect of mutations at met-88 and met-90 on the biotinylation of lys-89 of the
apo 1.3S subunit of transcarboxylase, FASEB J., 2, 2505, 1988.
72. Lamhonwah, A.-M., Quan, F., and Gravel, R. A., Sequence homology around the
biotin-binding site of human propionyl CoA carboxylase and pyruvate carboxylase, Arch.
Biochern. Biophys.. 254, 631, 1987.
73. Carlson, C. A. and Kim, K.-H., Regulation of hepatic acetyl CoA carboxylase by
phosphorylation and dephosphorylation, 1. Bioi. Chern., 248, 378, 1973.
74. Allred, J, B. and Roehrig, K. L., Inhibition of hepatic lipogenesis by cyclic-3', 5'-
nucleotide monophosphate, Biochern. Biophys. Res. Cornrnun., 46, 1135, 1972.
75. Allred, J, B. and Roehrig, K. L., Metabolic oscillations and food intake, Fed. Proc ..
32, 1727, 1973.
76. Allred, J, B. and Roehrig, K. L., Inhibition of rat liver acetyl CoA carboxylase in
N6, 0 2-dibutyryl cyclic-3', 5' -adenosine monophosphate in vitro, J. Bioi. Chern., 248,
4131, 1973.
77. Haystead, T. A., Moore, F., Cohen, P., and Hardie, D. G., Roles of the AMP-
activated and cyclic-AMP-dependent protein kinases in the adrenaline-induced inacti-
vation of acetyl-CoA carboxylase in rat adipocytes, Eur. J. Biochern., 187, 199, 1990.
78. Thampy, K. G. and Wakil, S. J,, Regulation of acetyl-coenzyme A carboxylase. II.
Effect of fasting and refeeding on the activity, phosphate content, and aggregation state
of the enzyme, J. Bioi. Chern., 263, 6454, 1988.
Chapter 13
APOLIPOPROTEIN B mRNA EDITING
Nicholas 0. Davidson
TABLE OF CONTENTS
l. Overview and Definition ......................................... 298
II. Tissue-Specific Distribution and Developmental

Regulation ....................................................... 299
A. Mammalian apo B mRNA Editing ....................... 299
B. Apo B mRNA Editing in non-apo B-Producing
Cells and Tissues ......................................... 301
C. Topology of apo B mRNA Editing ....................... 303
III. Modulation of apo B mRNA Editing by Hormonal and

Nutritional Factors ............................................... 304
A. Modulation by Thyroid Hormone ......................... 304
B. Modulation of apo B mRNA Editing Following
Changes in Hepatic Lipogenesis .......................... 307
C. Modulation of apo B mRNA Editing by
Other Manipulations ...................................... 308
IV. In Vitro Editing of apo B mRNA ................................ 309

A. Methodological Considerations ........................... 309
B. Sequence Requirements for apo B mRNA
Editing ................................................... 311
V. Summary and Conclusions ....................................... 312
References ............................................................... 313
0·8493-6961-4/93/$0.00 +$.50
© 1993 by CRC Pre». 1nc. 297
I. OVERVIEW AND DEFINITION
Apolipoprotein B (apo B) is a large hydrophobic protein synthesized in

the liver and small intestine of mammals. Apo B serves an essential although
incompletely understood role in the assembly and secretion of triglyceride-
rich lipoproteins (chylomicrons and very low density lipoproteins) and also
functions in the catabolic clearance of low density lipoprotein (LDL), the
major transport vehicle of plasma cholesterol in humans. Mammalian apo B
is the product of a single gene which maps to the p23-p24 region of chro-
mosome 2 1•1 (for detailed reviews see References 3 and 4). Apo B mRNA is
expressed and processed in a tissue-specific fashion, the molecular and cellular
basis for which will be the focus of this chapter. As currently viewed, one
form of apo B is synthesized in the human liver as a protein of 4536 amino
acids with a relative molecular mass of 512,000. This form of the protein is
referred to on a centile scale as apo B 100. 59 By contrast, the intestinal form
of the protein contains 2152 amino acids, is colinear with the amino terminal
half of apo B 100, and is referred to as apo B48. 10 • 11 Several important struc-
tural domains have been identified for apo B based upon eDNA sequence
and monoclonal epitope mapping. The most important of these are the LDL
receptor binding domain and the attachment site of apolipoprotein (a) which
both reside in the carboxyl terminal half of the protein. 12- 14 The biological
relevance of apo B48 may thus relate to the absence of the above domains
which are present only in the full length protein. Apo B48 is found in the
systemic circulation in association with intestinally derived lipoprotein par-
ticles, namely chylomicrons and chylomicron remnants. These particles are
cleared from the plasma compartment via a (principally hepatic) receptor
which has been partially characterized and is referred to as the LDL receptor-
related protein or LRP. 15 • 16 This receptor recognizes apolipoprotein E as its
major ligand. 17 Thus, lipoprotein particles containing apo B48 have a distinct
catabolic fate from those with apo B 100 which are cleared principally via the
LDL receptor. The implications for this observation in terms of atherosclerosis
susceptibility remain to be tested experimentally, but it suggests several models
whereby such characteristics of an intestinal particle may have evolved to
facilitate their function as an efficient delivery system for dietary triglyceride.
By contrast, the finer regulation of plasma cholesterol homeostasis may be
achieved through hepatic very low density lipoprotein (VLDL) secretion and
ultimately LDL uptake.
As alluded to above, this chapter will focus on the posttranscriptional
modification of apo B mRNA referred to as apo B mRNA editing. This
process is distinct from other co- or posttranscriptional processing events such
as capping, polyadenylation, and splicing and, until recently, was unique to
apo B as an example of posttranscriptional modification of the coding infor-
mation in a mammalian gene. Apo B mRNA editing was the first example
in vertebrates of this departure from one of the central tenets of molecular
Davidson 299
biology that DNA encodes an RNA template which is identical and which
subsequently specifies a predictable protein. Over the last few years, several
examples of RNA editing have been described in lower eukaryotes, and these
have been classified according to the underlying mechanism (for detailed
reviews, see References 18 and 19). These mechanisms include the posttran-
scriptional insertion or deletion of uridine residues in trypanosome mito-
chondrial genes which results in the production of a translationally competent
open reading frame. This form of RNA editing is of interest also because it
is mediated by small guide RNAs. 20 Other forms of RNA editing include the
posttranscriptional insertion of guanosine residues in paramyxovirus and cy-
tidine residues in physarum polycephalum. These examples are discussed in
detail in the reviews cited. 18 • 19
Apo B mRNA editing was described in 1987 when several groups si-
multaneously reported the site-specific modification of apo B mRNA as the
underlying mechansim for the production of distinct isoforms of the protein
from the human liver and small intestine. 21 . 23 Powell et al. 21 prepared intestinal
cDNAs and demonstrated that nucleotide 6666 in human and rabbit intestinal
apo B eDNA was changed from the genomically temp1ated cytidine to a
uridine residue. This change modified codon 2153 from a CAA which encodes
glutamine to UAA which specifies an in-frame stop codon. Chen et aJ.2 2
reported the identical conclusion but used a strategy consisting of tryptic
fragment analysis and direct RNA sequencing. These findings indicated that
intestinal apo B was the product of a single apo B gene in which codon 2153
is altered to produce a translational stop codon and thereby specify a truncated
apo B (apo B48) as the primary translation product. Recently, a second
example of mammalian RNA editing was described in which certain subunits
of the brain glutamate receptor transcripts undergo a single base change in
which a glutamine codon (CAG) is changed to an arginine codon (CGG). 24
The significance of the observation that both forms of RNA editing involve
a glutamine codon may be fortuitous since the base involved is different
(cytidine vs. adenosine) although the mechanism may involve a deamination
in both instances. The consequences following editing of the brain glutamate
receptor transcripts are profound since the arginine residue regulates the func-
tional characteristics of these gated receptor channels. Thus, in both situations
currently known of mammalian RNA editing, the process has enormous func-
tional consequences for gene expression. It is likely that an understanding of
RNA editing mechanisms and their evolution will provide important insight
into a novel area of regulation of gene expression and will reveal new functions
for this process.
II. TISSUE-SPECIFIC DISTRIBUTION AND

DEVELOPMENTAL REGULATION
A. MAMMALIAN APO B mRNA EDITING
Several studies have addressed the tissue-specific distribution of apo B
mRNA editing. The original descriptions 2123 were of intestinal apo B mRNA
editing since this was central to an understanding of the production of apo

B48 by the small intestinal enterocyte. Following this description, Davidson
et ai.2 5 cloned and sequenced the region of rat apo B eDNA corresponding
to the 3' terminus of apo B48 and demonstrated that apo B mRNA editing
occurs in both the rat liver and small intestine. This finding was confirmed
by others 26 and raises an important and presently unresolved question-namely,
the evolutionary and biological advantage of editing both the intestinal and
hepatic apo B transcripts. The major functional consequence of this obser-
vation is that the rat and mouse liver synthesize and secrete both apo B 100
and B48. As far as currently known, mammalian hepatic apo B synthesis and
secretion in all other species, including humans, is exclusively apo B 100. A
single report27 of apo B mRNA editing in the human liver has not been
confirmed by others and the validity of this observation is in doubt since
subsequent, independent investigation of this issue has determined that human
hepatic apo B mRNA is essentiallly all unedited. 28 Apo B mRNA editing in
the mammalian enterocyte occurs as a developmentally regulated event in
human, 28 rat, 29 and pig small intestine. 30 Teng et al. 28 demonstrated that human
intestinal apo B was more than 90% unedited in fetal small intestinal RNA
from late first trimester samples. This finding confirmed earlier preliminary
work suggesting that human fetal intestine synthesized both forms of apo B. 31
Teng et al. 28 also demonstrated that the early gestation fetal small intestine
synthesizes and secretes both apo B 100 and apo B48, indicating that the
unedited form of the apo B transcript is translationally competent in the small
intestine and, furthermore, leads to apo B100 secretion. As the small intestine
undergoes maturation during the late first and early second trimester, the
proportion of edited apo B mRNA increases such that at 19 to 20 weeks
gestation, small intestinal apo B mRNA is approximately 80 to 90% edited.
Adult small intestine was found to contain a variable quantity of unedited
apo B mRNA, varying from 3 to 19% in one series. Other studies have
independently established that the human small intestine indeed synthesizes
apo B 100. 32 •33 Of interest in this regard is the observation that although adult
small intestine synthesizes and secretes apo B100, the proportions of newly
synthesized apo B 100 to apo B48 in the tissue or the culture media do not
approach the proportions of unedited to edited apo B mRNA-there being
far less than 3% apo B 100 detectable as a percentage of total immunoreactive
apo B species. The implications of this are that the unedited transcript is
translationally active but that the determinants of apo B 100 synthesis and
secretion from the small intestine involve other, co- , or posttranslational
events which may influence the stability of the nascent apo B peptide. Such
events have been postulated to play a major role in the regulation of apo B
secretion from Hep G2 cells particularly in relation to altered lipid flux. 34
Other studies have established that rat apo B mRNA editing is devel-
opmentally regulated in both the small intestine and liver. Work by Wu et
al. 29 demonstrated that the temporal sequence of the developmental changes
Davidson 301
in apo B mRNA editing is distinct for the liver and small intestine with a
striking increase in intestinal editing prenatally while the hepatic transcript is
largely unedited until postnatal day 20. Thus, the emergence of edited apo
B mRNA in both the rat liver and intestine appears to coincide with devel-
opmental changes in triglyceride metabolism. Parallel conclusions cannot be
proposed for either human or porcine intestine since fetal hepatic and intestinal
triglyceride metabolism have yet to be characterized in these species. In regard
to the observation that apo B mRNA editing is developmentally induced in
the fetal small intestine, studies conducted in our laboratory have addressed
the question of whether apo B mRNA editing changes in the adult as enter-
ocytes migrate up the villus during their 3- to 5-day life span. Using adult
rats, sequential enterocyte isolation from villus tip to crypt was undertaken
by citrate-EDTA chelation and RNA extracted from isolated cells whose
position was inferred from sucrase activity enrichment (Figure 1). Apo B
mRNA editing was determined by reverse-transcription and coupled poly-
merase chain reaction amplification of apo B eDNA followed by single nu-
cleotide discrimination using differential oligonucleotide hybridization. As
can be seen in Figure I, the small intestinal enterocyte edits apo B mRNA
as soon as it is transcribed in cells emerging from the crypt.
As a model of the developing fetal small intestine, workers have used a
human colon cancer-derived cell line (Caco-2) which, in culture, undergoes
a form of' 'spontaneous differentiation'' and displays certain phenotypic char-
acteristics of developing enterocytes. Teng et al. 28 used this cell line to study
apo B mRNA editing and found that during the course of differentiation from
pre- to late postconfluency, apo B mRNA abundance increased 20-fold but
the proportions of edited to unedited transcript remained unaltered at less than
5% at all times studied. Thus, in this cell line apo B mRNA abundance
appears to be regulated by mechanisms distinct from those which influence
apo B mRNA editing. Other investigators 36 using this cell line have found
that apo B mRNA editing increases when the cells are grown on semiperme-
able filters rather than plastic. The specific cues which mediate this response
are presently unknown but may provide insight into the mechanisms under-
lying developmental changes in human small intestinal apo B mRNA editing.
B. APO B mRNA EDITING IN NON-APO B-PRODUCING

CELLS AND TISSUES
Mammalian apo B is expressed in a tissue-specific manner with gene
transcription and protein synthesis predominantly confined to the adult liver
and small intestine. In the fetus, however, apo B mRNA and protein bio-
synthesis have been found to occur in a number of extraintestinal, extrahepatic
sites. In the rat 29 •37 these include the placental and fetal membranes, while in
humans, we and others have found that numerous tissues, including lung,
kidney, stomach, colon, adrenal, and fetal membranes were all competent to
synthesize and secrete apo B. 28 ·38 These latter studies demonstrated the pres-
Jejunum Sucrase TAA CAA
•
68
•• • •
71
•• • •
36
•• •
TAA
••
CAA
•
Ileum Sucrase TAA CAA
••
••
6
••
4
•• "
FIGURE 1. Gradients of apo B mRNA editing in rat small intestine. Isolated rat enterocytes
were prepared from both jejunum and ileum by citrate-EDTA chelation with estimates of the
relative position of each fraction on the villus-crypt axis determined by sucrase-specific activity.
RNA was prepared from each fraction and subjected to reverse-transcription and polymerase
chain reaction amplification of apo B eDNA followed by differential oligonucleotide hybridization
to detect the single nucleotide substitution. It is evident that intestinal apo B mRNA is extensively
edited in all regions of the small intestine with no horizontal or vertical gradient detectable.
ence of apo B mRNA using reverse transcription and coupled polymerase

chain reaction amplification (RT-PCR) of low levels of endogenous apo B
mRNA. Of equal significance was the observation that a number of simul-
taneously analyzed tissue RNA samples, such as placenta, brain, and heart,
were negative. Analysis of these PCR amplified eDNA samples indicated that
apo B was edited to a varying extent ( 10 to 50%) in all the fetal tissues
examined with the notable exception of the liver, and, in all these sites, apo
Davidson 303
B eDNA was edited to a progressively greater extent during development.

Of importance was the observation that despite the presence of edited transcript
in these extraintestinal sites, the form of apo B synthesized and secreted was
exclusively apo B IOO. Thus, the regulation of apo B48 synthesis involves
other possibly co- or posttranslational events.
A different approach to this question was used by Bostrom et al. 39 .4° who
utilized a chimeric apo B expression construct to determine the presence of
apo B mRNA editing in a variety of cell lines. The advantage of this approach
is the ability to detect apo B processing events in the transfected construct
distinct from the endogenous transcript. These workers used either a 63- ,
186- , or a 354-base pair (bp) region of human apo B eDNA flanking the
edited base (6507 to 6860) annealed in midreading frame to an apo E eDNA
under the transcriptional control of the Moloney murine leukemia virus long
terminal repeat. Previous studies39 by these workers demonstrated that the
354- but not the 63-bp construct was competent to undergo editing in Caco-
2 cells, suggesting that the requisite length of apo B mRNA sequence to
direct editing of codon 2153 was over 63 bps. These workers transfected 18
different cell lines with the 354-bp construct, using cells that contained en-
dogenous apo B mRNA, such as Caco-2 and various rat and mouse liver cell
lines, in which the transcript is edited albeit to varying degrees, and HepG2
cells where apo B mRNA is expressed but not edited. In addition, these
workers transfected cells from a variety of embryonal origins where no en-
dogenous apo B mRNA is expressed. The findings of this study indicate that
the 354-bp construct was edited in cells that edit endogenous apo B mRNA,
although details on the relative efficiency of exogenous vs. endogenous apo
B mRNA editing and corresponding synthesis rates for apo B IOO vs. apo B48
were not provided. Of greater interest was the observation that a number of
cells of different embryonic origin, none of which express endogenous apo
B, were competent in editing the 354-bp construct and in some instances, to
secrete a truncated apo B protein. The implication of this finding is that
numerous cells possess the editing machinery, and that apo B mRNA editing
may represent one example of a process which is more widespread than
previously thought.
C. TOPOLOGY OF APO B mRNA EDITING

The mechanism involved in apo B mRNA editing is unknown but several
possibilities exist. The ready explanation for this single nucleotide change is
that it involves a site-specific cytosine deamination to a uracil. Other pos-
sibilities include some form of base exchange reaction or modification such
as occurs in certain tRNAs or the posttranscriptional insertion and deletion
of residues which occurs in trypanosomes. Bostrom et al. 40 demonstrated that
a synthetic radiolabeled apo B cRNA could be edited in vitro and the resulting
RNA digested to its 5'-monophosphates and resolved by two-dimensional thin
layer chromatography. The finding that the edited base comigrated with au-
thentic uridine 5' -monophosphate indicates strongly that the process involves
a cytosine deaminase. The question of where in the cell this modification
occurs was addressed in an imaginative series of experiments from Lau et
a!. 41 These workers addressed this issue using the rat liver which is known
to contain both edited and unedited apo B mRNA. The authors first prepared
nuclear and cytoplasmic RNA and demonstrated that nuclear RNA contained
less than 20% edited apo B mRNA while cytoplasmic apo B mRNA-either
total or polysomal-contained approximately 60% edited species. Lau et al.~ 1
then went on to fractionate nuclear RNA into poly A+ and poly A- RNA in
order to ascertain the relationship between polyadenylation and editing of apo
B mRNA. They found that poly A+ RNA contained approximately 50% edited
apo B mRNA while poly A- RNA contained only 10% edited species. These
workers then identified an intron in the rat apo B genomic DNA sequence
5% to the edited base in a similar position to that reported for the human apo
B gene. Using a primer downstream of the edited base and another primer
at the 3' terminus of the upstream exon 25, Lau et al. 41 were able to use the
polymerase chain reaction to amplify approximately 3 kb of apo B flanking
the intron-exon junction in both genomic DNA and unspliced RNA. Splicing
was shown to remove the 500-bp intron, and thus the amplification of a 2.5-
kb fragment could be distinguished from the 3-kb fragment alluded to above.
These workers demonstrated that unspliced nuclear pre-RNA contained less
than 10% edited apo B mRNA and that unspliced poly A- RNA contained
essentially no edited apo B mRNA. Thus, the conclusion from these exper-
iments is that apo B mRNA editing is not a cotranscriptional process, but
occurs predominantly in the nucleus and is completed by the time the RNA
is spliced and polyadenylated. These findings have important implications
for the identity and characteristics of the postulated editing machinery which
should presumably be nuclear, although it is well recognized that passage of
proteins and other macromolecules occurs through nuclear pores, which thus
does not preclude the possibility of the editing machinery originating in the
cytoplasmic compartment. The topology of apo B mRNA editing in other cell
types, particularly the small intestine, has not been evaluated to date, at least
in part because of difficulties associated with the preparation of intestinal
nuclei and intact nuclear RNA.
III. MODULATION OF APO B mRNA EDITING BY

HORMONAL AND NUTRITIONAL FACTORS
A. MODULATION BY THYROID HORMONE

As detailed above, studies by many workers demonstrated that the rat
liver synthesizes and secretes both forms of apo B, but Davidson et al. 42 were
the first to demonstrate that the proportions of these isoforms was subject to
modulation following alterations in thyroid hormone status. Using rats made
hypothyroid following dietary exposure to propylthiouracil, Davidson eta!. 42
Davidson 305
treated groups of these animals with increasing doses of T3 and examined

apolipoprotein gene expression in the liver and small intestine. These studies
revealed a tissue-specific effect of thyroid hormone with regard to the expres-
sion of apolipoprotein A-I and A-IV genes which were transcriptionally upre-
gulated in the liver but not the small intestine. Of interest was the effect of
T3 administration on apo B synthesis in the liver. In contrast to the euthyroid
and hypothyroid animals, animals treated with T3 had no detectable apo B 100
synthesis following immunoprecipitation of radiolabeled hepatic cytosolic
(225,000 X g) supernatants. Following up on these observations, Davidson
et a!. 25 used a reverse-transcription/polymerase-chain reaction based strategy
to amplify the region corresponding to the 3' terminus of rat apo B48 and
demonstrated, by differential oligonucleotide hybridization, that the mecha-
nism underlying the absence of hepatic apo BlOO synthesis in the T3 treated
animals was a switch from approximately 60/40 edited to unedited apo B
mRNA to greater than 90% edited apo B mRNA. Thus apo B mRNA editing
was demonstrated to be both a developmentally regulated event and to be
modulated in the adult animal following hormonal manipulation. It bears
emphasis that changes in rat hepatic apo B mRNA editing occur with only
minor (less than twofold) changes in apo B mRNA abundance in all the
settings that we have encountered. Subsequent studies 43 were undertaken to
establish the functional consequences for hepatic triglyceride-rich lipoprotein
assembly of the deletion of the carboxyl terminal half of apo B. Animals
were prepared as before and groups of euthyroid, hypothyroid, and hyper-
thyroid rats used to prepare hepatic Golgi for analysis of the content lipo-
proteins. Using nascent intracellular Golgi VLDLs prepared from control,
euthyroid animals demonstrated a polydisperse morphology with sizes aver-
aging 419 A while the analogous particles isolated from hypothyroid animals
averaged only 160 A. Previous work42 suggested that the hypothyroid rat has
a relative block in the secretion of hepatic triglyceride, and the findings of
this study extend these general observations by demonstrating that lipoproteins
isolated from a presecretory intracellular organelle are smaller and presumably
underlipidated. By contrast, Golgi lipoproteins from hyperthyroid animals
where apo B mRNA editing was greater than 90% demonstrated a normal
morphology and size distribution with an average diameter of 328 A. Analysis43
of the protein components of these Golgi VLDLs by Western blotting revealed
the presence of essentially only apo B48 in the hyperthyroid animals while
both hypothyroid and control euthyroid animals demonstrated a large pre-
ponderance of apo BIOO. Thus, as would be anticipated from the results of
studies of intestinal lipoprotein synthesis and secretion (where apo B48 is
synthesized almost exclusively), apo B48 is competent to direct the synthesis
and secretion of hepatic triglyceride-rich lipoproteins.
In regard to the mechanism of the T3-dependent modulation of hepatic
apo B mRNA editing in the rat, other studies have attempted to address the
role of the nuclear thyroid hormone receptor and potential intermediates of
T3 action, such as growth hormone, as the proximate mediators of the alter-

ations of apo B mRNA editing. These were important considerations in view
of the observation that, following T3 administration to hypothyroid rats, there
was a lag period of 12 to 18 h before any changes were detectable in either
hepatic apo B mRNA editing or apo BlOO synthesis. Studies were conducted
in hypothyroid rats treated with one of two synthetic analogs of T3. 44 These
analogs were developed as hypocholesterolemic agents since it appeared that
they were able to lower serum cholesterol levels in rats without precipitating
the cardiac side effects common with authentic T3. 45 Both analogs chosen
were able to bind to the hepatic (but not cardiac) nuclear thyroid hormone
receptor with reduced but comparable affinity to native T3. However, one
analog demonstrated approximately 20 times the bioactivity of the other.
Following administration of these agents to hypothyroid rats, hepatic apo B
mRNA editing was examined and found to be increased with both the "active"
and ''inactive'' analog. Thus, binding to the nuclear thyroid hormone receptor
was considered an important component of the induction of apo B mRNA
editing following thyroid hormone administration. 44 Induction of apo B mRNA
editing following T3 administration was accompanied by coordinate pretran-
slational increases in hepatic apo A-I and A-IV gene expression. Adminis-
tration of T3 analogs produced a less marked increase in hepatic mRNA
abundance for these genes (3- to 5-fold compared to 6- to 14-fold for native
T3) and a variable increase in hepatic mRNA abundance for several prototypic
T3-responsive genes, such as malic enzyme and spot 11. Thus the induction
of hepatic apo B mRNA editing occurs somewhat independently from the
hormonal induction of other classes of T3 responsive genes. 44 When hypo-
thyroid animals were treated with pharmacologic doses of growth hormone
there was no change in the proportions of edited and unedited apo B mRNA. 44
This was despite effective reversal of the growth arrest which accompanies
hypothyroidism in the rat. Thus growth hormone was considered, by itself,
insufficient to mediate the hormonal modulation of apo B mRNA editing.
Recent studies 46 using hypophysectomized rats treated with various combi-
nations of growth hormone and T4 have suggested a facilitative role for growth
hormone in mediating apo B mRNA editing both in the whole animal and in
isolated hepatocytes, but clearly more work is needed in this area. A possible
clue in this regard may come from preliminary work 47 where the nuclear
thyroid hormone receptor was transfected into Caco-2 cells and apo B mRNA
editing found to be modulated. However, this work requires additional con-
firmation. Finally, the role of thyroid hormone in the developmental regulation
of both intestinal and hepatic apo B mRNA editing needs clarification. In the
rat, hepatic apo B mRNA editing is predominantly a postnatal event, 29 and
it is tempting to speculate that thyroid hormone may undergo developmental
increases in parallel with the induction of apo B mRNA editing. This remains
to be demonstrated, however. It is unlikely that thyroid hormone plays a role
in the development or maintenance of intestinal apo B mRNA editing since
Davidson 307
the changes in apo B mRNA editing in both the rat and human small intestine
precede any developmental increase in fetal thyroid hormone activity and
maternal T3 does not cross the placenta. Additionally in the adult rat small
intestine, there was no change in apo B mRNA editing following changes in
thyroid hormone status. 25
B. MODULATION OF APO B mRNA EDITING FOLLOWING

CHANGES IN HEPATIC LIPOGENESIS
Thyroid hormone modulation of hepatic apo B mRNA editing was ac-
companied by striking alterations in hepatic lipid mobilization. 42 Since there
are well characterized examples of genes which display response to both T3
and high carbohydrate intake, such as malic enzyme and spot 11 and spot 14
(reviewed in Reference 44), it was of interest to examine the possibility that
the gene(s) responsible for mediating apo B mRNA editing in the rat liver
may represent an example of this general class of response. Against this idea
is the previous observation that changes in apo B mRNA editing show a longer
lag period following T3 administration (12 to 18 h) than is generally en-
countered with this class of responsive genes (which often show responses
within minutes to a few hours), and, secondly, that apo B mRNA editing has
not been reproduced in vitro following T3 administration. Baum et al. 48 studied
groups of rats fasted for 24 or 48 h and other groups fasted for 48 h and
subsequently fed a high carbohydrate, fat-free diet for 24 or 48 h. This
maneuver produced a 30-fold range of hepatic triglyceride concentration from
a nadir at 48 h fasting to a peak at 48 h refeeding a high carbohydrate diet.
In association with these changes, apo B synthesis rates were determined
following intraportal vein administration of tritiated leucine and quantitative
imrnunoprecipitation of apo B. In animals fasted for 48 h there was a decrease
in the ratio of apo B48 to apo B 100 synthesis and a corresponding decrease
in hepatic apo B mRNA editing. In animals fasted for 48 h and then refed a
high carbohydrate diet for either 24 or 48 h, there was no apo B100 synthesis
detectable and hepatic apo B mRNA was greater than 90% edited. In asso-
ciation with these findings, serum apo B isomorphs as demonstrated on West-
em blots, were found to be altered in parallel such that control animals and
animals fasted for 24 or 48 h demonstrated mostly apo B 100 in their serum
while animals fasted and refed a high carbohydrate diet demonstrated essen-
tially only apo B48. This latter finding 48 clearly reflects a number of events
including synthetic and catabolic processes, but as a maneuver to alter cir-
culating levels of apo B100, these observations suggest an important avenue
of investigation. When all animals were included in an analysis of the impact
of apo B mRNA editing on hepatic apo B 100 synthesis rates, there was a
striking correlation found, suggesting that hepatic apo B mRNA editing is an
extremely important determinant of this process in vivo. Studies48 were ex-
tended to an evaluation of the regulation of other hepatic apolipoprotein genes
in the setting of the changes in hepatic triglyceride metabolism alluded to
above. It was found that apo A-1 and A-IV mRNA abundance and protein
biosynthesis were increased approximately two- to fourfold in the animals
refed a high carbohydrate diet for 48 h. These changes are thus of a lesser
magnitude but in the same direction as encountered42 with T3 treatment. The
time course of these changes has not been fully characterized since the earliest
time point examined was 24 h after exposure to a high-carbohydrate diet and
the experiment only continued up to 48 h of refeeding. Thus, important
questions still remain, such as the earliest point that a response to carbohydrate
is detectable and whether equivalent responses can be obtained with different
carbohydrate sources such as fructose and glucose. Additionally, it will be
of interest to examine longer periods of exposure to a high-carbohydrate diet
to determine whether the liver is capable of effectively unloading the accu-
mulated lipid after sustained intake of this modified isocaloric diet. Finally
it will be important to evaluate the synthesis of nascent intrahepatic lipopro-
teins by isolating Golgi VLDLs and examining their morphology, lipid con-
tent, and apo B isomorph distribution. These studies are currently underway
in the laboratory.
C. MODULATION OF APO B mRNA EDITING BY

OTHER MANIPULATIONS
Since the first report that hepatic apo B mRNA editing was modulated
by thyroid hormone, several investigators have examined other potential routes
of manipulating this activity. Two recent reports49 ·50 are of interest in this
regard. Work from Glickman's laboratory 49 has examined the regulation of
both hepatic and intestinal apolipoprotein gene expression following the
administration of ethinyl estradiol to rats. This agent produces a profound
hypocholesterolemia in rats which is thought to be secondary to overexpression
of hepatic LDL receptors. These workers found that administration of this
agent to rats for 5 d produced a decrease in hepatic but not small intestinal
apo B mRNA editing and a corresponding increase in the relative synthesis
rates of apo B 100. In keeping with the findings of Baum et a!., 48 alluded to
above, these workers found a strong correlation between hepatic apo B mRNA
editing and apo B 100 synthesis. The possible contribution of prolonged fasting
to the decrease observed in hepatic apo B mRNA editing in this model was
explored and found not to be a factor in this response although the precise
mechanism remains obscure. A second report50 was presented in abstract form
recently and describes the changes in hepatic apo B mRNA editing in non-
insulin-dependent diabetes mellitus. This report describes a decrease in hepatic
apo B mRNA editing in uncontrolled noninsulin-dependent diabetes in rats.
The report raises the larger and more interesting possibility that insulin may
play a role in the regulation of hepatic apo B mRNA editing distinct from its
well characterized effects on the posttranslational processing of nascent apo
B 100 and B48.
Davidson 309
IV. IN VITRO EDITING OF APO B mRNA
A. METHODOLOGICAL CONSIDERATIONS
The major question for investigators examining the biology of apo B
mRNA editing is the underlying mechanism of this interesting process. As
indicated above the most likely explanation is a site-specific deamination of
the cytosine at nucleotide 6666, but the basis for this conclusion was attained
experimentally as a result of a technical advance pioneered by Driscoll et
al., 51 which described an in vitro system to examine apo B mRNA editing.
This sytem used cytoplasmic S 100 supernatants prepared from MeA 7777
cells, a rat hepatoma line which synthesizes and secretes both apo BlOO and
apo B48 and which therefore provides a ready source of "editing activity"
to examine possible in vitro RNA modification. Driscoll et al. 51 incubated
MeA 7777 extracts with various lengths of synthetic apo B RNA prepared
from the region flanking the edited base. In order to detect the single nucleotide
change, these workers devised an ingenious scheme whereby an antisense
apo B oligonucleotide (35-mer) is annealed downstream from the edited base
at position 6666 and extended with reverse transcriptase in the presence of
dATP, dCTP, dTTP, and dideoxy-GTP as a chain terminator. There being no
other "C" residues between the 5' end of the antisense oligonucleotide and
the edited base, the primer is extended until it reaches the first upstream "C".
If nucleotide 6666 is unedited the primer undergoes chain termination and
produces an extension product of 42 bp. If nucleotide 6666 is edited to a
"U" then the primer extends to the next upstream "C", at 6655 in human
apo B eDNA and 6661 in rat and mouse apo B eDNA. The extended and
terminated primers are resolved effectively by urea-acrylamide electropho-
resis. In the original description by Driscoll et al., 51 in vitro editing was
demonstrated with S I00 extracts from MeA 7777 cells, but the efficiency of
editing was less than 5% of the input RNA. The reasons for the low activity
are of interest since endogenous apo B mRNA from these cells is edited
approximately 20%. One possibility is that subtle sequence differences be-
tween human and rat apo B eDNA may be of importance (see below and
Figure 2). There may be additional factors, however, since an intriguing
observation is that MeA 7777 cells transfected with a human full-length apo
B eDNA edit a substantial proportion of the human apo B mRNA (greater
than 50%) while the endogenous transcript is edited less than half of this
value. 52 ·53 Thus issues relating to differential accessibility and transcript sta-
bility require consideration. The relatively low level of in vitro apo B mRNA
editing observed with MeA 7777 cells has been observed by others in nuclear
extracts prepared from rat liver. 41 The question arises as to whether this
particular source (i.e., rat hepatic or hepatoma-derived nuclear or SIOO ex-
tract) contains relatively low levels of editing activity or whether the puri-
fication scheme used is inappropriate. Investigators using rat liver nuclear
extracts to examine in vitro editing have had to optimize their methods for
A. Aligned nucleotide sequences
64 82 6522
Human CCCACAGCAAGCTAATGATTATCTGAATTCATTCAATTGGGAGAGACAAGTTTCACATGC
Pig ------------T--------------G-A------G------A----------TGAG---
Rabbit -----------T--------------G-A-------------------------CAG---
Rat T--T-----GAT-C--------------G---CTG-C--------------AG-TGG---
Mouse T--T-----GAT-C--C-----------G---CTG-C--------------AG-TGG---
6542 6582
Human CAAGGAGAAACTGACTGCTCTCACAAAAAAGTATAGAATTACAGAAAATGATATACAAAT
Pig ---AA------AT-G--A-T---TGG--G-T---------------------G---G---
Rabbit ------------A---A--T----------T----A---------G-------------C
Rat ------A---T-A---T--T---TGG----C--------------T------G---T---
Mouse ------A---A-A---T--T---TGG----T--------------T------G---T---
6602 6642
Human TGCATTAGATGATGCCAAAATCAACTTTAATGAAAAACTATCTCAACTGCAGACATATAT
Pig ---------CA--------------C--------------A----G--A--A------G-
Rabbit ------G---A----------------A-----------G--------T---------G-
Rat ---C------AG---------------G-----------C--------TG-------CGC
Mouse ---CA-----AG---------------C-----------C--------TG-------CGC
6662 6702
Human GATACAATTTGATCAGTATATTAAAGATAGTTATGATTTACATGATTTGAAAATAGCTAT
Pig -----------------------------A------------------T--G-C------
Rabbit -----------------------------A--T----C----------T-----------
Rat -----------------------G-----A-------GC---G--C--A----G-A----
Mouse -----------------------------A-------CC------C--A----G-A----
6722 6762
Human TGCTAATATTATTGATGAAATCATTGAAAAATTAAAAAGTCTTG
Pig -----GG-------------------C--C---G----T-----
Rabbit A----G------A---C-------G-------------T-----
Rat ----C-G---------AG------------GC------TG----
Mouse ----G-G---------CG------------G-------TG----
B. Aligned amino acid sequences

2092 2132
Human PQQANOYLN5FNWERQV5HAKEKLTALTKKYRITENOIQIALOOAKINFNEKL5QLQTYM
Pig ---V----5T-5-----L5--K-H50FMEO-------VR----N----L----T-----V
Rabbit ---V----5!--------5------TF--N-K-------T---N----L----------V
Rat ---IH----A50-----AG------5FMEN----0--VL----5----L-------E--A
Mouse ---IHH---A50-----AG----I-5FMEN----O--VL--I-5------------E--A
2152 2182
Human IQFOQYIK05YOLHOLKIAIANIIOEIIEKLK5L
Pig ---------N-----F-T---R------AT--I-
Rabbit ---------NF----F-----5---Q-M----I-
Rat -------R-N--AQ---RT--0---R------M-
Mouse ---------N--P----RT--E---R------M-
FIGURE 2. (A) Aligned nucleotide sequence of apo B eDNA flanking the edited base. Four
mammalian apo B cDNAs are contrasted to the human sequence with identity represented by a
hyphen. 30 (B) Aligned amino acid sequence deduced from four mammalian apo B cDNAs in
comparison to human apo B flanking the edited base. The single Jetter code is employed for
amino acid identity. 30
reliable detection of low levels of apo B mRNA editing which in some cases
are less than 1 to 2% of input RNA. 54 Studies from a number of labs indicate
that the yield and functional activity of small intestinal S 100 extracts is
considerably higher than that reported for the rat liver. Using isolated rabbit, 40
rat, 55 or baboon56 enterocyte S100 extracts, workers have reported in vitro
Davidson 311
conversion rates of~ 10 to 40% of input RNA. Whether this editing activity
is developmentally induced in the small intestine awaits further study. There
appear to be some inconsistencies between different reports concerning the
optimal conditions for in vitro editing with the original report describing a
cryolabile activity which required 50 mM EDTA. 51 Later reports have found
a number of differences from the earlier study, and most recently Greeve et
al. 55 have detailed extensive findings in regard to the characterization of rat
enterocyte SlOO editing activity. The major conclusions emerging from this
and other reports is that the editing activity is most likely a protein with no
cofactor or additional RNA requirements. Attempts to purify this editing
activity from crude enterocyte S 100 extracts have resulted in severalfold
enrichment, but no consensus has emerged as to its apparent molecular mass.
Driscoll has suggested that baboon enterocyte S 100 extracts contain an editing
activity of approximately 125 kDa. 56 Another controversial aspect of the in
vitro editing methodology is whether apo B mRNA assembles into a higher
order molecular complex following incubation with a source of editing ac-
tivity. Following an initial report of this phenomenon, 5 7 others55 have failed
to confirm the existence of a so-called ''editosome' '. Distinct from the editing
activity, there appears to be an apo B mRNA binding activity in rat liver
nuclear extracts. 5 8 This binding activity has an apparent molecular weight of
40 kDa following UV cross-linking and specifically binds a 26-nucleotide
apo B RNA fragment spanning the edited base. It will be of interest to
investigate the relationship of this apparent RNA binding protein to the editing
machinery by way of mutagenesis studies in which editing is abolished.
Species considerations may be relevant in this regard also, since studies59
from our laboratory have indicated that chick small intestinal enterocytes
contain an editing enhancing factor(s), and studies are currently underway to
purify and compare these factor(s) to the binding activity found in rat tissue.
B. SEQUENCE REQUIREMENTS FOR APO B mRNA EDITING

Several studies have addressed this issue following the technical advances
alluded to above. A general point of interest is that the nucleotide sequence
spanning the edited base at position 6666 is highly conserved in mammalian
apo B cDNAs (Figure 2). In regard to the minimal length of sequence in-
formation, there is some confusion in the literature most likely reflecting
experimental differences. Davies et al. 60 determined that a 26-nucleotide re-
gion of human apo B RNA flanking the edited base was as efficiently edited
as the endogenous (rat) transcript when transfected into MeA 7777 cells. This
result is at variance with data suggesting that Caco-2 cells edit a 354- and
186- but not a 63 bp chimeric construct following stable transfection. 39 .4° The
reasons for this discrepancy are unclear. Recent work from Scott's laboratory 61
has demonstrated that a human apo B RNA sequence of 26 nucleotides can
be edited in vitro using S l 00 extracts prepared from rat enterocytes. This is
of interest since earlier studies from this group 51 indicated that a 55- but not
a 26-nucleotide fragment could be edited in vitro using SlOO extracts prepared
from MeA 7777 cells. Thus it is likely that there are true biological differences
in editing activity within different cell types which may, additionally, be more
or less preserved during the preparation of S 100 extracts. In regard to the
nucleotide specificity for apo B mRNA editing, recent studies have provided
some important advances. Chen et al. 54 used a coupled transcription editing
reaction to demonstrate that in a series of mutations in a 9-nucleotide sequence
flanking the edited base, in only 2 of 22 such mutants (a transversion mutant
changing the wild type ATACAATIT to ATTCT ATIT and a deletion mutant
ATAC-TTTT) was editing abolished. The authors contend that this indicates
a relatively lax sequence specificity for apo B mRNA editing. An important
limitation of this study was the low level of editing demonstrated by rat liver
nuclear extracts which was less than 2% of the input apo B RNA. In order
to investigate whether more distant sequence information conferred important
specificity Shah et a!. 61 systematically mutagenized a 55-bp region of human
apo B eDNA and demonstrated that the minimal sequence information required
for in vitro editing by rat enterocyte S 100 extracts is contained with a 26-
nucleotide stretch flanking the edited base. However, these workers found
an 11-nucleotide region downstream of position 6666 in which all but one
point mutations abolished editing. This result is of interest since recent work
from our laboratory 59 has demonstrated that chick apo B mRNA is not edited
in vivo by small intestinal enterocytes, and that the chick apo B eDNA
sequence is divergent from mammalian apo B cDNAs with 3 changes from
the conserved mammalian sequence noted in the 11-nucleotide cassette de-
scribed by Shah et alY These changes would be predicted to abolish in vitro
apo B RNA editing according to the data of Shah et a!., and, indeed, no
editing could be detected when chick apo B RNA was incubated in vitro with
S 100 extracts prepared from rat enterocytes while human, rat, and porcine
apo B RNAs were all efficiently edited. A further point of interest in regard
to the sequence specificity of apo B RNA editing concerns the demonstration
that a second edited base is detectable in human apo B RNA at position 6802
which changes threonine (ACA) to isoleucine (AUA) at codon 2198. 62 This
is of no functional consequence since the transcript is terminated upstream
of this nucleotide in all instances examined. The finding is of interest in view
of the sequence homology flanking both edited sites, and it reinforces the
hypothesis that the requisite information for editing is contained within the
26 nucleotides flanking position 6666.
V. SUMMARY AND CONCLUSIONS
Apo B mRNA editing is the first example of a process which expands

the repertoire of vertebrate gene regulation. It is likely that the basis for this
interesting process will be of importance in other contexts particularly in view
of the demonstration that apo B mRNA editing is detectable in cells which
contain only trivial amounts of apo B mRNA. Areas that will provide inter-
Davidson 313
esting information in regard to the biology of apo B rnRNA editing will be

to determine the tissue- and cell-specific cues which regulate this process in
response to development and to hormonal and nutritional modulation. These
issues are currently under investigation and will form the focus of future
reports.
ACKNOWLEDGMENTS
Work cited from the author's laboratory was supported by Grants HL-
38180, HL K0-4 02166 from the NHLBI and DK-42086 from NIDDK. I
would like to recognize the contributions of BaBie Teng, Charles Baum, Ruth
Carlos, and Annalise Hausman to the work presented in this review. This
chapter is dedicated to my father, Jack Davidson, M.B. Ch.B., on the occasion
of his 75th birthday.
REFERENCES
1. Chan, L., VanTuinen, P., Ledbetter, D. H., Daiger, S. P., Gotto, A.M., and Chen,
S. H., The human apolipoprotein B-100 gene: a highly polymorphic gene that maps to
the short arm of chromosone 2, Biochem. Biophys. Res. Comm.• 133, 248, 1985.
2. Mehrabian, M., Sparkes, R. S., Mohandas, T., Klisak, I. J., Schumaker, V. N.,
Heinzmann, C., Zollman, S., Ma, Y., and Lusis, A. J., Human apolipoprotein B:
chromosmal mapping and DNA polymorphisms of hepatic and intestinal species, Somatic
Cell Mol. Genet., 12. 245, 1986.
3. Luo, C. C., Li, W. H., Moore, M. N., and Chan, L., Structure and evolution of the
apolipoprotein multigene family, J. Mol. Bioi., 187, 325, 1986.
4. Li, W. H., Tanimura, M., Luo, C. C., Datta, S., and Chan, L., The apolipoprotein
multigene family: biosynthesis, structure, structure-function relationships, and evolution,
J. Lipid Res.. 29, 245, 1988.
5. Knott, T. J., Pease, R. J., Powell, L. M., Wallis, S. C., Rail, S. C., Innerarity,
T. L., Blackhart, D., Taylor, W. H., Marcel, Y., Milne, R., Johnson, D., Fuller,
M., Lusis, A. J., McCarthy, B. J., Mahley, R. W., Levy-Wilson, B., and Scott, J.,
Complete protein sequence and identification of structural domains of human apolipo-
protein B, Nature (London), 323, 734, 1986.
6. Yang, C. Y., Chen, S. H., Gianturco, S. H., Bradley, W. A., Sparrow, J. T.,
Tanimura, M., Li, W. H., Sparrow, D. A., DeLoof, H., Roseneu, M., Lee, F. S.,
Gu, Z. W., Gotto, A.M., and Chan, L., Sequence, structure, receptor-binding domains
and internal repeats of human apolipoprotein B-100. Nature (London), 323, 738, 1986.
7. Chen, S. H., Yang, C. Y., Chen, P. F., Setzer, D., Tanimura, M., Li, W. H., Gotto,
A. M., and Chan, L., The complete eDNA and amino acid sequence of human apoli-
poprotein B-100, J. Bioi. Chern., 261, 12918, 1986.
8. Cladaras, C., Hadzopoulou-Cladaras, M., Nolte, R. T., Atkinson, D., and Zannis,
V. I., The complete sequence and structural analysis of human apolipoprotein B-100:
relationship between apoB-100 and apoB-48 forms, EMBO J., 5, 3495, 1986.
9. Ludwig, E. H., Blackhart, B. D., Pierotti, V. R., Caiati, L., Fortier, C., Knott, T.,
Scott, J., Mahley, R. W., Levy-Wilson, B., and McCarthy, B. J., DNA sequencing
of the human apolipoprotein B gene, DNA, 6, 363, 1987.
10. Innerarity, T. L., Young, S. G., Poksay, K. S., Mahley, R. W., Smith, R. S., Milne,
R. W., Marcel, Y. L., and Weisgraber, K. H., Structural relationship of human apo-
lipoprotein B48 to apolipoprotein BIOO, J. Clin. Invest., 80, 1794, 1987.
11. Hardman, D. A., Protter, A. A., Chen, G. C., Schilling, J. W., Sato, K. Y., Lau,
K., Yamanaka, M., Mikita, T., Miller, J., Crisp, T., McEnroe, G., Scarborough,
R. M., and Kane, J. P., Structural comparison of human apolipoproteins B-48 and B-
100, Biochemistry, 26, 5478, 1987.
12. Yang, C. Y., Kim, T. W., Weng, S. A., Lee, B., Yang, M., and Gotto, A. M., Jr.,
Isolation and characterization of sulfhydryl and disulfide peptides of human apolipoprotein
B-100, Proc. Nat!. Acad. Sci. U.S.A., 87, 5523, 1990.
13. Law, A. and Scott, J., A cross-species comparison of the apo1ipoprotein B domain that
binds to the LDL receptor, J. Lipid Res., 31, 1109, 1990.
14. Pease, R. J., Milne, R. W., Jessup, W. K., Law, A., Provost, P., Fruchart, J. C.,
Dean, R. T., Marcel, Y. L., and Scott, J., Use of bacterial expression cloning to
localize the epitopes for a series of monoclonal antibodies against apolipoprotein B-1 00,
J. Bioi. Chern., 265, 553, 1990.
15. Herz, J., Hamann, U., Rogne, S., Myklebost, 0., Gausepohl, H., and Stanley,
K. K., Surface location and high affinity for calcium of a 500-kd liver membrane protein
closely related to the LDL-receptor suggest a physiological role as lipoprotein receptor,
EMBO J., 7, 4119, 1988.
16. Kowal, R. C., Herz, J., Goldstein, J. L., Esser, V., and Brown, M.S., Low density
lipoprotein receptor-related protein mediates uptake of cholesterol esters derived from
apoprotein E-enriched lipoproteins, Proc. Nat!. Acad. Sci. U.S.A., 86, 5810, 1989.
17. Beisegel, U., Weber, W., Ihrke, G., Herz, J., and Stanley, K. K., The LDL-receptor-
related protein, LRP, is an apolipoprotein E-binding protein, Nature (London), 341, 162,
1989.
18. Cattaneo, R., Different types of messenger RNA editing, Annu. Rev. Genet., 25, 71,
1991.
19. Sollner-Webb, B., RNA editing, Curr. Op. Cell Bioi., 3, 1056, 1991.
20. Blum, B., Bakalara, N., and Simpson, L., A model for RNA editing in kinetoplastid
mitochondria: "Guide" RNA molecules transcribed from maxicircle DNA provide the
edited information, Cell, 60, 189, 1990.
21. Powell, L. M., Wallis, S.C., Pease, R. J., Edwards, Y. H., Knott, T. J., and Scott,
J., A novel form of tissue-specific RNA processing produces apolipoprotein B48 in
intestine, Cell. 50, 831, 1987.
22. Chen, S. H., Habib, G., Yang, C. Y., Gu, Z. W., Lee, B. R., Weng, S., Silberman,
S. R., Cai, S. J., Deslypere, J. P., Rosseneu, M., Gotto, A. M., Li, W. H., and
Chan, L., Apolipoprotein B-48 is the product of a messenger RNA with an organ-specific
in-frame stop codon, Science, 238, 363, 1987.
23. Hospattanker, A. V., Higuchi, K., Law, S. W., Meglin, N., and Brewer, H. B.,
Identification of a novel in-frame translational stop codon in human intestine ApoB
mRNA, Biochem. Biophys. Res. Commun., 148, 279, 1987.
24. Sommer, B., Kohler, M., Sprengel, R., and Seeburg, P. H., RNA editing in brain
controls a determinant of ion flow in glutamate-gated channels, Cell, 67, II, 1991.
25. Davidson, N. 0., Powell, L. M., Wallis, S. C., and Scott, J., Thyroid hormone
modulates the introduction of a stop codon in rat liver apolipoprotein B messenger RNA,
J. Bioi. Chern., 263, 13482, 1988.
26. Tennyson, G. E., Sabatos, C. A., Higuchi, K., Meglin, N., Brewer, H. B., Expression
of apolipoprotein B mRNAs encoding higher- and lower-molecular weight isoproteins in
rat liver and intestine, Proc. Nat[. Acad. Sci. U.S.A., 86, 500, 1989.
Davidson 315
27. Higuchi, K., Hospattankar, A. V., Law, S. W., Meglin, N., Cortright, J,, and
Brewer, H. B., Human apolipoprotein B (apo B) mRNA: identification of two distinct
apo B mRNAs, an mRNA with the apo B-100 sequence and an apoB mRNA containing
a premature in-frame translational stop codon, in both liver and intestine, Proc. Natl.
Acad. Sci. U.S.A., 85, 1772, 1988.
28. Teng, B., Verp, M., Salomon, J., and Davidson, N. 0., Apolipoprotein B messenger
RNA editing is developmentally regulated and widely expressed in human tissues, J.
Bioi. Chern., 265, 20616, 1990.
29. Wu, J, H., Semenkovich, C. F., Chen, S. H., Li, W. H., and Chan, L., Apolipoprotein
B mRNA editing: validation of a sensitive assay and developmental biology of RNA
editing in the rat, J. Bioi. Chern., 265, 12312, 1990.
30. Teng, B., Black, D. D., and Davidson, N. 0., Apolipoprotein B messenger RNA
editing is developmentally regulated in pig small intestine: nucleotide comparison of
apolipoprotein B editing regions in five species, Biochem. Biophys. Res. Comm., 173,
74, 1990.
31. Glickman, R. M., Rogers, M., and Glickman, J, N., Apolipoprotein B synthesis by
human liver and intestine in vitro, Proc. Nat!. Acad. Sci. U.S.A., 83, 5296, 1986.
32. Levy, E., Rochette, C., Londono, I., Roy, C. C., Milne, R. W., Marcel, Y. L., and
Bendayan, M., Apolipoprotein B-1 00: immunolocalization and synthesis in human in-
testinal mucosa, J. Lipid Res., 31, 1937, 1990.
33. Hoeg, J, M., Sviridov, D. D., Tennyson, G. E., Demosky, S. J,, Meng, M. S.,
Bojanovski, D., Safonova, I. G., Repin, V. S., Kuberger, M. B., Smirnov, V. N.,
Higuchi, K., Gregg, R. E., and Brewer, H. B., Both apolipoproteins B-48 and B-100
are synthesized and secreted by the human intestine, J. Lipid Res., 31, 1761, 1990.
34. Dixon, J, L., Furukawa, S., and Ginsberg, H. N., Oleate stimulates secretion of
apolipoprotein B-containing lipoproteins from Hep 02 cells by inhibiting early intracel-
lular degradation of apolipoprotein B, J. Bioi. Chern., 266, 5080, 1991.
35. Coleman, R. A., Haynes, E. B., Sand, T. M., and Davis, R. A., Developmental
coordinate expression of triacylglycerol and small molecular weight apoB synthesis and
secretion by rat hepatocytes, J. Lipid Res., 29, 33, 1988.
36. Jiao, S., Moberly, J, B., and Schonfeld, G., Editing of apolipoprotein B messenger
RNA in differentiated Caco-2 cells, J. Lipid Res .. 31, 695, 1990.
37. Demmer, L.A., Levin, M.S., Elovson, J., Reuben, M.A., Lusis, A. J,, and Gordon,
J, I., Tissue-specific expression and developmental regulation of the rat apolipoprotein
B gene, Proc. Nat!. Acad. Sci. U.S.A., 83, 8102, 1986.
38. Hopkins, B., Brice, A. L., Schofield, P. N., Baralle, F. E., and Graham, C. F.,
Identity of cells containing apolipoprotein B messenger RNA, in 6- to 12-week postfer-
tilization human embryos, Development, 100, 83, 1987.
39. Bostrom, K., Lauer, S. J., Poksay, K. S., Garcia, Z., Taylor, J. M., and Innerarity,
T. L., Apolipoprotein B48 RNA editing in chimeric apolipoprotein EB mRNA, J. Bioi.
Chern., 264, 15701, 1989.
40. Bostrom, K., Garcia, Z., Poksay, K. S., Johnson, D. F., Lusis, A. J,, and Innerarity,
T. L., Apolipoprotein B mRNA editing: direct determination of the edited base and
occurrence in non-apolipoprotein B producing cell lines, J. Bioi. Chern., 265, 22446,
1990.
41. Lao, P. P., Xiong, W., Zhu, H. J,, Chen, S. H., and Chan, L., Apolipoprotein B
mRNA editing is an intracellular event that occurs posttranscriptionally coincident with
splicing and polyadenylation, J. Bioi. Chern., 266, 20550, 1991.
42. Davidson, N. 0., Carlos, R. C., Drewek, M. J., and Parmer, T. G., Apolipoprotein
gene expression in the rat is regulated in a tissue-specific manner by thyroid hormone,
J. Lipid Res., 29, 1511, 1988.
43. Davidson, N. 0., Carlos, R. C., Sherman, H. L., and Hay, R. V., Modulation of
apolipoprotein B-100 mRNA editing: effects on hepatic very low density lipoprotein
assembly and intracellular apoB distribution in the rat, J. Lipid Res., 31, 899, 1990.
44. Davidson, N. 0., Carlos, R. C., and Lukaszewicz, A. M., Apolipoprotein B mRNA
editing modulated by thyroid hormone analogs but not growth hormone administration
in the rat, Mol. Endocrinol., 4, 779, 1990.
45. Underwood, A. H., Emmett, J. C., Ellis, D., Flynn, S. B., Leeson, P. D., Benson,
G. M., Novelli, R., Pearce, N. J., and Shah, V. P., A thyromimetic that decreases
plasma cholesterol levels without increasing cardiac activity, Nature (London), 324, 425,
1986.
46. Sjoberg, A., Oscarsson, J., Bostrom, K., Innerarity, T. L., Eden, S., and Olofsson,
S. 0., Growth hormone participates in the regulation of the editing of apolipoprotein B
mRNA in the rat liver, Arterioscler. Thromb., II, 1402a, 1991.
47. Patterson, A. P., Eggerman, T. L., Demosky, S. J., and Brewer, H. B., T3 Nuclear
receptor and growth hormone modulate apolipoprotein B editing, Arteriosclerosis and
Thrombosis, II , 1403a, 1991.
48. Baum, C. L., Teng, B., and Davidson, N. 0., Apolipoprotein B messenger RNA
editing in the rat liver: modulation by fasting and refeeding a high carbohydrate diet, J.
Bioi. Chon., 265, 19263, 1990.
49. Seishima, M., Bisgaier, C. L., Davies, S. L., and Glickman, R. M., Regulation of
hepatic apolipoprotein synthesis in the 17 alpha-ethinyl estradiol-treated rat, J. Lipid Res.,
32, 941, 1991.
50. Jiao, S., Yanagi, K., Shimomura, I., Kobatake, T., Keno, Y., Kobo, M., Tokunaga,
K., Matsuzawa, Y., and Tarui, S., Enhancement of apoB mRNA editing in the liver
of non-insulin-dependent diabetic rats, Arteriosclerosis and Thrombosis, II, 1424a. 1991.
51. Driscoll, D. M., Wynne, J. K., Wallis, S. C., and Scott, J., An in vitro system for
the editing of apolipoprotein B mRNA, Cell, 58, 519, 1989.
52. Blackhart, B. D., Yao, Z., and McCarthy, B. J., An expression system for human
apolipoprotein BIOO in a rat hepatoma cell line, J. Bioi. Chern., 265. 8358, 1990.
53. Yao, Z., Blackbart, B. D., Linton, M. F., Taylor, S.M., Young, S. G., and McCarthy,
B. J., Expression of carboxyl-terminally truncated forms of human apolipoprotein B in
rat hepatoma cells: Evidence that the length of apolipoprotein B has a major effect on
the buoyant density of the secreted lipoproteins, J. Bioi. Chern., 266, 3300, 1991.
54. Chen, S. H., Li, X., Liao, W. S., Wu, j. H., and Chan, L., RNA editing of apoli-
poprotein B mRNA: sequence specificity detem1ined by in vitro coupled transcription
editing, J. Bioi. Chern., 265, 6811, 1991.
55. Greeve, J., Navaratnam, N., and Scott, J., Characterization of the apolipoprotein B
mRNA editing enzyme: no similarity to the proposed mechanism of RNA editing in
kinetoplasid protozoa, Nucleic Acids Res., 19, 3569, 1991.
56. Driscoll, D. M. and Casanova, E., Characterization of the apolipoprotein B mRNA
editing activity in enterocyte extracts, J. Bioi. Chern., 265, 21401, 1990.
57. Smith, H. C., Kuo, S. R., Backus, J. W., Harris, S. G., Sparks, C. E., and Sparks,
j. D., In vitro apolipoprotein B mRNA editing: Identification of a 27S editing complex,
58. Lau, P. P., Chen, S. H., Wang, J. C., and Chan, L., A 40 ki1odalton rat liver nuclear
protein binds specifically to apolipoprotein B mRNA around the RNA editing site, Nucleic
Acids Res., 18, 5817, 1991.
59. Teng, B. B. and Davidson, Evolution of intestinal apolipoprotein B mRNA editing:
chicken apolipoprotein B mRNA is not edited but chicken enterocytes contain in-vitro
editing enhancement factor(s). J. Bioi. Chern .. in press, 1992.
60. Davies, M. S., Wallis, S. C., Driscoll, D. M., Wynne, J. K., Williams, G. W.,
Powell, L. M., and Scott, J., Sequence requirements for apolipoprotein B RNA editing
in transfected rat hepatoma cells, J. Bioi. Chern., 264, 13395, 1989.
Davidson 317
61. Shah, R. R., Knott, T. J., Legros, J. E., Navaratnam, N., Greeve, J. C., and Scott,
J., Sequence requirements for the editing of apo1ipoprotein B mRNA, J. Bioi. Chern.,
266, 16301, 1991.
62. Navaratnam, N., Patel, D., Shah, R. R., Greeve, J. C., Powell, L. M., Knott, T. J.,
and Scott, J., An additional editing site present in apo1ipoprotein B mRNA, Nucleic
Acids Res., 19, 1741, 1991.
Chapter 14
REGULATION OF THE 68-kDa HEPATIC CARNITINE

PALMITOYLTRANSFERASE
PaulS. Brady, Roderick A. Barke, and Linda J. Brady
TABLE OF CONTENTS
I. Introduction ...................................................... 320

A. General Aspects of Fatty Acid Oxidation in
Liver ..................................................... 320
B. General Aspects of CPT .................................. 320
C. Problems in the Study of CPT ............................ 321
D. Which CPT Are We Dealing With? ...................... 323
II. Regulation of Catalytic Efficiency of CPT ....................... 324

A. Regulation by Malonyl CoA .............................. 324
B. Regulation by Phosphorylation ........................... 326
C. Regulation by Alteration of Lipid Domain ............... 326
III. Regulation of Mass of 68-kDa CPT .............................. 327

A. Purification and Antibody Generation .................... 327
B. Cloning and Sequencing .................................. 328
C. Regulation of CPT Gene Expression ..................... 329
D. Promoter Analysis ........................................ 329
IV. Future Directions to Define the Regulation of Fatty

Acid Oxidation .................................................. 332
References ............................................................... 333
0-8493-6961-4/93/$0.00 +$.50
I. INTRODUCTION
A. GENERAL ASPECTS OF FATTY ACID OXIDATION IN LIVER

When a long-chain fatty acid enters the liver, a number of metabolic fates
are available to it (Figure 1). The major fate of a long-chain fatty acid, such
as palmitic acid, is activation by esterification with coenzyme A. This is
catalyzed by one member of the family of acyl CoA synthetases. A minor
fate of long chain fatty acids, which apparently does not require this initial
CoA activation, is microsomal w-oxidation. The main pathways of acyl CoA
metabolism are [3-oxidation in the mitochondria or esterification to either
triglyceride or phospholipid in the cytosol.
The mitochondrial [3-oxidative pathway accounts for 90% of total cellular
[3-oxidation of long chain fatty acids. Under most conditions, peroxisomal
chain shortening of long chain fatty acids represents a small fraction of long
chain fatty acid [3-oxidation. However, under certain conditions, such as high
dietary fat intake or administration of peroxisome proliferating drugs, the
peroxisomal contribution to long chain fatty acid [3-oxidation may increase
to 20 to 25% or more of total [3-oxidation. 1
The camitine acyltransferases are a family of enzymes involved in fatty
acid transport within the cell. The nomenclature of the camitine acyltrans-
ferases relies on the purported chain length specificity of the various enzymes.
Generally, short (camitine acetyltransferases), medium (camitine octanoyl-
transferases), and long chain (camitine palmitoyltransferases) acyltransferases
are recognized. However, the chain length specificity of the medium and long
chain enzymes is quite broad, and the specificity overlaps to such an extent
that chain length specificity per se cannot be used to distinguish these en-
zymes.2·3
A long chain camitine acyltransferase, carnitine palmitoyltransferase
(CPT), is located in the mitochondrion only. CPT regulates the flux of long
chain fatty acids into the mitochondria. This enzyme is the focus of the present
review. However, long to medium chain acyltransferases, camitine octan-
oyltransferases, are located in peroxisomes and endoplasmic reticulum. Short
chain acyltransferases, carnitine acetyltransferases, are located in mitochon-
dria and peroxisomes. Bieber has reviewed the functions and regulation of
these latter two acyltransferases. 2
B. GENERAL ASPECTS OF CPT

Camitine palmitoyltransferase (CPT; E.C. 2.3.1.23) catalyzes the re-
versible reaction:
long chain acyl-CoA

+ 1-camitine ~ long chain acyl-1-camitine + CoASH (1)
The inner mitochondrial membrane is impermeable to long chain acyl CoAs,

while long chain acylcamitines can readily traverse this barrier. Thus, CPT
Brady et al. 321
ro-Oxidation
(microsomal)
~ Mitochondrial ~-Oxidation ~ Desaturation, Elongation (microsomal)

~ Peroxisomal ~-Oxidation ~Triglyceride and Phospholipid
Synthesis (microsomal)
FIGURE 1. Pathways of hepatic fatty acid utilization. When a long chain fatty acid (here
represented as palmitic acid) enters the liver, two major metabolic fates are available, oxidation
via the mitochondria or peroxisomes and reesterification to triglyceride or phospholipid.
modulates the flux of long chain acyl CoA into the mitochondria. Figure 2
presents a classical view of CPT function within the mitochondria. It is clear
that CPT activity is found on the inner surface of the mitochondrial inner
membrane. This activity has been referred to as CPT-B, CPT-II, CPT2 , or
CPT; (for CPT inner). It is also clear that CPT activity is found on the cytosolic
side of the mitochondrion. The outer activity has been referred to as CPT-
A, CPT-I, CPT,, or CPTa (for CPT outer). Further, the outer CPT is inhibited
by malonyl CoA, a substrate in the fatty acid synthetic pathway. Unfortu-
nately, while localization of activities is possible, it remains very difficult to
determine how many CPT proteins are involved in producing these CPT
activities. The problem of definition of the number of distinct CPT proteins,
as well as the problems mentioned in the following section, has confounded
efforts to study the regulation of CPT at the gene level. Only recently has a
68-kDa CPT been purified, cloned, and its regulation studied.H
C. PROBLEMS IN THE STUDY OF CPT

Mitochondrial carnitine palmitoyltransferase activities can be demon-
strated both on the cytosolic and the matrix faces of the mitochondrion.
Recently, it has been demonstrated that the outer CPT activity may be as-
sociated with the inner face of the outer mitochondrial membrane. 7 However,
it has been impossible to determine whether these topologically distinct ac-
Palmitoylcarnitine
+
Co A
1-Carnitine
+
Palmitoyi-CoA
1-Carnitine
~-Oxidation
Outer Inner
Membrane Membrane
FIGURE 2. Mitochondrial camitine palmitoyltransferase. Mitochondrial camitine palmitoyl-

transferase activities are found on the outer and inner surfaces of the mitochondrion. The outer
activity, in situ, exhibits sensitivity to malonyl CoA inhibition.
tivities represent discrete proteins or are the product of a single protein in

different membrane environments. Studies with intact mitochondria have shown
that the apparent Kms and Vma.s are quite similar, given the constraint that
both long chain acyl CoAs and long chain acylcamitines are detergents and,
therefore, cannot be used at concentrations high enough to calculate accurate
kinetic constants. 2 The outer CPT activity is inhibited by malonyl CoA, while
the inner activity is not.
The first major difficulty encountered in defining the number of CPTs
is that there is no intrinsic difference in the reaction catalyzed. The reaction
(1) is freely reversible, and the kinetic constants observed on either face of
Brady et al. 323
the mitochondria are not sufficiently different to distinguish two enzymes.

Further, the mitochondrial inner membrane has an asymmetric lipid distri-
bution. 8 The specific lipid composition of the membrane face contributes
significantly to the measured characteristics of the enzyme activity. 9 . 12 His-
torically, a number of investigators have attempted to attribute the CPT ac-
tivities to separate proteins based on differential sensitivity to a variety of
inhibitors or differential solubilization by detergents. 13 • 14 However, such ap-
proaches generally compare intact mitochondria to mitochondria treated with
detergent or other mechanism to expose the inner CPT. Invariably, this leads
to some form of disruption of the mitochondrial membrane, which, itself,
alters CPT Kms, Vmax• and sensitivity to malonyl CoA.
Some years ago we employed the approach of inverting mitochondrial
vesicles. This approach avoids the more obvious artifacts. In this type of
study, the inner surface of the mitochondria exhibits sensitivity to bromo-
palmitoyl CoA and 2-tetradecylglycidyl CoA, inhibitors commonly proposed
to distinguish the two CPT proteins. In fact, the purified 68-kDa CPT protein,
which is clearly at least the inner CPT (whether it is the one and only CPT
is subject to debate), can be induced to recover at least partial 2-tetradecyl-
glycidyl CoA sensitivity by the addition of dimethyl sulfoxide. 12 Still, under
no experimental conditions has any investigator reported malonyl CoA sen-
sitivity of the purified 68-kDa CPT alone.
The second key problem is that removal of the protein from the membrane
alters the protein's characteristics. To date, only one CPT protein has been
isolated from mitochondria. 2 •15 • 16 The protein has aM, of 68 to 69 kDa and
lacks malonyl CoA sensitivity. Antibody generated to this protein reacts with
a 68 to 69-kDa band on Western blotting and precipitates only a single mRNA
translation product. All of the CPT activity solubilized from hepatic mito-
chondria by either Tween 20 or by sonication is precipitable by antibody
generated to this protein. 17• 18 Unfortunately (for the sake of simplicity), this
does not preclude the possibility that another CPT protein, the outer, malonyl
CoA-sensitive CPT, is lost due to the various solubilization schemes. Indeed,
this possibility has been proposed.
D. WHICH CPT ARE WE DEALING WITH?

To date a single mitochondrial CPT has been cloned, the precursor to the
68-kDa CPT. This protein appears to correspond to the inner CPT. A second
protein, which binds malonyl CoA or 2-tetradecylglycidyl CoA, has been
isolated from mitochondrial outer membrane. 19 - 21 This 85- to 95-kDa protein
has not been demonstrated to have catalytic activity. In studies described
below, combining this protein with the 68-kDa CPT results in malonyl CoA-
sensitive CPT activity. 22 •23
Ghadiminejad and Saggerson24 have shown cross-reactivity of the 68-
kDa CPT and a larger molecular weight protein associated with the mito-
chondrial inner membrane on Western blots. Our own blots of total mito-
chondria often show two bands, particularly where CPT activity is induced.
Calculation of M, led us to conclude that the less abundant, upper band
represented the 68-kDa precursor. In support of this, mRNA yields a single
immunoprecipitable band with M, corresponding to the higher weight band
observed on Western blots. Hoppel recently presented the possibility that
outer membrane malonyl CoA binding protein and inner membrane 68-kDa
CPT are associated at mitochondrial membrane pores or junctions. 25 Such
junctions are known to serve as sites for protein import. His suggestion may,
indeed, explain the enrichment of the outer mitochondrial membrane with a
68-kDa CPT precursor. Can we, then, be sure that there is a single mito-
chondrial CPT protein associated with a malonyl CoA-binding regulatory
protein? The answer remains an unfortunate no. The good fortune in this is
that the activity of the 68-kDa CPT is regulated. If this is the only mito-
chondrial CPT, we are well on our way to understanding its regulation. If
there is a second mitochondrial CPT, the data available from in situ studies
indicate that it too is regulated; that is, outer CPT activity changes. It would,
then, be important to determine how this change in activity is mediated. Such
data are currently available for the 68-kDa CPT.
Regulation of the activity of any enzyme can be divided into a number
of events (Figure 3). In the most general sense, these events can be subdivided
into (1) short-term events which alter catalytic efficiency but do not alter the
quantity of protein, and (2) longer term events which specifically alter protein
quantity. If protein quantity is regulated, this regulation can be accomplished
at a number of levels. As with many regulatory enzymes, CPT exhibits both
forms of regulation. The specific difficulty which arises with CPT is that the
majority of in situ studies examine the outer CPT activity, while all studies
of purified CPT involve the 68-kDa protein.
II. REGULATION OF CATALYTIC EFFICIENCY OF CPT
Several forms of regulation of catalytic activity of camitine palmitoyl-

transferase have been reported. Competitive inhibition (with palmitoyl CoA)
by malonyl CoA of camitine palmitoyltransferase was the first reported mech-
anism of control of this enzyme. 26 More recently, other events, such as
phosphorylation and membrane fluidity, have been suggested to exert some
level of control. Unfortunately, the impact of each event on the overall flux
through CPT is undefined.
A. REGULATION BY MALONYL CoA

Control of camitine palmitoyltransferase by malonyl CoA inhibition is
the oldest reported mechanism of control. However, in the 14 years or so
since the initial report, the mechanism remains controversial.
Malonyl CoA sensitivity has been presumed to be a characteristic of the
outer CPT activity. However, all studies employed some means of disrupting
Brady et al. 325
~no~ ••
• Transcription
••
•
Primary
RNA
Transcript
•
: Post-transcriptional
• Processing
••
mRNA
in
Cytosol
•
Translation :
•
Nascent
Protein
Mature
Protein
••
••
Degradation ~· • • • •• • • •••
FIGURE 3. Control of protein activity. Control of carnitine palmitoyltransferase activity (or

the activity of any enzyme) can be controlled at a variety of levels, from the rate of transcription
of RNA to interaction of the protein with specific inhibitors.
the mitochondrion, be it detergent lysis or sonication, in demonstrating that

the inner CPT activity is not inhibited by malonyl CoA. Our own studies with
inverted mitochondrial vesicles, which should represent the least disrupted
case, also failed to show malonyl CoA sensitivity of the inner CPT activity. 9
It may be that the outer and inner surfaces talk to each other through some
physical connection. In unpublished studies, exposure of intact mitochondria
to malonyl CoA, followed by inversion and washing, resulted in some (25
to 50%) decline in the inner CPT activity. We had previously suggested such
a connection in studies with 2-tetradecylglycidyl CoA (TDGA-CoA). 12
2-Tetradecylglycidyl CoA is presumed to inhibit only the outer CPT
activityY However, if one exposes the intact mitochondria to TGDA-CoA,
washes them, and then inverts them, the inner CPT activity is inhibited. 12
Still, we also reported that the inner CPT activity is, of itself, sensitive to
this inhibitor. Further, the purified 68-kDa CPT protein can be rendered
TGDA-CoA-sensitive under proper conditions. Thus, there was always the
possibility that TDGA-CoA leaked from the outer to the inner surface.
Current and ongoing work from two groups 22 •23 demonstrates that the
combining of 68-kDa CPT and outer mitochondrial membrane or malonyl
CoA-binding protein, under defined conditions, results in a malonyl CoA-
sensitive CPT activity. These same conditions neither result in CPT activity
of the outer membrane preparation in the absence of the 68-kDa CPT nor
malonyl CoA sensitivity of the 68-kDa CPT in the absence of the outer
membrane preparation. Whether they are observing regained CPT activity of
the outer membrane malonyl CoA-binding protein or recovered malonyl CoA
sensitivity of the 68-kDa CPT is unclear. Putting aside the question of one
vs. two CPT proteins, these data would serve to confirm the observation that
mitochondrial outside and inside could communicate relative to a malonyl
CoA inhibition of CPT activities on both faces.
While it remains uncertain how malonyl CoA sensitivity is conferred, a
recurring theme in CPT regulation has been that when CPT activity increases,
malonyl CoA sensitivity declines. This effect is likely quite important in
governing total flux through CPT.
B. REGULATION BY PHOSPHORYLATION
Harano and co-workers have reported that CPT is phosphorylated in
response to cAMP in cultured cells. 28 To our knowledge, this initial obser-
vation has not been pursued. It remains unclear if the phosphorylation has
physiologic significance, but the possibility of this form of regulation will
need to be further investigated.
C. REGULATION BY ALTERATION OF LIPID DOMAIN

As mentioned above, the mitochondrial membranes represent a highly
complex, highly structured, and incredibly asymmetric domain. That CPT
activity is regulated is unquestioned. However, we tend to neglect the mem-
brane association of the activities. Or, if we do consider membrane associ-
ation, it is as an inconvenience in the purification of the protein.
The membrane characteristics do appear to alter CPT activities on both
inner and outer mitochondrial surfaces. This can be demonstrated by two
means. First, altering membrane fluidity, with agents such as ethanol or
benzoyl alcohol, results in increased CPT activity. 9 As mentioned above, the
mitochondrial inner membrane has a very asymmetrical phospholipid distri-
Brady eta!. 327
bution, with a very high cardiolipin content on the inner face. Exposure of
the outer face to cardiolipin also results in increased CPT activity. 11 Second,
a number of physiologic states where CPT activity is elevated, such as star-
vation or severe streptozotocin-diabetes in the rat. also produce increased
mitochondrial fluidity. 12 Thus, membrane fluidity can and does have an effect
on CPT activity without an effect on malonyl CoA sensitivity. The extent to
which membrane fluidity does contribute to overall CPT regulation remains
uncertain. More recent observations on the regulation of the 68-kDa CPT
regulation show that a primary mechanism of control is via the control of
protein synthesis. At the same time, the largest observed increase in 68-kDa
CPT protein is fivefold with a two- to threefold increase a more typical
observation. In assessing the overall control of flux through CPT, the com-
bination of short-term and long-term effects must be considered.
III. REGULATION OF MASS OF 68-kDa CPT
A. PURIFICATION AND ANTIBODY GENERATION

To date, only a single 68-kDa mitochondrial camitine palmi toy !transferase
has been isolated. This protein is generated from a larger precursor, as is
typical of proteins inserted into the mitochondrial membrane. 2930 The puri-
fication of the protein has allowed the generation of antibody, which, in turn,
has allowed quantification of CPT protein. Antibody to the 68-kDa protein
has been generated in a number of laboratories, and the exchange of antibody
has introduced some uniformity into the study of CPT. 15 17 Antibody to the
68-kDa CPT is effective in precipitating the protein for turnover and translation
studies, and in quantifying the protein by Western blot. 17·29 The polyclonal
antibody, regardless of source, is not inhibitory to catalytic activity of the
CPT. This indicates that the catalytic site is not available, probably because
it represents a hydrophobic pocket.
Increases in CPT activity have been shown to closely follow CPT quantity.
Consequently, we believe that, although other factors such as phosphorylation
or changes in the lipid environment have a limited effect, the major mechanism
of control of the 68-kDa CPT must be protein synthesis. The generation of
an effective antibody has allowed us to directly measure protein turnover, as
well as protein quantity. In a number of studies, we have now shown that
the synthesis of the 68-kDa carnitine palmitoyltransferase is rapidly increased
by a variety of stimuli, including diabetes in vivo 31 and cAMP administration
in cell systems. 32 Alterations in degradation are not observed. This increase
in synthesis can be linked directly to increases in translation rate.
If communication of malonyl CoA sensitivity to the inner enzyme is
proven possible, or if the 68-kDa CPT proves to be the only mitochondrial
CPT protein, then malonyl CoA may share an equal role in the overall reg-
ulation of this enzyme.
Carnitine Octanoyltransferase
593
659
FIGURE 4. Relationship of camitine palmitoyltransferase and camitine octanoyltransferase

amino acid sequences. The amino acid sequences of CPT and COT were compared without
allowing conservative amino acid substitution and looking for 8 amino acid segments showing
at least 75% conservation. The specific amino acid sequences are shown.
B. CLONING AND SEQUENCING

The rat liver 68-kDa camitine palmitoyltransferase has now been cloned
by two groups 4 ·5 and the sequence of the corresponding human enzyme has
also been reported by a third. 6 All three clones are in generally good agree-
ment, although we observe a somewhat longer rat CPT sequence than do
Woeltje et al. The rat and human enzymes show a high degree of homology. 6
The peroxisomal camitine octanoyltransferase has also been cloned and se-
quenced. 33 The peroxisomal enzyme is not membrane associated. The com-
parison of the amino acid homology and hydrophobicity for these two amino
acid sequences are presented (Figure 4). The sequences are not strikingly
similar. There are three small regions with strong homology: an early region
of 26 amino acids, a small central region of 6 amino acids, and a region
toward the carboxy terminal of 9 amino acids. This is not too surprising,
since antibodies to these proteins do not cross-react. The hydrophobicity plots
(Kyte-Doolittle) are quite similar, reflecting the observation that the 68-kDa
camitine palmitoyltransferase membrane protein is not strikingly hydrophobic.
The value of a eDNA is twofold. First, as seen above, we can predict an
amino acid sequence, make structural predictions, and begin to evaluate the
protein configuration. The value of this is that we can assess the specific
critical catalytic components. Second, we can employ the eDNA as a tool to
examine changes in mRNA levels and transcriptional activity in various states.
As we have seen, CPT activity, CPT protein level, and CPT translation rate
are increased in response to starvation, 4 •33 diabetes, 31 and peroxisome prolif-
erator administration, 17•30 and decreased in response to sepsis. 34 We can pro-
Brady et al. 329
ceed to determine if these changes are the product of alterations in the quantity
of mRNA and, ultimately, the product of alterations in transcription rates.
C. REGULATION OF CPT GENE EXPRESSION

It has been possible to determine CPT activity in situ for approximately
20 years. It is only recently that we have been able to determine the specific
effects of various perturbations on the inner CPT protein through the use of
antibody. The inner CPT level responds to a wide variety of agents. Agents
which increase inner CPT protein include starvation and diabetes, mediated
through glucagon/cAMP and insulin. However, in addition to this relatively
well-studied control axis, inner CPT protein is also increased in response to
a variety of agents which proliferate peroxisomes or which interfere with
normal fatty acid oxidation (e.g., clofibrate treatment, 30 riboflavin defi-
ciency).4 CPT activity and protein levels are decreased in sepsis. 34 The septic
state represents a complex set of hormone-mediated responses, not only to
glucagon/insulin, but potentially to cytokines, prostaglandins, and other met-
abolic mediators.
The determination of the eDNA sequence for the 68-kDa CPT allows
direct quantification of the mRNA levels for this protein (by dot or Northern
blot), as well as the transcription rate (by nuclear run-on assay). A selection
of available data is presented in Table 1. In all cases studied to date, changes
in CPT activity are directly associated with changes in the quantity of protein.
The quantity of protein is a product of the level of mRNA, which, in turn,
is a product of the nuclear transcription rate. Simply stated, the regulation of
68-kDa protein levels must be controlled at the transcriptional level. Subse-
quent analysis is aimed at determining how regulation is accomplished.
D. PROMOTER ANALYSIS
The regulation of the 68-kDa CPT protein must, therefore, involve in-
teraction of specific nuclear regulatory proteins with the promoter region for
the CPT gene. Such genomic regulatory elements are generally found up-
stream or 5' to the structural elements. Therefore, we have undertaken the
isolation of promoter regions for both the rat and human 68-kDa CPTs. 35
Given the high degree of homology of the eDNA for these two proteins, the
probability of successful simultaneous isolation seemed high. Using poly-
merase chain reaction approaches to the isolation of these promoters, we were
able to initially isolate regions of approximately 500 bp upstream from the
mRNA coding region. The 68-kDa CPT, and CPT activity, in general, are
quite interesting because not only do they respond to expected stimuli, such
as cAMP-mediated induction, but, also, they respond to apparently unrelated
stimuli, such as the peroxisomal proliferating agents. We have taken three
simultaneous approaches to the study of the promoter. First, we have examined
the promoter region directly using sequence comparison to known consensus
nuclear protein binding elements, footprinting with purified nuclear binding
TABLE 1 w
w
0
Regulation of the Hepatic Mitochondrial 68-kDa Carnitine Palmitoyltransferase
68-kDa CPT Protein 68-kDa CPT

State Model (immunoreactive) 68-kDa CPT mRNA Transcription Ref.
Starvation Rat i t i 17
Diabetes Rat ti i i i t 17,31
Riboflavin deficiency Rat ii i i NO 4
cAMP Rat i t i t ii 36
H4IIE Cell ii i t NO 32
Sepsis Rat t t NO 34
Interleukin-6 H4IIE Cell t i NO 37
DEHP Rat ti t i ti i t t 5,15,17
Clofibrate Rat t i ii t 38
Note: The quantity of 68 kDa carnitine palmitoyltransferase protein is regulated primarily at the transcriptional level.
Increases in the protein, determined by densitometric scanning of Western blots. increases in mRNA. determined
by blot or translational activity, and transcription rates correlate quite well. The table illustrates the approximate
degree of response ( j = 1.5- to 2-fold increase; j t = 2- to 3-fold increase; i i i 2: 4-fold; L = 50% ~
decrease; NO = not determined). The table also indicates something of the variety of stimuli to which the 68- ~-
kDa CPT responds.
c;·
:::::
1::>
:::::
1::>...
c:J
~
:::::
~
~
""\::i
~
"'c;·
"':::::
Brady eta/. 331
-500 +1
A.
B.
Acute
Phase
c.
FIGURE 5. Mapping the 68-kDa camitine palmitoyltransferase promoter. We have isolated
an approximately 600-bp region of the CPT promoter: (A) shows the areas of protection by rat
liver nuclear extract in footprint analysis; (B) represents regions of protection by purified nuclear
binding proteins CREB, C/EBPa, and C/EBP~; (C) shows the predicted binding regions for a
variety of binding proteins based on consensus nucleotide sequences in the promoter.
proteins where available, and gel mobility shift assays. These studies have
indicated that the 68-kDa CPT promoter contains a single cAMP response
element (CRE), binding sites for C/EBPa (CCAAT/enhancer binding protein),
and for C/EBPI3 (the DBP-IL6 dependent binding protein). Studies with rat
liver nuclear extracts also define a large region of binding which incudes a
C/EBPa and a C/EBP binding region, but one which is much broader and
more complex, containing consensus binding elements for other proteins,
which we have not directly tested. A map of the CPT promoter is presented
(Figure 5).
The second major approach to promoter analysis is functional. The CPT
promoter has been introduced into a plasmid containing a reporter gene,
chloramphenicol acetyltransferase. The CPT promoter is then introduced into
HepG2 cells and its response to various stimuli examined. These can include
hormonal factors as cAMP or thyroid hormone, peroxisomal proliferators,
such as clofibrate, or expression vectors for nuclear regulatory proteins, such
as c-fos, c-jun, or C/EBPa. 35
The final approach represents a descriptive analysis of changes in the
level of nuclear binding proteins. Rat liver or cell nuclear extracts are prepared
following stimulus with cAMP or peroxisomal proliferators. The changes in
specific nuclear proteins can be assayed by Western blotting. Preliminary
data indicate that c-fos, c-jun, and C/EBPI3 increase in sepsis.
The difficulty with the analysis of any promoter is that what may bind
can be easily defined. What does bind is more difficult to determine. Current
and ongoing work in this area is aimed at answering this question.
IV. FUTURE DIRECTIONS TO DEFINE THE

REGULATION OF FATTY ACID OXIDATION
In our initial enthusiasm, many of us had assumed that the isolation of

the eDNA for the 68-kDa CPT would allow us to resolve the question of how
many mitochondrial CPTs truly exist. So far, this has not been the case. What
we have resolved is that changes in 68-kDa CPT activity can be directly
attributed to changes in transcription of 68-kDa CPT mRNA. However, if a
second mitochondrial CPT exists, it may be sufficiently different from the
68-kDa CPT that the antibody to one does not cross-react with the other. This
degree of difference exists between the 68-kDa mitochondrial CPT and the
peroxisomal COT, and molecular probes for one protein do not detect the
other protein. Have the tools of molecular biology let us down in our desire
to distinguish the events which produce outer and inner CPT activity? We
would hold that this is not so. We have not yet asked the correct question
for the available tools; we need to begin to apply the tools that we have. A
number of groups remain in the search for the still elusive outer CPT protein.
However, the immediately approachable question is whether the 68-kDa CPT
represents the only mitochondrial CPT protein. Since we cannot yet find the
other CPT protein, perhaps the approach should be to remove the protein that
we do recognize and to see what activity remains. Methods are available for
this, using antisense probes to the 68-kDa CPT. While we do not yet have
an answer, we are currently using this approach to contribute to our under-
standing.
In a much broader sense, the mitochondrial J3-oxidative pathway must
interact with peroxisomal and microsomal fatty acid oxidative events. CPT
activities are part of a family of enzymes including several other long chain
and short chain carnitine acyltransferase activities located in these other sub-
cellular compartments. The long chain carnitine acyltransferases are all ma-
lonyl CoA-sensitive and are all increased by starvation and diabetes. However,
the metabolic reasons for, and the consequences of, this subcellular com-
partmentation are not known. The proteins have not all been successfully
isolated. Nonetheless, isolation and subsequent cloning appear imminent.
With the development of molecular tools for this entire cast of carnitine
acyltransferases, we may be able to examine their interaction by selective
overexpression or suppression. That is, by placing the coding region for the
specific protein under the control of a strong transfectable promoter, we can
generate high levels of protein in cell systems. Alternatively, use of antisense
probes may allow us to eliminate the activity of a specific protein in cell
systems. Such approaches will allow us to examine both the response of the
other related enzymes to this perturbation and the response of the overall
metabolic processes of fatty acid oxidation and esterification. Through such
approaches we can begin to study the role of each enzyme in the whole
organism.
Brady et al. 333
REFERENCES
I. Kondrup, J. and Lazarow, P. B., Peroxisomal 13-oxidation in intact rat hepatocytes:
quantitation of its flux, Ann. N.Y. Acad. Sci., 386, 404, 1982.
2. Bieber, L. L., Carnitine, Annu. Rev. Biochem., 57, 261, 1988.
3. Ramsay, R. R., Microsomal, mitochondrial and peroxisomal carnitine palmitoyltrans-
ferase activities change differently in starvation and diabetes, Biochim. Biophys. Acta,
(submitted.)
4. Brady, P. S., Feng, Y.-X., and Brady, L. J., Transcriptional regulation of carnitine
palmitoyltransferase synthesis in riboflavin deficiency in rats, J. Nutr., 118, 1128, 1988.
5. Woeltje, K. F., Esser, V., Weis, B. C., Sen, A., Cox, W. F., McPhaul, M. J.,
Slaughter, C. A., Foster, D. W., and McGarry, J.D., Cloning, sequencing, and
expression of a eDNA encoding rat liver mitochondrial carnitine palmitoyltransferase II,
1. Bioi. Chern., 265, 10720, 1990.
6. Finocchiaro, G., Taroni, F., Rocchi, M., Martin, A. L., Colombo, I., Tarelli, G. T.,
and DiDonato, S., eDNA cloning, sequence analysis, and chromosomal localization of
the gene for human carnitine palmitoyltransferase, Proc. Nat/. Acad. Sci. U.S.A., 88,
661, 1991.
7. Murthy, M.S. R. and Pande, S. V., Malonyl-CoA binding site and the overt carnitine
palmitoyltransferase activity reside on the opposite sides of the outer mitochondrial mem-
brane, Proc. Nat/. Acad. Sci. U.S.A., 84, 378, 1987.
8. Nicolay, K., Timmers, R., Spoelstra, E., VanderNeut, R., Fok, J., Huigen, Y.,
Verkleij, A., and deKruiff, B., Biochim. Biophys. Acta, 778, 359, 1984.
9. Brady, L. J,, Silverstein, L. J., Hoppel, C. L., and Brady, P. S., Hepatic mitochondrial
inner membrane properties and carnitine palmitoyltransferase A and B: effect of diabetes
and starvation, Biochem. 1., 232, 445, 1985.
10. Brady, L. J., Hoppel, C. L., and Brady, P. S., Hepatic mitochondrial inner membrane
properties, 13-oxidation and carnitine palmitoyltransferases A and B, Biochem. 1., 233,
427, 1986.
II. Brady, L. J, and Brady, P. S., Hepatic and cardiac carnitine palmitoyltransferase
activity: effects ofadriamycin and galactosamine, Biochem. Pharmacal., 36, 3419, 1987.
12. Brady, P. S. and Brady, L. J., Action in vivo and in vitro of 2-tetradecylglycidic acid,
2-tetradecylglycidyl-CoA and 2-tetradecylglycidylcarnitine on hepatic carnitine palmi-
toyltransferase, Biochem 1., 238, 801-809, 1986.
13. Yates, D. W. and Garland, P. B., Carnitine palmitoyltransferase activities
(EC 2.3.1.-) of rat liver mitochondria, Biochem. 1., 119,547, 1970.
14. Hoppel, C. L. and Brady, L., Carnitine palmitoyltransferase and transport of fatty acids,
in The Enzymes of Biological Membranes, Martonosi, A., Ed., Vol. 2, Plenum Press,
NY, 1985, 139.
15. Miyazawa, S., Ozasa, H., Osumi, T., and Hashimoto, T., Purification and properties
of carnitine octanoyltransferase and carnitine palmitoyltransferase from rat liver, 1. Biochem.
(Tokyo), 94, 529, 1983.
16. Clarke, P.R. H. and Bieber, L. L., Isolation and purification of mitochondrial carnitine
octanoyltransferase from beef heart, 1. Bioi. Chern., 256, 9861, 1981.
17. Brady, P. S. and Brady, L. J,, Hepatic carnitine palmitoyltransferase turnover and
translation rates in fed, starved, streptozotocin-diabetic and diethylhexylphthalate-treated
rats, Biochem. 1., 246, 641, 1987.
18. Kerner, J, and Bieber, L. L., Studies on the association of b-oxidation enzymes with
CPTi and CPTo of heart mitochondria, FASEB 1., 4 (Abstr. 3104), A802, 1990.
19. Declerq, P. E., Falck, J. R., Kuwajima, M., Tyminski, H., Foster, D. W., and
McGarry, J. D., Characterization of the mitochondrial carnitine palmitoyltransferase
enzyme system: use of inhibitors, 1. Bioi. Chern., 262, 9812, 1987.
20. Woeltje, K. F., Kuwajima, M., Foster, D. W., and McGarry, J.D., Characterization
of the mitochondrial camitine palmitoyltransferase enzyme system: use of detergents and
antibodies, J. Bioi. Chem., 262, 9822, 1987.
21. Zammit, V. A., Corstorphine, C. G., and Kolodziej, M. P., Evidence for distinct
functional molecular sizes of camitine palmitoyltransferases I and II in rat liver mito-
chondria, Biochem. J., 263, 89, 1989.
22. Ghagiminejad, I. and Saggerson, E. D., Camitine palmitoyltransferase (CPT2) from
mitochondrial inner membrane becomes inhibitable by malonyl-CoA if reconstituted with
outer membrane malonyl-CoA binding protein, FEBS, 269, 406, 1990.
23. Chung, C., Woldegiorgis, G., and Bieber, L. L., Restoration of malonyl-CoA sen-
sitivity to purified rat heart mitochondrial camitine palmitoyltransferase by addition of
protein fraction(s) from an 86 kD malonyl-CoA binding affinity column, FASEB J., 5
(Abstr. 1290), A592, 1990.
24. Ghagiminejad, I. and Saggerson, E. D., The relationship of rat liver overt camitine
palmitoyltransferase to the mitochondrial malonyl-CoA binding entity and to the latent
palmitoyltransferase, Biochem. J., 270, 787, 1990.
25. Hoppel, C. L., Camitine palmitoyltransferases, in Current Concepts in Carnitine Re-
search, Carter, A., Ed., CRC Press, Boca Raton, FL, 1992, 153.
26. McGarry, J, D., Leatherman, G. F., and Foster, D. W., Camitine palmitoyltransferase.
I, J. Bioi. Chem., 253, 4128, 1978.
27. Kiorpes, T. C., Hoerr, D., Ho, W., Weaner, L. E., Inman, M. G., and Tutwiler,
G. F., Identification of 2-tetradecylglycidyl coenzyme A as the active form of methyl 2-
tetradecylglycidate (methyl palmoxirate) and its characterization as an irreversible, active
site-directed inhibitor of camitine palmitoyltransferase A in isolated rat liver mitochondria,
J. Bioi. Chem., 259, 9750, 1984.
28. Harano, Y., Kashiwago, H., Kojima, H., Suzuki, M., Hashimoto, T., and Shigeta,
Y., Phosphorylation of camitine palmitoyltransferase and activation by glucagon in iso-
lated rat hepatocytes, FEBS Lett., 188, 267-272, 1985.
29. Ozasa, H., Miyazawa, S., and Osumi, T., Biosynthesis of camitine octanoyltransferase
and camitine palmitoyltransferase, J. Biochem. (Tokyo), 94, 543, 1983.
30. Brady, P. S., Marine, K. A., Brady, L. J., and Ramsay, R. R., Co-ordinate induction
of hepatic mitochondrial and peroxisomal camitine acyltransferase synthesis by diet and
drugs, Biochem. J., 260, 93, 1989.
31. Brady, L. J, and Brady, P. S., Regulation of camitine palmitoyltransferase synthesis
in spontaneously diabetic BB Wistar rats, Diabetes, 38, 65,
32. Wang, L., Brady, P. S., and Brady, L. J., Turnover of camitine palmitoyltransferase
mRNA and protein in H4IIE cells: effect of cyclic AMP and insulin, Biochem. J., 263,
703, 1989.
33. Chatterjee, B., Song, C. S., Kim, J, M., and Roy, A. K., Cloning, sequencing and
regulation of rat liver camitine octanoyltransferase: transcriptional stimulation of the
enzyme during peroxisome proliferation, Biochemistry, 27, 9000, 1988.
34. Birklid, S.D., Brady, P. S., Brady, L. J., and Barke, R. A., Hepatic gene expression
in a rat intra-abdominal sepsis model: fatty acid oxidation, FASEB J., 4 (Abstr. 2250),
A654, 1990.
35. Brady, P. S., Liu, J.-S., Park, E. A., Hanson, R. W., and Brady, L. J., Isolation
and characterization of the promoter for the gene coding for camitine palmitoyltransferase,
Biochem. J., 286, 779, 1992.
36. Brady, L. J, and Brady, P. S., Regulation of camitine palmitoyltransferase by glucagon
and insulin, Biochem. J., 258, 677, 1989.
37. Barke, R. A., Brady, L. J., and Brady, P. S., The Ca2+ second messenger system
and IL-l-alpha modulation of hepatic gene transcription and mitochondrial fatty acid
oxidation, Surgery, 110, 285, 1991.
38. Brady, L. J., Ramsay, R. R., and Brady, P. S., Regulation of camitine acyltransferase
synthesis in lean and obese Zucker rats by DHEA and clofibrate, J. Nutr., 121, 525,
1991.
Chapter 15
LONG- TERM REGULATION OF HEPATIC

GLUTAMINASE AND THE UREA CYCLE ENZYMES
Malcolm Watford
TABLE OF CONTENTS
I. Introduction ...................................................... 336
II. Sources and Roles of Glutamine ................................. 336
III. Pathways of Glutamine Metabolism .............................. 337

A. Glutamine Synthetase .................................... 337
B. Glutaminase .............................................. 337
C. Hepatic Glutamine Metabolism and the Urea
Cycle ..................................................... 338
IV. Regulation of Glutaminase Activity .............................. 340

A. Short-Term Regulation ................................... 340
8. Long-Term Regulation ................................... 341
V. Regulation of Expression of Glutaminase Genes ................. 343

A. Glutaminase cON As ........................ " ............ 343
8. Heterogeneity of Expression in the Liver ................. 343
C. Regulation of Expression ................................. 344
D. Coordinate Expression with the Urea Cycle
Enzyme Genes ........................................... 345
VI. Summary ........................................................ 346
References ............................................................... 34 7
0-8493-6961-4/93/$0.00 +$.50
1993 by CRC Press. Inc. 335
I. INTRODUCTION
Glutamine is not a dietary essential amino acid, but it does play an

important role within the body, and recent developments in the treatment of
several pathological states have resulted in the provision of large amounts of
dietary glutamine, either enterally or parenterally. 1"3 A major fate of glutamine
within the body is catabolism via the hepatic pathways of gluconeogenesis
and urea synthesis. This review focuses on the long-term regulation of glu-
tamine catabolism within the liver through changes in the activities of phos-
phate-activated glutaminase and the urea cycle enzymes.
II. SOURCES AND ROLES OF GLUTAMINE
It is widely believed that ingested glutamine, at normal dietary intake

levels ( ~4 g per day), is metabolized by the small intestine and therefore
does not enter the bloodstream intact. 4 •5 Feeding of large (up to 40 g per day)
quantities of glutamine in pathological and experimental situations possibly
overloads the capacity of the intestine and may result in dietary glutamine
reaching the portal blood. Little is known regarding the fate of such excess
glutamine, but it is likely that most of it is taken up by the liver and catabolized
to urea and glucose.
Glutamine is the most abundant amino acid in the body, being present at
>20 mM in human skeletal muscle 6 and 0.5 to 0.8 mM in plasma; 6
·~ fur-
thermore, this glutamine pool is turning over rapidly, at a rate of 130 tJ-mol/
min per 100-g body weight, in the adult rat. 9 This implies a continuous and
high rate of glutamine production within the body. The major sites of glu-
tamine synthesis are skeletal muscle, 10 lungs, 11 · 12 and adipose tissue. 13 The
liver can also be a site of net glutamine production, 8 • 14 • 15 and in the rat. and
some herbivorous species, the kidneys may exhibit net glutamine synthesis. 16
For this chapter the utilization of glutamine will be reviewed without
regard to the site of production. Possibly the major, quantitative site of glu-
tamine utilization is the small intestine where glutamine is an important res-
piratory fuel for enterocytes. the end products being ammonia, citrulline.
alanine, proline. lactate, and carbon dioxide. 4 • 17 Another major site of glu-
tamine utilization is a group of "active" cells, such as lymphocytes, thy-
mocytes. macrophages, reticulocytes, and tumor cells. 5 · 1R· 19 Again, the role
of glutamine in such cells is as a respiratory substrate and these cells may
represent the largest net "organ" of glutamine catabolism within the body.
Other sites of net glutamine utilization are the kidneys in metabolic acidosis, 16
the mammary gland during lactation, 20 •21 and under some conditions the
liver. B.lo.Is.z2.23
The liver possesses both the potential for glutamine utilization and glu-
tamine syntheses. However, net exchange of glutamine across the liver is not
always observed. There is extensive evidence that both synthesis and hy-
Watford 337
drolysis of glutamine occur in the liver concomitantly and the net flux is,
therefore, the difference between these two pathways. 24 Although species
differences exist, it appears that the liver is a site of glutamine synthesis in
the normal physiological conditions 14 but changes to a site of net utilization
in diabetes, 25 •26 high protein feeding, 15 •22 exercise, 23 and during some stages
of feeding and starvation. 27 This is in keeping with its role as a direct substrate
for gluconeogenesis and urea synthesis. Glutamine also serves as an important
precursor for liver urea synthesis by the indirect pathway of glutamine ca-
tabolism, in extrahepatic cells, such as enterocytes, to alanine, ammonia, and
other amino acids which are then taken up by the liver. 8
III. PATHWAYS OF GLUTAMINE METABOLISM

A. GLUTAMINE SYNTHETASE
Glutamine synthesis is catalyzed by glutamine synthetase (EC 6.3.1.2,
L-glutamate:ammonia ligase [ATP]) which utilizes ATP to add ammonia to
glutamate (Equation 1). The activity was first described by Krebs 28 in 1935
and has recently been reviewed?9 It will not be discussed further in this
chapter.
Glutamate + NH 3 + ATP ~Glutamine + ADP + Pi (l)
B. GLUTAMINASE
The degradation of glutamine can occur by many reactions, including
many amidotransferases. However, there is only one true glutaminase, giving
rise to stoichiometric amounts of glutamate and ammonia. 30 Unfortunately,
many enzymes will produce ammonia from glutamine and therefore reliable
methods for the assay of glutaminase activity in nonpurified preparations must
rely on the measurement of glutamate production. Phosphate activated glu-
taminase, sometimes known as phosphate-dependent glutaminase or glutam-
inase l (EC 3.5.1.2, L-glutamine amidohydrolase), is the only true glutam-
inase found in mammalian tissues (Equation 2). It is the major enzyme
responsible for glutamine catabolism in the tissues mentioned in this review,
with the possible exception of lactating mammary gland. 21 ·31 It is hereafter
referred to simply as glutaminase. However, there are two isozymes of glu-
taminase, described here as liver-type and kidney-type; the latter is sometimes
referred to as brain-type in the literature.
Glutamine + H2 0 ~ Glutamate + NH: (2)
Krebs 28 was the first to describe the glutamine-hydrolyzing activity of

mammalian tissues. He also noted the existence of the two isozymes differing
in pH optima and sensitivity to glutamate inhibition. Kidney-type glutaminase
has a relatively low Km for glutamine, a low affinity for phosphate, and is
subject to inhibition by low concentrations of glutamate. 2 1.3 2 ·33 Liver-type glu-

taminase exhibits a higher Km for glutamine, higher affinity for phosphate,
and is not affected by low concentrations of glutamate but has an absolute re-
quirement for activation by ammonia (half maximal at 0.25 to 0.5 rnM). 32 · 35
Both enzymes are located in the mitochondrial matrix in loose association
with the inner mitochondrial membrane. 35 •36 There is evidence that the kinetic
properties of the enzymes are considerably altered by association with the
inner mitochondrial membrane and therefore the properties listed cannot be
assumed to be maintained in vivo.37
Kidney-type glutaminase has been purified from kidney/ 8 •39 brain, 40 and
Ehrlich ascites tumor cells. 41 Antibodies raised against the purified kidney
enzyme cross-react with the enzyme in kidney, brain, and small intestine but
not with that of adult rat liver. 42 Purification procedures developed for the
kidney enzyme were based on the reversible polymerization of the protein in
the presence of phosphate or borate. Since the liver-type enzyme does not
undergo such polymerization, alternative purification methods were devel-
oped. 43 . 45 The liver enzyme appears to possess only one type of polypeptide,
58 kDa in size, but the size of the active enzyme is not known. 45 The kidney
enzyme is active as a heterotetramer comprised of three 65-kDa subunits and
one 68-kDa subunit. 46 In addition, antibodies raised to liver-type glutaminase
show very low reactivity towards the enzyme of brain, kidney, and small
intestine. 45 The recent cloning of cDNAs to both kidney- and liver-type
glutaminases 4648 has revealed other differences (see Section V).
The two isozymes are found in different tissues. 32 The liver type is only
seen in liver parenchymal cells after birth and is not found in any other
tissue. 45 •48 Kidney-type glutaminase is present in all other glutamine-utilizing
tissues with the possible exception of mammary gland. 2 1.3 1 By characterization
of different kinetic properties, Linder-Horowitz49 examined the appearance
of the two types of glutaminase activity in liver during development. Fetal
liver contains only kidney-type activity until about 2 days before birth at
which time this activity decreases sharply and is not detectable in postnatal
liver. Conversely liver-type activity first appears around birth and rapidly
(within 2 to 4 d) reaches adult levels. The recent availability of antibody 45
and cDNA 48 probes to the glutaminases has extended these results to show
that liver-type glutaminase protein and mRNA first appear about l day before
birth and reach adult levels within 2 days after birth.
C. HEPATIC GLUTAMINE METABOLISM AND

THE UREA CYCLE
Glutamine is catabolized by the liver by the pathway outlined in Figure
I . Glutamine is taken up by parenchymal cells by transport system N. 50 There
is also evidence of a facilitative system/ 5 •51 perhaps system L, 52 which trans-
ports glutamine out of the cell. Glutamine then enters the mitochondria to be
Watford 339
ARGININO- ASL
A S / SUCCINATE~
ASPARTATE --1 t--- FUMARATE

ARGININE
CITRULLINE ~ ARGINASE
OTC
~ ORNITHINE
_j '-II UREA I
CARBAMOYL PHOSPHATE
HCO) ~CPS I
GLUTAMATE+ NH 3
t GLUTAMINASE
II GLUTAMINE II
FIGURE 1. Hepatic glutaminase and the urea cycle.
irreversibly hydrolyzed by phosphate-activated glutaminase. The ammonia

and glutamate formed can then be used for urea synthesis and gluconeogenesis.
Little, or no, glutamine is completely oxidized53 •54 and evidence from a careful
analysis of dietary protein suggests that this is true for most amino acids. 54
This review takes the opinion that glutamine catabolism and urea synthesis
within the liver are part of the same pathway and focuses on the coordinate
regulation of hepatic glutaminase with the urea cycle enzymes. The urea cycle
is therefore described in some detail.
The urea cycle is usually described as beginning within the mitochondrial
matrix where NH 3 and bicarbonate condense to form carbamoyl phosphate
via the action of carbamoyl phosphate synthetase 1 (CPS 1). Ornithine trans-
carbamoylase (OTC) then catalyzes the combination of carbamoyl phosphate
with ornithine to form citrulline. Citrulline exits the mitochondria to the
cytosol where argininosuccinate synthetase (AS) adds an aspartate moiety to
form argininosuccinate. This is then hydrolyzed by argininosuccinate lyase
(ASL) to give arginine and fumarate. The arginine is hydrolyzed by arginase
to urea and at the same time regenerates ornithine. 55 There is considerable
evidence that the cycle efficiently channels intermediates, and this possibly
extends to the channeling of substrates into the cycle itself. 55 - 59 For example
MeijerW has shown that ammonia generated by the action of glutaminase is
preferentially used for the carbamoyl phosphate synthetase reaction in liver
mitochondria.
The function of the urea cycle is ammonia detoxification but since it also
removes bicarbonate it has been proposed to play a significant role in the
regulation of systemic pH. The complete cycle is found only in the liver
although partial reactions are found in the kidney and small intestine consti-
tuting tissue-compartmented system for arginine synthesis. 61 These aspects
have been the subject(s) of many recent reviews. 55 •62 · 65
IV. REGULATION OF GLUTAMINASE ACTIVITY
A. SHORT-TERM REGULATION
Both hepatic glutamine metabolism and the urea cycle are subject to
regulation by both short-term and long-term mechanisms. Short-term regu-
lation can be inferred from the observation that the concentration of glutamine
within the liver is held constant at about 5 1-Lmoles/g wet weight. The con-
centrations of urea cycle substrates, ammonia, and bicarbonate are also not
reduced to zero and do not vary widely. A number of key regulatory points
have been identified in the pathway of hepatic glutamine metabolism, transport
into the cell, glutaminase and a-ketoglutarate dehydrogenase activities, and
entry into the urea cycle. 34 •35 •52 A variety of evidence has been cited to
demonstrate a regulatory role for hepatic glutaminase activity. Recently Pog-
son et al. 52 have applied the calculations of quantitative regulatory theory to
this problem in isolated rat hepatocytes. They report that glutaminase exhibits
a control strength of close to 1.0 (indicative of a major regulatory role), while
that of transport system N is 0.31 and system Lis -0.4 (indicating a negative
effect). Despite problems with the application of this theory to enzymes within
the mitochondrial matrix the results are further indication that glutaminase
activity is of regulatory significance.
Although glutamine at high concentrations (>2 mM) is readily utilized
by experimental systems in vitro, such as the perfused liver and isolated
hepatocytes, no net glutamine utilization is observed at physiological gluta-
mine concentrations ( < 1 mM). K A number of substances will stimulate
glutamine utilization in such systems, for example, glucagon, ammonia, leu-
cine, cysteine, bicarbonate, low pH, cAMP, catecholamines, vasopressin,
angiotensin II, and thyroxine. 34 ·66 . 69 The mechanisms by which these agents
exert their effects is not well understood and their relevance in vivo is not
clear. They may act by changing the interaction of glutaminase with the inner-
mitochondrial membrane. 69 Recent work has suggested that hormonal regu-
lation of cell (and mitochondrial) volume may play a role. 70 •71 Ochs72 has
proposed that glucagon and cAMP act at the level of glutaminase activity but
that hormones acting via changes in calcium (a-adrenergic, vasopressin, and
angiotensin II) act at the level of a-ketoglutarate dehydrogenase. As mentioned
in Section III, hepatic glutaminase activity is activated by one of its products,
ammonia. This was first described by Charles 73 and subsequently shown to
be an absolute requirement for the enzyme. 34 •74 Welboume 75 has proposed an
Watford 341
intricate regulatory system involving the feed-forward activation of hepatic

glutaminase by ammonia produced from intestinal glutamine catabolism.
Kaiser et al. 76 have found flux through liver glutaminase to be increased
sixfold in human cirrhotic liver, and in rats bearing ascites tumor cells,
Quesada et al. 77 report a dramatic decrease in hepatic glutaminase activity
within 24 h of injection of the cells into the peritoneal cavity. It is not clear
what mechanism is responsible for these effects, but they probably reflect
acute regulation rather than long-term changes in the amount of enzyme
protein.
Glutamine is also reported to affect a number of hepatic pathways, i.e.,
glycogen synthesis, proteolysis, and glycolysis. Many, if not all, of these
effects are due to cell volume (swelling) changes resulting from the cotransport
of sodium with glutamine. 70 •71 •78
Kidney-type glutaminase has also been proposed to be of regulatory im-
portance in the short-term regulation of renal ammoniagenesis. This has been
extensively reviewed 79 •80 recently and will not be considered further here.
Short-term regulation of kidney-type glutaminase in other tissues has not been
extensively characterized.
The short-term regulation of the urea cycle is proposed to be exerted
primarily at the level of carbamoyl phosphate synthetase I which is subject
to allosteric activation by N-acetyl glutamate. The hydrolysis of glutamine to
glutamate, which can be used for N-acetyl glutamate synthesis, is in accord
with the major fate of hepatic glutamine catabolism being urea synthesis.
Meijer et al. 55 have recently reviewed the role of carbamoyl phosphate syn-
thetase I, and other enzymes, in the acute regulation of the cycle.
B. LONG-TERM REGULATION
Since the pathways of glutamine catabolism and urea synthesis within the
liver are tightly regulated in the short-term, the capacity of these pathways
is, by definition, in excess under normal circumstances. But there is also
extensive evidence of long-term regulation, although for hepatic glutaminase
this has only recently been investigated in depth.
As pointed out by Folin81 in 1905, urinary urea represents the most
variable nitrogenous excretion product. In what must now be considered a
potentially dangerous series of experiments, Polin showed that urea production
was variable and regulated by diet. He fed himself, and others, a very low
protein diet of arrowroot starch and cream for 14 d and found that urea
excretion fell to low levels, but it rose again within 24 h of resuming his
normal diet. Although based on anecdotal evidence, this downregulation of
urea production has been reported to be of pathological importance in the
refeeding of high protein diets to severely malnourished prisoners at the end
of World War 11. 82
The long-term regulation of hepatic glutaminase has only been recently
recognized. Some earlier reports indicated that the activity was increased after
feeding high protein diets 83 and decreased on feeding low protein diets. 84
TABLE 1
Long-Term Regulation of Glutaminase in Rat Liver
Activity.. mRNA48 Transcription rate""
Control 100 ± 10 100 ± 28 100 ± 29

Diabetes (6 days) 402 ± 12 476 ± 36 448 ± 160
Note: All values reported as% of control values and are means ± SEM
for 4 to 6 animals in each study. Control activity = 4.1 f.Lmol!
min/g wet wt. Other values are based on arbitrary densitometry
units.
These results were based on questionable assay methods but did reflect changes
in activity although the absolute values were incorrect. Examination of the
regulation of hepatic glutaminase required the establishment of an accurate
and reliable assay procedure based on that of Curthoys and Lowry. 85 The
original method was modified to accommodate the kinetic properties of the
hepatic enzyme. This meant the inclusion of high substrate concentration
(100 mM glutamine), limited phosphate (50 mM), and the inclusion of am-
monium chloride (2 rnM) to ensure complete activation of the enzyme. 86
Using this method, it has been shown that hepatic glutaminase activity is
increased during diabetes, 86 high protein feeding, 87 and starvation, 87 and de-
creased during low protein feeding 88 (Table 1). These changes are unrelated
to changes in acid-base status, the major determinant of renal glutaminase
activity. Using a protein-free diet, McGivan et a!. 89 found reduced activity
of hepatic glutaminase. Similar results are seen with a 5% protein diet 88 which
gives a more consistent response since rats consuming a protein-free diet lose
weight and are obviously in a catabolic state. The observed pattern of reg-
ulation of hepatic glutaminase activity is similar to the pattern observed for
net hepatic glutamine catabolism in vivo.
Long-term changes in hepatic glutaminase activity parallel those observed
in the activity of all five urea cycle enzymes and in the activity of a key
regulatory enzyme of gluconeogenesis, phosphoenolpyruvate carboxyki-
nase. 65 Thus changes in the maximal capacity of hepatic glutaminase are in
keeping with the role of glutamine as a substrate for urea synthesis and
gluconeogenesis. It is well established that hormones, such as glucagon,
insulin, and glucocorticoids, are the agents acting at the level of the liver cell
to regulate expression of the gluconeogenic and urea cycle enzymes. 65 ·90 To
date the question of hormonal regulation of liver glutaminase has received
little attention. In rat hepatocytes in primary culture, the activity is reported
to increase in response to glucagon and to exogenous ammonia. 89 However,
the activity decreases in cells in primary culture in the absence of hormones,
and even with hormones the activity does not rise above that seen in freshly
isolated hepatocytes. It is therefore not clear what mechanisms are involved
in this response or if it has any bearing on the situation in vivo.
Watford 343
Kidney-type glutaminase is subject to long-term regulation, but this has

only been extensively characterized in kidney, with regard to metabolic aci-
dosis. The activity is increased three- to fourfold in the kidney of acidotic
rats, 85 •91 but changes in the glutaminase activity in other tissues, lung, skeletal
muscle, adipose tissue, and intestine are not seen. 91 There is a large increase
observed in glutaminase activity of the small intestine during streptozotocin
diabetes in the rat, 86 •92 •93 but this is the result of an increased number of
enterocytes, not to changes in glutaminase activity per enterocyte. 87
V. REGULATION OF EXPRESSION OF
GLUTAMINASE GENES
A. GLUTAMINASE cDNAs
Liver glutaminase is encoded by a rnRNA of approximately 2.8 kb in
length which is expressed only in postnatal liver. 48 It is absent until a few
hours prior to birth but rapidly reaches adult levels within 2 days of birth. 48
As mentioned elsewhere, kidney-type glutaminase is comprised of two dif-
ferent sized subunits. Curthoys 46 has recently shown that both subunits are
the product of the same gene, located on chromosome 9 in the rat. 94 Tran-
scription of the gene gives rise to two species of mRNA, 6.0 and 3.3 kb in
length. Both encode a 72-kDa precursor polypeptide which is then differ-
entially processed to give the 65- and 68-kDa subunits. 46
Comparison of the known eDNA sequences of liver- and kidney-type
glutaminases shows little similarity at the nucleotide level. When the predicted
amino acid sequences are compared a region of -150 residues with high
identity (>85%) has been found. 46 •48 The complete sequence of liver-type
glutaminase has not yet been determined. The results to date indicate that the
two isozymes are the products of different but evolutionary related genes.
Interestingly, the family of CPS enzymes has been shown to have evolved
from the fusion of a glutaminase with a synthetase gene, 95 and many of the
amidotransferases exhibit a conserved active site triad of cysteine/histidine/
aspartate. 96 The presence of such a site has not been detected in the mammalian
glutaminase sequences determined to date46 .48 and may therefore indicate a
separate line of evolution.
B. HETEROGENEITY OF EXPRESSION IN THE LIVER

Liver possesses both the capacity for glutamine utilization (glutaminase)
and glutamine synthesis (glutamine synthetase). It has been shown that these
enzymes are simultaneously active and that the net utilization or production
of glutamine by the liver is the difference in activity. 24 •64 The elegant im-
munochemical studies of Gebhardt and Mecke 97 showed that glutamine syn-
thetase protein was located in a restricted population of cells perhaps only 1
to 3 cells in thickness surrounding the venous exit. This has been confirmed
in numerous publications and is now the standard index for identity of per-
ivenous cells. 55 •98 •99 Haussinger and co-workers, 64 using the isolated perfused
rat liver, proposed an intercellular glutamine cycle, with glutamine being
degraded in the periportal cells and synthesized in perivenous cells. The model
required that glutaminase be located predominantly in periportal cells. Studies 100
of enzyme activity and mRNA abundance in isolated periportal and perivenous
cells show glutaminase to be mainly located in the periportal cells along with
other enzymes of glutamine catabolism, phosphoenolpyruvate carboxykinase,
and the urea cycle enzymes. The pattern of expression of the carbamoyl
synthetase I changes during different physiological conditions, 101 and the
availability of eDNA probes will allow experiments to examine if this is also
true for liver-type glutaminase.
C. REGULATION OF EXPRESSION
Long-term increases in the activity of enzymes may be the result of the
modification of existing protein or due to changes in the amount of the protein.
In the rat the activity of hepatic glutaminase is increased fourfold in strep-
tozotocin diabetes. 86 Using equivalence point titration and Western blotting,
the increase in glutaminase activity during diabetes was shown to be com-
pletely accounted for by an increase in the amount of glutaminase protein
with no change in the specific activity of the protein. 45 The availability of a
eDNA has allowed determination of the relative abundance of glutaminase
mRNA and this too increases fourfold. 48 Recent work has established that
this is due to a fourfold increase in the rate of transcription of the hepatic
glutaminase gene (Table 1). 88 The abundance of the mRNA is also increased
in response to injection of large doses of dibutyryl cAMP. 101 After high (60%)
protein feeding or starvation, there is an increase in liver glutaminase activity,
mRNA abundance, and rate of transcription. 88 Conversely, low (0 or 5%)
protein feeding results in decreases in glutaminase activity, mRNA abundance,
and transcription rate. 88 These results establish that liver glutaminase is reg-
ulated in the long-term primarily by changes in the rate of transcription of
the gene, and no changes in protein or mRNA turnover need be postulated.
This is in striking contrast to the expression of kidney-type glutaminase
in rat kidney. Expression is regulated by changes in the amount of enzyme
and rate of enzyme synthesis and this correlates with changes in the relative
abundance of the mRNA. 103 However, it appears that mRNA turnover, rather
than changes in transcription rate, is the predominant mechanism of regula-
tion. 103 • 104 In metabolic acidosis the abundance of the message increases five-
to sevenfold and appears to be very stable. On recovery from acidosis, the
abundance of the message falls rapidly and no changes in the rate of tran-
scription are seen. In LLPC-PK 1-F+ cells in culture the abundance of the
mRNA is increased >20-fold at pH 6.9 compared to pH 7.4, and again no
changes in the rate of transcription of the gene are seen. The apparent half-
life of the mRNA changes from 2.5 hat pH 7.4 to 15 hat pH 6.9 and thus
accounts for the rise in abundance. 104 The long (6 kb) kidney-type glutaminase
Watford 345
mRNA contains many sequences in the noncoding region which could be

involved in mRNA stabilization.
D. COORDINATE EXPRESSION WITH THE UREA CYCLE

ENZYME GENES
The enzymes of the urea cycle are considered as classics for the study of
enzyme protein turnover and adaptation. In the 1930s it was demonstrated
that arginase activity changed with dietary protein intake, and in the 1960s
Schimke 105 demonstrated coordinate changes in the activity of all five en-
zymes. He extended the field by using the response to develop quantitative
aspects of enzyme protein turnover and adaptation which became the basis
for all subsequent studies.
The five urea cycle enzymes are subject to coordinate regulation, as is
the activity of N-acetyl glutamate synthetase. 55 •65 The enzymes are all located
primarily in the periportal region of the liver acinus. 55 Activities are highest
under conditions of high protein feeding or starvation and lowest in low protein
feeding. Changes also occur in uncontrolled diabetes and during hypercata-
bolic states such as sepsis and trauma. 65 The apparent half-lives of the proteins
are long (3 to 6 d), and therefore changes in the amount of enzyme activity
are not manifested rapidly and several days may be required to reach new
steady states. Since all of the enzymes respond to hormones, such as glucagon,
glucocorticoids, thyroxine and insulin, it is likely that the changes observed
with diet are mediated by hormonal signals. 65 •90 · 105 Changes in activity appear
to be due to increased amounts of enzyme protein resulting from increased
abundance of the specific mRNAs and increased rates of transcription of the
genes. Exceptions are an increase in arginase activity in diabetes without a
change in the amount of enzyme, 107 changes in the rate of arginase enzyme
protein degradation during starvation, 108 and evidence of translational regu-
lation of carbamoyl phosphate synthetase I. 109 Furthermore, in contrast to the
other enzymes, ornithine transcarbamoylase does not show dramatic responses
to glucagon, cAMP, or glucocorticoids in tissue culture and short-term ex-
periments in vivo. 110
Morris, 65 and others, 111 • 112 have reviewed the molecular biology of the
long-term regulation of the urea cycle enzymes, and it will not be covered
in great detail here. The evidence is not complete. Much of the work is from
computer analysis of DNA sequence or is inferred from the pattern of re-
sponses seen in vivo and in vitro. Very little work has been done with reporter
genes fused to urea cycle enzyme gene promoter sequences. The pattern that
is emerging is that the urea cycle enzyme genes possess 5' regulatory se-
quences which allow regulation of expression by cAMP and glucocorticoids.
In addition, argininosuccinate synthase is subject to regulation by arginine
which acts to inhibit transcription through a region 5' to the gene. 113 • 114 There
is also evidence of the presence of tissue-specific elements. 65 It is likely,
given the coordinate regulation of expression, that the genes possess similar
regulatory elements. For CPS I and OTC and arginase, the presence of ele-
ments recognized by C/EBP or related factors has been demonstrated, and it
is likely that CRE and GRE will be present. 65 A unique sequence has been
found, "UCE" (urea cycle element), 65 • 115 in the regions 5' to at least 4 of
the genes (OTC, AS, ASL, and arginase) and this may play a role in the
coordinate regulation. It has been hypothesized that this element may be
involved in the coordinate expression of other genes involved in nitrogen
metabolism.
The common expression of glutaminase with urea cycle enzymes (and
gluconeogenic enzymes) in periportal cells, and the similar ontogenic devel-
opment pattern, appearing around birth in the liver, is in keeping with the
role of hepatic glutamine metabolism. However, differences do exist. The
urea cycle enzymes are expressed in other tissues, CPS I and OTC in the
intestine, and ASS and ASL in kidney, while hepatic glutaminase is only
expressed at measurable levels in postnatal liver. Evidence from experiments
in vivo and in vitro indicate that glutaminase will have similar elements for
the tissue-specific and hormonal regulation of expression. The presence or
absence of these, and the UCE, await the availability of genomic clones and
sequence analysis.
VI. SUMMARY
This chapter has centered on the long-term regulation of expression of

hepatic glutaminase gene and how this relates to expression of the urea cycle.
These enzymes are proteins with relatively long (days) half-lives, and changes
in activity take some time to be manifested. This could mean that the changes
are simply adaptive in nature and do not themselves play any role in the
regulation of metabolism. This may be true in subjects eating regular diets.
However, the decrease in activity observed in times of low rates of hepatic
amino acid catabolism certainly limits the capacity of the body to deal with
a sudden nitrogen load, while the increased capacity observed in hypercata-
bolic states allows the body to deal with the high rates of nitrogen disposal.
There is considerable evidence that the urea cycle acts as a metabolon;
its substrates and intermediates are efficiently channeled along the pathway.
This concept may even extend to the enzymes involved in the provision of
substrates, such as glutaminase. Many enzymes involved in the degradation
of excess amino acids are also subject to coordinate long-term regulation with
those of the urea cycle. Perhaps this reflects the integrated nature of the
system. Application of molecular biological techniques to the comparison of
expression of these various hepatic genes is leading to an understanding of
how hepatic amino acid metabolism is regulated.
Watford 347
ACKNOWLEDGMENTS
Work done in the author's laboratory was supported by Grant DK-37301
from the National Institutes of Health. The author is grateful to Drs. Brosnan,
Curthoys, and Morris for providing manuscripts in press.
REFERENCES
l. Lacey, J. M. and Wilmore, D. W., Is glutamine a conditionally essential amino acid?,
Nutr. Rev., 48, 297, 1990.
2. Lowe, D. K., Benfell, K., Smith, R. J., Jacobs, D. 0., Murawski, B., Ziegler, T. R.,
and Wilmore, D. W., Safety of glutamine-enriched parenteral nutrient solutions in
humans, Am. J. Clin. Nutr., 52, 1101, 1990.
3. Furst, P., Alpers, S., and Stehle, P., Dipeptides in clinical nutrition, Proc. Nutr. Soc.,
49, 343, 1990.
4. Windmueller, H. G. and Spaeth, A. E., Uptake and metabolism of plasma glutamine
by the small intestine, J. Bioi. Chern., 249, 5070, 1974.
5. Newsholme, E. A. and Parry-Billings, M., Properties of glutamine release from skeletal
muscle and its importance to the immune system, J. Parent. Ent. Nutr., 14, 63S, 1990.
6. Souba, W. W., Glutamine. A key substrate for the splanchnic bed, Annu. Rev. Nutr.,
11,285, 1991.
7. Souba, W. W., Smith, R. J., and Wilmore, D. W., Glutamine metabolism by the
intestinal tract, J. Parent. Ent. Nutr., 9, 608, 617, 1985.
8. Lund, P. and Watford, M., Glutamine as a precursor of urea, in The Urea Cycle,
Grisolia, S., Baguena, R., and Mayor, F., Eds., John Wiley & Sons, New York, 1976,
479.
9. Squires, E. J. and Brosnan, J. T., Measurement of the turnover rate of glutamine in
normal and acidotic rats, Biochem. J., 210, 277, 1983.
10. Lund, P. and Williamson, D. H., Inter-tissue nitrogen fluxes, Brit. Med. Bull., 41,
251, 1985.
11. Welbourne, T. C., Role of the lung in glutamine homeostatis, Contr. Nephrol., 63, 178,
1988.
12. Souba, W. W., Herkowitz, K., and Plumley, D. A., Lung glutamine metabolism, J.
Parent. Ent. Nutr., 14, 68S, 1990.
13. Frayn, K. N., Khan, K., Coppack, S. W., and Elia, M., Amino acid metabolism in
human subcutaneous adipose tissue, Clin. Sci., 80, 471, 1991.
14. Aikawa, T., Matsuka, H., Yamamoto, H., Okuda, T., Ishikawa, E., Kawano, T.,
and Matsumura, E., Gluconeogenesis and amino acid metabolism. Interorganal relations
and roles of glutamine and alanine in the amino acid metabolism of fasted rats, J.
Biochem., 74, 1003, 1974.
15. Fafournoux, P., Remesy, C., and Demigne, C., Fluxes and membrane transport of
amino acids in rat liver under different protein diets, Am. J. Physiol., 259, E614, 1990.
16. Tannen, R. L., Ammonia metabolism, Am. J. Physiol., 235, F265, 1978.
17. Watford, M., Lund, P., and Krebs, H. A., Isolation and metabolic characteristics of
rat and chicken enterocytes, Biochem. J., 178, 589, 1979.
18. Ardawi, M. S. M. and Newsholme, E. A., Metabolism in lymphocytes and its im-
portance in the immune response, Essays Biochem., 21, I, 1985.
19. Newsholme, E. A., Crabtree, B., and Ardawi, M.S. M., The role of high rates of
glycolysis and glutamine metabolism in rapidly dividing cells. Biosci. Rep., 5, 393, 1985.
20. Watford, M., Erbelding, E. J., and Smith, E. M., Glutamine metabolism in rat small
intestine: response to lactation. Biochem. Soc. Trans., 14. 1058, 1986.
21. Horowitz, M. L. and Knox, W. E., A phosphate activated glutaminase in rat liver
different from that in kidney and other tissues, Enrymol. Bioi. Clin., 9, 214, 1968.
22. Yamamoto, H., Aikawa, T., Matsumura, H., Okuda, T., and Ishikawa, E., Inter-
organa! relationships of amino acid metabolism in fed rats, Am. J. Physiol., 244, 1428,
1974.
23. Wasserman, D. H., Greer, R. J., Williams, P. E., Becker, T., Lacy, D. B., and
Abumrad, N. N., Interaction of gut and nitrogen metabolism during exercise, Metab-
olism, 40, 307, 1991.
24. Vincent, N., Martin, G., and Baverel, G., Simultaneous synthesis and degradation of
glutamine in isolated rat liver cells. Effect of vasopressin. Biochim. Biophys. Acta, 1014,
184, 1989.
25. Schrock, H. and Goldstein, L., Interorgan relationships for glutamine metabolism in
normal and acidotic rats, Am. J. Physiol., 240, E519. 1981.
26. Brosnan, J. T., Man, K.-C., Hall, D. E., Colbourne, S. A., and Brosnan, M. E.,
Interorgan metabolism of amino acids in streptozotocin-diabetic ketoacidotic rat, Am. J.
Physiol., 244, El51, 1983.
27. Yamamoto, H., Aikawa, T., Matsutaka, H., Okuda, T., and Ishikawa, E., Inter-
organa! relationships of amino acid metabolism in fed rats, Am. J. Physiol., 226, 1428,
1974.
28. Krebs, H. A., Metabolism of amino acids. The synthesis of glutamine from glutamic
acid and ammonia, and the enzymic hydrolysis of glutamine in animal tissues, Biochem.
J., 29, 1951, 1935.
29. Cooper, A. j. L., Glutamine synthetase, in Glutamine and Glutamate in Mammals,
Kvamme, E., Ed., CRC Press. Boca Raton, FL, 1988, 7.
30. Meister, A., Function of glutathione in kidney via the gamma-glutamyl cycle, Med.
Clin. N. Am., 59, 649, 1975.
31. Erbelding, E. J., Glutamine Metabolism in the Small Intestine. MS thesis. Cornell
University, Ithaca, NY, 1985.
32. Huang, Y.-Z. and Knox, W. E., A comparative study of glutaminase isozymes in rat
tissues, Enzyme, 21, 408, 1976.
33. Curthoys, N. P., Cellular distribution and induction of the enzymes of renal ammoni-
agenesis and gluconeogenesis, in pH Homeostasis, Haussinger, D .. Ed. , Academic Press,
New York, 323, 1988.
34. McGivan, j. D., Metabolism of glutamine and glutamate in liver-regulation and phys-
iological significance, in Glutamine and Glutamate in Mammals, Kvamme. E., Ed., CRC
Press, Boca Raton, FL, 1988, 183.
35. Kovacevic, Z. and McGivan, J. D., Mitochondrial metabolism of glutamine and glu-
tamate and its physiological significance, Physiol. Rev., 63, 547, 1983.
36. Kalra, J. and Brosnan, J. T., Localization of glutaminase in rat liver, FEBS Lett., 32,
325, 1973.
37. McGivan, J. D., Lacey, J. H., and Joseph, S. K., Localization and some properties
of phosphate-dependent glutaminase in disrupted liver mitochondria. Biochem. J., 192,
537, 1980.
38. Kvamme, E., Tveit, B., and Svenneby, G., Glutaminase from pig renal cortex. I.
Purification and general properties, J. Bioi. Chern., 245. 1871, 1970.
39. Curthoys, N. P., Kuhlenschmidt, T., and Godfrey, S. S., Regulation of renal am-
moniagenesis. Purification and characterization of phosphate-dependent glutaminase from
rat kidney, Arch. Biochem. Biophys., 174. 82, 1976.
40. Svenneby, G., Pig brain glutaminase: purification and identification of different enzyme
forms. J. Neurochem., 17, 1591, 1970.
Watford 349
41. Quesada, A. R., Sanchez-Jimenez, F., Perez-Rodriguez, J., Marquez, J., Medina,
M. A., and Nunez de Castro, I., Purification of phosphate-dependent glutaminase from
isolated mitochondria of Ehrlich ascites-tumour cells, Biochem. J., 255, 1031, 1988.
42. Curthoys, N. P., Kuhlenschmidt, K., Godfrey, S. S., and Weiss, R. F., Phosphate-
dependent glutaminase from rat kidney. Cause of increased activity in response to acidosis
and identity with glutaminase from other tissues, Arch. Biochem. Biophys., 172, 162,
1976.
43. Patel, M. and McGivan, j. D., Partial purification and properties of rat liver glutaminase,
Biochem. J., 220, 583, 1984.
44. Heini, H. G., Gebhardt, R., and Mecke, D., Purification and characterization of rat
liver glutaminase, Eur. J. Biochem., 162, 541, 1987.
45. Smith, E. M. and Watford, M., Rat hepatic glutaminase: purification and immuno-
chemical characterization, Arch. Biochem. Biophys., 260, 740, 1988.
46. Shapiro, R. A., Farrell, L., Srinivasan, M., and Curthoys, N. P., Isolation, char-
acterization, and in vitro expression of a eDNA that encodes the kidney isoenzyme of
the mitochondrial glutaminase, J. Bioi. Chern., 266, 18792, 1991.
47. Banner, C., Hwang, J.-J., Shapiro, R. A., Wenthold, R. J., Nakatani, Y., Lampe!,
K. A., Thomas, J. W., Huie, D., and Curthoys, N. P., Isolation of a eDNA for rat
brain glutaminase, Mol. Brain Res., 3, 247, 1988.
48. Smith, E. M. and Watford, M., Molecular cloning of a eDNA for rat hepatic glutam-
inase. Sequence similarity to kidney-type glutaminase, J. Bioi. Chern., 265, 10631, 1990.
49. Under-Horowitz, M., Changes in glutaminase activities of rat liver and kidney during
pre- and post-natal development, Biochem. J., 114,65, 1969.
50. Kilberg, M. S., Handlogten, M. E., and Christensen, H. N., Characteristics of an
amino acid transport system in rat liver for glutamine, asparagine, histidine and closely
related analogs, J. Bioi. Chern., 255,4011, 1980.
51. Farfournoux, P., Demigne, C., Remesy, C., and Le Cam, A., Bidirectional transport
of glutamine across the cell membrane in rat liver, Biochem. 1 .. 216. 401, 1983.
52. Pogson, C. 1., Low, S. Y., Knowles, R.N., Salter, M., and Rennie, M. J., Application
of metabolic control theory to amino acid metabolism in liver, Alfred Benson Symp., 30,
262, 1991.
53. Blackburn, E. H. and Hird, F. J. R., Metabolism of glutamine and glutamate by rat
liver mitochondria, Arch. Biochem. Biophys., 152, 258, 1972.
54. Jungas, R. L., Halperin, M. L., and Brosnan, J. T., A quantitative analysis of amino
acid oxidation and related gluconeogenesis in man, Physiol. Rev., in press.
55. Meijer, A. J., Lamers, W. H., and Chamuleau, R. A. F. M., Nitrogen metabolism
and ornithine cycle function, Physiol. Rev., 70. 701, I990.
56. Cohen, N. S., Cheung, C.-W., and Raijman, L., Channeling of extramitochondrial
ornithine to matrix ornithine transcarbamoylase, J. Bioi. Chern., 262, 203, 1987.
57. Cheung, C.-W., Cohen, N. S., and Raijman, L., Channeling of urea cycle intermediates
in situ in permeabilized hepatocytes, J. Bioi. Chern., 264, 4038, 1989.
58. Watford, M., Channeling in the urea cycle: a metabolon spanning two compartments,
Trends Biochem. Sci., 14, 313, 1989.
59. Watford, M., The urea cycle: a two-compartment system, Essays Biochem., 26, 49,
1990.
60. Meijer, A. j., Channeling of ammonia from glutaminase to carbamoyl-phosphate syn-
thetase in liver mitochondria, FEBS Lett., 191. 249, 1985.
61. Dhanakoti, S. N., Brosnan, J. T., Herzberg, G. R., and Brosnan, M. E., Renal
arginine synthesis: studies in vitro and in vivo, Am. J. Physioi .. 259. E437, 1990.
62. Atkinson, D. E. and Bourke, E., The role of ureagenesis in pH homeostasis, Trends
Biochem. Sci., 9, 291, 1984.
63. Brosnan, J. T., Lowry, M., Vinay, P., Gougoux, A., and Halperin, M. L., Renal
ammonium production-une vue canadienne. Can. J. Physioi. Pharamacoi., 65, 489,
1987.
64. Haussinger, D., Nitrogen metabolism in liver: structural and functional organization and
physiological relevance, Biochem. J., 267, 281. 1990.
65. Morris, S. M., Regulation of enzymes of urea and arginine synthesis, Annu. Rev. Nutr.,
12, in press.
66. Corvera, S. and Garcia·Sainz, j. A., Hormonal stimulation of mitochondrial glutam-
inase. Effects of vasopressin, angiotensin II, adrenaline and glucagon, Biochem. J., 210,
957, 1983.
67. Kashiwagura, T., Erecinska, M., and Wilson, D. F., pH Dependence of hormonal
regulation of gluconeogenesis and urea synthesis from glutamine in suspensions of he-
patocytes, J. Bioi. Chern., 260, 407. 1985.
68. Verhoeven, A. j., Estrela, j. M., and Meijer, A. j., Adrenergic stimulation of glu-
tamine metabolism in isolated rat hepatocytes, Biochem. J., 230, 457, 1985.
69. McGivan, j. D., Vadher, M., Lacey, J., and Bradford, N., Rat liver glutaminase.
Regulation by reversible interaction with the mitochondrial membrane. Eur. J. Biochem.,
148, 323, 1985.
70. Haussinger, D. and Lang, F., The mutual interaction between cell volume and cell
function: a new principle of metabolic regulation, Biochem. Cell. Bioi., 69, I, 1991.
71. vonDahl, S., Hallbrucker, C., Lang, F., and Haussinger, D., Regulation of liver cell
volume and proteolysis by glucagon and insulin, Biochem. J., 278, 771, 1991.
72. Ochs, R. S., Glutamine metabolism in isolated hepatocytes. Evidence for catecholamine
stimulation of a-ketoglutarate dehydrogenase, J. Bioi. Chern., 259, 13004, 1984.
73. Charles, R., Mitochondriale Citrulline Synthese: een Ammonik Fixerend en ATP Ver-
bruikend Proces, Ph.D. thesis, University of Amsterdam, Rototype, Amsterdam, 1968.
74. McGivan, j. D. and Bradford, N. M. Characteristics of the activation of glutaminase
by ammonia in sonicated rat liver mitochondria. Biochim. Biophys. Acta, 759, 241, 1983.
75. Welbourne, T. C., lnterorgan glutamine flow in metabolic acidosis, Am. J. Physiol.,
253, F1069, 1987.
76. Kaiser, S., Gerok, W., and Haussinger, D., Ammonia and glutamine metabolism in
human liver slices. New aspects of the pathogenesis of hyperammonemia in chronic liver
disease, Eur. J. Clin. Invest., 18, 535, 1988.
77. Quesada, A. R., Medina, M.A., Marquez, J., Sanchez-Jimenez, F. M., and Nunez
de Castro, I., Contribution by host tissue to circulating glutamine in mice inoculated
with Ehrlich ascites tumor cells. Can. Res., 48, 1551, 1988.
78. Watford, M., A swell way to regulate metabolism, Trends Biochem. Sci., 15, 329,
1990.
79. Tannen, R. L. and Sastrasinh, S., Response of ammonia metabolism to acute acidosis,
Kidney Int., 25, I, 1984.
80. Schoolwerth, A. C. and LaNoue, K., Transport of metabolic substrates into renal
mitochondria, Annu. Rev. Physiol., 47, 143, 1985.
81. Folio, 0., Laws governing the chemical composition of urine. Am. J. Physiol., 13, 66,
1905.
82. Freedland, R. A. F. and Briggs, S., A Biochemical Approach to Nutrition, Chapman
and Hall, London, 1977, 46.
83. Freedland, R. A. F. and Taylor, A. R., Studies on glucose-6-phosphatase and glutam-
inase in rat liver and kidney, Biochim. Biophys. Acta, 92, 567, 1964.
84. Simell, 0., Effect of low protein diet and hyperammonemia on liver glutaminase activity
in the rat, Experientia, 30, 324, 1974.
85. Curthoys, N. P. and Lowry, 0. H., The distribution of glutaminase isoenzymes in the
various structures of the nephron in normal. acidotic, and alkalotic rat kidney, J. Bioi.
Chern., 248, 162, 1973.
86. Watford, M., Smith, E. M., and Erbelding, E. j., Regulation of phosphate-activated
glutaminase activity and glutamine metabolism in the streptozotocin-diabetic rat, Biochem.
J., 224, 207, 1984.
Watford 351
87. Watford, M., Erbelding, E. J., Shapiro, A. C., Zakow, A. C., and Smith, E. M.,
The adaptive response of phosphate-activated glutaminase in the rat. Contr. Neph., 47,
140, 1985.
88. Watford, M., Zhan, Z., and Vincent, N. C., unpublished.
89. McGivan, J. D., Boon, K., and Doyle, F. A., Glucagon and ammonia influence the
long-term regulation of phosphate-dependent glutaminase activity in primary cultures of
rat hepatocytes, Biochem. J., 274, 103. 1991.
90. Snodgrass, P. J., Lin, R. C., Muller, W. A., and Aoki, T. T., Induction of urea cycle
enzymes of rat liver by glucagon, J. Bioi. Chern., 253, 2748, 1978.
91. Watford, M., Regulation of expression of the genes for glutaminase and glutamine
synthetase in the acidotic rat, Contr. Neph., 92, 211, 1991.
92. Watford, M., Erbelding, E. J., and Smith, E. M., The regulation of glutamine and
ketone body metabolism in the small intestine of the long-term (40-day) streptozotocin-
diabetic rat, Biochem. J., 242, 61, 1987.
93. Nagy, L. E. and Kretchmer, N., Effect of diabetic ketosis on jejunal glutaminase,
94. Mock, B., Kozak, C., Seldin, M. F., Ruff, N., D'Hoostelaere, L., Szpirer, C.,
Seuanez, H., O'Brien, S., and Banner, C., A glutaminase (Gls) gene maps to mouse
chromosome l, rat chromosome 9, and human chromosome 2. Genomics, 5. 291, 1989.
95. Nyunoya, H., Broglie, K. E., and Lusty, C. J., The gene coding for carbamoyl-
phosphate synthetase I was formed by fusion of an ancestral glutaminase gene and a
synthetase gene, Proc. Nat/. Acad. Sci. U.S.A., 82, 2244, 1985.
96. Zalkin H., Argos, P., Narayana, S. V. L., Tiedeman, A. A., and Smith, j. M.,
Identification of a trpG-related glutamine amide transfer domain in Escherichia coli GMP
synthetase, J. Bioi. Chern., 260, 3350. 1985.
97. Gebhardt, R. and Mecke, D., Heterogeneous distribution of glutamine synthetase among
rat liver parenchymal cells in situ and in primary culture, EMBO J., 2, 567. 1983.
98. Jungermann, K. and Katz, N., Functional specialization of different hepatocyte pop-
ulations, Physiol. Rev., 69, 708, 1989.
99. Gaasbeek-Janzen, J. W., Lamers, W. H., Moorman, A. F. M., deGraaf, A., Los,
J. A., and Charles, R., Immunohistochemical localization of carbamoylphosphate syn-
thetase (ammonia) in adult rat liver. Evidence for a heterogeneous distribution, J. His-
tochem. Cytochem., 32, 557, 1984.
100. Watford, M. and Smith, E. M., Distribution of hepatic glutaminase activity and mRNA
in perivenous and periportal hepatocytes, Biochem. J., 267, 265, 1990.
101. Moorman, A. F. M., deBoer, P. A. j., Charles, R., and Lamers, W. H., Diet and
hormone induced reversal of the carbamoylphosphate synthetase mRNA gradient in the
rat liver lobulus, FEBS Leu., 276, 9, 1990.
102. Watford, M. and Smith, E. M., Regulation of hepatic glutaminase mRNA levels in
the rat, Biochem. Soc. Trans., 17, 175, 1988.
103. Tong, J., Shapiro, R. A., and Curthoys, N. P., Changes in the levels of translatable
glutaminase mRNA during onset and recovery from metabolic acidosis, Biochemistry,
26, 2773, 1987.
104. Kaiser, S. and Curthoys, N. P., Effect of pH and bicarbonate on phosphoenolpyruvate
carboxykinase and glutaminase mRNA levels in cultured renal epithelial cells, J. Bioi.
Chern., 266, 9397, 1991.
lOS. Schimke, R. and Doyle, D., Control of enzyme levels in animal tissues, Annu. Rev.
Biochem., 39, 929, 1970.
106. Morris, S. M., Moncman, C. L., Rand, K. D., Dizikes, G. J., Cederbaum, S. D.,
and O'Brien, W. E., Regulation of mRNA levels for five urea cycle enzymes in rat
liver by diet, cyclic AMP and glucocorticoids, Arch. Biochem. Biophys., 256. 343, 1987.
107. Bond, J. S., Failla, M. L., and Unger, D. F., Elevated manganese concentration and
arginase activity in livers of streptozotocin-induced diabetic rats, J. Bioi. Chern., 258,
8004, 1983.
108. Schimke, R. T., The importance of both synthesis and degradation in the control of
arginase levels in rat liver, J. Bioi. Chern .. 239. 3808, 1964.
109. vanRoon, M.A., Zonneveld, D., Charles, R., and Lamers, W. H., Accumulation of
carbamoylphosphate synthetase and phosphoenolpyruvate carboxykinase mRNA in em-
bryonic rat hepatocytes. Evidence for translational control during the initial phases of
hepatocyte specific gene expression in vitro. Eur. J. Biochern., 178. 191, 1988.
110. Nebes, V. L. and Morris, S. M., Regulation of messenger ribonucleic acid levels for
five urea cycle enzymes in cultured rat hepatocytes. Requirements for cyclic adenosine
monophosphate, glucocorticoids and ongoing protein synthesis, Mol. Endocrinol., 2,
444, 1988.
Ill. Jackson, M. J., Beaudet, A. L., and O'Brien, W. J., Mammalian urea cycle enzymes,
Annu. Rev. Genet., 20, 431, 1986.
112. Takiguchi, M., Matsubasa, T., Amaya, Y., and Mori, M., Evolutionary aspects of
urea cycle genes, BioEssays, 10, 163, 1989.
113. Boyce, F. M., Anderson, G. M., Rusk, C. D., and Freytag, S. 0., Human argini-
nosuccinate synthetase minigenes are subject to arginine mediated repression but not to
trans induction, Mol. Cell. Bioi., 6, 1244, 1986.
114. Jackson, M. J., O'Brien, W. E., and Beaudet, A. L., Arginine-mediated regulation
of an argininosuccinate synthetase minigene in normal and canavanine resistant human
cells, Mol. Cell. Bioi., 6, 2257, 1986.
115. Engelhardt, J, F., Steel, G., and Valle, D., Transcriptional analysis of the ornithine
aminotransferase promoter, J. Bioi. Chern., 266, 752, 1990.
Chapter 16
EFFECTS OF DIETARY PROTEIN ON

GENE EXPRESSION
David K.-C. Chan and James L. Hargrove
TABLE OF CONTENTS
I. Opposing Effects of Dietary Protein and Carbohydrate .......... 354
II. Dietary Protein as a Stimulus to Organ Growth .................. 355
III. Hormonal Secretion during Fasting and Refeeding ............... 356

A. Peptides from the Gut and Pancreas ...................... 357
B. Pituitary Hormones, Thyroxine, and
Glucocorticoids ........................................... 359
IV. Physiological Effects of Amino Acids ........................... 360
V. Regulation of Enzyme Synthesis by Dietary Protein ............. 361

A. Effects of Dietary Composition on Enzyme
Responses ................................................ 361
B. Dietary Regulation of Tyrosine Aminotransferase ........ 364
C. Regulation of Serine-Threonine Dehydratase by
Protein and Carbohydrate ................................. 367
VI. Conclusions ...................................................... 369
References ............................................................... 370
0-8493-6961-4193/$0.00 + $.50
I. OPPOSING EFFECTS OF DIETARY PROTEIN

AND CARBOHYDRATE
When meals that contain protein are ingested, the liver is exposed to
elevated concentrations of amino acids and hormones in the blood of the
hepatic portal vein. This produces two important effects that alter the regu-
lation of specific genes. The first of these effects is an increase in the size
of the liver and kidneys. Hepatic and renal cells proliferate, and this involves
an increase in DNA synthesis, RNA synthesis, and the activation of enzymes
needed for synthesis of cellular constituents. '·2 In addition to cell division,
cell growth occurs with accumulation of protein. The second effect is an
increase in the concentration of enzymes needed to catabolize the excess
amino acids, dispose of the unneeded amino group, and utilize the resulting
carbon chain. 3 . 6 Due to this adaptation, the concentration of amino acids in
the general circulation does not increase to the degree noted in the portal
circulation.
Although amino acids can be converted to glucose through the processes
of trans- or deamination and gluconeogenesis, fatty acids do not fuel glu-
coneogenesis efficiently in liver. This causes dietary protein and carbohydrate
to have a special, opposing relationship on the synthesis of enzymes involved
in amino acid catabolism. Fasting or consuming diets that contain relatively
large amounts of protein causes adaptive increases in the concentrations of
enzymes in liver and kidney that convert amino acids to precursors for syn-
thesis of glucose and fatty acids; 4 · 6 conversely, dietary carbohydrate decreases
the activity of enzymes that participate in gluconeogenesis and amino acid
catabolism. 7•8 The ability of carbohydrates to decrease the synthesis of en-
zymes that participate in catabolism of amino acids is superficially similar to
catabolite repression in bacteria and has been referred to as glucose repression.
These responses have been characterized best in animals fed experimental
diets that provide extreme levels of protein or carbohydrate and produce large
changes in enzyme levels. Although these diets are extreme in comparison
to the mixed diets typical of North Americans, adaptive changes in gene
expression probably occur in response to diets of any composition. In the
words of Helen Teppermann et a!., 9 "the business of differentiation in the
liver is eternally unfinished", and this comment no doubt applies to cells of
the endocrine system and central nervous system that contribute to nutrient-
initiated signaling.
The means by which the diet affects genetic activity probably differs for
individual organs, and depends on the duration of any prior fast and the
composition of the ensuing meals. Levels of nutrients in the blood plasma
are significant even during starvation, and their concentrations increase only
moderately during the absorptive phase. For example, it is unusual for the
concentration of glucose or amino acids in the general circulation to increase
by more than 50% after a meal in nondiabetic subjects.' 0 · 13 The concentration
Chan and Hargrove 355
in the splanchnic circulation may increase two- or threefold. Similarly, eating

a single, high-fat meal increases the level of fatty acids in the blood of human
subjects by 20 to 40% compared to fasting levels. 14 The intestinal secretion
of chylomicrons causes turbidity of the blood plasma to increase noticeably,
even though the degree of lipemia is not extreme. Therefore, changes in gene
expression in response to dietary modification take place during periods when
levels of nutrients differ from values found in fasted subjects by relatively
small amounts, but the expression of particular genes may increase or decrease
by factors of tenfold or more as a consequence of signal amplification within
responsive cells. This chapter recounts how dietary protein and carbohydrate
affect hormonal secretion and genes that encode several inducible enzymes.
II. DIETARY PROTEIN AS A STIMULUS

TO ORGAN GROWTH
The liver and kidneys both play important roles in regulating levels of
amino acids in the blood plasma and in synthesizing glucose from excess
amino acids. Perhaps for this reason, dietary protein promotes growth of these
organs more strongly than it stimulates growth of the rest of the body. This
topic has received considerable attention due to the concern that excessive
amino acids may cause kidney function to deteriorate; 15 in addition, excess
amino acids stimulate kidney growth, and this could be beneficial in some
cases. 2 Depending on the level of protein, the weight of these organs in rats
may increase from 10 to 50% over a period of several days, and the rate of
DNA replication increases, as judged by incorporation of radiolabeled de-
oxythymidine. 16 The amount of protein relative to DNA also increases, sug-
gesting that protein or amino acids promote both hypertrophy and hyperplasia;
hyperplasia predominates in young rats and hypertrophy predominates in older
onesY In addition to their effects on organ growth, amino acids rapidly
increase the rate of renal blood flow (hyperemia) and glomerular filtration
rate (GFR). 15 ·18 When the subjects are refed diets containing lower levels of
protein, the blood flow and organ weights return to normal values.
Amino acids are thought to affect the function of kidney and liver by
direct and indirect mechanisms. Mixtures of amino acids that cause renal
hyperemia also stimulate secretion of insulin, glucagon, and growth hor-
mone.18 When infused alone, glucagon increases GFR. However, a mixture
of branched-chain amino acids increased secretion of insulin and growth
hormone without affecting GFR. 18 This evidence suggests that glucagon par-
ticipates in the hyperemic response to dietary protein. In addition to the
hormonal responses, amino acids may generate other mediators that affect
organ function. Arginine not only stimulates glucagon secretion, but is con-
verted to nitric oxide by the vascular endothelium. Nitric oxide increases
blood flow and alters levels of cyclic guanosine 3' ,5 '-monophosphate in
glomeruli. 19 It counteracts the action of vasoconstrictors, such as the local
vascular hormone, endothelin. Other mediators may also be involved, for

example, the prostaglandins. Inhibitors of cyclooxygenase prevent the syn-
thesis of prostaglandins, which can dilate renal blood vessels. In the relative
absence of eicosanoids, amino acids do not induce hyperemia. Moreover,
dietary protein activates the renin-angiotensin system by increasing plasma
renin activity; the quantity of mRNA encoding renin increases in kidney but
the message for angiotensinogen is unchanged. 20
What promotes the growth response of liver and kidney to amino acids?
The answer is probably complex, because both organs contain receptors for
a variety of steroidal and peptide hormones, as well as for growth factors and
cytokines. Stimuli that are associated with cell growth almost invariably
increase the activity of ornithine decarboxylase (ODC), and dietary protein
potently stimulates activity of this enzyme in both organs. Farwell et a!.
observed that diets that contained 26% casein caused a 20-fold increase in
ODC activity in liver and a 10-fold increase in kidneys in comparison to
responses of rats fed a protein-free diet. 16 Peak values occurred during the
nocturnal feeding period and declined with continued feeding of the protein-
containing diet. The protein-containing diet stimulated hepatic DNA synthe-
sis, but this repsonse was prevented by propane- I ,3-diamine, an inhibitor of
ODC activity. Kaysen eta!. 21 observed that the increase in glomerular filtration
rate occurred before the ODC response, and that ODC activity returned to
low values after three days of feeding a high protein diet. The changes in
GFR and ODC activity preceded changes in kidney weight and were impli-
cated in the cell proliferation which was needed for kidney growth. The effect
of amino acids on ODC activity could be reproduced by injecting casein
hydrolyzate or mixtures of amino acids, and a mixture of glutamate, aspartate,
and alanine was the smallest group that gave the full response. 22 All effective
mixtures increased the content of polyamines in kidney, and it was postulated
that effective groups also stimulated ureagenesis. Although other genes in the
kidney also respond to dietary protein, the response is much less than the
magnitude noted for ODC activity. For instance, the mRNAs for arginosuc-
cinate synthetase and lyase increase only about threefold in animals fed a
60% protein diet. 23 Ornithine decarboxylase activity is regulated at multiple
levels, and the points affected by dietary protein have not yet been explored
in the kidney. It will be interesting to learn more about mediators of the
growth-promoting effects of dietary protein on these two organs.
III. HORMONAL SECRETION DURING

FASTING AND REFEEDING
Many nutritional and biochemical studies concerning expression of genes

for enzymes have utilized an experimental design in which animals that have
been starved for 2 or 3 d are then refed a diet of specified composition. This
starvation-refeeding paradigm produces reproducible changes in hormone se-
TABLE I
Effects of Energy-Yielding Nutrients on Hormone Secretion
Organ/Hormone Glucose Amino acids Fatty acids Ref.
Gut
Neurotensin o· 0 ++ 25
Peptide YY + + ++ 26
Gastric inhibitory peptide ++ 01+ ++ 27
Cholecystokinin 01+ ++ ++ 28
Pancreas
Insulin ++ + +10 29-32
Glucagon ++ 24, 32-34
Pancreatic polypeptide +1-b ++ 26
Anterior pituitary
Growth hormone ++ 23, 34-37
Adrenocorticotropin 0/-b ++ 0 38-42
Thyroid stimulating hormone 43
Adrenal
Cortisol + ++ 0 32, 39--41
Thyroid gland
T, + 0 32, 40, 44
T. + 0 44
T,IT4 + 0
Definition of symbols: 0, no effect; +,stimulatory effect; -,inhibitory effect; +10 or

+I -, variable.
Hypoglycemia is a strong stimulus for secretion of these hormones.
cretion during the initial phase, but changes that occur after refeeding depend
on composition of the diet, and the hormones participate in activating specific
genes. Table I summarizes effects of glucose, amino acids, and fatty acids
on secretion of a variety of hormones from the gut, pancreas, and anterior
pituitary gland. It is evident that each class of energy-yielding macronutrient
produces a distinct pattern of hormone secretion from the intestinal mucosa,
pancreas, and pituitary. Peptide hormones are gene products of the endocrine
system, and nutrients frequently affect gene expression in hormone-secreting
cells; glucose regulation of transcription from the insulin gene is an important
example. 24
A. PEPTIDES FROM THE GUT AND PANCREAS

Peptide hormones from the enteroendocrine system, such as cholecys-
tokinin, gastric inhibitory peptide, and pancreatic polypeptide, are secreted
into the hepatic portal circulation in response to nutrients present in the lumen
of the gut (Table l ), 25 -44 and starvation causes their output to decline. Glucose
stimulates secretion of gas tic inhibitory peptide and insulin with weaker effects
on other hormones. Amino acids stimulate secretion of pancreatic polypeptide
from the intestine and pancreas. and augment secretion of glucagon and
insulin; dietary protein or amino acids also increase secretion of growth hor-
mone, adrenocorticotropin (ACTH), and cortisol. Fatty acids have minor
effects on insulin and glucagon but strongly stimulate secretion of neurotensin,
peptide YY, gastic inhibitory peptide, and cholecystokinin.
Tissues that contain receptors for these peptide hormones may respond
to diet-induced secretion of the gut peptides. For example, the endocrine and
exocrine pancreas contains receptors for many peptides. 45 Insulin output is
increased by gastric inhibitory peptide, pancreatic polypeptide, neurotensin,
and bombesin, and may be decreased by galanin. The liver contains receptors
for secretin, pancreatic polypeptide, and vasoactive intestinal peptide, 46 and
starvation would be expected to decrease the activity of signaling pathways
linked to these receptors. At present, it is not clear to what extent peptides
from the intestinal mucosa affect gene expression in the liver. However,
vasoactive intestinal peptide activates adenylate cyclase, 47 and cyclic adeno-
sine-3'5'-monophosphate (cAMP) regulates numerous hepatic genes. During
the initial phase of starvation in humans and rats, secretion of insulin from
the pancreas declines as levels of glucose and other secretagogues decrease
in the blood plasma. 24 .48 - 51 The output of pancreatic glucagon changes less
than insulin, but fasting causes an initial increase in glucagon secretion in
the human. 50 In both species, the ratio of insulin to glucagon in the blood
plasma declines. Fasting increases the activity of the sympathetic nervous
system and leads to increased secretion of epinephrine from the adrenal med-
ulla. 52 Rising levels of catecholamines and ACTH stimulate output of glu-
cocorticoids from the adrenal cortex and antagonize the action of insulin while
enhancing the action of glucagon on expression of genes that participate in
gluconeogenesis.
Studies done with experimental animals suggest that the relative levels
of carbohydrate and protein in the diet modify the ratio of insulin to glucagon
secreted into the plasma. 53 Glucose is the primary stimulus to insulin secretion
and is necessary for secretion in response to most amino acids; in contrast,
glucose inhibits glucagon secretion, whereas amino acids stimulate this pro-
cess. 33 In portal blood, the ratio of glucagon to insulin was correlated with
the concentrations of total amino acids and branched-chain plus aromatic
amino acids. 53 Diets containing high carbohydrates produce different meta-
bolic effects on liver than diets with high protein content. With mixed or high
carbohydrate diets, the level of glycogen in liver increases during the feeding
period, and cyclic AMP levels remain low. However, diets containing elevated
protein stimulate production of cAMP, and little glycogen is deposited. 54 This
is confirmed by measurements of enzyme activities that are regulated in
opposing fashion by insulin compared to glucagon or cAMP. Diets containing
high protein caused phosphoenolpyruvate carboxykinase (PEPCK) levels to
increase, whereas pyruvate kinase declined, and carbohydrates promoted the
opposite effect. 55
The ability of different amino acids to promote hormone secretion differs

among hormones and across species. In general, amino acids are weaker
secretagogues for insulin than is glucose, but leucine, glycine, serine, and
alanine all enhanced insulin secretion after intravenous administration to sheep. 56
While sheep are ruminants, monogastric animals have similar responses with
respect to insulin release. In many species, arginine stimulates insulin secre-
tion more potently than other amino acids. Secretion of glucagon in sheep
was elevated most by alanine, glycine, serine, and arginine, whereas aspartic
acid, phenylalanine, and glutamate were the most effective releasers of growth
hormone (GH or somatotropin). 56 The hormonal response caused by refeeding
diets that contain elevated protein may depend on the relative level of different
amino acids in the protein source.
B. PITUITARY HORMONES, THYROXINE, AND

GLUCOCORTICOIDS
Nutrient status strongly modifies hormonal secretion from the pituitary
gland, and this effect is thought to be mediated by altered synthesis and
secretion of neurotransmitters and releasing factors from specific areas in the
hypothalamus. Fasting in humans and experimental animals causes secretion
of most pituitary hormones to decline, and brings about a reversible condition
with symptoms similar to effects of hypophysectomy. ~ Growth is inhibited,
48 51
reproductive cycles cease, and the metabolic rate declines because the secre-
tion of thyroid-stimulating hormone, prolactin, luteinizing hormone, and fol-
licle-stimulating hormone is depressed. Decreased secretion of gonadotropic
hormones represses the preovulatory surge of LH secretion in the female and
causes a decline in output of gonadal steroids in both sexes. Sex steroids are
anabolic agents, and their decreased secretion reduces the capacity to maintain
synthesis of proteins that are essential to normal functions of muscle and
bone. In the rat, secretion of GH declines, whereas in humans output of GH
increases during fasting or hypoglycemia.
The response of the thyroid gland to stimulation by thyrotropin diminishes
with food deprivation in man and experimental animals. 44 •57 Although plasma
concentrations of thyroid-stimulating hormone (TSH) may not differ greatly,
levels of 3,5,3' -triiodothyronine (T3 ) and T4 decline during long-term food
deprivation, and TSH levels do not rise to compensate. Pituitary function is
affected, as indicated by a lower level of mRNA for thyroid-stimulating
hormone in the anterior pituitary gland of rats. The mRNA for thyrotropin-
releasing hormone in the hypothalamus declines in rats. 58
Acute hypoglycemia, which can be caused by injecting insulin, increases
the secretion of counter-regulatory hormones from the pituitary gland, in-
cluding ACTH, GH, and prolactin. This effect has been postulated to involve
glucoreceptors in specific areas of the brain, and is thought to reflect the lack
of glucose and not the presence of insulin. For instance, administering suf-
ficient glucose to prevent hypoglycemia inhibited the release of ACTH and
GH caused by insulin. 38 During starvation, elevated output of ACTH causes

the weight of the adrenal glands to increase, and increases plasma levels of
corticosterone three- to fivefold in rats. 39-41
Effects of GH on metabolism partly result from increased synthesis and
secretion of the somatomedin, insulin-like growth factor I (IGF-I) from liver
and other tissues, which is initiated through binding of GH to its receptor.
In rats, the level of IGF-I declines during fasting or restriction of dietary
protein, and this is paralleled by decreased levels of IGF-I mRNA in liver. 59
Fasting also causes resistance to the action of GH, and this may be a con-
sequence of decreased levels of the GH receptor in liver. 60 Fasting causes the
concentration of this receptor and its mRNA to decline in liver.
The secretion of GH is controlled by GH-releasing factor (GHRF), so-
matostatin, and neurotransmitters such as dopamine, 61 and the cause for the
changes due to fasting and refeeding have not been fully explored. However,
Bruno et al. 62 •63 have found that food deprivation in rats decreased the con-
centration of the mRNA for prepro-GHRF in the hypothalamus without af-
fecting the mRNA for somatostatin. Upon refeeding, low levels of protein (4
to 12% of the diet by weight) were found to increase the concentration of
prepro-GHRF mRNA. Feeding meals containing protein increases the con-
centration of amino acids in the brain, 64 and GH secretion is stimulated by
an intravenous infusion of amino acids in man and other animals. 36 •65 In
sheep, the response of GH to arginine and leucine was smaller than responses
to the acidic amino acids. Several amino acids caused a mild hyperglycemia,
with glycine, asparagine, serine, alanine, threonine, glutamine, arginine,
methionine, proline, and phenylalanine being most effective. Leucine caused
plasma glucose levels to decline. In humans, the amino acids that best stim-
ulated secretion of growth hormone were arginine, histidine, lysine, leucine,
and phenylalanine. 65
During refeeding, metabolic effects produced by the diet are most extreme
when the majority of calories are derived from one nutrient. Matzen et al. 36
compared effects in human subjects who ingested 500 kcal of carbohydrate
(dextrin-maltose) or an isocaloric diet with 80% of calories from protein.
Plasma glucose and insulin did not rise in subjects receiving the protein diet,
but levels of GH and glucagon were both elevated, and TSH gradually de-
clined. During this period, levels of T3 , reverse T3 , and T4 did not change
significantly. In evaluating studies that make use of the fasting-refeeding
paradigm, it is important to be mindful of these complex effects of nutrients
on hormonal secretion.
IV. PHYSIOLOGICAL EFFECTS OF AMINO ACIDS
Although amino acids serve a common purpose during protein synthesis,

each amino acid has unique properties and functions in interorgan metabolism,
some of which may influence gene expression. For example, alanine and
glutamine are released from extrahepatic tissues (mainly skeletal muscle)

under gluconeogenic conditions. 5 •6 Alanine is an important glucose precursor
through the glucose-alanine cycle, whereas glutamine serves as an important
energy source for the intestine and kidney. 66 Besides serving as a glucose
precursor, alanine facilitates gluconeogenesis by inhibiting pyruvate
kinase, 67•68 thus preventing a futile cycle from developing between glycolysis
and gluconeogenesis. In addition, alanine inhibits proteolysis and stimulates
protein synthesis in liver, 69 ·70 which may help prevent excessive loss of protein
during starvation. Finally, alanine serves as a potent stimulus to release of
glucagon/' which is elevated relative to insulin by fasting, diabetes, or inges-
tion of diets containing protein. Leucine, or a product of leucine metabolism,
inhibits protein degradation in muscle and liver, 72 and is a potent secretagogue
for glucagon and insulin. 73 Aspartic acid and glutamate are excitatory neu-
rotransmitters, 74 and both stimulate secretion of GH from the anterior pituitary
after intravenous infusion. It is feasible that some central effects of acidic
amino acids could be mediated through receptors in circumventricular organs.
In addition to serving as a precursor for neurotransmitter synthesis, tryptophan
exerts numerous metabolic effects including stabilization of several hepatic
enzymes that do not participate in its metabolism, including tyrosine ami-
notransferase and phosphoenolpyruvate carboxykinase. 75 Tryptophan or are-
lated metabolite activates PEP carboxy kinase by a mechanism that may require
glutathione peroxidase. 76 •77 Pharmacological doses of tryptophan stimulate
protein synthesis, and evidence suggests that it binds to a receptor on the
nuclear envelope that may facilitate translocation of mRNA into the cyto-
plasm.78 Arginine stimulates secretion of insulin, glucagon, and somatotropin,
and is the substrate used for synthesis of nitric oxide, a relaxant of smooth
muscle in the vascular endothelium. 79
After protein-containing meals have been ingested, the level of amino
acids increases in the portal and systemic blood. The partial list of functions
described above suggests that hyperaminoacidemia may have numerous met-
abolic consequences that are unrelated to protein synthesis, but that may be
mediated by altered hormone secretion, activation of the autonomic nervous
system, formation of ammonia, or by several other processes. Therefore, it
is essential that the level of catabolic enzymes for amino acids adapt to the
dietary intake and prevent large changes in their concentration.
V. REGULATION OF ENZYME SYNTHESIS BY

DIETARY PROTEIN
A. EFFECTS OF DIETARY COMPOSITION ON

ENZYME RESPONSES
The metabolic response of the liver to refeeding diets with different
compositions has been studied extensively, and it is clear that nutrient levels
as well as hormonal effects are important in the induction or repression of
Q)
"'
c::
0
100
•••••
•• ......
............
.......... r/:l.
h\
I
Q.
Ill
Q)
•o ............... 4····· Serine
80
a: ••••• •••• De hydratase
. ·•· ··...
.... . ·· ··.•.·.
E
::l
60
E #' ••
)(
ca IS. ••••••
--
,i'
:::E
0
40
. ······ ·•···•. •
0
c::
Q)
20
A~
. /
/ Glucose 6-Phosphate~··.
Dehydrogenase •••• .b
...
Q) 0 ···············IS.
a.. 0 20 40 60 80 100
Dietary Protein (Per Cent by Weight)
FIGURE 1. Patterns of induction by dietary protein differ for various enzymes and peptides.
The threshold of induction is quite low for xanthine oxidase (squares) and glucose-6-phosphate
dehydrogenase (circles), but is substantially higher for serine dehydratase (triangles) and most
transaminases (not shown). Maximal expression of xanthine oxidase occurs at a dietary casein
level of about 20%, whereas serine dehydratase continues to increase in proportion to protein
consumed, and glucose-6-phosphate declines at high levels of protein. Data were replotted from
articles by Harper,' Potter and Ono," and Ogawa et al 82
various enzymes. The synthesis of lipogenic enzymes is increased by diets

containing high carbohydrate content, 80 and these conditions suppress syn-
thesis of several transaminases, dehydratases, and enzymes of the urea cycle
that are sensitive to dietary amino acids. As a group, the concentrations of
these enzymes increase with dietary protein and decrease with dietary car-
bohydrate.4-8 However, the level of protein required to induce a response
differs for various enzymes. As shown in Figure 1, refeeding diets that contain
as little as 5% protein by weight causes induction of xanthine oxidase and
glucose-6-phosphate dehydrogenase, 81 whereas a threshold of about 18% pro-
tein is required to induce serine dehydratase and its mRNA. 82 Glucose-6-
phosphate dehydrogenase is interesting in the requirement for both amino
acids and carbohydrate to achieve full induction, and dietary fat inhibits
induction of the enzyme. 83 Whereas xanthine oxidase reaches a plateau beyond
which it is not further induced at about 20% protein, and very high levels of
protein cause glucose-6-phosphate dehydrogenase to decline, 81 serine dehy-
dratase continues to increase in proportion to the amount of amino acids that
is ingested.
It is interesting to compare the response of gluconeogenic enzymes during
fasting and refeeding to that of glucokinase. During starvation, the liver must
maintain euglycemia, and glucokinase hinders this by trapping glucose in the
liver as glucose-6-phosphate. Levels of glucokinase decline during starvation,
-...
c 80
Q) A. Glucokinase
0 70
Q,
60
Dl
-
E 50
Iii
c 40
::::>
-
>- 30
-
>
u
c(
20
10
Fasting
0 2 3 4
-...
Q)
0
Q,
200
B. PEP Carboxykinase
150
E Fasting
...
as
-
Dl
._
ltl 100
c
::::>
->- 50
->
u
c( 0
Refeeding
0.0 0.5 1.0 1.5 2.0
Time (Days)
FIGURE 2. Enzyme induction in response to refeeding depends on composition of the diet.

Glucokinase concentration in liver declines during fasting and is induced to high levels in response
to mixed diets that contain a large proportion of carbohydrate, but not by diets that contain
extreme levels of protein (Panel A). PEP carboxykinase levels increase during fasting in support
of gluconeogenesis, and diets containing high levels of carbohydrate rapidly inhibit mRNA
transcription and cause the enzyme concentration to decline (Panel B. open squares). However,
refeeding diets that contain high levels of protein increase synthesis of PEP carboxy kinase, and
the enzyme concentration increases. The figure was prepared from data reported by Blumenthal
et al., 84 Hopgood et al.. 85 and Seitz et al.""
whereas levels of PEP carboxykinase increase (Figure 2). 84 ·85 After feeding,
the level of glucokinase increases and PEP carboxykinase decreases in pro-
portion to dietary carbohydrate. In contrast, dietary protein fails to produce
an increase in glucokinase in fasted rats, and increases the activity of PEP
carboxykinase (Figure 2). The decline in PEP carboxykinase activity results
from the ability of insulin and carbohydrate to inhibit mRNA production. 87

Recent studies in transgenic mice have shown that the stimulatory effect of
protein on PEPCK expression in liver and kidney is also due to increased
mRNA accumulation. 88
When high-protein diets are first introduced, the level of amino acids in
the bloodstream increases greatly after feeding. However, continual feeding
of diets with elevated protein leads to adaptive increases in the capacity of
the liver to metabolize amino acids. As a result, the concentration of amino
acids that are degraded in the liver actually decreases in well-adapted animals.
No comparable increase occurs for the branched-chain amino acid amino-
transferase, which is located in the mitochondria of skeletal muscle and liver,
but the keto acid dehydrogenase activity increases in proportion to dietary
protein. 89 Therefore, the level of leucine, valine, and isoleucine increases
linearly as dietary protein content increases. This may serve several useful
purposes, such as permitting the adaptation to high levels of protein to continue
despite the increased capacity for degradation of amino acids. When high
protein diets are fed continually, the activity of enzymes involved in lipo-
genesis and glycolysis decreases and remains at a low level.
B. DIETARY REGULATION OF TYROSINE

AMINOTRANSFERASE
Tyrosine aminotransferase (TyrAT; E.C. 2.6.1.5) provides a useful ex-
ample of a mammalian gene that is regulated by dietary protein and carbo-
hydrate. This enzyme is found only in liver, and the gene is inactive during
most of fetal life. Transcription does not begin until the last third of gestation
in the rat (Figure 3A). Just prior to birth, the promoter becomes active in
hepatocytes but not elsewhere, and the level of TyrAT reaches a value that
is less than one tenth of adult levels. 90 •91 In the earliest period, the rate of
transcription appears to be fixed at a low level, and no response is obtained
for hormones that augment transcription. However, just prior to birth this
changes, as the gene acquires the ability to respond weakly to glucagon or
cyclic AMP, and marginally to hydrocortisone.
Just after birth, the response ofTyrATchanges impressively. At this time,
animals that fail to replenish their stores of blood glucose must initiate glu-
coneogenesis or risk death, and synthesis of a set of enzymes that supports
this process begins at a high rate. The enzyme levels increase manyfold within
a few hours of birth, surge past the adult levels, and then fall back to typical
values that are roughly twentyfold higher than were found in the fetal liver. 90 •91
Not only does transcription from the promoter increase to a higher "basal"
rate, 92 but the capacity to respond to glucocorticoid hormones becomes prom-
inent (Figure 3A). Whereas the fetal hepatocyte responds with at best a twofold
increase in enzyme after steroid treatment, in the adult, responses of tenfold
are usual at optimal doses of hormone.
?:-
'
.:: 100
u
C( A
I
I
I
80 Birth I
I
Promoter Status: I
1
I
I
60 I
I
Inactive Active Inducible I
I
I
40
~ 20
Ill
~
~ o6-~~-o~~~~~~~~~~~~~~~~-r~~~
·10 .g ·8 ·7 ·6 ·5 ·4 ·3 ·2 ·1 0 2 3 4
Days
?:- Fasted
,.,.
·;: 200
60%
u
C( B h.
I \ I \
I \ I \
I \ I \
I \ I \
I \ I \
I \
-.
I \
I \
I \ I \
I \
I \
,.
100 I 30% \
I \ I \
: ....'f~ '\ i. ...........~ ~
I
~
.....•
.•· o •.....
~
...._.._.._.
1 e 12Yo •.. ,. 1
1
rl
................. :.~ ... ..
Ill I •" ... ._.._ I ;
c ~-··
I •
"'0
>.
1-
0 6 12 18 24 30 36 42 48
Hours Elapsed
FIGURE 3. Regulation of the gene encoding the enzyme, tyrosine aminotransferase, is typical
of mammalian genes that are induced by dietary protein. Panel A: Acquisition of the capacity
to synthesize this enzyme and its mRNA occurs in liver during the latter third of gestation.
Enzyme activity is indicated by the open circles. At the time of birth (arrow), synthesis begins
at a high rate in support of gluconeogenesis, and the capacity to respond to diet and hormones
begins. Symbols indicate the following treatments: injection of fetuses with cyclic AMP (closed
circles) of hydrocortisone (open squares); injection of neonates with hydrocortisone (closed
squares). Panel B: The quantity of enzyme in adult liver undergoes a diurnal cycle because
mRNA synthesis responds to feeding and the level of protein in the diet. Tissue from animals
that are fed a diet containing 12% protein contains very little TyrAT, and the diurnal variation
is small (open circles). The variation becomes much larger when dietary protein makes up 30%
(closed circles) or 60% (triangles) of the diet by weight. Diurnal rhythms of this magnitude
require that the enzyme and its mRNA are both unstable. as predicted by the kinetic theory
discussed in Chapter I. The figure is a composite drawn from data reported in References. 90· 9 '
~ 6
>-.::' A. Serine Dehydratase
_.c
u >. 5
<-o0
.a 4
CD
:: E
~ Cl 3
-gO
>-0 Casein hydrolyzate
.c!:
CD !II 2
c=c: Casein hydrolyzate
::::)
CD
plus glucose
...c:
CD
~ ~
(/)
0 6 12 1B 24
Time (Hours)
120
CD B. Ornithine Aminotransferase
!II
...cu
-
CD
!II
100
-... -...
c:
cu ~BO
0 c:
-
c: 00 60
E
<
-
0
'#. 40
CD
c:
-...
.c
c:
0
20
0
0 5 10 15 20 25
Concentration (mM)
FIGURE 4. Glucose interferes with the ability of protein to increase the synthesis of several
enzymes of amino acid catabolism in liver. (A) Serine dehydratase is induced by dietary protein
or by hydrolyzed casein, administered by gastric incubation to intact rats at the times indicated
by arrows (closed circles). If I g of glucose is included along with the casein, the induction is
prevented (open circles). (B) Carbohydrate repression of ornithine aminotransferase can be
observed in isolated cells. Hepatocytes were prepared from rats fed a low protein diet and cultured
in the presence of hexoses with insulin added to the culture medium. After 28 h, fresh medium
containing the hexoses and 0.1 rnM dibutyryl cyclic AMP was added, and samples were taken
for assay at 52 h. (Data replotted from Peraino et al. 96 and Merrill and Pitot 101 .)
A second prominent feature of the gene for TyrAT in adult liver is that
it generates enzyme in a diurnal rhythm that is driven by the feeding cycle.
Shown in Figure 4B are results of allowing rats access to diets of different
protein content (12, 30, or 60% by weight) under an inverted light/dark
cycle. 93 The enzyme level begins to rise at the onset of darkness after feeding
is initiated, but begins to decline soon after food is removed. This type of
circadian rhythm is a prominent feature of short-lived gene products that
respond to nutrients. In the case ofTyrAT, the immediate stimulus that causes
the increased rate of synthesis is not totally clear; it is likely to involve changes
in levels of insulin, glucagon, and adrenal steroids in the blood plasma in
addition to elevated amino acid content. Recent evidence suggests that the
nutrient-related induction of TyrAT that occurs during feeding is a result of
altered production of mRNA, as judged by quantitative Northern analysis. 94
The low amplitude of the rhythm observed when animals are fed low protein
diets may indicate that carbohydrates decrease the rate at which this enzyme
is synthesized; this effect is noted for a variety of enzymes that initiate
degradation of amino acids. 7
C. REGULATION OF SERINE-THREONINE DEHYDRATASE BY

PROTEIN AND CARBOHYDRATE
Serine Dehydratase (SDH; E.C. 4.2.1.13) is a pyridoxal phosphate-de-
pendent enzyme that catalyzes deamination of serine and threonine. SDH
expression is limited to the liver and kidney cortex and is developmentally
controlled. Levels of enzyme begin to increase in liver shortly after birth,
but adult levels are not attained until the end of the first month in the rat. 95
Dietary protein and carbohydrate regulate its synthesis in opposing fashion,
with protein causing a large increase in the rate of enzyme synthesis, whereas
carbohydrate inhibits enzyme synthesis. 96 - 98 The concentration of SDH begins
to increase after feeding of high protein diets is initiated, and increases up to
100-fold or more, with the final concentration almost directly proportional to
the proportion of protein in the diet. By comparison, most transaminases
increase no more than tenfold under this regimen. Synthesis of SDH in isolated
hepatocytes is under the control of glucagon and glucocorticoids, which must
be present simultaneously for optimal induction to occur. 99 Insulin strongly
inhibits synthesis of the enzyme in cultured cells. Hormones do not affect
the stability of the mRNA or the enzyme, which have half-lives of about 5
to 10 h and 13 h, respectively.
The opposing effects of dietary protein and carbohydrate on the synthesis
of SDH and ornithine aminotransferase can be observed after intubation of
hydrolyzed casein into the stomach at 6 h intervals (Figure 4A). 96 This causes
a large increase in enzyme activity as a result of increased synthesis of protein.
However, the induction is prevented if glucose or fructose is administered
with the casein hydrolyzate. Insulin secretion does not fully account for the
ability of glucose to inhibit induction by dietary protein because glucose
represses the synthesis of SDH in livers of streptozotocin-diabetic rats whereas

insulin had relatively little effect. 7 Glucose repression of SDH is due to
inhibition of transcription, for SDH mRNA virtually disappears when rats are
fed a protein free diet. 82 Although the intracellular mediators of this effect
are unknown, Sudilovsky et a!. showed that intubation of glucose inhibited
induction of SDH and TyrAT by glucagon without preventing the increase
in cAMP caused by this hormone. 100
Repression of enzyme synthesis by carbohydrates can be demonstrated
in cultured cells. Merrill and Pitot 101 isolated hepatocytes from livers of rats
that had previously been fed a low-protein diet. The dibutyryl analog of cAMP
increased the synthesis and concentration of ornithine aminotransferase sev-
eralfold in the presence of culture medium containing insulin and glucose.
However, addition of fructose (Figure 4B), sorbose, sorbitol, or glycerol to
the culture medium prevented the induction. The inhibitory effect of these
carbohydrates required their metabolism and was associated with a decrease
in mRNA concentration. This evidence suggested that part of the inhibitory
effect of carbohydrates on enzyme synthesis was due to a direct action of the
sugar or a metabolite of the sugar on production of specific mRNAs. Direct
effects of glucose in stimulating mRNA production in hepatocytes has been
observed by Towle et a!. for pyruvate kinase and a protein called S14. 102 In
these cases, glucose causes a concentration-dependent increase in mRNA
levels and transcription rate when insulin is present at a fixed concentration.
Ogawa et a!. 82 showed that the concentration of SDH mRNA achieved
after feeding levels of dietary casein ranging from 0 to 91% of total calories
was proportional to the amount of enzyme activity, and presumably protein
concentration. When the protein-fed animals were switched back to a protein-
free diet, enzyme activity declined with a half-life of about 18 h, and mRNA
declined with a half-life of about 3 h. It was not clear whether half-life of
the mRNA was affected by diet, but earlier evidence suggested that glucose
might destabilize the mRNA.
Recent studies have demonstrated that when the gene for SDH is activated
by protein, an upstream region becomes sensitive to digestion in vitro with
the nuclease, DNAse I. Four DNAse !-hypersensitive sites (I to IV in Figure
5) were identified in the region located within 6 kb upstream from the structural
gene. Sites II and III were especially sensitive in samples from rats fed the
high-protein diet, indicating that diet altered structure of the chromatin in
these areas. This change in chromatin structure occurred only in liver, although
a second hypersensitive area near the start site of transcription was observed
in kidney (Figure 5). Another interesting feature observed in this study was
that diet contributed to a pronounced, diurnal cycle in mRNA production that
was not as evident for the enzyme level, possibly because of the mRNA had
a much shorter half-life. Region II was most affected by the diurnal cycle,
and sensitivity in region I decreased. Ogawa et a!. 82 concluded that nutritional
status could activate or inactivate transcription of the SDH gene by controlling
regulatory factors that act in these regions of the SDH gene.
Regulatory Regions in the Promoter

of the Serine Dehydratase Gene
-10 kb -8 kb -6 kb -4 kb -2 kb 0
Region IV III II I
.ill .IIIII II
Dietary Protein * + + *
Dietary Carbohydrate * +
Circadian Variation +
Active in Kidney +
FIGURE 5. DNA elements that respond to dietary protein and that promote circadian variation
in mRNA synthesis have been identified in the promoter for the gene encoding serine dehydratase
by testing for their sensitivity to digestion by DNAse I. DNA that was prepared from rats fed
diets containing elevated casein was readily digested by the nuclease at sites that mapped to
about -3200 and -3800 nucleotides 5' to the start site of transcription (regions II and III).
These sites were not sensitive to nuclease digestion in DNA isolated from animals fed a protein-
free diet, but a site located at positions - 100 to - 200 was more sensitive to digestion (region
I). Sensitivity of DNA in regions I and II varied in circadian fashion, with region II being more
sensitive in the nocturnal feeding period. The figure is based on data reported by Ogawa et al ' 2
VI. CONCLUSIONS
Even though fasting causes reproducible changes in hormone secretion

and gene expression, the consequences of refeeding depend on the composition
of the diet. During fasting, the rate of gluconeogenesis is elevated, and this
becomes inappropriate if an animal is refed a diet that contains abundant
carbohydrates. However, if an animal is refed a diet containing high protein
or lipid but little carbohydrate, that animal must continue to produce glucose.
Therefore, it is not surprising that hormonal responses differ according to
dietary content of the energy-yielding nutrients.
Studies of changes in gene expression after fasting and refeeding have
already yielded several striking results. First, fasting and refeeding alter the
structure of chromatin in regions that flank structural genes involved in met-
abolic regulation, and the alterations depend on the amount of carbohydrate,
protein, and fat in the diet. 82 •103 This confirms the idea that second messenger
systems are activated in specific cells in response to dietary constituents,
either directly or indirectly as a result of diet-related hormone secretion. It
should prove very interesting to determine the components of these transducing
systems and compare them with systems used in microorganisms. Second,
some effects of nutrients take place in the absence of hormones, whereas
others require hormones or are secondary to hormone secretion. In most cases,

nutrients and hormones must both be present to obtain an optimum response.
This makes sense because the point of modifying gene transcription in such
cases is to permit adaptation to nutrient supply. Third, it is likely that the
genetic apparatus responds more to nutritional supply than is apparent from
measures based on enzyme activities. Regulated mRNAs generally have shorter
half-lives and much lower concentrations than proteins, and are much more
sensitive indices of responses to diet. It will be important to survey changes
in mRNA levels in various organs and determine the extent to which nutrients
reprogram the genetic apparatus not just in the liver and pancreas, but also
in the brain, kidney, skeletal muscles, and other organs.
REFERENCES
I. Harper, A. E., Effect of variations in protein intake on enzymes of amino acid metab-
olism, Can. J. Biochem. Physiol., 43, 1589, 1965.
2. Fine, L. G., The role of nutrition in hypertrophy of renal tissue, Kidney Int., 32 (Suppl.
22), S2, 1987.
3. Szepesi, B. and Freedland, R. A., Dietary effects on rat liver enzymes in meal-fed
rats, J. Nutr., 96, 382. 1968.
4. Peters, J. C. and Harper, A. E., Adaptation of rats to diets containing different levels
of protein: effects on food intake, plasma and brain amino acid concentrations and brain
neurotransmitter metabolism, J. Nutr .. 115, 382, 1985.
5. Munro, H. N. and Crim, M. C., The proteins and amino acids, in Modern Nutrition
in Health and Disease, Shils, M. E. and Young, V. R., Eds .. 7th ed., Lea and Febiger,
Philadelphia, 1988, I.
6. Young, V. R. and Marchini, J. S., Mechanisms and nutritional significance of metabolic
responses to altered intakes of protein and amino acids. with reference to nutritional
adaptation in humans, Am. J. Clin. Nutr .. 51. 270, 1990.
7. Soling, H., Kaplan, J., Erbstoeszer, M., and Pitot, H. C., The role of hormones in
glucose repression in rat liver, Adv. Enzyme Regul., 7, 171, 1969.
8. Goldberg, M. L., The glucose effect: carbohydrate repression of enzyme induction.
RNA synthesis, and glucocorticoid activity-a role for cyclic AMP and cyclic GMP.
Life Sci .. 17, 1747, 1975.
9. Teppermann, H. M., Teppermann, J., Pownall, J. D., and Branch, A., On the
response of hepatic glucose-6-phosphate dehydrogenase activity to changes in diet com-
position and food intake pattern, Adv. Enzyme Regul., I. 121. 1963.
10. Harper, A. E., Diet and plasma amino acids, Am. J. Clin. Nutr .. 21. 358, 1968.
II. Ashley, D. V., Barclay, D. V., Chauffard, F. A., Moennoz, D., and Leathwood,
P. D., Plasma amino acid responses in humans to evening meals of differing nutritional
composition, Am. J. Clin. Nutr., 36, 143, 1982.
12. Eisenstein, A. B., Srack, 1., Gallo-Torres, H., Georgiadis, and Miller, 0. N., In-
creased glucagon secretion in protein-fed rats: lack of a relationship to plasma amino
acids, Am. J. Physiol., 236, E20, 1979.
13. Felig, P., Wahren, J., and Hendler, R., Influence of oral glucose ingestion on splanch-
nic glucose and gluconeogenic substrate metabolism in man, Diabetes, 24, 468, 1975.
14. Dole, V. P., James, A. T., Webb, P. W., Rizack, M. A., and Sturman, M. F., The
fatty acid patterns of plasma lipids during alimentary lipemia, J. Clin. Invest., 38, 1544,
1959.
15. Klahr, S., Effects of protein intake on the progression of renal disease, Annu. Rev. Nutr.,
9, 87, 1989.
16. Farwell, D. C., Miguez, J, B., and Herbst, E. J,, Ornithine decarboxylase and po-
lyamines in liver and kidneys of rats on cyclical regimens of protein-free and protein-
containing diets, Biochem. J., 168, 49, 1977.
17. Jakobsson, B., Celsi, G., Lindblad, B., and Aperia, A., Influence of different protein
intake on renal growth in young rats, Acta Paediatr. Scand., 76, 293, 1987.
18. Wada, L., Don, B. R., and Schambelan, M., Hormonal mediators of amino acid-
induced glomerular hyperfiltration in humans, Am. J. Physiol., 260, F787, 1991.
19. King, A., Troy, J,, Anderson, S., Neuringer, J,, Gunning, M., and Brenner, B. M.,
Nitric oxide: a potential mediator of renal hyperemia and hyperfiltration, J. Am. Soc.
Nephrol., !, 1271, 1991.
20. Rosenberg, M. E., Chmielewski, D., and Hostetter, T. H., Effect of dietary protein
on rat renin and angiotensinogen gene expression, J. Clin. Invest., 85, 1144, 1990.
21. Kaysen, G. A., Rosenthal, C., and Hutchinson, F. N., GFR increases before renal
mass or ODC activity increase in rats fed high protein diets, Kidney Int., 36, 441, 1989.
22. Sens, D. A., Levine, J, H., and Buse, M.G., Stimulation of hepatic and renal ornithine
decarboxylase activity by selected amino acids, Metabolism, 32, 787, 1983.
23. Morris, S. M., Jr., Moncman, C. L., Holub, J, S., and Hod, Y., Nutritional and
hormonal regulation of mRNA abundance for arginine biosynthetic enzymes in kidney,
24. Efrat, S., Surana, M., and Fleischer, N., Glucose induces insulin gene transcription
in a murine pancreatic J3-cellline, J. Bioi. Chern., 266, 11141, 1991.
25. Ferris, C. F., Neurotensin, in Handbook of Physiology, Section 6, The Gastrointestinal
System, Vol. II, Neural and Endocrine Biology, Schultz, S. G., Makhlouf, G. M., and
Rauner, B. B., Eds., Waverly Press, Inc., Baltimore, MD, 1989, 559.
26. Taylor, I. L., Pancreatic polypeptide family: pancreatic polypeptide, neuropeptide Y,
and peptide YY, Op. cit., 1989, 475.
27. Brown, J, C., Buchan, A.M. J,, Mcintosh, C. H. S., and Pederson, R. A., Gastric
inhibitory polypeptide, Op. cit., 1989, 403.
28. Rehfeld, J, F., Cholecystokinin, Op. cit., 1989, 337.
29. Mayhew, D. A., Wright, P. H., and Ashmore, J., Regulation of insulin secretion,
Pharmacal. Rev., 21, 183, 1969.
30. Martin, C. R., Endocrine Physiology, Oxford University Press, Oxford, 1985, 159.
31. Unger, R. H., Dobbs, R. E., and Orci, L., Insulin, glucagon, and somatostatin secretion
in the regulation of metabolism, Annu. Rev. Physiol., 40, 307, 1978.
32. Edozien, J. C., Niehaus, N., Mar, M.-H., Makoui, T., and Switzer, B. S., Diet-
hormone interrelationships in the rat, J. Nutr., 108, 1767, 1978.
33. Rorsman, P., Ashcroft, F. M., and Berggen, P.-O., Regulation of glucagon release
from pancreatic A-cells, Biochem. Pharmacal., 41, 1783, 1991.
34. Castellino, P., Giordano, C., Perina, A., and DeFronzo, R. A., Effects of plasma
amino acid and hormone levels on renal hemodynamics in humans, Am. J. Physiol., 255,
F444, 1988.
35. Pecile, A. and Muller, E. E., Control of growth hormone secretion, in Neuroendocri-
nology, Vol. I, Martini, L. and Ganong, W., Eds., Academic Press, New York, 1966,
537.
36. Matzen, L. E., Andersen, B. B., Jensen, B. G., Gjessing, H. J,, Sindrup, S. H.,
and Kvetny, J,, Different short-term effect of protein and carbohydrate intake on TSH,
growth hormone, (GH), insulin, C-peptide, and glucagon in humans, Scand. J. Clin.
Lab. Invest., 50, 801, 1990.
37. Alvarez, C. V., Mallo, F., Burguera, B., Cacicedo, L., Dieguez, C., and Casanueva,
F. F., Evidence for a direct pituitary inhibition by free fatty acids of in vivo growth
hormone responses to growth hormone-releasing hormone in the rat, Neuroendocrinology,
53, 185, 1991.
38. Vigas, M., Tatar, P., Jurcovicova, J,, and Jezlova, D., Glucoreceptors located in
different areas mediate the hypoglycemia-induced release of growth hormone, prolactin,
and adrenocorticotropin in man, Neuroendocrinology, 51, 365, 1990.
39. Munro, H. N., Steele, M. H., and Hutchison, W. C., Action of dietary proteins and
amino acids on the rat adrenal gland, Br. J. Nutr., 19, 137, 1965.
40. Rabolli, D. and Martin, R. J,, Effects of diet composition on serum levels of insulin,
thyroxine, triiodothyronine, growth hormone, and corticosterone in rats, J. Nutr., 107,
1068, 1977.
41. Slag, M. F., Ahmed, M., Gannon, M. C., and Nuttall, F. Q., Meal stimulation of
cortisol secretion: a protein induced effect, Metabolism. 30, 1104, 1981.
42. Aizawa, T., Yasuda, N., and Greer, M.A., Hypoglycemia stimulates ACTH secretion
through a direct effect on the basal hypothalamus, Metabolism, 30, 996, 1981.
43. Jepson, M. M., Bates, P. C., and Millward, D. J,, The role of insulin and thyroid
hormones in the regulation of muscle growth and protein turnover in response to dietary
protein in the rat, Br. J. Nutr., 59, 397, 1988.
44. Danforth, E., Jr. and Burger, A. G., The impact of nutrition on thyroid hormone
physiology and action, Annu. Rev. Nutr., 9, 201, 1989.
45. Gardner, J, D. and Jensen, R. T., Receptors for gut peptides and other secretagogues
on pancreatic acinar cells, in The Gastrointestinal System, Handbook of Physiology,
Section 6, Vol. 2, Schultz, S. G., Makhlouf, G. M., and Rauner, B. M., Eds., American
Physiological Society, 1989, 171.
46. Rosselin, G., Liver receptors for regulatory peptides, Op. cit., 245.
47. Sanchez, V., Governa, R., and Calvo, J. K., Glycogenolytic effect of vasoactive
intestinal peptide in the rat in vivo, Experientia, 47, 625, 1991.
48. Becker, D. J,, The endocrine responses to protein calorie malnutrition, Annu. Rev. Nutr.,
3, 187, 1983.
49. Goodman, M. N., Larsen, P. R., Kaplan, M. M., Aoki, T. T., Young, V. R., and
Ruderman, N. B., Starvation in the rat. II. Effect of age and obesity on protein sparing
and fuel metabolism in the rat, Am.]. Physiol., 239, E277, 1980.
50. Xie, Q.-W., Experimental studies on changes of neuroendocrine functions during star-
vation and refeeding, Neuroendocrinology, 53 (Suppl. 1), 52, 1991.
51. Thomas, G. B., Mercer, J, E., Karalis, T., Rao, A., Cummins, J, T., and Clarke,
I. J., Effect of restricted feeding on the concentrations of growth hormone, gonadotropins,
and prolactin in plasma, and on the amounts of messenger ribonucleic acid for GH,
gonadotropin subunits, and PRL in the pituitary glands of adult ovariectomized ewes,
52. Sakaguchi, T., Takahashi, M., and Bray, G. A., Diurnal changes in sympathetic
activity, J. Clin. Invest., 82, 282, 1988.
53. Jarrousse, C., Lardeux, B., Bourdel, G., Girard-Globa, A., and Rosselin, G., Portal
insulin and glucagon in rats fed proteins as a meal: immediate variations and circadian
modulations, J. Nutr., 110, 1764, 1980.
54. Tiedgen, M. and Seitz, H. J., Dietary control of circadian variations in serum insulin,
glucagon, and hepatic cyclic AMP, J. Nutr., 110, 876, 1980.
55. Peret, J., Foustock, S., Chanez, M., Bois-Joyeux, B., and Assani, R., Plasma glu-
cagon and insulin concentrations and hepatic phosphoenolpyruvate carboxyk:inase and
pyruvate kinase activities during adaptation of rats to a high protein diet, J. Nutr., Ill,
1173, 1981.
56. Kuhara, T., Ikeda, S., Ohneda, A., and Sasaki, Y., Effects of intravenous infusion
of 17 amino acids on the secretion ofGH, glucagon, and insulin in sheep, Am. J. Physiol.,
260, E21, 1991.
57. Hugues, J.-N., Enjalbert, A., Burger, A., Voirol, M.-J., Sebaoun, J,, and Epelbaum,
J., Sensitivity of thyrotropin (TSH) secretion to 3,5 .3 '-triiodothyronine and TSH-releasing
hormone in rat during starvation, Endocrinology. 119, 253, 1986.
58. Blake, N. G., Eckland, D. J, A., Foster, 0. J, F., and Lightman, S. L., Inhibition
of hypothalamic thryrotropin-releasing hormone messenger ribonucleic acid during food
deprivation, Endocrinology, 129, 2714, 1991.
59. Straus, D. S. and Takemoto, C. D., Effect of fasting on insulin-like growth factor-!
(IGF-1) and growth hormone receptor mRNA levels and IGF-1 gene transcription in rat
liver, Mol. Endocrinol., 4, 91, 1990.
60. Thissen, J, P., Triest, S., Underwood, L. E., Maes, M., and Ketelslegers, J, M.,
Divergent responses of serum insulin-like growth factor-! and liver growth hormone
receptors to exogenous GH in protein-restricted rats, Endocrinology, 126, 908, 1990.
61. Plotsky, P. M. and Vale, W. W., Patterns of growth hormone-releasing factor and
somatostatin secretion into the hypophyseal-portal circulation of the rat. Science, 230,
461' 1986.
62. Bruno, J, F., Olchovsky, D., White, J, D., Leidy, J, W., Song, J,, and Berelowicz,
M., Influence of food deprivation in the rat on hypothalamic expression of growth
hormone-releasing factor and somatostatin, Endocrinology, 127, 2111, 1990.
63. Bruno, J. F., Song, J,, and Berelowitz, M., Regulation of rat hypothalamic prepro-
growth hormone-releasing factor messenger ribonucleic acid by dietary protein, Endo-
crinology, 129, 1226, 1991.
64. Glaeser, B. S., Maher, T. J,, and Wortman, R. J,, Changes in brain levels of acidic,
basic, and neutral amino acids after consumption of single meals containing various
proportions of protein, J. Neurochem., 41, 1016, 1983.
65. Knopf, R. F., Conn, J. W., Fajans, S. S., Floyd, J. C., Guntsche, E. M., and Rull,
J. A., Plasma growth hormone response to intravenous amino acids, Am. J. Endocrinol.,
26, 1140, 1965.
66. Felig, P., Progress in endocrinology and metabolism. The glucose-alanine cycle, Me-
tabolism, 22, 179, 1973.
67. Kayne, F. J, and Price, N.C., Amino acid effector binding to rabbit muscle pyruvate
kinase, Arch. Biochem. Biophys., 159, 292, 1973.
68. Dong, F. M. and Freedland, R. A., Effects of alanine on gluconeogenesis in isolated
rat hepatocytes, J. Nutr., 110, 2341, 1980.
69. Perez-Sala, D., Parrila, R., and Ayuso, M. S., Key role of L-alanine in the control
of hepatic protein synthesis, Biochem. J., 241, 491, 1987.
70. Poso, A. R. and Mortimore, G. E., Requirement for alanine in the amino acid control
of hepatic protein degradation, Proc. Nat/. Acad. Sci. U.S.A., 81, 2039, 1984.
71. Muller, W. A., Faloona, G. R., and Unger, R. H., The effect of alanine on glucagon
secretion, J. Clin. Invest., 50, 2215, 1971.
72. Walser, M., Role of branched-chain ketoacids in protein metabolism, Kidney Int., 38,
595, 1990.
73. MacDonald, M. J,, McKenzie, D. I., Kaysen, J, H., Walker, T. M., Moran, S.M.,
Fabien, L. A., and Towle, H. C., Glucose regulates leucine-induced insulin release
and the expression of the branched chain ketoacid dehydrogenase E Ia subunit gene in
pancreatic islets, J. Bioi. Chern., 266, 1335, 1991.
74. Garthwaite, J, and Garthwaite, G., Mechanisms of excitatory amino acid neurotoxicity
in rat brain slices, Adv. Exp. Bioi. Med., 268, 505, 1990.
75. Pestana, A., Dietary and hormonal control of enzymes of amino acid catabolism in liver,
Eur. J. Biochem., 11, 400, 1969.
76. Lardy, H. A. and Merryfield, M. L., Ferroactivator and the regulation of gluconeo-
genesis, Curr. Top. Cell. Reg., 18, 243, 1981.
77. Punekar, N. S. and Lardy, H. A., Phosphoenolpyruvate carboxykinase ferroactivator

I. Mechanism of action and identity with glutathione peroxidase, J. Bioi. Chern., 262,
6714.
78. Kurl, R. N., Verney, E., and Sidransky, H., Identification and immunohistochemical
localization of tryptophan binding protein in nuclear envelopes of rat liver, Arch. Biochem.
Biophys., 265, 286, 1988.
79. Palmer, R. M. J., Ashton, D. S., and Moncada, S., Vascular endothelial cells syn-
thesize nitric oxide from L-arginine. Nature (London), 333, 664, 1989.
81. Potter, V. R. and Ono, T., Enzyme patterns during metabolic transitions in rat liver
and Morris hepatoma 5123, Cold Spring Harbor Symp. Quant. Bioi., 26, 355, 1961.
82. Ogawa, H., Fujioka, M., Su, Y., Kanamoto, R., and Pitot, H. C., Nutritional reg-
ulation and tissue-specific expression of the serine dehydratase gene in rats, J. Bioi.
Chern., 266, 20412, 1991.
83. Perez, N., Clark-Turri, L., Rabajille, E., and Niemeyer, H., Regulation of rat liver
enzymes by natural components of the diet, J. Bioi. Chern., 239, 2420, 1964.
84. Blumenthal, M. D., Abraham, S., and Chaikoff, I. L., Dietary control of liver glu-
cokinase activity, Arch. Biochem. Biophys., 104. 215, 1964.
85. Hopgood, M. F., Ballard, F. J., Reshef, L., and Hanson, R. W., Synthesis and
degradation of phosphoenolpyruvate carboxylase in rat liver and adipose tissue. Changes
during a starvation-re-feeding cycle, Biochem. J., 134, 445, 1975.
86. Seitz, H. J., Tiedgen, M., and Tarnowski, W., Regulation of hepatic phosphoenol-
pyruvate carboxykinase (GTP). Role of dietary proteins and amino acids in vivo and in
the isolated perfused rat liver. Biochim. Biophys. Acta, 632, 473, 1980.
87. Beale, E. G., Hartley, J, L., and Granner, D. K., N",O''-dibutyryl cyclic AMP and
glucose regulate the amount of messenger RNA coding for hepatic phosphoenolpyruvate
carboxykinase (GTP), J. Bioi. Chern., 257, 2022, 1982.
88. Short, M. K., Clouthier, D. E., Schaefer, I. M., Hammer, R. E., Magnuson, M. A.,
and Beale, E. G., Tissue-specific. developmental, hormonal, and dietary regulation of
rat phosphoenolpyruvate-human growth hormone fusion genes in transgenic mice, Mol.
Cell. Bioi., 12, 1007, 1992.
89. Block, K. P., Aftring, R. P., and Buse, M.G., Regulation of rat liver branched-chain
a-keto acid dehydrogenase activity by meal frequency and dietary protein, J. Nutr., 120,
793, 1990.
90. Sereni, F., Kenney, F. T., and Kretschmer, N., Factors influencing the development
of tyrosine a-ketoglutarate transaminase activity in rat liver, J. Bioi. Chern., 234, 609,
1959.
91. Greengard, 0., Enzymic differentiation in mammalian liver, Science, 163, 891, 1969.
92. Johnson, A. C., Lee, K.-L., Isham, K. R., and Kenney, F. T., Gene-specific acqui-
sition of hormonal responsiveness in rat liver during development, J. Cell. Biochem.,
37, 243, 1988.
93. Watanabe, M., Potter, V. R., and Pitot, H. C., Systematic oscillations in tyrosine
transaminase and other metabolic functions in liver of normal and adrenalectomized rats
on controlled feeding schedules, J. Nutr., 95, 207, 1968.
94. Bartels, H., Herbort, H., and Jungermann, K., Predominant periportal expression of
the phosphoenolpyruvate carboxykinase and tyrosine aminotransferase genes in rat liver.
Dynamics during the daily feeding rhythm and starvation-refeeding cycle demonstrated
by in situ hybridization, Histochemistry, 95, 637, 1990.
95. Snell, K., Enzymes of serine metabolism in normal, developing, and neoplastic rat tissues,
Adv. Enz. Regul., 22, 325, 1984.
96. Peraino, C., Lamar, C., Jr., and Pitot, H. C., Studies on the induction and repression
of enzymes in rat liver. IV. Effects of cortisone and phenobarbital, J. Bioi. Chem., 241,
2944, 1966.
97. Anderson, H. L., Benevenga, N. J., and Harper, A. E., Associations among food
and protein intake, serine dehydratase, and plasma amino acids, Am. J. Physiol., 214,
1008, 1968.
98. Jost, J, P., Khairallah, E. A., and Pitot, H. C., Studies on the induction and repression
of enzymes in rat liver. Regulation of the rate of synthesis and degradation of serine
dehydratase by dietary amino acids and glucose, J. Bioi. Chem., 243, 3057, 1968.
99. Noda, C., Yakiyama, M., Nakamura, T., and lchihara, A., Requirements of both
glucocorticoids and glucagon as co-inducers for activation of transcription of the serine
dehydratase gene in cultured rat hepatocytes, J. Bioi. Chem., 263, 14764, 1988.
100. Sudilovsky, 0., Pestana, A., Hinderaker, P. H., and Pitot, H. C., Cyclic adenosine
3' ,5'-monophosphate during glucose repression in rat liver, Science, 174, 142, 1971.
101. Merrill, M. J, and Pitot, H. C., Inhibition of cyclic AMP-dependent induction of
ornithine aminotransferase by simple carbohydrates in cultured hepatocytes, Arch. Biochem.
Biophys., 259, 250, 1987.
102. Towle, H. C., Thompson, K. S., Liu, J., and Shih, H.-M., Carbohydrate metabolism
regulates gene transcription in liver: identification of response elements and nuclear
factors, FASEB J., 6, A402, 1992.
103. Castro, E. C., Nutrient effects on DNA and chromatin structure, Annu. Rev. Nutr., 7,
407, 1987.
Chapter 17
VITAMIN D-DEPENDENT CALCIUM BINDING

PROTEIN, CALBINDIN-D: REGULATION
OF GENE EXPRESSION
Rajbir K. Gill and Sylvia Christakos
TABLE OF CONTENTS
I. Introduction ...................................................... 378
II. Regulation of Intestinal and Renal Calbindin Gene

Expression ....................................................... 379
III. Regulation of Calbindin Gene Expression in Bone ............... 382
IV. Regulation of Calbindin Gene Expression in the

Nervous System .................................................. 383
V. Analysis of Calbindin Promoters ................................. 384
References ............................................................... 386
0-8493-6961-4/93/$0 ()()+$.50
© 1993 by CRC Pre!'.s, Inc. 377
I. INTRODUCTION
The vitamin 0 endocrine system is a major regulator of calcium home-

ostasis. The primary function of vitamin 0 in maintaining calcium homeostasis
is through direct stimulation of intestinal transport. effects on calcium reab-
sorption in the kidney, and mobilization of mineral from bone. 1 •2 The bio-
logical responses of the hormonally active form of vitamin 0, 1,25-dihy-
droxyvitamin 0 3 (l ,25(0H}:0 3 ), are initiated at its target tissues by the
interaction of the hormone with its specific high affinity intracellular receptor
and the association of the steroid receptor complex with specific sequences
of DNA nucleotides resulting in the activation and expression of genes. 1 ·2
The protein in intestine and kidney whose synthesis is induced by 1,25(0H) 2 0 3
is the calcium-binding protein, calbindin. 3 It has been suggested that the renal
and intestinal calbindin are involved in vitamin 0-dependent intracellular
translocation of calcium ions. 3 · 5 In addition, calbindin has been reported in
many other tissues including bone (chrondrocytes of rat growth plate cartilage
and ameloblasts of rodent teeth) 6 - 10 and tissues that are not regulators of serum
calcium, such as pancreas, "·15 uterus, placenta, 19 ·20 and brain, 21 - 25 and in a
variety of species/·26 ·27 thus suggesting multiple effects of the vitamin 0
endocrine system. Biochemically the calbindins are divided into two classes
of proteins which do not share amino acid homology; the 28,000 M, protein
(calbindin-0 28 k, which has four high affinity calcium binding sites, which is
conserved during evolution, and is present in avian intestine and in avian and
mammalian kidney, pancreas, bone, and brain) and the 9000 M, protein
(calbindin-09 k which has two calcium binding domains, is less conserved
evolutionarily, and is present in mammalian intestine, placenta, uterus, bone
and in mouse and neonatal rat kidney). 3 Although the exact role of the cal-
bindins in many of these tissues remains to be determined, the importance of
the discovery of the calbindins is that key advances in our understanding of
the diversity of the vitamin 0 endocrine system have been made through the
study of the tissue distribution of the calbindins and their colocalization with
the 1,25(0H) 2 0 3 receptor. 28 In addition the biosynthesis of calbindin has
provided an important approach for studying the molecular mechanism of
1,25(0H)z0 3 action in major target tissues such as intestine and kidney. This
chapter will focus on how an understanding of the regulation of calbindin
(regulation by l ,25(0H)z0 3 , tissue-specific regulation and regulation by fac-
tors other than 1,25(0H) 2 0 3 ) has resulted in a better understanding of the
possible functional significance of calbindin and has enabled us to ask ap-
propriate questions so as to define specific genetic regulatory elements in the
calbindin promoter.
Gill and Christakos 379
II. REGULATION OF INTESTINAL AND RENAL

CALBINDIN GENE EXPRESSION
In the intestine, calbindin (calbindin-D28k in avian and calbindin-D9k in

mammals) is localized in columnar epithelial cells 29 ·30 and has been postulated
to facilitate intracellular calcium translocation. 4 Avian and mammalian renal
calbindin-D 28 k is present in the principal cells of the distal tubule in which
selective reabsorption of calcium is known to occur. 31 . 34 In order to obtain a
better understanding of the molecular mechanisms of 1,25(0H) 2D3 action in
these tissues, DNAs complementary to both rat and chick intestinal calbind-
ins35.37 and to mammalian calbindin-D 28 / 8·39 have been cloned, the deduced
amino acid sequences reported, and the probes have been used to examine
the regulation of calbindin gene expression. 40"52 Administration of l ,25(0H)2D3
to vitamin D-deficient animals has been shown to induce rapid transcription
of the calbindin gene in chick42 and rat intestine 43 and in rat kidney. 46 The
peak of calbindin gene transcription occurs at 1 to 3 h which precedes the
peak of accumulation of calbindin mRNA (12 h) and calbindin protein (48
h) (Figure 1). These results indicate that the early increases in calbindin mRNA
and calbindin protein may be due to the stimulation of transcription. However,
the long lag time between transcription and the peak of mRNA and protein
accumulation reflects the involvement of posttranscriptional mechanisms. The
accumulation of calbindin-D mRNA long after 1,25(0H) 2 D 3 injection suggests
an effect on stabilization of RNA by hormone. Recent in vitro studies using
actinomycin D and cycloheximide have also suggested that posttranscriptional
mechanisms are involved in the regulation of the renal and intestinal calbindins
by 1,25(0H) 2 D3.53 ·54 Thus, calbindin mRNA is regulated by the D hormone
at the transcriptional and posttranscriptional level and stabilization of tran-
scribed RNA may be an important mechanism in the regulation of calbindin
gene expression by 1,25(0H) 2 D 3. Developmental studies have indicated that
renal calbindin-D 28 k mRNA increases sharply between birth and 1 week of
age, the period of rapid nephron differentiation, suggesting a possible role
for renal calbindin in the differentiation process. 47 In rat intestine, calbindin-
D9k mRNA is induced in the third postnatal week, coinciding with weaning
and the period of increased duodenal active transport of calcium (Figure 2). 47
Thus, calbindin may be an essential factor involved in the development of
intestinal active calcium transport. It is of interest that the developmental
induction of 1,25(0H) 2 D 3 receptor (VDR) mRNA in intestine (Figure 2) and
kidney coincides with the induction of calbindin mRNA in these tissues,
indicating that the induction of VDR mRNA has a role in regulating the
developmental expression of the calbindin gene. 47
Further studies in intestine and kidney have provided evidence that the
calbindin gene is not always regulated by 1,25(0H) 2 D 3 and that factors other
than 1,25(0H) 2 D 3 can modulate calbindin gene expression. Glucocorticoid
4
A TRANSCRIPTION RATE
3
~
0 6 12 18 24 48 72
I.J..J
(J)
<1:
I.J..J
a:
u
z:
a_.
0
LL.
0 6 12 18
CALBINOIN-D28K
5 PROTEIN
4
0 6 12 18 24 72
TIME AFTER 1,25(0Hl 2D3 INJECTION
FIGURE 1. Analysis of rat renal calbindin-D 28 k transcription, mRNA, and protein levels in
kidney after 1,25(0H),D, treatment. (A) Calbindin-D gene transcription at various times after
1,25(0H),D3 administration (200 ng/100 g body weight to vitamin D-deficient rats). The tran-
scription rate was measured using the nuclear transcription assay46 and expressed as fold increase
compared to control. The results shown are the mean of four different experiments; (B) The
time course of increase in calbindin-D 28 k mRNA levels following 1,25(0H)2 D, administration.
Northern blots of poly(A +)RNA isolated from kidneys of vitamin D-deficient rats at various
times after single injection of 1,25(0H) 20 3 were quantitated by densitometric analysis. 44 The
signal intensities were adjusted for variations in J3-actin densities and the fold increase in mRNA
levels compared to control was calculated. The values shown are the mean of three separate
Northern hybridization experiments: (C) The time course of calbindin-D 28 k accumulation after
1,25(0H)2 0 3 treatment. The levels of calbindin-D at various times were measured by immu-
nobinding assay 44 and expressed as fold increase compared to control. Each value represents the
mean of three separate groups of rats.
A B
AGE (Wks)
2 3 4 5
CALBINOIN- OgK mRNA
,i
I I
I I
I I
I I
CALBINOIN- OgK I I
I
~I I
1
>- I
..... I
Vi I
.........z .I
VDR I
~
-' I
<[ I
z I
'"'Vi I
....> I
;::: I
<[
I
....0::-' I
I
I
I
I
I
IS S rRNA I
I VDR mRNA
I
I
I
/
ft
/
0
0 2 3 4 5
AGE (Wks)
FIGURE 2. Developmental changes in intestinal calbindin-D 9, mRNA and VDR mRNA as-
sessed by Northern blot hybridization. (A) Rat intestinal poly(A +)RNA (5 j.Lg per lane) was
fractionated through a formaldehyde-agarose gel. transferred onto a nylon membrane, and hy-
bridized with calbindin-D9,, VDR, and 18S rRNA probes: (B) densitometric quantitation of
Northern analysis. The results represent the mean :!: SE of three separate Northern hybridization
analyses from different groups of rats. The data were normalized based on results obtained upon
rehybridization with 18S rRNA eDNA. (From Huang, et al., J. Bioi. Chern .. 264, 17454, 1989.
With permission.)
administration has been shown to result in an inhibition of rat and mouse

intestinal calbindin-D9 k gene expression 47•52 which may be related to the pre-
viously reported glucocorticoid-mediated inhibition of intestinal calcium ab-
sorption. 55 Hall et al. 56 and Corradino et al. 57 have noted stimulatory as well
as inhibitory effects of dexamethasone on chick intestinal calbindin-D28k.
Renal calbindin-D28k is unaffected by glucocorticoids. Other tissues are, how-
ever, suggesting that tissue-specific regulation by glucocorticoids of calbindin
gene expression exists. 47•52 Studies concerning the regulation of renal calbindin
have suggested modulation by calcium of calbindin gene expression. 45 •53 In
addition, regulation of calbindin-D28 k by estrogen (in mouse and chick uterus)

and regulation of neuronal calbindin-D 28 k by retinoic acid have also been
observed. 58 ' 60 These findings suggest multiple, tissue-specific interactions in
the regulation of calbindin gene expression. Similar to calbindin, a number
of other steroid responsive genes have also been reported to be under the
control of more than one steroid. 1,25-Dihydroxyvitamin D 3 , dexamethasone,
and retinoic acid have been reported to decrease procollagen gene expression,
and receptors for both vitamins A and D recognize a common 11 nucleotide
response element in the human osteocalcin gene. 61 •62 The DNA binding do-
main of the steroid receptor mediates the specific recognition of the hormone
response elements (HREs) and includes a cysteine-rich region that comprises
two zinc finger-like structures. The first three amino acids in the stem of the
first zinc finger (P box) specify the half-site sequence. 63 Based on the P box
sequence, receptors are classified as members of either the glucocorticoid
receptor (including glucocorticoid, mineralocorticoid, androgen, and proges-
terone receptors) or the estrogen/thyroid hormone receptor subfamily (in-
cluding estrogen, thyroid hormone, retinoic acid, and vitamin D receptors). 63 -65
On the basis of its distinct P box, the members of each class display cross-
recognition of HRE. Umesono et al. recently reported that the retinoic acid
receptor, the thyroid hormone receptor, and the vitamin D receptor recognize
a common consensus half-site sequence and that this site is also the binding
site for the estrogen receptor (but in the form of an inverted repeat). 66lt is of
interest that 1,25(0H) 2 D3 , estrogen, and retinoic acid (members of the estro-
gen receptor/thyroid receptor subfamily) have all been reported to regulate
calbindin-D 28k. Thus, it is possible that receptors for these ligands may rec-
ognize a common sequence in the calbindin-D 28k promoter. It is possible that
the glucocorticoid regulation of calbindin involves the recognition of a dif-
ferent sequence. Understanding the multiple steroid interactions in the reg-
ulation of calbindin gene expression should help clarify the mechanism of
regulation of calbindin function and should aid in our understanding of the
molecular interactions of the 1 ,25(0H) 2D 3 receptor with other steroid hormone
receptors in the control of gene expression.
III. REGULATION OF CALBINDIN GENE

EXPRESSION IN BONE
The calbindins have been localized immunocytochemically in the chon-
drocytes of the growth plate cartilage of the rat and chick and in the ameloblasts
of rodent teeth. 7 - 10 The presence of the calbindins in chondrocytes suggests
that these proteins may be involved in the transport of calcium toward extra-
cellular sites of calcification in the growth plate. The presence of the calbindins
in ameloblasts suggests a role for these proteins in mineralization. From the
immunocytochemical studies it had not been clear whether the calbindins in
chondrocytes and ameloblasts were vitamin D dependent. Recent studies by
Berdal et al., using Northern analysis of total RNA from microdissected
epithelial mesenchymal tooth germ of erupting rat incisor, indicated the pres-
ence of both calbindin-D 9 k and calbindin-D 28 k mRNA in this tissue. 10 The
mRNA for both calbindins was nearly undectectable in vitamin D deficient
rats but was upregulated by I ,25(0H) 2 D 3 injection. These findings, although
suggestive of a direct role of I ,25(0H)zD 3 regulated calbindin gene expression
in the mineralization process, are still preliminary. Further studies are needed
to verify the vitamin D dependence of calbindin in chondrocytes and ame-
loblasts.
IV. REGULATION OF CALBINDIN GENE

EXPRESSION IN THE NERVOUS SYSTEM
Calbindin-D 28 k is one of the major calcium binding proteins in brain. It

is present in some cells but not others throughout the brain and constitutes
between I and 2% of the total soluble protein in certain brain areas. 12 • 13 •21 - 25
Studies by Stumpf et a!. demonstrated specific nuclear retention of
[ H] l ,25(0H) 2 D 3 in many of the same areas of rat brain in which calbindin
3
was localized immunocytochemically, suggesting central modulation by

I ,25(0H) 2 D3 Y·68 It should be noted however that there is only a partial
correlation between calbindin-D 28 k and I ,25(0H)zD 3 receptors in the brain
and that I ,25(0H)zD 3 administration has not been found to regulate calbindin
or calbindin gene expression in this tissue. It is possible that I ,25(0H)zD 3
has a role in the regulation of other neuronal-related processes such as bio-
rhythms, behavior, stress, and reproduction. 68 Although neuronal calbindin
is not altered by I ,25(0H) 2 D3 administration, specific decreases in calbindin
gene expression have been observed in brain areas particularly affected in
aging and in neurodegenerative disease processes (substantia nigra, Parkin-
son's disease; nucleus basalis, Alzheimers disease, and corpus striatum, Hun-
tington's disease). 49 •51 In the aging cerebellum, there is a 73% reduction in
calbindin mRNA and only a 15 to 20% loss in calbindin containing Purkinje
cells, suggesting that the reduction in calbindin mRNA is primarily due to a
decrease of calbindin transcription with age. 51 These findings suggest that a
decrease in calbindin may lead to a failure of calcium buffering or intraneu-
ronal calcium homeostasis which may contribute to calcium-mediated irre-
versible cytotoxic events during aging and the pathological process. This
hypothesis is supported by our recent findings using cell cultures of embryonic
rat hippocampus. 69 A direct relationship between calbindin immunoreactivity
and resistance to neurotoxicity induced by glutamate or calcium ionophore
was documented. In addition calbindin positive neurons were better able to
reduce free intracellullar calcium levels than calbindin negative neurons. These
studies indicate excitoprotective and calcium-reducing roles for calbindin and
are one of the first studies providing evidence for the functional significance
of neuronal calbindin. Due to the significance of calbindin in the nervous
system, it was important to determine the factors which could modulate the
expression of this protein in brain. Recently, we reported an 80 to 85%
0 0 0
200
A B
2001-
c
2001-
."'
;z
::>
::t:
400 f-
."'
;z
::>
::t:
4001-
><
u
'r
u
400
600 ~
6001- 600 f-
800 1- \ \
8001- 800f-
1000 I I I I I I I I I I I I
1000 1000
0 250 500 750 1000 1250 0 250 500 750 1000 1250 0 250 500 750 1000 1250
MOUSE CHICK MOUSE
FIGURE 3. Dot matrix analysis showing the homologies between human (- 840/ +!54),
mouse ( -1075/ + 154), and chick (- 850/ + 154) calbindin-D 28 k gene promoter regions. Each
dot indicates a minimum of 17 base pair matches out of 20. The coding regions are in the lower
right comer.
decrease in the levels of calbindin and calbindin mRNA in the hippocampus

of adrenalectomized rats. 50 Administration of corticosterone to normal and
adrenalectomized rats resulted in the upregulation of calbindin-D28k protein
and mRNA in hippocampus. These studies present the first evidence of a
regulator of calbindin gene expression in brain. Additionally, we recently
found that calbindin-D 28 k can be induced by retinoic acid in neuroblastoma
and medulloblastoma cells. 60 Thus, we are only now beginning to understand
the regulation of neuronal calbindin. If, as our results indicate, brain calbindin
plays a critical role in neuronal calcium homeostasis, then future studies using
the chromosomal gene for calbindin will be important for understanding the
basic mechanisms involved in regulating neuronal calcium homeostasis.
V. ANALYSIS OF CALBINDIN PROMOTERS
To elucidate the molecular mechanism of calbindin gene regulation and

to identify cis-acting elements in the promoter region, the chromosomal genes
for human, 73 mouse, 74 and chick calbindin-D 28 / 5 - 77 and rat calbindin-D9 k 78
have been isolated. The chicken calbindin-D 28 k gene (18.5 kb) is split into
II coding exons by 10 intervening sequences. 76 The calbindin-D9 k gene (2.5
kb) contains three exons interrupted by two introns. In the promoter region,
a TATA box sequence is found 30 bp upstream from the Cap site of the rat
calbindin-D 9 k and the chick, mouse, and human calbindin-D28k genes. 73 -78 The
consensus for the CAT box is present upstream from the TATA box at various
lengths. The distal promoter (surrounding the TATA box) for each of these
genes is highly GC rich. 73 -78 Determination of the sequence homology among
human, mouse, and chick calbindin-D28 k promoters was done by dot matrix
analysis (Figure 3). When considering comparisons between chick-human and
chick-mouse promoters there was no strongly conserved sequence (although

avian and mammalian calbindin-D 28 k share 79% amino acid identity). How-
ever, comparison between human-mouse (which share 98% amino acid se-
quence identity) indicated that the distal promoter was strongly conserved
while the proximal part of the promoter diverged in the two species as shown
in Figure 3. The lack of conserved sequences between the avian and mam-
malian promoters may be due to the evolutionary difference between the
species as well as to the difference in tissue distribution (calbindin-D 28 k is
present in avian intestine but not in mammalian intestine).
Functional studies have been done by linking the mouse calbindin-Dm
promoter to a reporter gene coding for bacterial chloramphenicol acety !trans-
ferase (CAT). 79 A fragment of the mouse calbindin-D 28 k promoter with co-
ordinates - 1075/ + 34 was capable of promoting the expression of CAT in
rat osteosarcoma cells (Ros 17/2.8), insulinoma cells (13 cell line RINR; clone
1046-38, in which calbindin is present endogenously), and in LLC-PK 1 cells
(pig kidney cells). The highest levels of expression were in the Ros 17/2.8
cells which contained the highest concentration of I ,25(0H) 2 D3 receptors.
These cells were used for all further studies. Deletion mutant analysis of the
promoter indicated that the level of basal CAT expression increased fivefold
with deletion of 587 bp of upstream sequences. This suggests the presence
of an upstream repressor sequence. Upon treatment with sodium butyrate (2
mM) there occurred an eightfold increase in basal CAT activity. One segment
located between - 702 and - 488 was found to confer a 1,25(0H) 2 D3 de-
pendent increase in CAT activity in the presence of butyrate (incubation of
cells in the presence of 1 ,25(0H) 2 D 3 and butyrate [0.4 and 0.8 rnM] resulted
in a three- to fourfold stimulation in CAT activity above the stimulation
observed with butyrate alone). These results suggest the presence of a reg-
ulatory repressor which needs to be released before activation by 1,25(0H)zD 3
can be observed or that butyrate is activating some other factor which is
involved in the observed vitamin D mediated response. In addition, another
segment located between - 390 and - 175 consistently contributed a signif-
icant twofold induction of CAT activity in the presence of 1,25(0H)zD3 alone,
without butyrate. Sequence - 198/- 169 has homology to the rat osteocalcin
vitamin D response element (VDRE) (Figure 4). 8081 To further characterize
this downstream region gel-retardation assays were done using a 32 P labeled
probe (- 248/ -78) and nuclear extract from 1,25(0H)zD3 -treated Ros 17/
2. 8 cells. Incubation of the probe with the nuclear extract resulted in a specific
DNA-protein interaction as indicated by a band of retarded mobility. Addition
of cold oligonucleotide representing the rat osteocalcin VDRE specifically
competed with the probe, and this competition was demonstrated by abolishing
the band of retarded mobility. These initial findings represent the first analysis
of the mouse calbindin-D28k promoter. It should be noted that, at this time,
little is understood about sequences important in the regulation of the calbindin
gene. With the availability of the calbindin gene we can now initiate studies
RAT OSTEOCALCIN -458 GGTGA ATGA GGACA TTAC TGACC

MOUSE CALBINDIN-D.,m, -198 GGGGA TGTGA GGAGA 8bp TGAGC
FIGURE 4. Sequence homology between the VDRE of the rat osteocalcin gene and sequence
- 198/169 of the mouse calbindin-D 28 , promoter.
concerning the basic mechanisms, whereby calbindin is trancribed in only

specific cells and the multiple interactions involved in the regulation of the
calbindin gene. In addition, in order to obtain an understanding of the multiple
mechanisms of 1,25(0H)zD 3 action, the study not only of transcriptional but
also of posttranscriptional mechanisms is needed. These studies should pro-
vide new insight concerning 1,25(0H)zD 3-regulated gene expression and should
help clarify the molecular interaction of the 1,25(0H) 2 D3 receptor with other
factors in the control of gene expression.
REFERENCES
I. DeLuca, H. F., The vitamin D story: a collaborative effort of basic science and clinical
medicine, FASEB J., 2, 224, 1988.
2. Henry, H. L. and Norman, A. W., Vitamin D: metabolism and biological action, Annu.
Rev. Nutr., 4, 493, 1984.
3. Christakos, S., Gabrielides, C., and Rhoten, W. B., Vitamin D-dependent calcium
binding proteins: chemistry, distribution, functional considerations and molecular biology,
Endocrine Rev., 10, 3, 1989.
4. Feher, J, J,, Facilitated calcium diffusion by intestinal calcium binding protein, Am. J.
Physiol., 244, C303, 1983.
5. Bronner, F., Pansu, D., and Stein, W. D., An analysis of intestinal calcium-transport
across the rat intestine, Am. J. Physiol., 250, G561, 1986.
6. Christakos, S. and Norman, A. W., A vitamin D-dependent calcium binding protein
in bone, Science, 202, 70, 1978.
7. Balmain, N., Brehier, A., Cuisinier-Gieizes, P., and Mathieu, H., Evidence for the
presence of calbindin-D 28 , (CaBP28•) in the tibial growth cartilage of rats, Cell Tissue
Res., 245, 331, 1986.
8. Zhou, X. Y., Dempster, D. W., Marion, S. L., Pike, J, W., Haussler, M. R., and
Clemens, T. L., Bone vitamin D-dependent calcium binding protein is localized in
chondrocytes of growth plate cartilage, Calcif. Tissue Int., 38, 244, 1986.
9. Elms, T. N. and Taylor, A. N., Calbindin-D 28 k localization in rat molars during oden-
togenesis, J. Dent. Res., 66, 1431, 1987.
10. Berdal, A., Dupret, J, M., Pike, J, W., and Mathieu, H., Vitamin D controlled
expression of calbindin genes in teeth, Abstracts, 8th Workship on Vitamin D, 1991, 206.
II. Morrisey, R., Bucci, T., Empson, R., and Lutkin, E., Calcium binding protein: its
cellular localization in jejunum, kidney and pancreas, Proc. Soc. Exp. Bioi. Med., 149,
56, 1975.
12. Christakos, S., Friedlander, E. J., Frandsen, B. R., and Norman, A. W., Devel-
opment and application of a radioimmunoassay for chick intestinal calcium binding protein
and tissue distribution, Endocrinology, 104, 1495, 1979.
13. Sonnenberg, J., Pansini, A. R., and Christakos, S., Vitamin D-dependent rat renal
calcium binding protein: development of a radioimmunoassay tissue distribution and
immunological identification, Endocrinology, 115, 640, 1984.
14. Roth, J., Bonner-Weir, S., Norman, A. W., and Orci, L., Immunocytochemistry of
vitamin D-dependent calcium binding protein in chick pancreas: exclusive localization in
~cells, Endocrinology, Ill, 2216, 1982.
15. Rhoten, W. B. and Christakos, S., Vitamin D dependent calcium binding protein is
present in mammalian~ cells, Diabetes, 32, 130A, 1983.
16. Corradino, R. A., Wasserman, R. H., Pubolos, M. H., and Chang, S. 1., Vitamin
0 3 induction of a calcium binding protein in the uterus of the laying hen, Arch. Biochem.
Biophys., !25, 378, 1968.
17. Delorme, A. C., Dana, J. L., Acker, M.G., Ripoche, M.A., and Mathieu, H., In
rat uterus, 17~ estradiol stimulates a calcium binding protein similar to duodenal vitamin
D dependent calcium binding protein, Endocrinology, 113, 1340, 1983.
18. Bruns, M. E., Overpeck, J. G., Smith, G. C., Hirsch, G. N., Mills, S. E., and
Bruns, D. E., Vitamin D-dependent calcium binding protein in rat uterus: differential
effects of estrogen, tamoxifen, progesterone and pregnancy on accumulation and cellular
localization, Endocrinology, 122, 2371, 1988.
19. Delorme, A. C., Cassier, P., Geny, B., and Mathieu, H., Immunocytochemical lo-
calization of vitamin D-dependent calcium binding protein in the yolk sac of the rat,
Placenta, 4, 263, 1983.
20. Bruns, M. E., Kleeman, E., and Bruns, D. E., Vitamin D-dependent calcium binding
protein in mouse yolk sac: biochemical and immunochemical properties and response to
I ,25-dihydroxycholecalciferol, J. Bioi. Chem., 261, 7485, 1986.
21. Taylor, A. N., Chick brain calcium binding protein: response to cholecalciferol and some
developmental aspects, J. Nutr., 107, 480, 1977.
22. Jande, S. S., Maler, L., and Lawson, D. E. M., Immunohistochemical mapping of
vitamin D dependent calcium binding protein in brain, Nature, 294, 765, 1981.
23. Roth, J., Baeten, D., Norman, A. W., and Garda-Segura, L. M., Specific neurons
in chick central nervous system stain with an antibody against intestinal vitamin D-
dependent calcium binding protein, Brain Res., 222, 452, 1981.
24. Feldman, S. C. and Christakos, S., Vitamin D-dependent calcium binding protein in
rat brain: biochemical and immunocytochemical characterization, Endocrinology, 112,
290, 1983.
25. Celio, M., Calbindin-D 28k and parvalbumin in the rat nervous system, J. Neuroscience,
35, 375, 1990.
26. Parmentier, M., Ghysens, M., Rysen, F., Lawson, D. E. M., Pasteels, J. L., and
Pochet, R., Calbindin in vertebrate classes: immunohistochemical localization and West-
em blot analysis, Gen. Comp. Endocrinol., 65, 399, 1987.
27. Christakos, S., Malkowitz, L., Sori, A., Sperduto, A., and Feldman, S.C., Calcium
binding protein in squid brain: biochemical similarity to the 28,000 M, vitamin D de-
pendent calcium binding protein, J. Neurochem., 49, 1427, 1987.
28. Christakos, S. and Norman, A. W., Vitamin D dependent calcium binding protein and
its relation to 1,25-dihydroxyvitamin D receptor localization and concentration, in Cal-
cium Binding Proteins and Calcium Function, Siegel, F. et al., Eds., Elsevier, New
York, 1980, 371.
29. Thorens, B., Roth, j., Norman, A. W., Perrelet, A., and Orci, L., Immunocyto-
chemical localization of vitamin D-dependent calcium binding protein in chick duodenum,
J. Cell Bioi., 94, 201, 1982.
30. Taylor, A. N., Gleason, W. A., and Lankford, G. L., Vitamin D dependent calcium
binding. Immunocytochemical localization of the rat intestinal protein, J. Histochem.
Cytochem., 32, 153, 1984.
31. Christakos, S., Brunette, M. G., and Norman, A. W., Localization of immunoreactive
vitamin D-dependent calcium binding protein in chick nephron, Endocrinology, 109, 322,
1981.
32. Roth, J., Thorens, B., Hunziker, W., Norman, A. W., and Orci, L., Vitamin D-
dependent calcium binding protein: immunocytochemical localization in chick kidney.
Science, 214, 197, 1981.
33. Rhoten, W. B. and Christakos, S., Immunocytochemical localization of vitamin D-
dependent calcium binding protein in mammalian nephron, Endocrinology, 109, 981,
1981.
34. Roth, J., Brown, D., Norman, A. W., and Orci, L., Localization of the vitamin D-
dependent calcium binding protein in mammalian kidney, Am. J. Physiol., 243, F243,
1982.
35. Desplan, C., Thomasset, M., and Moukhatar, M. S., Synthesis, molecular cloning
and restriction analysis of DNA complementary to vitamin D dependent calcium binding
protein mRNA from rat intestine, J. Bioi. Chern., 258, 2762, 1983.
36. Hunziker, W., Siebert, P., King, M., Strucki, P., Dugaiczk, A., and Norman, A. W.,
Molecular cloning of a vitamin D-dependent calcium binding protein mRNA sequence
from chick intestine, Proc. Nat/. Acad. Sci. U.S.A., 80, 4228, 1983.
37. Hunziker, W., The 28kDa vitamin D dependent calcium binding protein is a six domain
structure, Proc. Nat/. Acad. Sci. U.S.A., 83, 7578, 1986.
38. Wood, T. L., Kobayashi, Y., Frantz, G., Varghese, S., Christakos, S., and Tobin,
A. J., Calcium binding protein (calbindin-D 28k): expression of calbindin-D28k RNAs in
rodent kidney and brain, DNA, 7, 585, 1988.
39. Hunziker, W. and Schrickel, S., Rat brain calbindin-D28 k: six domain structure and
extensive amino acid homology with chicken calbindin-D 28. , Mol. Endocrinol., 2, 465,
1988.
40. Perret, C., Desplan, C., Brehier, A., and Thomasset, M., Characterization of rat 9-
kDa cholecalcin (CaBP) messenger RNA using a complementary DNA, Eur. J. Biochem.,
148, 61, 1983.
41. Theofan, G. and Norman, A., Effects of a amanitin and cycloheximide on 1,25-
dihydroxyvitamin D, dependent calbindin-D 28k and its mRNA in vitamin D3 replete chick
intestine, J. Bioi. Chern., 261, 7311, 1986.
42. Theofan, G., Nguyen, A. P., and Norman, A. W., Regulation of calbindin-D 28k gene
expression by I ,25-dihydroxyvitamin D3 is correlated to receptor occupancy, J. Bioi.
Chern., 261, 16943, 1986.
43. Dupret, J., Brun, P., Perret, C., Lomri, N., Thomasset, M., and Cuisinier-Gieizes,
P., Transcriptional and posttranscriptional regulation of vitamin D-dependent calcium
binding protein in rat duodenum by I ,25-dihydroxycholecalciferol, J. Bioi. Chern., 262,
16553, 1987.
44. Varghese, S., Lee, S., Huang, Y.-C., and Christakos, S., Analysis of rat vitamin D-
dependent calbindin-D28k gene expression, J. Bioi. Chern., 263, 9776, 1988.
45. Huang, Y.-C. and Christakos, S., Modulation of rat calbindin-D,.. gene expression by
1,25-dihydroxyvitamin D3 and dietary alteration, Mol. Endocrinol., 2, 928, 1988.
46. Varghese, S., Deaven, L. L., Huang, Y. C., Gill, R., Iacopino, A. M., and Chris-
takos, S., Transcriptional regulation and chromosomal assignment of the mammalian
calbindin-D28k gene, Mol. Endocrinol., 3, 495, 1989.
47. Huang, Y.-C., Lee, S., Stolz, R., Gabrielides, C., Pansini-Porta, A., Bruns, M. E.,
Bruns, D., Mifflin, T., Pike, J, W., and Christakos, S., Effect of hormones and
development on the expression on the rat I ,25-dihydroxyvitamin D, receptor gene and
comparison to calbindin gene expression, J. Bioi. Chem., 264, 17454, 1989.
48. Rhoten, W. B. and Christakos, S., Cellular gene expression for calbindin-D 28 k in mouse
kidney, Anat. Rec., 227, 145, 1990.
49. Iacopino, A. M. and Christakos, S., Reduction of neuronal calcium binding protein
(calbindin-D2,.) gene expression in aging and neurodegenerative diseases, Proc. Nat/.
Acad Sci. U.S.A., 87, 4078, 1990.
50. Iacopino, A. M. and Christakos, S., Corticosterone regulates calbindin-D 28 • mRNA
and protein levels in rat hippocampus, J. Bioi. Chern., 265, 10177, 1990.
51. Iacopino, A. M. and Christakos, S., Calcium binding protein (calbindin-D2,.) gene
expression in the developing and aging mouse cerebellum, Mol. Brain Res., 8, 283,
1990.
52. Li, H. and Christakos, S., Differential regulation by 1,25 dihydroxyvitamin D3 of
calbindin-D9 • and calbindin-D2,. gene expression in mouse kidney, Endocrinology, 128,
2844, 1991.
53. Enomoto, H., Hendy, G. N., and Clemens, T. L., Regulation of calbindin-D,.. (CaBP)
mRNA in chick kidney cells: critical importance of posttranscriptional mechanisms and
calcium ion concentration, J. Bone Min. Res., 6 (Suppl. 1), abstract 390, Sl83, 1991.
54. Meyer, Galligan, M., Jones, G., Komm, B. S., and Haussler, M. R., 1 ,25(0H)zD 3
dependent regulation of calbindin-D 2,. mRNA apparently requires the synthesis of an
uncharacterized protein in chick duodenal organ culture, J Bone Min. Res., 6 (Suppl. 1),
abstract 406, Sl85, 1991.
55. Kimberg, D. V., Baerg, R. D., Gershone, E., and Graudusius, R. T., Effect of
cortisone treatment on the active transport of calcium by the small intestine, J. Clin.
Invest., 50, 1309, 1971.
56. Hall, A. K., Bishop, J. E., and Norman, A. W., Inhibitory and stimulatory effects of
dexamethasone and 1,25dihydroxyvitamin D3 on chick intestinal calhindin-D28• and its
mRNA, Mol. Cell. Endocrinol., 51, 25, 1987.
57. Corradino, R. A. and Fullmer, C. S., Positive cotranscriptional regulation of intestinal
calbindin-D 2,. gene expression by 1,25dihydroxyvitamin D3 and glucocorticoids, Endo-
crinology, 128, 944, 1991.
58. Navikis, R. J., Katzenellenbogen, B. S., and Nalbandov, A. V., Effects of the sex
steroids and vitamin D 3 on calcium binding protein in the chick shell gland, Bioi. Reprod.,
21, 1153, 1979.
59. Opperman, L. A., Sandes, T. j., Mills, S. E., Christakos, S., Bruns, D. E., and
Bruns, M. E., The control of calbindin-D9k and calbindin-D28• in mouse uterus by
1,25(0H),D3 and estrogen, J. Bone Min. Res., 5 (Suppl. 2), S264, 1990.
60. Hall, A. K., Wang, Y.-Z., Gill, R. K., and Christakos, S., Modulation of calbindin-
D2,.-like gene expression by retinoic acid in neuronal cells overexpressing a retinoic acid
receptor, J. Bone Min. Res., 6 (Suppl. 1), Sl84, 1991.
61. Kim, H. T. and Chen, T. L., 1,25Dihydroxyvitamin D3 interaction with dexamethasone
and retinoic acid: effects on procollagen messenger ribonucleic acid levels in rat osteoblast
like cells, Mol. Endocrinol., 3, 97, 1989.
62. Schute, R., Umesono, K., Mangelsdorf, D. J., Bolado, J,, Pike, J. W., and Evans,
R. M., Jun-fos and receptors for vitamin A and D recognize a common response element
in the human osteocalcin gene, Cell, 61, 497, 1990.
63. Umesono, K. and Evans, R. M., Determinants of target gene specificity for steroid/
thyroid hormone receptor, Cell, 57, 1139, 1990.
64. Mader, S., Kuman, U., de Verneuil, H., and Chambon, P., Three amino acids of
the esterogen receptor are essential to its ability to distinguish an estrogen from gluco-
corticoid response element, Nature, 338, 271, 1989.
65. Danielsen, M., Hinck, L., and Ringold, A.M., Two amino acids within the knuckle
of the first zinc finger specify DNA response element activation by glucocorticoid re-
ceptor, Cell, 57, 1131, 1989.
66. Umesono, K., Murakami, K. K., Thompson, C., and Evans, R. M., Direct repeats
as selective response elements for the thyroid hormone, retinoic acid and vitamin D,
receptors, Cell, 65, 1255, 1991.
67. Stumpf, W. E. and O'Brien, L. P., 1,25(0H), Vitamin D, sites of action in the brain,
Histochemistry, 87, 393, 1987.
68. Stumpf, W. E. and Bidman, H.-J., Sites and mechanisms of action of vitamin Din
the nervous system, Abstracts, Eighth Workshop on Vitamin D, 23, 1991, 23.
69. Mattson, M. P., Rychlik, B., Chu, C., and Christakos, S., Evidence for calcium
reducing and excitoprotective roles for calcium binding protein, calbindin-D"' in cultured
hippocampal neurons, Neuron, 6, 41, 1991.
70. Christakos, S. and Norman, A. W., Evidence for a specific high affinity binding protein
for 1,25 dihydroxyvitamin D3 in chick kidney and pancreas, Biochem. Biophys. Res.
Commun., 89, 56, 1979.
71. Christakos, S. and Norman, A. W., Biochemical characterization of 1,25dihydroxyvitamin
D, receptors in chick pancreas and kidney cytosol, Endocrinology, 108, 140, 1981.
72. Clark, S. A., Stumpf, W. E., Sar, M., DeLuca, H. F., and Tanaka, Y., Target cells
for 1,25dihydroxyvitamin D3 in rat pancreas, Cell Tissue Res., 109, 515, 1980.
73. Parmentier, M., DeViglder, J. J. M., Muri, E., Szisirer, C., Islam, M. Q., Guerts
Van Kessel, A., Lawson, D. E. M., and Vassart, G., The human calbindin 27 kDa
gene: structural organization of the 5' and 3' region, chromosomal assignment and re-
striction fragment length polymorphism, Genomics, 4, 309, 1989.
74. Gill, R. K. and Christakos, S., Isolation of genomic clones harboring sequences ho-
mologous to mammalian calbindin-DZKk and their characterization, J. Bone Mineral Res.,
4, 1989, S291.
75. Minghetti, P. P., Cancela, L., Fujisawa, Y., Theofan, G., and Norman, A. W.,
Molecular structure of the chicken vitamin-D induced calbindin-D28k gene reveals eleven
exons. six Cal+ binding domains and numerous promoter regulatory elements, Mol.
Endocrinol .. 2, 355. I988.
76. Wilson, P. W., Rogers, J., Harding, M., Pohl, V., Pattyn, G., and Lawson, D. E. M.,
Structure of chick chromosomal gene for calbindin and calretinin, J. Mol. Biol.. 200,
615, 1988.
77. Ferrari, S., Drusian, E., Baltini, R., and Fregni, M., Nucleotide sequence of the
promoter region of the gene encoding chicken calbindin-DZKk• Nucleic Acid Res., 16,
353, 1988.
78. Perret, C., Lomri, N., Gouhier, N., Auffrey, C., and Thomasset, M., The rat vitamin-
D dependent calcium binding protein (9-KDa CaBP) gene: complete nucleotide sequence
and structural organization, Eur. J. Biochem., 172, 43, 1988.
79. Gill, R. and Christakos, S., Identification of sequence elements in the mouse calbindin-
D28k gene which confer basal activation and hormonal inducible response, Abstracts,
Eighth Workshop on Vitamin D, 1991, 81.
80. Markose, E. R., Stein, J. L., Stein, G. S., and Lian, J, B., Vitamin D-mediated
modifications in protein-DNA interaction at two promoter elements of the osteocalcin
gene, Proc. Nat/. Acad. Sci. U.S.A.. 87, 1701, 1990.
81. Demay, M. B., Gerardi, J. M., DeLuca, H. F., and Kronenberg, H. M., DNA
sequence in the rat osteocalcin gene that bind the I ,25 dihydroxyvitamin D3 receptor and
confer responsiveness to I ,25-dihydroxyvitamin D,, Proc. Nat!. Acad. Sci. U.S.A., 87,
369, 1990.
Chapter 18
VITAMIN D REGULATION OF OSTEOBLAST

GROWTH AND DIFFERENTIATION
Jane B. Lian and Gary S. Stein
TABLE OF CONTENTS
I. Introduction ...................................................... 392
II. Vitamin D and Bone Physiology ................................. 393
III. Vitamin D Modulation of Osteoblast Cell Growth

and Differentiation ............................................... 394
A. Role of I ,25(0H) 2 D3 in Commitment to Cell
Differentiation ............................................ 395
B. Role of 1,25(0H) 2 D3 on Osteoblast Activity and
Bone Matrix Formation ................................... 395
I. The Normal Sequelae of Osteoblast
Differentiation .................................... 395
2. Vitamin D Modification of Osteoblast
Growth and Differentiation ....................... 399
C. Influence of 1,25(0H) 2 D3 on Osteoclast
Differentiation and Bone Resorption ..................... 404
IV. Vitamin D Regulation of the Osteocalcin Gene .................. 406

A. Sequence Organization of the Osteocalcin Gene
Reflects Vitamin D-Mediated Control .................... 407
B. Regulation of Vitamin D-Responsive Element
(VDRE) Activity ......................................... 409
C. Phenotype Suppression: A Postulated Mechanism
for the Coordinate Regulation of Vitamin
D-Mediated Enhancement and Basal
Transcription of the Osteocalcin Gene by Fos-
Jun Interactions at AP-I Sites within the VDRE
and Other Promoter Elements ............................ 412
V. Potential Therapeutic Applications of Vitamin D ................ 4I5
VI. Summary ........................................................ 419
References ............................................................... 420
l).g493-f>9f>J-4i9V$0 !XJ+ $.50

,~, llJ~.l hy CRC Pre~~. Inc 391
I. INTRODUCTION
Vitamin D is a steroid (pro)-hormone that undergoes several transfor-

mations to the active hormone 1,25-dihydroxyvitamin D3 ( 1,25(0H) 2 D3 ). The
active 1,25-dihydroxy D 3 is produced in the kidney and has three major target
tissues (intestine, kidney, and bone) that contribute to the regulation of calcium
homeostasis. Figure 1 schematically illustrates the vitamin D endocrine sys-
tem. At the physiologic level, in response to a fall in circulating calcium
levels, parathyroid hormone induces synthesis of the renal hydroxylase en-
zyme (25-hydroxycholecalciferol !-hydroxylase). Then 1,25(0H) 2 D3 pro-
motes the absorption of calcium and phosphorus across the intestinal epithelial
cells and mobilizes mineral from bone. Normal serum calcium levels are
maintained, since high serum calcium indirectly and 1,25(0H) 2 D 3 directly
suppresses parathyroid hormone secretion and function. 1 Calcitonin, also se-
creted in response to hypercalcemia, inhibits osteoclastic activity. At the
cellular level, the interaction of 1 ,25(0H) 2 D 3 and its receptor is analogous
to the mode of action of other steroid hormones and involves changes in gene
expression in the target cells; 2 however, rapid "non-genomic" actions of
vitamin D, possibly potentiating the regulation of calcium metabolism, are
also being observed. 3 - 5
For several decades, vitamin D has been recognized as functionally im-
portant to regulatory mechanisms operative in skeletal development (cartilage
and bone formation) and the lifetime remodeling of bone involving resorption
and bone formation. The active metabolite 1,25(0H) 2 D3 mediates growth and
differentiation of chondrocytes, 6 - 8 osteoblasts, 9 • 10 and progenitor cells that
develop the osteoclast phenotype, both directly and/or as a consequence of
the effects of the hormone on a broad spectrum of integrative cellular signaling
mechanisms (reviewed in References 11, 12). Not to be dismissed are the
well documented physiological influences of vitamin D on calcium metab-
olism that result in modifications of the development and activities of skeletal
cells (reviewed in References 10, 13). In addition to these central roles of
vitamin Din bone cell development and maintenance of functional properties,
it is well established that vitamin D-related aberrations in metabolism of the
hormone and/or receptor binding and the association of hormone-receptor
complexes with vitamin D-responsive elements in gene regulatory sequences
are the principal components of skeletal disorders. 14 Since vitamin D levels
are a determinant of the physiological regulation and pathological status of
bone, the nutritional contribution to skeletal development and maintenance
must be recognized. Here, the dietary levels of vitamin D and factors that
affect the metabolism or utilization of the hormone are key considerations.
This chapter focuses on the combined application of molecular, biochem-
ical, and morphologic approaches that have contributed to defining the mul-
tiple effects of vitamin D on bone cell growth and differentiation. The os-
teocalcin gene is representative of a series of genes that are vitamin D-
Lian and Stein 393
SKC, uv
LIVER
250HD~ URINE IP, lea
KIDNEY/ ,---_.INTESTINE --~

tea ABSORPTION
1 ,25(0H)2D3 tea BINDING PROTEIN
I SERUM CALCIUM
I PTH
t
SECRETION L -_ _ . BONE ----....J
t
HYPOCALCEMIA
I RESORPTION
\e
CALCITONIN
FIGURE I. Interrelationships of the vitamin D endocrine system in response to low serum

calcium. The final step in synthesis of the active hormone metabolite, 1,25(0H)2 D3 , occurs in
the kidney and can be stimulated by parathyroid hormone (PfH). Organs responsive to 1,25(0H)zD3
to increase serum calcium are kidney, intestine, and bone. Calcitonin will inhibit excessive
osteoclastic resorption in bone.
responsive, and stringently regulated during progressive development of the

osteoblast phenotype. We will, therefore, address the influences of the hor-
mone at the transcriptional and at several posttranscriptional levels. At each
of these levels, osteocalcin expression can be modulated. Osteocalcin expres-
sion depends on the developmental stage of the osteoblast, both independently
and synergistically or antagonistically, with other physiological effectors of
this expression. It is also important to complement assessments of transcrip-
tional activity or mRNA levels with in situ hybridization and immunohisto-
chemistry where gene expression can be established at the single cell level.
We will illustrate how such experimental approaches have been effectively
integrated to address the structural and functional properties of vitamin D
regulation of osteocalcin in bone.
II. VITAMIN D AND BONE PHYSIOLOGY
Many advances in our understanding of vitamin D regulation of skeletal

tissues are attributable to availability of in vivo and in vitro model systems
that support bone growth and differentiation. The foundation for our present
knowledge of bone cell growth and differentiation under conditions where
cell and tissue relationships are maintained, was provided by in vivo studies
of growth and formation of bone tissue principally utilizing histology and
histochemistry to examine skeletal development. Similarly, the effects of
vitamin D depletion and repletion on bone growth were studied in vivo, where
changes in serum parameters were correlated with perturbations in bone mor-
phology, histochemistry and composition during avian and rodent skeletal
development. 15 - 20
Both inadequate or excessive vitamin D intakes have dramatic effects

on bone morphology and metabolism. Vitamin D deficiency increases the
amount of nonmineralized bone matrix (osteoid tissue), resulting in rickets
in children and osteomalacia in adults. Administration of 1,25(0H) 2 D3 to
vitamin D-deficient animals stimulates osteoclast bone resorption, then cures
rickets and osteomalacia. 21 Experimental studies in rodents have demonstrated
that bone mineralization is not dependent upon 1 ,25(0H) 2D 3 and can be
maintained if an adequate supplementation of calcium and phosphate in the
diet is used.Z2 - 25 However, other studies have demonstrated that vitamin D
can stimulate bone mineralization independently of changes in serum mineral
concentrations. 21 Yet, pharmacologic doses of 1,25(0H) 2 D3 administered in
vitro in primary osteoblast cultures 9 and in vivo can inhibit mineralization of
the newly formed bone matrix. 19 •26 In vivo morphometric studies have dem-
onstrated that high doses of 1,25(0H) 2 D3 decrease the amount and rate of
formation of new bone matrix in mice. 17 Thus, the effects of this metabolite
on skeletal tissue are indeed extremely complex, encompassing multiple pa-
rameters related to several osteoblast gene products that are involved in tissue
development and mineralization. 9 • 13
Vitamin D's primary action in bone is to stimulate bone resorption as a
calcitropic regulating hormone, and this has been well documented in vitro 21
and in vivo.28 In vivo repletion of hormone to vitamin D-deficient rats results
in an initial osteoclastic bone resorption. The target cell for this primary
function of the hormone is the osteoblast, which contains receptors for dihy-
droxyvitamin D3 • 29 •30 It has been proposed 31 that the osteoblast secretes factors
that stimulate osteoclastic activity in response to hormone. 32 Vitamin D also
contributes to bone resorption by promoting differentiation of the mononuclear
progenitor cell to form the multinucleated osteoclast. 33 •34 Thus, the hormone
appears to have both anabolic effects and antianabolic effects on osteoblasts
to promote bone resorption. Indeed, this dual function likely relates to the
balance necessary to maintain the coupling of bone resorption and bone for-
mation. Only recently have the responses of isolated osteoblasts and cocultures
of these cells with osteoclasts to vitamin D been characterized; these studies
show that the multiple actions of this hormone are highly complex.
III. VITAMIN D MODULATION OF OSTEOBLAST

CELL GROWTH AND DIFFERENTIATION
From the earliest in vivo studies, it appeared that vitamin D receptors

were confined to the osteoprogenitor cells that differentiate to osteoblasts and
chondrocytes. 35 Several in vitro cellular systems are currently being utilized
to examine the cascade of events in gene expression that supports the inter-
dependent sequence of biochemical and structural aspects of growth and
differentiation that are mediated by I ,25(0H) 2D 3 during the progressive de-
velopment of the osteoblast.
Lian and Stein 395
A. ROLE OF 1,25(0H) 2D3 IN COMMITMENT

TO CELL DIFFERENTIATION
Bone marrow cultures offer the possibility of examining events associated
with stem cell differentiation to osteoblasts. 36 ·37 Clonal lines from single cells
derived, for example, from fetal rat calvaria also provide an opportunity for
examining selective expression of genes during commitment to the bone cell
lineage. 38 •39 These are promising systems, offering the potential to study
growth and differentiation of osteoprogenitor cells. To date, our knowledge
of this aspect of osteoblast differentiation is incomplete, largely due to the
limited characterization of such model systems. The marrow tissue comprises
the stroma which can be differentiated to osteoblast-like cells largely under
the influence of glucocorticoids (e.g., dexamethasone). Glucocorticoids can
regulate and increase the availability of vitamin D receptors in some osteoblast
cell lines. 40 Marrow also contains cells of the hematopoietic lineage which
differentiate into osteoclasts, and this occurs largely under the influence of
1,25(0H) 2 D3 • In early studies of marrow cell cultures treated with vitamin
D, both increases in multinucleated cells (reflecting osteoclasts) and the pres-
ence of alkaline phosphatase positive cells (reflecting osteoblasts) were ob-
served. 10
B. ROLE OF 1,25(0H) 2D3 ON OSTEOBLAST ACTIVITY

AND BONE MATRIX FORMATION
1. The Normal Sequelae of Osteoblast Differentiation
Osteoblastic cells released from bone as outgrowth cultures or by a se-
quential series of combined trypsin and collagenase digestions41 .4 2 are a viable
cellular model for studying osteoblast growth and differentiation. The bone-
derived cells are enriched in cells committed to the osteoblast lineage, ex-
hibiting properties which include PTH and vitamin D responsiveness and
synthesis of bone-related proteins. Although this precludes investigation of
the process by which commitment to the osteoblast phenotype is regulated,
the primary diploid osteoblast cultures offer an effective system to examine
osteoblastic properties from a proliferating cell to a cell which becomes en-
veloped in a mineralized extracellular bone-like matrix. Normal diploid os-
teoblasts undergo an ordered development over several weeks in culture,
whether isolated from human bone43 •44 or fetal calvaria from 21-d rat bone, 45 •46
mouse, 47 16-d embryonic chick48 or bovine bone-derived cells. 49 Proliferating
cells differentiate to osteoblasts forming nodules of multilayered cells with a
mineralized, type I collagen extracellular matrix (Figure 2). 46 •50 •51 A bone-
tissue-like organization of these nodules formed in culture is supported by a
comparison of the ultrastructure of the mineralized regions of the culture to
sections through an intact 21-d fetal rat calvarium; both exhibit ordered de-
position of crystals within and between the orthogonally organized bundles
of collagen fibrils (Figure 2). No evidence for cell necrosis or intracellular
calcification is indicated in the cultures, particularly where osteoblasts are
embedded in mineralized matrix. 46 •51 - 53
FIGURE 2. Morphology of normal rat diploid osteoblast cultures. (A through E) Light mi-
croscopy of cultured osteoblasts isolated from 21-d fetal calvaria by sequential collagenase
digestion 5° (A and B) Cells labeled with 'H thymidine show greater than 95% of the cells are
proliferating after plating (A, day 5). The cells multilayer, fom1ing nodules (B, day 12), and
the onset of expression of the osteoblast phenotype is indicated by alkaline phosphatase staining
of cells where proliferation has ceased in the center of the nodule (C. day 18 alkaline phosphatase
histochemistry). Every cell has ceased to proliferate and greater than 90% of the cells are alkaline
phosphatase positive (0, day 28). It is within the multilayered nodule that mineralization is
initiated; von Kossa silver staining shows mineralized nodules throughout the dish. A through
E, photographed at 100 x. (F) Scanning electron micrograph of a nodule on day 28 at 20 x,
with mineralized matri!( enveloping osteoblasts. (G and H): Transmission electron microscopy
(bars are 5 and 2.5 microns). (G) shows orthogonal organization of the collagen matrix in thick
layers between the osteoblasts in vitro. (H) Higher magnification of the collagen bundles shows
mineral deposition within the collagen fibrils and absence of intracellular calcification.
In these cultures of mammalian diploid osteoblasts, cell growth and tissue-

specific gene expression have been mapped during the progressive devel-
opment of the bone cell phenotype from a proliferating cell to a mature
osteocytic-like cell within the context of a bone tissue-like organization (Figure
3). 50. 53 Temporally, genes encoding osteoblast phenotype markers are ex-
pressed during three distinct periods, including a growth period (proliferation),
a period of matrix development, and a mineralization period. During the first
10 to 12 d, the period of active proliferation is reflected by mitotic activity
Lian and Stein 397
MATRIX
PROLIFERATION MATURATION MINERALIZATION
z cRl3
0 H4 AP CP ([; CA+2
iii 100
en
w
a:
a. 80
X
w
..J
<(
60
:!!!
x
<( 40
:!!!
0~ 20
0
0 7 14 21 28 35
DAYS IN CULTURE
(AP-1 SITES)
e
PHENOTYPE SUPPRESSION-,~--------,
ALK 0:: I
PHOS (£)
(£)
PROLIFERATION - ECM ECM

'\ MATURATION - MINERAUZATION
!lEN 8 Dow.-.,..u-. j ~ Do~-reguL

FIBRONECTIN
BIOSYNTHESIS
cell growth ECM Maturation
FIGURE 3. Temporal expression of cell growth and osteoblast phenotype related genes during
the development of in vitro formed bone-like tissue by normal diploid rat osteoblasts. Top panel:
isolated primary cells were initially cultured in MEM with 10% FCS, then after confluence, in
BGJb medium supplemented with 10% FCS, 50 fLglml ascorbic acid and 10 mM 13-glycerol
phosphate. Cellular RNA was isolated at the times indicated (3, 5, 7, 10, 12, 16, 21, 28, and
35 d) during the differentiation time course and assayed for the steady-state levels of various
transcripts by Northern blot analysis. The resulting blots were quantitated by scanning densi-
tometry and the results plotted relative to the maximal expression of each transcript. Three
periods of gene expression are represented: (A) Genes characterizing the proliferation period
include H4 histone (which reflects DNA synthesis), c-myc and c-fos, plus genes expressed for
formation of the type I collagen, fibronectin (FN), and transforming growth factor-13 (TGFI3);
(B) alkaline phosphatase (AP) is expressed postproliferatively and associated with extracellular
matrix maturation; (C) genes induced to high levels with onset of extracellular matrix mineral-
ization; represented are osteopontin (OP) and osteocalcin (OC). Calcium (Ca2 +) accumulation
is indicated. This temporal pattern defines three periods, proliferation, matrix maturation, and
mineralization reflecting a developmental sequence of osteoblast maturation. Lower panel: the
interrelationship and feedback signals between the periods are indicated. AP-I activity in pro-
liferating osteoblast, that is, fos-jun protein binding to AP-I sites, suppresses activity of post-
proliferative expressed genes.
with expression of cell cycle (e.g., histone) and cell growth (e.g., c-myc, c-
fos and AP-1 activity) regulated genes. It is fundamental to development of
the bone cell phenotype that several genes associated with formation of the
extracellular matrix (type I collagen, fibronectin, and TGF~) are actively
expressed in this proliferation period. Immediately following the downreg-
ulation of proliferation, as reflected by the decline in DNA synthesis eH-
thymidine incorporation) and histone gene expression, the expression of al-
kaline phosphatase (enzyme activity and mRNA) increases greater than ten-
fold. At this time, a differentiation-specific histone gene is expressed. 54 •55
During this period (from 12 to 18 d) the extracellular matrix undergoes a
series of modifications in composition and organization that renders it com-
petent for mineralization. During this matrix maturation phase, every cell has
become alkaline phosphatase positive as the cultures progress into the mi-
neralization stage. In heavily mineralized cultures, cellular levels of alkaline
phosphatase mRNA decline.
With the onset of extracellular matrix mineralization, other bone-specific
genes are induced. For example, osteopontin or SPPI, a 60-kDa calcium
binding protein enriched in 0-phosphoserine, 56 and osteocalcin, a -y-carbox-
yglutamate-containing protein, 57•56 are upregulated in concert with the accu-
mulation of mineral. Notably, osteopontin is expressed during the period of
active proliferation (at 25% of maximal levels), consistent with its high level
of expression following oncogene transformation or phorbal ester treatments
of cells. 59 Osteopontin mRNA is decreased postproliferatively and then ex-
hibits an induction at the onset of mineralization achieving peak levels of
expression during the mineralization period (days 16 to 20). In contrast to
osteopontin, the vitamin K-dependent protein, 60 osteocalcin, is expressed only
postproliferatively with the onset of nodule formation. 51 This 5. 7-kDa calcium
binding protein is characterized by 3 -y-carboxyglutamic acid residues (Gla),
binds tightly to hydroxyapatite, and is maximally expressed during active
mineralization of the extracellular matrix (ECM) in vivo51 and in vitro. 51 A
related vitamin K-dependent protein in bone, matrix Gla protein (MGP), 61 is
characterized by the presence of 5 Gla residues in a 10-kDa molecule. MGP
is present in proliferating cells and is increased in relation to accumulation
of collagen, but levels of expression are not correlated with the mineralization
phase. 62 Several manipulations of the culture system51 demonstrate that in-
duction of high levels of osteocalcin and osteopontin mRNA are dependent
upon formation of a mineralized extracellular matrix whereas MGP levels are
sustained independent of mineralization. 62 Both osteocalcin and osteopontin
are expressed in mature osteoblasts within the nodules 63 and the proteins are
localized to the mineralized bone extracellular matrix. 64 Osteocalcin has been
shown to contribute to regulation of the mineral phase in bone both in vitro
as a potential inhibitor of mineral nucleation 65 and in vivo as a bone matrix
signal that promotes osteoclast differentiation and activation. 16 •66 •67 Osteo-
pontin has been localized to the region of osteoclast contact. 66 More recently,
Lian and Stein 399
others have used synthetic peptides containing osteopontin's Arg-Gly-Asp

sequence region, to show stimulation of osteoclast activity. 69 While the in-
duction of osteopontin and osteocalcin during the period of active mineral
accumulation may support the onset and progression of extracellular matrix
mineralization, these proteins may also function in the regulation of turnover
of the mineral phase of bone in vivo. This potential function would be con-
sistent with vitamin D regulation of these proteins during bone resorption.
The observed temporal expression of osteoblast phenotype markers during
the developmental sequence is a reflection of functional activities necessary
for the progressive formation of bone tissue. The expression of alkaline phos-
phatase mRNA and enzyme activity prior to the initiation of osteoblast mi-
neralization suggests that alkaline phosphatase may be involved in the prep-
aration of the extracellular matrix for the ordered deposition of mineral. In
the normal temporal expression of osteoblast genes, alkaline phosphatase peak
levels precede maximal deposition of mineral and in heavily mineralized
cultures, alkaline phosphatase mRNA levels and enzyme activity are down-
regulated. This is consistent with in vivo observations of vitamin D deficiency
where serum and tissue alkaline phosphatase levels are high in the absence
of mineralized matrix in rachitic bone. In this situation, with the absence of
matrix mineralization, alkaline phosphatase is apparently not downregulated.
The coinduction of genes, such as osteocalcin and osteopontin with active
accumulation of mineral, may support the onset and progression of extracel-
lular matrix mineralization. Alternatively, the induction of these minerali-
zation-associated genes may reflect an acquisition of osteoblast properties
necessary for signaling bone turnover in vivo as discussed above. The sequence
of expression of genes that encode osteoblast phenotype markers in culture
follows a pattern of gene expression and tissue distribution that occurs in
neonatal bones determined by in situ hybridization 70 .7 1 and during fetal cal-
varial development in vivo,12 supporting the biological relevance of the os-
teoblast culture system as a model for intramembranous bone differentiation.
Taken together, the patterns of expression of these genes and the synthesis
of the encoded proteins demonstrate that a temporal sequence of gene expres-
sion during the culture period is associated with development of the extra-
cellular matrix and reflects maturation of the osteoblast phenotype in vitro.
2. Vitamin D Modification of Osteoblast Growth and Differentiation

Vitamin D affects numerous osteoblast products, 11 including cell growth
related genes/ 3 - 75 nuclear and cytoskeletal proteins, 76 ·77 structural proteins
involved in bone matrix formation, and mineralization, 9 enzymes, 78 and plasma
membrane proteins 79 such as alkaline phosphatase, 80 ·81 Many responsive pro-
teins have not yet been characterized. 82 Vitamin D gene regulation has been
studied for only a few proteins, but these represent some of the major con-
stituents of the bone matrix. These include, for example, collagen, 83 osteo-
pontin,84 osteocalcin, 85 -87 and matrix Gla protein. 61 ·62 The development of
MATRIX
MINERALIZATION
PROLIFERATION MATURATION
120 400
100 H4 AP OP
300
n.
:::!!:
:::> 75 80
:::!!: 200
x
c(
50
II I•
40 100
:::!!:
:.!!
0
25
0 0 I
0
7 14 21 28 35 14 21 7 14 21 28 35
7 28 35
400
100 CXll
900 ~
300
oc
:::!!:
:::> 75
:::!!: 600
200
xc( 50
:::!!: 300 100
25
:.!!
0
0 0
I• I• II I• 0
7 14 21 28 35 7 14 21 28 35
7 14 21 28 35
DAYS DAYS DAYS
FIGURE 4. Examples of vitamin D-regulated genes during the three periods of osteoblast
phenotype development. Solid bars are controls and open bars are cells treated 48 h prior to
harvest with oo-•
M) I ,25(0H) 2D 3 . Samples during the time course were hybridized together
for quantitation of the control and hormone-treated samples as percent maximal expression for
each gene. Note the pleiotropic effects on proliferation related genes. Histone (H4) and type I
collagen (COLL) are inhibited early (day 8 through 16). then stimulated in mature osteoblasts
in a mineralized matrix (day 28 through 35); alkaline phosphatase (AP) is also stimulated.
normal diploid osteoblast cultures that produce a bone-like mineralized ex-

tracellular matrix has allowed for studies that can evaluate the effects of the
hormone at several stages of osteoblast differentiation from the proliferating
cell to the mature osteocyte-like cell enveloped in a mineralized matrix.
Vitamin D influences the extent to which osteoblast genes are expressed, in
most cases modulating the extent of expression, rather than inducing or sup-
pressing expression (Figure 4). For example, osteocalcin is not expressed
during proliferation and is not induced by vitamin D. Other genes that are
expressed in proliferating cells, such as OP and MGP, are vitamin D re-
sponsive, indicating that the vitamin D receptor is available, along with a
subset of accessory factors that is required for hormone binding and activity
at vitamin D responsive promoter elements.
Biphasic effects of vitamin D on gene expression are observed (Figure
4) that reflect differences in regulation dependent upon the stage of osteoblast
maturation. 9 •88 •89 For example, in normal diploid osteoblast cultures, vitamin
D inhibits proliferation when the cells are actively dividing, but stimulates
proliferation in quiescent osteocyte-like cells in a mineralized matrix. 9 The
antiproliferative or proliferative effects may be a consequence of vitamin D
regulation of growth factors, such as skeletal growth factor/ 3 •74 epidermal
growth factor, and TGFI3. 75 Genes, such as TGFI3 and collagen, are expressed
Lian and Stein 401
at high levels during proliferation and are downregulated by vitamin D in

proliferating rat osteoblasts, 83 ·90 but are upregulated in mature osteoblasts. 9 A 3
TGFI3 levels are also stimulated when expressed at low levels and inhibited
when expressed at high levels (unpublished observations, this laboratory).
Alkaline phosphatase also exhibits biphasic responsiveness to vitamin D con-
sistent with both positive or negative regulation observed in other cell culture
systems. 43 •80 •91 . 93 Such biphasic responsiveness can be a consequence of the
maturation state and matrix environment of the osteoblast. The inhibition of
collagen production and alkaline phosphatase enzyme activity by vitamin D
in vitro may relate to the pharmacologic effect in animals where an inhibition
of mineralization and a decrease in bone formation are observed.' 9 · 21 •26 Vi-
tamin D both positively and negatively regulates expression of genes encoding
osteoblast phenotype markers as a function of the differentiated stage of the
cell and basal levels of expression. The hormone can thus serve as a physi-
ological mediator of bone formation and remodeling, which are functionally
related to bone matrix competency and the differentiated state of the osteo-
blasts.
A reciprocal and functionally coupled relationship between the expression
of growth-related and differentiation-related genes exists, as schematically
illustrated in Figure 5 for development of the osteoblast phenotype. Several
lines of evidence support this growth-differentiation relationship; e.g., growth
supports collagen biosynthesis, collagen matrix formation contributes to ces-
sation of proliferation, and this is required for increased expression of alkaline
phosphatase and osteocalcin. 51 •53 The acute effect of 1 ,25(0H) 2 D 3 , if added
to the culture for periods up to 48 h, is to shift the profile of expressed genes
to one that represents a more differentiated cell, while maintaining the growth-
differentiation relationship. This may be a consequence of its antiproliferative
effect since the downregulation of proliferation promotes expression of phen-
otypic genes. 53 However, if J0- 8 M l ,25(0H) 2 D3 is present throughout the
culture period (continuously added from day 5, proliferating period), the
resulting inhibition of collagen and alkaline phosphatase gene expression
prevents formation of a mineralized extracellular matrix. 9 In conditions where
mineralization is prevented, there is no induction of osteopontin and osteo-
calcin by l ,25(0H) 2 D3 . This defines the second transition point and illustrates
the role of the mineralized extracellular matrix in promoting osteoblast dif-
ferentiation. Thus, the program of bone cell maturation is blocked because
of the inhibition of proliferation and collagen synthesis. Although vitamin D
is well documented to increase osteocalcin transcription and synthesis, 86 ·94
osteocalcin cannot be expressed nor modulated by vitamin D if osteoblast
differentiation is blocked by high concentrations and long exposure of pro-
liferating cells to vitamin D in this system. This concept is further supported
by experiments in which the cultures were treated with I ,25(0H) 2 D3 contin-
uously at various times after formation of the mineralized matrix. 9 Since the
OC gene is expressed under these conditions, stimulation by hormone is
MATRIX
MATURATION
OA YS AFTER PLA T1 NG
FIGURE 5. Model of the reciprocal relationship between proliferation and differentiation in

normal diploid cells during the rat osteoblast developmental sequence (top panel) modifications
in vitamin D treated cells with acute (middle) and chronic (lower) treatment. The proliferation-
differentiation relationships are schematically illustrated as arrows representing changes in expres-
sion of cell cycle and cell growth regulated genes (proliferation arrow) and genes associated
with the maturation (differentiation arrow) of the osteoblast phenotype as the extracellular matrix
develops and mineralizes in normal diploid cell cultures (top panel). Here, the three principal
periods of the osteoblast developmental sequence are designated within broken vertical lines
(proliferation, matrix development and maturation, and mineralization). These broken lines
indicate the two experimentally established principal transition points in the developmental
sequence exhibited by normal diploid osteoblasts during the progressive acquisition of the bone
cell phenotype: the first at the completion of proliferation when genes associated with matrix
development and maturation are upregulated, and the second at the onset of extracellular matrix
mineralization. The middle panel schematically illustrates the acceleration of the relationship
between growth and differentiation by vitamin D in short-term treatment to 48 h. Hormone
accelerates proliferation and differentiation (broken arrows) by changing levels of gene expression
in one period relative to that found in the following period. In the lower panel the inhibitory
effects of vitamin D on differentiation of osteoblasts in vitro (broken arrows) are illustrated for
continuous hormone treatment initiated on day 5 or day 20 reflected by inhibition (day 5) or
cessation (from day 20) of the formation and growth of mineralized nodules. H4, histone; COL-
I, type cxl collagen; FN, fibronectin; ALK PHOS, alkaline phosphatase; MGP, matrix Glaprotein;
OP, osteopontin; OC, osteocalcin.
Lian and Stein 403
observed related to extent of mineralized matrix that was formed at the time
of vitamin D exposure (Figure 5, lower panel). Continuous presence of hor-
mone blocked further maturation of the cultures (e.g., growth and mineral-
ization of nodules were inhibited). It has been observed that chronic exposure
of rat osteosarcoma cells (Ros 17/2.8) to I ,25(0H) 2 D 3 can result in altered
phenotypic expression. After 5 d, osteocalcin mRNA was decreased and MGP
expression, which is not normally detected in this Ros line, was induced. 61
One factor contributing to the variability of response of the osteoblast to
I ,25(0H) 2 D3 is potentially related to changes in the receptor levels of vitamin
D. Receptor level is regulated by proliferative status of the cell, presence of
glucocorticoids, and homologous upregulation of its receptor, as shown in
the Ros 17/2.8 rat osteosarcoma osteoblast-like cellline. 29 •95 However, it can
be seen in Figure 4 that genes expressed both during proliferation and dif-
ferentiation (e.g., OP and MGP) are regulated by the hormone throughout
the developmental sequences. Thus, receptor availability is not rate limiting.
Calcium is yet another contributing factor to the regulation of the vitamin D
receptor in osteoblast-like cells. This was suggested by the report96 that the
upregulation of the vitamin D receptors in UMR 106 cells is inhibited by two
different calcium channel blockers, nitrendipine and verapamil. Such obser-
vations raise the possibility that osteoblasts within a matrix that is undergoing
extracellular mineralization, may show variable levels of vitamin D respon-
siveness if the receptor level is closely regulated by the extracellular calcium
concentration.
In addition to the direct modulation of osteoblast gene activity that results
from interaction of the vitamin D receptor complex with specific gene se-
quences, new evidence has accumulated to indicate several nongenomic ef-
fects. Changes in cellular calcium occur in response to 1,25(0H) 2 D 3 in a
matter of seconds and these have been classified as "rapid" or nongenomic
effects. 1-a-25(0HhD3 rapidly increases cytosolic calcium and alters mem-
brane phospholipid metabolism in hepatocytes, 97 cultured osteoblasts, 3 .4·98
chondrocytes 7 and isolated enterocytes. 4 The rapid transmembrane influx of
extracellular calcium after stimulation by 1,25(0H) 2 D 3 is mediated largely
by the opening of voltage-gated calcium channels. 3 •99 The rapid, "nongen-
omic'' effects of vitamin D include prolonged opening of the voltage-sensitive
calcium channels present in the plasma membrane, 99 phosphoinositide break-
down,4 altered enzymatic activity in matrix vesicles, 7 and calcium transport
in chick intestine. 100 A fundamental question raised by these observations is
whether the responses are mediated by a different type of receptor for
1,25(0H) 2 D3 4·101 and whether there are interactions with the steroid-receptor
signaling pathway.
Thus, many aspects of osteoblast metbolism are regulated by vitamin D.
Several factors, however, cloud an understanding of the precise effects of
vitamin D on the osteoblast in relation to bone formation and resorption.
First, other than an understanding of collagen as a bone matrix structural
TABLE 1
Sequence Comparison of Known Vitamin D-Responsive Elements in
the Promoters of Vitamin D-Regulated Genes
API
-512 -483
Human osteocalcin 5'GGTGACTCA cc GGGTGA ACG GGGGCA TT 3'
API
-470 -443
Rat osteocalcin 5'TGCCCTGCA CT GGGTGA ATG AGGACA TT 3'
-761 -741
Mouse osteopontin 5' AC AA GGTTCA CGA GGTTCA CG 3'
- 198 -180
Mouse calbindin-D28k 5' GGGGGA TGTG AGGAGA 3'
-489 -475
Rat calbindin-D9, 5' GGGTCT CGG AAGCCC 3'
Note: Homologous half steroid elements are underlined.
protein, the functions of other products that are regulated by vitamin D in

osteoblasts are not clearly defined. 65 Second, the transcriptional effects of the
hormone on osteoblast expressed genes are not straightforward. As indicated
above, studies have reported either stimulation or inhibition of gene expression
and protein synthesis dependent upon the level of vitamin D in vivo or con-
centration of hormone in vitro. Effects vary with the cell culture system being
used, the density of the cells, and the duration of hormone treatment. From
our most recent studies in normal osteoblasts9 and numerous reports in trans-
formed cells, one explanation for the apparent discrepancies is that the ability
of l ,25(0H) 2D 3 to modulate gene transcription depends upon the basal levels
of expression of the gene, which in tum relates to the maturational state of
the cell being studied. The question arises as to whether changes in these
osteoblast parameters are primary or secondary to the hormone's effect on
changes in the differentiated state of the cell or the proliferative capacity of
the cell, since a change in proliferation alters the levels of nuclear proto-
oncogenes and other factors that regulate gene transcription. An analysis of
the vitamin D responsive element (VDRE) in the genes regulated by vitamin
D should lead to clarification of these complexities. To date, VDREs have
been identified in only five genes (Table 1): mouse osteopontin, 84 calbindin-
D28k, 102 calbindin-D9k, 103 and human and rat osteocalcins. 85 ·87·94 ·104
C. INFLUENCE OF 1,25(0H) 2 D3 ON OSTEOCLAST

DIFFERENTIATION AND BONE RESORPTION
When vitamin D-deficient animals are treated with l-a-25(0HhD 3 , os-
teoclastic bone resorption is enhanced and the number of osteoclasts in-
Lian and Stein 405
creases. 21 Osteoclastic bone resorption is induced by two different mecha-

nisms: first the recruitment of osteoclast mononuclear progenitors and their
transition to multinucleated osteoclasts and second, the activation of quiescent
osteoclasts. 10 Numerous studies have established that osteoclasts are formed
by the differentiation and fusion of progenitors derived from hematopoietic
stem cells. 105 • 106 Equally important, many studies have demonstrated that
1,25(0H) 2 D 3 is a potent mediator of differentiation of such progenitors; for
example, the human leukemia HL60 cell line can be induced to differentiate
to adherent macrophages by vitamin D. 107 As mentioned above, marrow
mononuclear cells can be promoted to differentiate into cells having osteoclast
characteristics such as ruffled borders and clear zones, tartrate resistant acid
phosphatase activity (TRAP), calcitonin receptors, and the ability to resorb
bone matrix. 34 Thus, while the fully differentiated multinucleated osteoclast
itself does not have receptors for 1,25(0H) 2 D 3 , the hormone can promote
formation of this bone resorbing cell.
Vitamin D also contributes to activation of the mature osteoclast via its
effects on osteoblasts. Indeed morphologic changes in osteoclasts are evident
in response to 1 ,25(0H) 2 D3 in cultured long bones. 108 Osteoclast-like mul-
tinucleated cells can be formed in cocultures of spleen cells together with
osteoblastic cells derived from the calvaria in the presence of 1,25(0H) 2 D 3 • 109
TRAP is a marker of active osteoclasts. In mouse marrow cultures, TRAP
positive multinucleated cells were formed only near clusters of alkaline phos-
phatase-positive stromal cells, indicative of an osteoblast-like cell. 110 Thus,
osteoblastic cells appear to be involved in osteoclast formation induced by
1,25(0H) 2 D3 . These and several other studies have provided evidence con-
sistent with the model of Rodan and Martin 31 which proposed that the effects
of 1,25(0H) 2 D3 on bone resorption occurs by a direct effect on bone that
could be elicited by an initial effect on the osteoblast to produce a local
humoral mediator which would then activate the osteoclast. While many
osteoblast products, such as osteopontin and osteocalcin, are secreted into the
extracellular matrix in response to vitamin D, the specific protein or factor
that interacts with the osteoclast to promote activation to a resorbing cell is
not yet defined. Thus, vitamin D induces bone resorption through differen-
tiation of osteoclast progenitors in a manner that involves marrow-derived
cells and by activation of existing osteoclasts mediated by its effects on the
osteoblast.
The effect of vitamin D on osteoblast characteristics is also consistent
with the hormone's ability to stimulate bone resorption. Collagen synthesis
in mature osteoblasts is inhibited by 1,25(0H) 2 D 3 in vitro and parallels the
effects for bone resorption. This downregulation of the major structural com-
ponent of bone tissue, but the upregulation of noncollagenous proteins, os-
teocalcin, and osteopontin, may relate to resorptive activities in bone. Yet
anabolic effects of vitamin D on collagen synthesis are also known. These
appear to be limited to the periosteal cells or cloned osteosarcoma lines ( e.g.,
MG63 cells) that exhibit phenotypic properties of immature osteoblasts in-

dicated by significant levels of type III collagen synthesis. 111 Here, the effect
of hormone to stimulate proliferation and collagen production in osteoblast
progenitor-like cells may serve as a coupling phenomenon to balance bone
resorption. Thus, it is possible for 1,25(0HhD 3 to meet the demands required
to maintain serum calcium by replacing the resorbed bone with newly dif-
ferentiated osteoblasts that have been stimulated to produce matrix.
IV. VITAMIN D REGULATION OF THE

OSTEOCALCIN GENE
It is well known that vitamin D anabolically and catabolically modulates

bone cell metabolic activities. More recently, it has become apparent that this
occurs through selective expression of a series of vitamin D responsive genes.
Therefore, several fundamental questions involve: (1) the identification and
characterization of the regulatory elements in skeletal-associated gene pro-
moters that respond to the hormone; (2) the mechanisms by which vitamin
D-responsive sequences of specific skeletal genes are selectively rendered
competent to bind the vitamin D receptor complex in a cell and tissue specific
manner and/or at particular stages of differentiation; and (3) the combined
influence of multiple physiological mediators on vitamin D responsiveness
at the level of transcriptional control. The VDRE of the osteocalcin gene was
the first vitamin D receptor binding sequence to be identified; and, as we
gain further insight into the structural and functional properties of the osteo-
calcin gene VDRE, the complexity of regulatory events at this transcription
control element is becoming increasingly apparent. Perhaps this is an indi-
cation of the basis for involvement of vitamin D as a mediator of osteocalcin
gene expression within the context of a broad spectrum of physiological
responses of the osteoblast.
Osteocalcin is regarded as a bone-specific protein. The 5. 7-kDa poly-
peptide (46 to 50 amino acids depending on the species) is synthesized by
osteoblasts as a 10,000-kDa precursor that undergoes posttranscriptional con-
version of glutamic acids 17, 21, and 24 to the calcium binding -y-carboxy-
glutamic acid that is catalyzed by a vitamin K-dependent carboxylase. While
the functions of osteocalcin remain to be definitively established, and the
protein may indeed have multiple biological activities, several lines of evi-
dence convincingly support involvement of osteocalcin both in skeletal de-
velopment and in bone resorption. Here, it should be noted that while osteo-
calcin is primarily associated with the bone extracellular matrix, serum levels
of osteocalcin provide an extremely important clinical indication of osteoblast
activity resulting from bone turnover (resorption). This is reflected by changes
in serum osteocalcin levels in several skeletal disorders. 57 But perhaps more
Lian and Stein 407
importantly, serum osteocalcin levels are modified in response to and/or

causally related to a broad spectrum of physiologic circumstances where
homeostatic mechanisms controlling calcium metabolism are operative. As a
consequence. the calcitropic hormones (vitamin D. parathyroid hormone)
participate in regulation of cellular osteocalcin levels and the extent to which
osteocalcin is secreted. 112 Osteocalcin synthesis and rnRNA levels are altered
in response to vitamin D, 85 . 87·94 ·Hl4 estrogen, 113 · 114 glucocorticoids, 104 •115 reti-
noic acid, 116 · 117 TGF !3, 118 cyclic nucleotides, 119 · 120 and -y-interferon. 121 The
picture that is emerging is one of a protein whose regulation must be mediated
by multiple factors and signaling mechanisms to accommodate a diverse series
of physiological requirements.
A. SEQUENCE ORGANIZATION OF THE OSTEOCALCIN GENE

REFLECTS VITAMIN D-MEDIATED CONTROL
Understanding the mechanisms controlling vitamin D responsiveness of
the osteocalcin gene necessitates examining hormone-mediated regulation dur-
ing the progressive development of the osteoblast phenotype. In rat osteoblast
cultures where the influence of vitamin D on gene expression has been sys-
tematically studied (Figure 4), vitamin D upregulates osteocalcin gene tran-
scription only by enhancing basal levels of expression that are initiated in
mature osteoblasts during extracellular matrix mineralization. Vitamin D stim-
ulation of osteocalcin was not observed in proliferating osteoblasts. Thus,
vitamin D does not function as an inducer of osteocalcin gene transcription,
but as an enhancer.
Sequence analysis of the osteocalcin gene provides an indication of the
complexity of transcriptional and posttranscriptional regulation of the gene
(Figure 6). The sequence organization of the mRNA coding region of the
gene reflects the biochemical events associated with processing of the initially
synthesized 10,000-kDa precursor. The complexity of vitamin D-mediated
control of osteocalcin transcription is illustrated by direct interactions of the
vitamin D receptor with the VDRE in the osteocalcin gene promoter, which
is accompanied by and functionally related to modifications in protein/DNA
interactions at other regulatory sequences of the osteocalcin gene. Identifi-
cation and characterization of proteins that interact in a sequence-specific
manner with their cognate regulatory elements in the osteocalcin gene pro-
moter further establishes the physiological mediators of osteocalcin gene tran-
scription and the circumstances under which they participate in regulation.
Additionally, the cloned gene permits direct quantitation of the rate of osteo-
calcin gene transcription and cellular levels of osteocalcin mRNA, permitting
a direct assessment of the involvement of transcriptional and posttranscrip-
tional control during development, tissue maintenance, and in skeletal dis-
orders.
.. ·r-----1·.
AP-1
.
GCACTGGGTGAA TGAGGACA TT ACTGACCGCT
AP- 1
r------1
AP-1
r--1
ATGACCCCCM TT AGTCCTGGCAG
GRE GRE
..----, r;---1
GilTIITMMGGT IITTGCAGAACAGACMGTCCC
c:QtS~\C"~
-1100 -464 -437 -99 -76 -33 -1 50
mRNA
'
FIGURE 6. Rat osteocalcin gene promoter indicating sequences for physiological regulatory
elements. Structural organization of the rat osteocalcin gene. (A) Sequences of those regulatory
elements in the proximal promoter that have been defined and partially characterized. These
include the vitamin D responsive element (VDRE); the osteocalcin box (OC box), which is a
primary proximal transcription regulatory element containing the CCAAT motif as a central core;
and the glucocorticoid response elements (GLUC) in the rat gene 173 or overlapping the TATA
motif as defined in the human gene. 121 Within the OC box and VDRE elements are also found
active AP-1 sites which bind the oncogene-encoded Fos/Jun protein complex. The solid circles
above or below G residues indicate vitamin D receptor protein-DNA interactions defined at single
nucleotide resolution within the vitamin D responsive element. 94 (B) With reference to the OC
box and TATA promoter elements, intron and exon organization and location of the pre-, pro-
peptide, and mature osteocalcin forms are illustrated.
In contrast to other bone-related genes (e.g., type I collagen and alkaline

phosphatase), only a single mRNA transcript has been observed from the
osteocalcin gene. It appears that 1 transcription initiation site is utilized, and
the 4 splicing events that fuse exons 1 through 4 into an mRNA encoding
the osteocalcin prepropeptide are not known to vary. The presence of only a
single osteocalcin mRNA additionally supports one functional poly A site and
a fixed length to the 3' poly A sequence. Although regulation of osteocalcin
expression does not appear to be modulated by changes in the organization
of the mRNA transcripts, this does not preclude the presence of sequences
in the transcribed region of the osteocalcin gene which contribute to the
regulation of transcription.
Consensus sequences for responsiveness to several physiological media-
tors of osteocalcin expression reside in the osteocalcin gene promoter. In
addition to a VDRE, sequences bearing homology to other steroid receptor
binding complexes, for example, estrogen, glucocorticoids, retinoic
acid, 86 • 116 • 121 and regulatory factors such as -y-interferon 122 and cyclic AMP,
are found in the 5' flanking sequences of the osteocalcin gene. As with many
genes transcribed by RNA polymerase II, a TATA motif is located at -42
to -39, a CCAAT element at -92 to -88, and AP-I and AP-2 sites are
Lian and Stein 409
present at several sites in the osteocalcin gene promoter. However, while

such consensus sequences provide an indication of potential regulatory mech-
anisms that may be operative, several lines of evidence are necessary to
establish function, among which are (1) demonstration of an influence on
transcriptional activity by deletion, substitution, or site-specific mutagenesis;
(2) identification and characterization of sequence-specific regulatory element
occupancy by cognate transcription factors; and (3) modifications in protein!
DNA interactions as a function of biological activity.
In the case of the osteocalcin gene promoter, to date only four regulatory
sequences, which are schematically illustrated in Figure 6, have been estab-
lished by the above criteria. Interestingly, they all contribute to vitamin D-
mediated transcriptional control of osteocalcin gene expression. The "osteo-
calcin box", a 24-nucleotide element with a CCAAT motif as the central
core, is a highly conserved regulatory sequence that is required for basal
expression. The TATA sequence region is also required for rendering the
osteocalcin gene transcribable. For both of these elements, a direct involve-
ment in transcriptional control has been provided by experimental approaches
in vivo, and sequence-specific protein/DNA interactions have been demon-
strated which exhibit variations with modifications in biological activity,
notably during development of the osteoblast phenotype and/or following
vitamin D treatment. The VDRE (- 466 to - 437) and multiple glucocorticoid
regulatory elements (GRE), one associated with the TATA domain, modulate
steroid hormone effects on osteocalcin gene transcriptional activity; both
elements exhibit protein/DNA interactions that influence vitamin D-mediated
control of transcription from the osteocalcin gene. Not to be overlooked is
evidence for potential interaction of these regulatory elements. Dexametha-
sone can, for example, inhibit or enhance the vitamin D stimulation of OC
gene transcription. 115 This may occur through protein-protein interactions of
transcription factors at each element or indirectly by either hormone inducing
synthesis of a transcription factor that resides at the DNA binding domain of
the other hormone. Preliminary evidence for both mechanisms is indicated.
The first can occur if the three-dimensional structure of the promoter changes,
bringing different regulatory elements into close proximity for protein-protein
interactions. Recent studies 123 have shown the presence of nucleosomes be-
tween the VDRE and proximal promoter elements (TATA region and OC
Box). Evidence for the second mechanism is indicated by the enhancement
of protein-DNA interactions in the TATA region when nuclear extracts from
vitamin D treated bone cells are examined. 124
B. REGULATION OF VITAMIN D RESPONSIVE ELEMENT

(VDRE) ACTIVITY
The VDRE of the rat 87•94 •125 and human 85 • 104 osteocalcin genes have been
identified and characterized by two approaches: (1) the determination of the
ability of systematically introduced nucleotide deletions or substitutions to

influence vitamin D upregulation of osteocalcin gene transcription 85 ·87 and
(2) the definition at single nucleotide resolution of the binding of vitamin D
receptor complexes. 94 As shown in Table I. the osteocalcin gene VDRE of
several mammals contains steroid half-element motifs with a 3-nucleotide
spacer. Of interest, the osteopontin VDRE contains a minimal 9-bp motif,
AGGTTCACG, that can confer vitamin D responsiveness 84 compared to the
24-bp osteocalcin VDRE within the osteocalcin gene promoter. Additionally,
consensus sequences for another vitamin-derived steroid hormone, retinoic
acid, 126 and for the nuclear proto-oncogene-encoded Fos and Jun proteins,
have been identified.SZ Results suggest that retinoic acid 116 • 117 and the Fos-
Jun complex 127 modulate osteocalcin gene transcription, thus providing a
functional basis for synergistic and/or antagonistic control of multiple regu-
latory activities within the VDRE. Such activities within the VDRE reflect
the contribution of multiple promoter binding factors competing for or com-
plexing with and thereby promoting and/or inhibiting interactions at a single
regulatory element. These relationships may provide an explanation for pos-
itive or negative activity of a single regulatory sequence under different
biological conditions.
Competency for occupancy of the VDRE by the vitamin D - vitamin D
receptor complex is requisite for the enhancement of osteocalcin gene tran-
scription by the hormone. Equally important is the mechanism by which the
VDRE is rendered refractory to binding of the regulatory complex in immature
osteoblasts not expressing osteocalcin and in nonosteoblastic cells. The an-
swers to these long-standing biological questions may be provided by the
complement of proteins that complex with the VDRE. Protein-DNA inter-
actions and protein-protein interactions must both be considered. Markose et
al. 94 mapped the protein-DNA interactions at the VDRE at specific guanine
residues using nuclear extracts from osteosarcoma cells that express osteo-
calcin under the influence of vitamin D. These results provided the first
indication of the complexity of transcription factor binding at the VDRE.
Subsequent studies 128 • 129 demonstrated the requirement of an accessory factor
for binding of the vitamin D receptor complex to the osteocalcin gene VDRE.
Additionally, the finding that vitamin D receptor phosphorylation is essential
for hormone-receptor complex formation 130 • 131 and upregulation of
transcription 103 reflects the complexity of events associated with activity of
the VDRE. Modifications that mediate or reflect activity at the osteocalcin
gene VDRE under various biological circumstances are suggested by striking
differences in VDRE protein-DNA interactions in normal diploid osteoblasts
compared with osteosarcoma cells and during progression of osteoblast dif-
ferentiation (Figure 7). Such differences in the VDR complex may relate to
mechanisms by which osteocalcin synthesis and regulation by vitamin D can
occur in proliferating transformed cells but is prevented in proliferating normal
diploid osteoblasts.
Lian and Stein 411
.D.D
~~ 0
<<
II
(.)
II
a: a.. + a: a..
PROBE:VDRE PROBE:VDRE
ROS 17/2.8 ROB
FIGURE 7. Characterization of vitamin D-mediated enhancement of protein-DNA interactions
at the rat OC gene VDRE in normal osteoblasts and Ros 17/2.8 cells. Gel mobility shift assays
using nuclear proteins from untreated (C) or w-• M vitamin D-treated (D) confluent (day 8)
Ros 17/2.8 cells or mature (day 32) normal diploid osteoblasts show enhancement of protein-
DNA interactions with a "P-labeled rat osteocalcin VDRE oligonucleotide DNA probe following
treatment of the cells with vitamin D receptor (VDR) binding is indicated by arrows. Monoclonal
antibodies against the porcine VDR which detect only the porcine VDR (P-Ab) or which cross-
react with the rat VDR (R-Ab) were preincubated with the nuclear proteins for 20 min prior to
DNA probe addition. 124 These antibodies demonstrate that the vitamin D enhanced protein-DNA
interactions observed in both cell types are dependent on the binding of the vitamin D receptor.
Note that two VDR-DNA complexes are formed in the osteosarcoma cells and only the fast-
mobility complex is detectable in differentiated ROB cells expressing osteocalcin.
C. PHENOTYPE SUPPRESSION: A POSTULATED MECHANISM

FOR THE COORDINATE REGULATION OF VITAMIN
D-MEDIATED ENHANCEMENT AND BASAL TRANSCRIPTION
OF THE OSTEOCALCIN GENE BY FOS-JUN INTERACTIONS
AT AP-1 SITES WITHIN THE VDRE AND OTHER
PROMOTER ELEMENTS
A mechanism by which vitamin D modulates the extent to which the
osteocalcin is transcribed postproliferatively in osteoblasts may involve reg-
ulation of Fos-Jun binding at the VDRE and at the OC box. We identified a
series of consensus sequences for transcription factor AP-1 86 •94 • 127 and dem-
onstrated sequence-specific interactions of these promoter binding sequences
with a stable heterodimeric Fos-Jun complex. 127 The possibility therefore
arises that Fos and Jun proteins expressed in proliferating osteoblasts could
suppress osteocalcin gene transcription until late in the development of the
bone cell phenotype.
Several experimental results support the concept of coordinate occupancy
of the Fos-Jun protein complex at the VDRE and OC Box regulatory elements,
providing a potential molecular mechanism to account for the absence of
osteocalcin gene expression in proliferating osteoblasts. Expression of c-fos
and c-jun occurs primarily during the proliferative period of the osteoblast
developmental sequence. 51 •54 Second, AP-1 binding activity is observed pri-
marily in proliferating osteoblasts and dramatically decreases after the down-
regulation of proliferation and the initiation of extracellular matrix maturation
and mineralization at which time osteocalcin gene transcription is initiated. 127
Clearly, distinct patterns of protein-DNA interactions occur at the VDRE as
well as the OC Box, with nuclear protein extracts from proliferating (day 5)
osteoblasts which do not express osteocalcin compared to differentiated cells
where AP-1 activity has significantly declined. In tumor cells (Ros 17/2.8),
osteocalcin expression is deregulated, in that transcription continues in pro-
liferating cells. Even so, a reciprocal relationship between AP-1 activity and
OC gene expression is maintained. In confluent osteoblast cultures AP-1
activity is markedly diminished and osteocalcin mRNA and transcription
increase tenfold. The loss of AP-1 activity during the developmental sequence
of osteoblast maturation coincides with marked changes in protein-DNA inter-
actions in osteocalcin gene promoter elements having AP-1 sites. For example,
we observed that protein-DNA contacts do not occur at G residues within the
AP-1 consensus sequences of the VDRE and OC Box when the gene is actively
transcribed. 94 Additionally, transfection of c-fos and c-jun into cells expressing
osteocalcin resulted in the downregulation of osteocalcin gene transcription
and further supports a Fos-Jun mediated suppression ofthe osteocalcin gene. 116
These results are consistent with a model in which coordinate occupancy of
the AP-1 sites in the VDRE and osteocalcin box in proliferating osteoblasts
may suppress both basal and vitamin D-enhanced osteocalcin gene transcrip-
tion. This phenomenon, in which osteocalcin gene expression is prevented
during proliferation, is called phenotype suppression. mm
Lian and Stein 413
AP -1 SITES WITHIN THE HUMAN VDRE

RAT
T G C C C T G C A C T G G G T G A A T GA G G A C A T T A C T G A C C
A C G G G AC G T G A C C C A C T T A C T C C T G T A A T G AC T G G
HUMAN
G G T G A C T C A C C G G G T G A AC G G G G G C A T T G C G A G G C
C C A C T G A G T G GC C C A C T T G C C C C C G T A A C G C T C C G
0
z :z~
.0
<(
.0
<( ....,....,o
:::> :::> _J
,, 0
"-J
I
+ a..
I
a: ++
Cl) z Cl) Cl) I
0:::>0 Oa...
u......, u.. u.. <(
Recombinant
Fos/Jun
Extracts from Proteins
Vitamin D
Treated Cells
FIGURE 8- Vitamin D receptor and Fos-Jun protein complex binding to the human VDRE
sequences. Nuclear extracts from 48-h vitamin D-treated Ros cells were used in gel mobility
shift assays (left panel) with an oligonucleotide probe to the human VDRE as indicated by the
dotted underline in the top panel sequence. The vitamin D receptor complex is blocked by
preincubation with antibody recognizing the rat receptor (R-AB) but not the pig-specific antibody
(9P-AB), provided by H. DeLuca, Madison, WI. The probe used for these protein-DNA inter-
actions has the classic AP-I consensus deleted (solid underline) but retains the AP-I site ho-
mologous to the rat VDRE (bold letters) that resides within the core of this regulatory element.
The right panel demonstrates that this is an active AP-I site by binding of only the Fos-Jun
protein complex which can be competed by the AP-1 oligonucleotide.
Mutual exclusion of occupancy by fos-jun and element-specific factors

at the VDRE and OC Box is further supported by interactions of fos-jun
binding using nuclear extracts from osteoblasts expressing the OC gene only
when mutations are introduced that preclude VDR and CAAT motif factor
interactions. 174 Using both recombinant fos and jun proteins, 127 and nuclear
extracts, we have demonstrated fos-jun complex binding (Figure 8) to the
internal AP-1 site (bold letters) in the human osteocalcin gene VDRE. 132 The
human VDRE, in contrast to the rat VDRE, has an additional upstream AP-
1 site (solid underline in Figure 8). While the internal AP-I site may function
in suppression of transcription, the upstream AP-I site in the human VDRE
has been shown to enhance vitamin D-stimulated gene transcription, but is
not necessary for vitamin D regulation. 128 This suggests that responsiveness
of the human and rat osteocalcin genes to factors that influence AP-I activity
may differ. Yet, the similar organization of the internal AP-I sites of the
VDRE and the identical AP-I organization in the osteocalcin box for the
human and rat osteocalcin genes suggests that there are some similar functional
properties of the elements, perhaps in regulating expression in relation to the
proliferative state of the cells.
A remaining question is the mechanism by which the osteocalcin gene is
rendered transcribable and vitamin D responsive following the downregulation
of proliferation. Here, the possibilities include:
I. Release of the Fos-Jun complex from the AP-I sites to permit sequences
to be available for occupancy by the vitamin D receptor complex and/
or by tissue-specific osteocalcin box transcription factors;
2. Modifications of the Fos-Jun complex that facilitate binding of acti-
vation related factors;
3. Association with other DNA binding proteins or protein-protein inter-
actions.
With respect to the latter possibility, binding of the Fos-Jun complex may
play a dual positive and negative role in the regulation of transcription. The
Fos-Jun complex may suppress osteocalcin gene transcription when prolif-
eration is ongoing by directly or indirectly modulating sequence-specific in-
teractions at the vitamin D receptor binding domain. Then postproliferatively,
the Fos-Jun complex may facilitate vitamin D receptor binding to support the
sequential upregulation of osteocalcin and other vitamin D responsive genes.
Such regulation could be accomplished by different fos-jun protein complexes.
Both fos and jun are members of a family of related proteins, and a recent
study indictes differential stimulation of c-jun and jun-B by I ,25(0H) 2 D 3 in
MC3T3-EI osteoblastic cells. 133
Further experimental results are necessary to determine whether the or-
ganization of steroid receptor binding domains and AP-I sites within pro-
moters of genes that are hormone responsive can provide a general mechanism
Lian and Stein 415
for the phenotype suppression and/or activation in different periods of cell

and tissue differentiation. However, support for such a model is provided by
the association of AP-1 sites within consensus sequences for other steroid
responsive elements, for example, glucocorticoids. 116 · 127•134 Undoubtedly the
phenotype suppression model represents a simplification of an extremely
complex series of protein-DNA interactions whereby multiple physiological
signals are transduced to the nucleus resulting in modifications in transcription
that progressively alter phenotypic properties of cells leading to structural and
functional events associated with differentiation. However, the strength of
such a model is that it provides a basis for experimentally addressing the
functional significance of organization of regulatory elements for phenoptypic
marker genes of cell differentiation within the context of gene expression
associated with proliferation.
It is necessary to account not only for precise control of basal transcription
from the osteocalcin gene, but also to explain the sensitivity of vitamin D
modulation to basal levels of expression. In focusing on this component of
transcriptional regulation, we obtained results consistent with the possibility
that the TATA box sequences (the primary transcriptional regulatory element)
and the VDRE may interact to control the level of expression, possibly by
cooperative interactions. We have shown sequence-specific, vitamin D-de-
pendent, protein-DNA interactions at both the VDRE and TATA box using
nuclear extracts from both Ros 17/2.8 cells 94 and normal diploid osteoblast
cultures. 124 • 135 Thus, vitamin D-mediated regulation at the VDRE is integrated
with activities at other basal regulatory elements via vitamin D-dependent
protein-DNA or protein-protein complexes.
V. POTENTIAL THERAPEUTIC APPLICATIONS

OF VITAMIN D
An increased understanding of molecular mechanisms by which vitamin
D-mediated genes are responsive to the hormone can provide the basis for a
broad spectrum of therapeutic uses. Here, knowledge of how VDREs of
specific genes and tissues are selectively rendered operative will allow for
development of treatment regimens that incorporate targeted responses. Spe-
cifically, we should take advantage of the complexity of regulatory events
that include a series of protein-DNA interactions at the VDRE which render
the hormone responsive element competent for binding the vitamin D receptor
and, equally important, the observed and potential variations in responsiveness
elicted by metabolites and analogs of the hormone.
Clinical disorders related to vitamin D can arise from multiple factors,
either (1) altered availability of the parent molecule vitamin D (e.g., by dietary
deficiency); (2) altered conversion of vitamin D to its principal active me-
tabolite 1,25(0H) 2 D3 , produced in the kidney, (3) conditions that may be due
to variations in organ responsiveness to the active metabolites; or (4) pertur-
TABLE 2
Examples of Human Diseases Related to
Disturbances in the Vitamin D Endocrine System
Anticonvulsant therapy
Cirrhosis
Diabetes
Drug-induced metabolism
Glucocorticoid antagonism
Hyperparathyroidism
Hypoparathyroidism
Leukemia
Malabsorption
Osteomalacia
Osteoporosis
Psoriasis
Renal disease
Sarcoidosis
Steatorrhea
bations in the integrated actions of these metabolites on selected responsive

genes. Table 2 indicates some of the major disorders associated with vitamin
D abnormalities or disturbances, including those secondary to alterations of
PTH and/or calcitonin as components of the vitamin D endocrine system, or
secondary to drug induced inactivation of vitamin D. Even today some groups
are at risk for vitamin D deficiency, including breast-fed infants who do not
receive vitamin D supplementation, and people in an environment with re-
duced exposure to sun and inadequate intake of fortified foods, e.g., Asian
women (particularly primigravidas) who wear veils and who may also con-
sume nonsupplemented foods. Disorders of the intestine, such as malabsorp-
tion syndrome, steatorrhea or tropical sprue, or liver disorders (e.g., cirrhosis
or obstructive jaundice), lead to vitamin D deficiency and thereby altered
bone formation.
Several hereditary disorders of the vitamin D endocrine system are
known. 136 One group, vitamin D-dependent Rickets Type II, involves mu-
tations that cause the vitamin D receptor gene to code for an abnormal receptor
and lead to defects that disturb all tissues responsive to vitamin D. In these
patients there is a marked heterogeneity of clinical manifestations. 137 This is
likely a consequence of at least five phenotypically different defects which
are related to variations in hormone-receptor binding capacity, affinity or
nuclear localization and defects in binding of the hormone-receptor complex
to DNA. In another inborn error of vitamin D metabolism, Vitamin D-de-
pendent Rickets Type I, the autosomal recessive defect is presumed to involve
the renal 25(0H)D 1 hydroxylase enzyme. These patients can be treated with
the active metabolite 1 ,25(0H) 2D 3 for normalization of disturbed parameters
of the vitamin D endocrine system. 138 In an X-linked dominant disorder,
Hypophosphatemic Vitamin D Resistant Rickets of unknown etiology, high
Lian and Stein 417
doses of phosphate and 1,25(0H) 2D3 have allowed for fairly normal growth
of these children. 139 In addition to these inborn errors of vitamin D metabolism
are those disorders involving a tissue-selective resistance to I ,25(0H) 2 D3-
but here there is, as yet, no understanding of the cellular basis which relates
to a direct effect of vitamin D actions either at the genomic or nongenomic
levels. For example, elevated 1,25(0H) 2D3 levels occur in human and rat
osteopetrosis, a hereditary disorder having a spectrum of osteoclast abnor-
malities.140 The hormone level may be elevated to maintain normal serum
calcium in the absence of active bone resorption. Alterations in some vitamin
D-regulated proteins (osteocalcin, osteopontin, collagen, alkaline phospha-
tase) indicate perturbations in bone responsiveness. 141
Vitamin D (1 ,25(0H) 2D 3) is being considered for the treatment of osteo-
porosis, 142 a growing problem. 143 Type I postmenopausal osteoporosis occurs
predominantly in women within 20 years after the menopause and is related
to estrogen deficiency. It is characterized by accelerated bone loss which leads
to suppressed PTH secretion and a trend toward low serum I ,25(0H) 2 D
concentrations and inefficient calcium absorption. 144. 146 In clinical trials in
which calcium intake was limited and dosages of 1,25(0H)zD 3 of 0.5 J.Lg/d
were employed, hypercalcemia and hypercalcuria occurred uncommonly. 147
Short-term rather than continuous treatment preferentially stimulates osteo-
blast function but not osteoclast activity. 148 Type II is the age related osteo-
porosis resulting in defects in the vitamin D endocrine system that includes
primary failure by the kidney and defective 25(0H)D1a-hydroxylase activ-
ity, 149 a decrease in vitamin D receptor concentration in intestinal mucosa and
possibly in bone cells, and, in some elderly patients, vitamin D deficiency.
With age there is decreased intestinal absorption of calcium, and this is a
major determinant in the etiology of the disease. 144
Although some investigators have documented reduced levels of
1,25(0H)zD3 in osteoporotic subjects when compared to age matched controls,
several studies document vitamin D levels in the normal range. 150 Thus, the
value of calcitriol treatment for established postmenopausal osteoporosis re-
mains controversial. Findings of several major clinical trials that have used
calcitriol in the treatment of osteoporosis are conflicting. There are studies,
e.g., Aloia et al. 151 and Caniggia et al., 152 that indicate calcitriol treatment
reduces bone loss in women with postmenopausal osteoporosis by increasing
calcium absorption and reducing bone resorption. Gallagher et al. 153 has re-
cently demonstrated that the treatment of postmenopausal osteoporotic women
with calcitriol for 2 years was associated with an increase in spine density
and total body calcium when compared to placebo. In the recent Tilyard
studyl 76 (n = 600), a 50% reduction of spinal deformation compared to
calcium-treated controls was observed. However, the Falch et al. 154 and the
Ott and Chestnut 155 studies indicate that calcitriol is not effective in the
treatment of established postmenopausal osteoporosis. The different obser-
vations may relate to the patient population and/or perhaps the stage of the
disease.
More promising, however, are applications of the antiproliferative actions

of l ,25(0H) 2 D 3 , particularly effective in cells of the immune system 156 and
keratinocytes, for the treatment of autoimmune disorders, 157 psoriasis, 158- 160
cancer, and graft vs. host reactions. 161 However, the calcemic activity of the
hormone, that is, its effects on bone, kidney, and intestine in elevating serum
calcium, can be a serious problem. It is for this reason that analogs of
1,25(0H) 2 D3 are being developed that have specific activities in particular
target tissues and cells, e.g., leukemic cells, with little effect on bone calcium
mobilization. 162 • 163 One compound, calcipotriol (MC903, Leo Pharmaceutical
Products Company, DK-2750, Ballerup, Denmark) has low calcemic activity 164
and trial studies have demonstrated that it may be useful for treating pso-
riasis. 158 The reason for its low calcium mobilizing effect in vivo is not clear.
Studies indicate that MC903 has good binding activity to the vitamin D
receptor but low affinity for the vitamin D binding protein in the serum. 165
When tested in human osteoblast culture systems, one study found that cal-
cipotriol and l ,25(0H) 2 D 3 resulted in similar stimulation of alkaline phos-
phatase and osteocalcin. 166 However, studies of Marie et al. 167 indicate that
MC903 was less effective than l ,25(0H) 2 D 3 dependent upon the basal level
of expression of osteocalcin. Another analog, 22-oxa-calcitriol (OCT), also
promotes differentiation without the calcemia. However, the anticalcemic
effect is not related to its inability to stimulate osteoblast activity. In two
studies, in human MG63 cells 168 and in rat osteosarcoma Ros 17/2.8 cells, 169
the response of hormone and analog to osteocalcin, alkaline phosphatase, and
induction of vitamin D receptors was similar. However, OCT was reported
not to mediate the rapid nongenomic effects such as increasing cytosolic
calcium. 3
These observations lead one to several speculations regarding the mech-
anism of the differential actions of the analogs on specific gene products or
cellular responses. Table 3 illustrates the variable degrees of mRNA stimu-
lation of two vitamin D-regulated genes by several 1,25(0H) 2 D3 analogs. The
properties of the hormone-receptor complex's interaction with the vitamin D
responsive gene segment may be altered. This hypothesis is supported by
observations of the fluorinated analogs that are more potent than 1,25(0H) 2 D3
in vivo 170 and in vitro in osteoblasts, with respect to alkaline phosphatase and
osteocalcin stimulation and inhibition of DNA synthesis. 166 Studies show,
however, that the increased potency is not due to receptor binding, as the
fluorinated derivatives have less or similar affinity constants as
1,25(0H) 2 D 3 • 171 • 172 Thus, it may be the specific interaction of the hormone-
receptor complex which promotes interactions with other transactivating fac-
tors and the DNA sequences of the genes that accounts for gene activation
or responsiveness. For example, PTH is downregulated by vitamin D, and
the conformation of the receptor-hormone analog complex may promote an
interaction with a particular transactivating factor that is specific for one gene
and not another. Based upon the knowledge that the 1,25(0H)zD 3 receptor
Lian and Stein 419
TABLE 3
Stimulation of Osteocalcin (OC) and
Matrix Gla Protein (MGP) mRNA by
Vitamin D Analogs in Osteoblasts
Ratio of Vitamin
D/controt•
oc MGP
R0-5535 1,25(0H),D 3 5.3 5.2

R0-6218 1,25(0H),D 2 4.6 2.4
R0-6709 1,25(0H) 2 24-epi.,-0 22 3 7 5
R0-8525 1,25R,26(0H) 3D3 6 1.5
R0-6890 1,25(0H) 2-26F,D 3 3.4 9.8
Rat -derived osteob1asts were cultured 16 d for the de-

velopment of mineralized bone nodules. Cells were
harvested 24 h after exposure to JQ- 8 M metabolite
for preparation of total cellular RNA and quantitation
of mRNA. Note differences in the extent of stimulation
among metabolities and between the two genes. 175
complex can either up- or downregulate different genes in association with

other gene regulatory elements and/or transcription factors, a similar mech-
anism may explain differential effects of the analogs.
VI. SUMMARY
The nutritional aspects of vitamin D on bone formation and turnover are

appreciated within the context of the active metabolite of vitamin D,
1,25(0H) 2 D 3 , that functions as a steroid hormone in modulating cellular
differentiation and the production of specific gene products of bone cells.
The effects of vitamin D at the genomic level and nongenomic levels are
discussed in relation to the central role of vitamin D in maintaining normal
bone metabolism. The concept of vitamin D as a physiologic factor in both
promoting bone resorption and bone formation is presented in relation to
positive and negative regulatory effects on the transcription of osteoblastic
genes. Using the bone-specific protein, osteocalcin, as a model of a gene
regulated by the hormone, the vitamin D responsive element is analyzed.
Several concepts of transcriptional regulation are presented to account for ( 1)
vitamin D modulating gene expression as an enhancer of basal, transcriptional
activity, and (2) the synergistic and/or antagonistic interactions of vitamin D
with other regulatory factors. In light of the very recent discoveries on gene
regulation by vitamin D and consequent effects on bone formation and re-
sorption, we have addressed current understanding of abnormalities of the
vitamin D endocrine system and therapeutic applications of synthetic analogs
of 1,25(0H) 2 D3 that produce highly selective effects.
REFERENCES
I. Naveh-Many, T., Marx, R., Keshet, E., Pike, J. W., and Silver, J., Regulation of
I ,25-dihydroxyvitamin 03 receptor gene expression by I ,25-dihydroxyvitamin 03 in the
parathyroid in vivo, J. Clin. Invest., 86, 1968, 1990.
2. Ozono, K., Sone, T., and Pike, J. W., The genomic mechanism of action of 1,25-
dihydroxyvitamin 0 3 , J. Bone Min. Res., 6, 1021, 1991.
3. Civitelli, R., Kim, Y. S., Gunsten, S. L., Fujimori, A., Huskey, M., Avioli, L. V.,
and Hruska, K. A., Nongenomic activation of the calcium message system by vitamin
0 metabolites in osteoblast-like cells, Endocrinology, 127, 2253, 1990.
4. Lieberherr, M., Grosse, B., Duchambon, P., and Drueke, T., A functional cell surface
type receptor is required for the early action of 1,25-dihydroxyvitamin 0 3 on the phos-
phoinositide metabolism in rat enterocytes, J. Bioi. Chern., 264, 20403, 1989.
5. Baran, D. T., Sorensen, A. M., Owen, T., Shalhoub, V., Stein, G., and Lian, J.,
lo.,25-0ihydroxyvitamin 0 3 rapidly increases cytosolic calcium in clonal rat osteosarcoma
cells lacking the vitamin 0 receptor, J. Bone Min. Res., 6, 1269, 1991.
6. Harmond, M. F., Thomasset, M., Rouais, F., and Ducassou, D., In vitro stimulation
of articular chondrocyte differentiated function by 1,25-dihydroxycholecalciferal or 24,R,25-
dihydroxycholecalciferol, J. Cell. Physiol., 119, 359, 1984.
7. Boyan, B. D., Schwartz, 1., Carnes, D. L., and Ramirez, V., The effects of vitamin
0 metabolites on the plasma and matrix vesicle membranes of growth and resting cartilage
cells in vitro, Endocrinology, 122, 2851, 1988.
8. Gerstenfeld, L. C., Kelly, C. M., von Deck, M., and Lian, J. B., Effect of 1,25-
dihydroxyvitamin D, on induction of chondrocyte maturation in culture: extracellular
matrix gene expression and morphology, Endocrinology, 126, 1599, 1990.
9. Owen, T. A., Aronow, M.A., Barone, L. M., Bettencourt, B., Stein, G., and Lian,
J. B., Pleiotropic effects of vitamin Don osteoblast gene expression are related to the
proliferative and differentiated state of the bone cell phenotype: dependency upon basal
levels of gene expression, duration of exposure and bone matrix competency in normal
rat osteoblast cultures, Endocrinology, 128, 1496, 1991.
10. Soda, T., Shinki, T., and Takahashi, N., The role of vitamin Din bone and intestinal
cell differentiation, Annu. Rev. Nutr., 10, 195, 1990.
11. DeLuca, H. F., The vitamin D story: a collaborative effort of basic science and clinical
medicine, FASEB J., 2, 224, 1988.
12. Minghetti, P. P. and Norman, A. W., 1,25(0H) 2-vitamin D3 receptors: gene regulation
and genetic circuitry, FASEB J., 2, 3043, 1988.
13. Stern, P. H., Vitamin 0 and bone, Kidney Int., 29, Sl7, 1990.
14. Marx, S. J. and Barsony, J., Tissue-selective 1,25-dihydroxyvitamin D3 resistance:
novel applications of calciferols, J. Bone Min. Res., 3, 481, 1988.
15. Toverud, S. U., Hammarstrom, L. E., and Kristoffersen, U. M., Quantitative studies
on acid phosphatase in developing rat bones and teeth during hypervitaminosis D, Arch.
Oral Bioi., 20, 175, 1975.
16. Lian, J. B., Carnes, D. L., and Glimcher, M. j., Bone and serum concentrations of
osteocalcin as a function of 1,25 (OH) 2 D3 circulating levels in rats with bone disorders,
17. Marie, P. J., Hott, M., and Garba, M.-T., Contrasting effects of 1,25-dihydroxyvi-
tamin D3 on boRe matrix and mineral appositional rates in the mouse, Metabolism, 34,
777, 1985.
18. Weisbrode, S. E., Capen, C. C., and Norman, A. W., Ultrastructural evaluation of
the effect of 1,25-dihydroxyvitamin D3 on bone of thyroparathyroidectomized rats fed a
low-calcium diet, Am. J. Pathol., 92, 459, 1978.
19. Hock, J. M., Gunnes-Hey, M., Poser, J., Olson, H., Bell, N.H., and Raisz, L. G.,
Stimulation of underminera1ized matrix formation by 1,25 dihydroxyvitamin D, in long
bones of rats, Calcif. Tissue Int., 28, 79, 1986.
Lian and Stein 421
20. Gallagher, J. A., Beneton, M., Harvey, L., and Lawson, D. E. M., Response of
rachitic rat bones to I ,25-dihydroxyvitamin D,: biphasic effects on mineralization and
lack of effect on bone resorption, Endocrinology, 119, 1603, 1986.
21. Marie, P. J, and Travers, R., Continuous infusion of I ,25-dihydroxyvitamin D3 stim-
ulates bone turnover in the normal young mouse, Calcif. Tissue Int., 35, 418, 1983.
22. Underwood, J, L. and DeLuca, H. F., Vitamin D is not directly necessary for bone
growth and mineralization, Am. J. Physiol., 246, E493, 1984.
23. Holtrop, M. E., Cox, K. A., Carnes, D. L., and Holick, M. F., Effects of serum
calcium and phosphorus on skeletal mineralization in vitamin D-deficient rats, Am. J.
Physiol., 251, E234, 1986.
24. Clark, S. A., Boass, A., and Toverud, S. U., Effects of high dietary contents of calcium
and phosphorus on mineral metabolism and growth of vitamin D-deficient suckling and
weaned rats, Bone Min., 2, 257, 1987.
25. Kollenkirchen, U., Walters, M. R., and Fox, J,, Plasma Ca influences vitamin D
metabolite levels as rats develop vitamin D deficiency, Am. J. Physiol., 260, E447, 1991.
26. Wronski, T. J,, Halloran, B. P., Bikle, D. D., Globus, R. K., and Morey-Holton,
E. R., Chronic administration of I ,25-dihydroxyvitamin D,: increased bone but imparied
mineralization, Endocrinology, 119, 2580, 1986.
27. Raisz, L. G., Trummel, C. L., Holick, M. F., and DeLuca, H. F., 1,25-dihydrox-
ycholecalciferol, a potent stimulator of bone resorption in tissue culture, Science, 175,
768, 1972.
28. Tanaka, Y. and DeLuca, H. F., Bone mineral mobilization activity of I ,25-dihydroxy-
cholecalciferol, a metabolite of vitamin D., Arch. Biochem. Biophys., 146, 574, 1971.
29. Chen, T. L., Cohn, C. M., Morey-Holton, E., and Feldman, D., lo:,25-Dihydroxy-
vitamin D 3 receptors in cultured rat osteoblast-like cells, J. Bioi. Chem., 258, 4350,
1983.
30. Chen, T. L., Li, J, M., van Ye, T., Cone, C. M., and Feldman, D., Hormonal
responses to 1,25-dihydroxyvitamin D3 in cultured mouse osteoblast-like cells-modulation
by changes in receptor level, J. Cell. Physiol., 126,21, 1986.
31. Rodan, G. A. and Martin, T. J,, Role of osteoblasts in hormonal control of bone
resorption-a hypothesis, Ca/cif. Tissue Int., 33, 349, 1981.
32. McSheehy, P. M. J. and Chambers, T. J., 1,25-Dihydroxyvitamin D 3 stimulates rat
osteoblastic cells to release a soluble factor that increases osteoclastic bone resorption,
J. Clin. Invest., 80,425. 1987.
33. Abe, E., Miyaura, C., Tanaka, H., Shina, Y., Kuribayashi, T., Suda, S., Nishii,
Y., DeLuca, H. F., and Suda, T., lo:,25-Dihydroxyvitamin D, promotes fusion of
mouse alveolar macrophages both by a direct mechanism and by a spleen cell-mediated
indirect mechanism, Proc. Nat!. Acad. Sci. U.S.A., 80, 5583, 1983.
34. Roodman, G. D., Ibbotson, K. J., MacDonald, B. R., Kuehl, T. J., and Mundy,
G. R., I ,25-Dihydroxyvitamin D, causes formation of multinucleated cells with several
osteoclast characteristics in cultures of primate marrow, Proc. Nat!. Acad. Sci. U.S.A.,
82, 8213, 1985.
35. Narbaitz, R., Stumpf, W. E., Sar, M., Huang, S., and DeLuca, H. F., Autoradio-
graphic localization of target cells for lo:,25-dihydroxyvitamin D, in bones from fetal
rats, Calcif. Tissue Int., 35, 177, 1983.
36. Owen, M. and Friedenstein, A. J., Stromal stem cells: marrow-derived osteogenic
precursors, in Cell and Molecular Biology of Vertebrate Hard Tissues, Evered, D. and
Harnett, S., Eds., John Wiley & Sons, Chichester, 1988, 42.
37. Benayhahu, D., Kletter, Y., Zipori, D., and Weintraub, S., Bone marrow-derived
stromal cell line expressing osteoblastic phenotype in vitro and osteogenic capacity in
vivo, J. Cell. Physiol., 140, I, 1989.
38. Bernier, S. M., Desjardina, J., Sullivan, A. K., and Gollzman, D., Establishment
of an osseous cell line from fetal rat calvaria using an immunocytolytic method of cell
selection: characterization of the cell line and of derived clones, J. Cell. Physiol., 145,
274, 1990.
39. Grigoriadis, A. E., Heersche, J, N. M., and Aubin, J, E., Differentiation of muscle,
fat, cartilage, and bone from progenitor cells present in a bone-derived clonal cell pop-
ulation: effect of dexamethasone, J. Cell Bioi., 106, 2139, 1988.
40. Chen, T. L., Cone, C. M., Morey-Hollon, E., and Feldman, D., Glucocorticoid
regulation of I ,25(0H) 2D 3-vitamin D, receptors in cultured mouse bone cells, J. Bioi.
Chern., 257, 13564, 1982.
41. Peck, W. A., Birge, S. J., and Fedak, S. A., Bone cells: biochemical and biological
studies after enzymatic isolation, Science, 146, 1476, 1964.
42. Wong, G. L., Luben, R. A., and Cohn, D. V., 1,25-Dihydroxycholecalciferol and
parathormone: effects on osteoclast-like and osteoblast-like cells, Science, 197, 663, 1977.
43. Beresford, J, N., Gallagher, J, A., and Russell, R. G. G., 1,25-Dihydroxyvitamin
D, and human bone-derived cells in vitro: effects on alkaline phosphatase, Type I collagen
and proliferation, Endocrinology, 119, 1176, 1986.
44. Robey, P. G. and Termine, L. J., Human bone cells in ~·itro, Calc. Tissue Int., 37,
453, 1985.
45. Bellows, C. G., Aubin, J. E., Heersche, H. N. M., and Antosz, M. E., Mineralized
bone nodules formed in vitro from enzymatically released rat calvaria cell populations,
Calcif. Tissue Int., 38, 143, 1986.
46. Bhargava, U., Bar-Lev, M., Bellows, C. G., and Aubin, J, E., Ultrastructural analysis
of bone nodules formed in vitro by isolated fetal rat calvaria cells, Bone, 9, 155, 1988.
47. Ecarot-Charrier, B., Glorieux, F. H., van der Rest, M., and Pereira, G., Osteoblasts
isolated from mouse calvaria initiate matrix mineralization in culture, J. Cell Bioi., 96,
639, 1983.
48. Gerstenfeld, L. C., Chipman, S. D., Glowacki, J,, and Lian, J, B., Expression of
differentiated function by mineralizing cultures of chicken osteoblasts, Dev. Bioi., 122,
49, 1987.
49. Ibaraki, K., Termine, J, D., Whitson, S. W., and Young, M. F., Bone matrix mRNA
expression in differentiating fetal bovine osteoblasts, J. Bone Miner. Res., 7, 743, 1992.
50. Aronow, M. A., Gerstenfeld, L. C., Owen, T. A., Tassinari, M. S., Stein, G. S.,
and Lian, J, B., Factors that promote progressive development of the osteoblast phe-
notype in cultured fetal rat calvaria cells, J. Cell Physiol., 143, 213, 1990.
51. Owen, T. A., Aronow, M., Shalhoub, V., Barone, L. M., Wilming, L., Tassinari,
M. S., Kennedy, M. B., Pockwinse, S., Lian, J. B., and Stein, G. S., Progressive
development of the rat osteoblast phenotype in vitro: reciprocal relationships in expression
of genes associated with osteoblast proliferation and differentiation during formation of
the bone extracellular matrix, J. Cell. Physiol., 143, 420, 1990.
52. Gerstenfeld, L. C., Chipman, S.D., Kelly, C. M., Hodgens, K. J,, Lee, D. D., and
Landis, W. J., Collagen expression, ultrastructural assembly, and mineralization in
cultures of chicken embryo osteoblasts, J. Cell Bioi., 106, 979, 1988.
53. Stein, G. S., Lian, J. B., and Owen, T. A., Relationship of cell growth to the regulation
of tissue-specific gene expression during osteoblast differentiation, FASEB J., 4, 3111,
1990.
54. Collart, D., Ramsey-Ewing, A., Bortell, R., Lian, J., Stein, J., and Stein, G.,
Isolation and characterization of a eDNA from a human histone H2B gene which is
reciprocally expressed in relation to replication-dependent H2B histone genes during HL60
cell differentiation, Biochemistry, 30, 1610, 1991.
Lian and Stein 423
55. Shalhoub, V., Gerstenfeld, L. C., Lian, J. B., Stein, J. L., and Stein, G. S., Down-
regulation of cell growth and cell cycle regulated genes during chick osteoblast differ-
entiation with the reciprocal expression of histone gene variants, Biochemistry, 28, 5318,
1989.
56. Oldberg, A., Franzen, A., and Heinegard, D., Cloning and sequence analysis of rat
bone sialoprotein (osteopontin) eDNA reveals an Arg-Gly-Asp cell-binding sequence,
57. Hauschka, P. V., Lian, J. B., Cole, D. E. C., and Gundberg, C. M., Osteocalcin
and matrix Gla protein: vitamin K-dependent proteins in bone, Physiol. Rev., 69, 990,
1989.
58. Price, P. A., Role of vitamin K-dependent proteins in bone metabolism, Annu. Rev.
Nutr., 8, 565, 1988.
59. Craig, A. M., Smith, J. H., and Denhardt, D. T., Osteopontin, a transformation-
associated cell adhesion phosphoprotein, is induced by 12-0-tetradecanoylphorbol 13-
acetate in mouse epidermis, J. Bioi. Chern., 264, 9682, 1989.
60. Lian, J. B. and Friedman, P. A., The vitamin K-dependent synthesis of gamma-
carboxyglutamic acid by bone microsomes, J. Bioi. Chern., 253, 6623, 1978.
61. Fraser, J.D. and Price, P. A., Induction of matrix Glaprotein synthesis during prolonged
I ,25-dihydroxyvitamin D, treatment of osteosarcoma cells, Calcif. Tissue Int., 46, 270,
1990.
62. Barone, L. M., Owen, T. A., Tassinari, M. S., Bortell, R., Stein, G. S., and Lian,
J. B., Developmental expression and hormonal regulation of the rat matrix Gla protein
(MGP) gene in chondrogenesis and osteogenesis, J. Cell. Biochem., 46, 000-000, 1991.
63. Pockwinse, S., Wilming, L., Conlon, D., Stein, G. S., and Lian, J. B., Expression
of cell growth and bone specific genes at single cell resolution during development of
bone tissue-like organization in primary osteoblast cultures, submitted.
64. Nagata, T., Bellows, C. G., Kasugai, S., Butler, W. T., and Sodek, J., Biosynthesis
of bone proteins [SPP-1 (secreted phosphoprotein-!, osteopontin), BSP (bone sialoprotein)
and SPARC (osteonectin)] in association with mineralized-tissue formation by fetal rat
calvarial cells in culture, Biochem. J., 274, 513, 1991.
65. Boskey, A. L., Wians, F. H., Jr., and Hauschka, P. V., The effect of osteocalcin on
in vitro lipid-induced hydroxyapatite formation and seeded hydroxyapatite growth, Calc.
Tissue Int., 37, 57, 1985.
66. Lian, J. B., Tassinari, M., and Glowacki, J., Resorption of implanted bone prepared
from normal and warfarin-treated rats, J. Clin. Invest., 73, 1223, 1984.
67. Glowacki, J. and Lian, J., Impaired recruitment and differentiation of osteoclast pro-
genitors by osteocalcin-depleted bone implants, Cell Differ., 21, 247, 1987.
68. Reinbolt, F. P., Hultenby, K., Oldberg, A., and Heinegard, D., Osteopontin-a
possible anchor of osteoclasts to bone, Proc. Nat!. Acad. Sci. U.S.A., 87, 4473, 1990.
69. Miyauchi, A., Alvarez, J., Greenfield, E. M., Teti, A., Grano, M., Colucci, S.,
Zambonin-Zallone, A .. , Ross, F. P., Teitelbaum, S. L., Cheresh, D., and Hruska,
K. A., Recognition of osteopontin and related peptides by an etvf3 3 integrin stimulates
immediate cell signals in osteoclasts, J. Bioi. Chern., 266, 20369, 1991.
70. Weinreb, M., Shinar, D., and Rodan, G. A., Different pattern of alkaline phosphatase,
osteopontin, and osteocalcin expression in developing rat bone visualized by in situ
hybridization, J. Bone Min. Res., 5, 831, 1990.
71. Sandberg, M., Autio-Harmainen, H., and Vuorio, E., Localization of the expression
of types I, Ill, and IV collagen, TGF-131 and c-fos genes in developing human calvarial
bones, Dev. Bioi., 130, 324, 1988.
72. Yoon, K., Buenaga, R., and Rodan, G. A., Tissue specificity and developmental
expression of osteopontin, Biochem. Biophys. Res. Comm., 148, 1129, 1987.
73. Scharla, S. H., Strong, D. D., Mohan, S., Baylink, D. J., and Linkhart, T. A.,
1,25-Dihydroxyvitamin D, differentially regulates the production of insulin-like growth
factor I (IGF-1) and IGF-binding protein-4 in mouse osteoblasts, Endocrinology. 129,
3139, 1991.
74. Kurose, H., Yamaoka, K., Okada, S., Nakajima, S., and Seino, Y., 1,25-Dihy-
droxyvitamin D, [1,25-(0H) 2D 3] increases insulin-like growth factor I (IGF-I) receptors
in clonal osteoblastic cells. Study on interaction of IGF-I and 1,25-(0H),D,, Endocri-
nology. 126, 2088, 1990.
75. Petkovich, P.M., Wrana, J, L., Grigoriadis, A. E., Heersche, J. N. M., and Sodek,
J., 1,25-Dihydroxyvitamin D3 increases epidermal growth factor receptors and trans-
forming growth factor 13-like activity in a bone-derived cell line, J. Bioi. Chern., 262,
13424, 1987.
76. Gronowicz, G., Egan, J. J., and Rodan, G. A., The effect of 1,25-dihydroxyvitamin
D, on the cytoskeleton of rat calvaria and rat osteosarcoma (ROS 17/2. 8) osteoblastic
cells, J. Bone Min. Res., I, 441, 1986.
77. Lomri, A. and Marie, P. J., Changes in cytoskeletal proteins in response to parathyroid
hormone and I ,25-dihydroxyvitamin D in human osteoblastic cells, Bone Min., 10, I,
1990.
78. Hong, M. H., Jin, C. H., Toshiyuki, S., Ishimi, Y., Abe, E., and Suda, T., Tran-
scriptional regulation of the production of the third component of complement (C3) by
lcx,25-dihydroxyvitamin D 3 in mouse marrow-derived stromal cells (ST2) and primary
osteoblastic cells, Endocrinology, 129, 2774, 1991.
79. Shull, S., Tracy, R. P., and Mann, K. G., Identification of a vitamin D-responsive
protein on the surface of human osteosarcoma cells, Proc. Nat/. Acad. Sci. U.S.A., 86,
5405, 1989.
80. Majeska, R. J, and Rodan, G. A., The effect of 1,25(0H),D3 on alkaline phosphatase
in osteoblastic osteosarcoma cells, J. Bioi. Chern., 257, 3362, 1982.
81. Mulkins, M.A., Manolagas, S.C., Deftos, L. H., and Sussman, H. H., Dihydroxy-
vitamin D3 increases bone alkaline phosphatase enzyme levels in human osteogenic sar-
coma cells, J. Bioi. Chern., 258, 6219, 1983.
82. Murray, E. J., Murray, S. S., and Manolagas, S. C., Two-dimensional gel autora-
diographic analyses of the effects of 1.25-dihydroxycholecalciferol on protein synthesis
in clonal rat osteosarcoma cells, Endocrinology, 126, 2679, 1990.
83. Harrison, J. R., Petersen, D. N., Lichtler, A. C., Mador, A. T., Rowe, D. W., and
Kream, B. E., I ,25-Dihydroxyvitamin D3 inhibits transcription of type I collagen genes
in the rat osteosarcoma cell line ROS 17/2.8, Endocrinology, 125, 327, 1989.
84. Noda, M., Vogel, R. L., Craig, A. M., Prahl, J., DeLuca, H. F., and Denhardt,
D. T., Identification of a DNA sequence responsible for binding of the I ,25-dihydroxy-
vitamin D 3 receptor and I ,25-dihydroxyvitamin D, enhancement of mouse secreted phos-
phoprotein I (Spp-1 or osteopontin) gene expression, Proc. Nat/. Acad. Sci. U.S.A.,
87, 9995, 1990.
85. Kerner, S. A., Scott, R. A., and Pike, J, W., Sequence elements in the human osteo-
calcin gene confer basal activation and inducible response to hormonal vitamin D 3 , Proc.
Nat/. Acad. Sci. U.S.A., 86, 4455, 1989.
86. Lian, J. B., Stewart, C., Puchacz, E., Mackowiak, S., Shalhoub, V., Collart, D.,
Zambetti, G., and Stein, G., Structure of the rat osteocalcin gene and regulation of
vitamin D-dependent expression, Proc. Nat/. Acad. Sci. U.S.A., 86, 1143, 1989.
87. Demay, M. B., Gerardi, J. M., DeLuca, H. F., and Kronenberg, H. M., DNA
sequences in the rat osteocalcin gene that bind the I ,25-dihydroxyvitamin D, receptor
and confer responsive to I ,25-dihydroxyvitamin D,, Proc. Nat/. Acad. Sci. U.S.A., 87,
369, 1990.
88. Spiess, Y. H., Price, P. A., Deftos, J, L., and Manolagas, S.C., Phenotype-associated
changes in the effects of I ,25-dihydroxyvitamin D3 on alkaline phosphatase and bone
GLA-protein of rat osteoblastic cells, Endocrinology, 118, 1340, 1986.
Lian and Stein 425
89. Kim, Y. S., Birge, S. J,, Avioli, L. V., and Miller, R., Cell density-dependent vitamin
D effects on calcium accumulation in rat osteogenic sarcoma cells (ROS 17/2), Calcif.
Tissue Int., 41, 218, 1987.
90. Canalis, E. and Lian, J, B., I ,25-dihydroxyvitamin 0 3 effects on collagen and DNA
synthesis in periosteum and periosteum-free calvaria, Bone, 6, 457, 1985.
91. Franceschi, R. T. and Young, J,, Regulation of alkaline phosphatase by I ,25-dihy-
droxyvitamin D3 and ascorbic acid in bone-derived cells, J. Bone Min. Res., 5, 1157,
1990.
92. Kyeyune-Nyombi, E., Lao, W. K.-H., Baylink, D. J., and Strong, D. D., Stimulation
of cellular alkaline phosphatase activity and its messenger RNA level in a human osteo-
sarcoma cell line by I ,25-dihydroxyvitamin D3 , Arch. Biochem. Biophys., 275, 363,
1989.
93. Canalis, E., Effect of hormones and growth factors on alkaline phosphatase activity and
collagen synthesis in cultured rat calvariae, Metabolism, 32, 14, 1983.
94. Markose, E. R., Stein, J, L., Stein, G. S., and Lian, J, B., Vitamin D-mediated
modifications in protein-DNA interactions at two promoter elements of the osteocalcin
gene, Proc. Nat/. Acad. Sci. U.S.A., 87, 1701, 1990.
95. Pan, L. and Price, P. A., Ligand-dependent regulation of the I ,25-dihydroxyvitamin
0 3 receptor in rat osteosarcoma cells, J. Bioi. Chern., 262, 4670, 1987.
96. van Leeuwen, J, P. T. M., Birkenhager, J, C., Boorman, C. J,, Schilte, J, P., and
Pols, H. A. P., Functional involvement of calcium in the homologous up-regulation of
the 1,25-dihydroxyvitamin D, receptor in osteoblast-like cells, FEBS, 270, 165, 1990.
97. Baran, D. T., Sorensen, A. M., Honeyman, T. W., Ray, R., and Holick, M. F.,
la,25-dihydroxyvitamin D 3-induced increments in hepatocyte cytosolic calcium and ly-
sophosphatidylinositol: inhibition by pertussis toxin and la,25-dihydroxyvitamin D,, J.
Bone Min. Res., 5, 517, 1990.
98. Leiberherr, M., Effects of vitamin D3 metabolites on cytosolic free calcium in confluent
mouse osteoblasts, J. Bioi. Chern., 262, 13168, 1987.
99. Caffrey, J, M. and Farach-Carson, M. C., Vitamin D, metabolites modulate dihy-
dropyridine-sensitive calcium currents in clonal rat osteosarcoma cells, J. Bioi. Chern.,
264, 20265, 1989.
100. Nemere, I. and Norman, A. W., Rapid action of 1,25-dihydroxyvitamin D, on calcium
transport in perfused chick duodenum: effect of inhibitors, J. Bone Min. Res., 2, 99,
1987.
101. Farach-Carson, M. C., Sergeev, I., and Norman, A. W., Nongenomic actions of
I ,25-dihydroxyvitamin D, in rat osteosarcoma cells: structure-function studies using li-
gand analogs, Endocrinology, 129, 1876, 1991.
102. Gill, R. K. and Christakos, S., Identification of sequence elements in the mouse
calbindin-D28 k gene which confer basal activation and hormone inducible response, in
Vitamin D Gene Regulation Structure-Function Analysis and Clinical Application, Nor-
man, A. W., Bouillon, R., and Thomasset, M., Eds., Walter de Gruyter, Berlin, 1991,
36.
103. Darwish, H. M. and DeLuca, H. F., Identification of a I ,25-dihydroxyvitamin D,-
response element in the 5'-flanking region of the rat calbindin D-9k gene, Proc. Natl.
Acad. Sci. U.S.A., in press.
104. Morrison, N. A., Shine, J,, Fragonas, J.-C., Verkest, V., McMenemy, L., and
Eisman, J, A., 1,25-Dihydroxyvitamin D-responsive element and glucocorticoid repres-
sion in the osteocalcin gene, Science, 246, 1158, 1989.
105. Burger, E. H., Van Der Meer, J, W. M., Van De Gevel, J, S., Gribnau, J, C.
Thesingh, C. W., and Van Furth, R., In vitro formation of osteoclasis from long-term
culture of bone marrow mononuclear phagocytes, J. Exp. Med., 156, 1604, 1982.
106. Walker, D. G., Control of bone resorption by haematopoietic tissue. The induction and
reversal of congenital osteopetrosis in mice through use of bone marrow and splenic
transplants, J. Exp. Med., 142, 651, 1975.
107. Bar-Shavit, Z., Teitelbaum, S. L., Reitsma, P., Hall, A., Pegg, L. E., Trial, J,, and
Kahn, A., Induction of monocytic differentiation and bone resorption by I ,25-dihy-
droxyvitamin 0 3 , Proc. Nat/. Acad. Sci. U.S.A., 80, 5907, 1983.
108. Holtrop, M. E. and Raisz, L. G., Comparison of the effects of I ,25-dihydroxychole-
calciferol, prostaglandin E2 and osteoclast-activating factor with parathyroid hormone on
the ultrastructure of osteoclasis in cultured long bones of fetal rats, Calc. Tissue Int.,
29, 201, 1979.
109. Takahashi, N., Akatsu, T., Sasaki, T., Yamaguchi, A., Kodama, H., Matin, T. J,,
and Soda, T., Osteoblastic cells are involved in osteoclast formation, Endocrinology,
123, 2600, 1988.
110. Udagawa, N., Takahashi, N., Akatsu, T., Sasaki, T., Yamaguchi, A., Kodama, H.,
Matin, T. J,, and Soda, T., The bone marrow-derived stromal cell lines MC3T3-G2/
PA6 and ST2 support osteoclast-like cell differentiation in co-culture with mouse spleen
cells, Endocrinology, 125, 1805, 1989.
Ill. Franceschi, R. T., Romano, P.R., and Park, K.-Y., Regulation of type I collagen
synthesis by I ,25-dihydroxyvitamin 0 3 in human osteosarcoma cells, J. Bioi. Chern.,
263, 18938, 1988.
112. Lian, J, B., Couttes, M. C., and Canalis, E., Studies of hormonal regulation of
osteocalcin synthesis in cultured fetal rat calvariae, J. Bioi. Chern., 60, 8706, 1985.
113. Aronow, M.A., Owen, T. A., Stein, G. S., and Lian, J. B., Estrogen inhibition of
osteocalcin mRNA expression occurs in cultured rat osteoblasts only after formation of
a mineralized extracellular matrix, J. Bone Min. Res., 5, S273, 1990.
114. Turner, R. T., Colvard, D. S., and Spelsberg, T. C., Estrogen inhibition of periosteal
bone formation in rat long bones: down-regulation of gene expression for bone matrix
proteins, Endocrinology, 127, 1346, 1990.
115. Schepmoes, G., Breen, E., Owen, T. A., Aronow, M. A., Stein, G. S., and Lian,
J, B., Influence of dexamethasone on the vitamin D-mediated regulation of osteocalcin
gene expression, J. Cell. Biochern., 47, 184, 1991.
116. Schiile, R., Umesono, K., Mangelsdorf, D. J., Bolado, J,, Pike, J, W., and Evans,
R. M., Jun-Fos and receptors for vitamins A and D recognize a common response element
in the human osteocalcin gene, Cell, 61, 497, 1990.
117. Nishimoto, S. K., Salka, C., and Nimni, M. E., Retinoic acid and glucocorticoids
enhance the effect of I ,25-dihydroxyvitamin 0 3 on bone 'Y-carboxyglutamic acid protein
synthesis by rat osteosarcoma cells, J. Bone Min. Res., 2, 571, 1987.
118. Noda, M., Transcriptional regulation of osteocalcin production by transforming growth
factor-[3 in rat osteoblast-like cells, Endocrinology, 124, 612, 1989.
119. Noda, M., Yoon, K., and Rodan, G. A., Cyclic AMP-mediated stabilization ofosteo-
calcin mRNA in rat osteoblast-like cells treated with parathyroid hormone, J. Bioi. Chern.,
263, 18574, 1988.
120. Theofan, G. and Price, P. A., Bone Gla protein messenger ribonucleic acid is regulated
by both 1,25-dihydroxyvitamin 0 3 and 3',5'-cyclic adenosine monophosphate in rat os-
teosarcoma cells, Mol. Endocrinol., 3, 36, 1989.
121. Stromstedt, P. E., Poellinger, L., Gustaffson, J, A., and Cralstedt-Duke, J,, The
glucocorticoid receptor binds to a sequence overlapping the TATA box of the human
osteocalcin promoter: a potential mechanism for negative regulation, Mol. and Cell Bioi.,
II, 3379, 1991.
122. Nanes, M.S., Rubin, J,, Titus, L., Hendy, G. N. and Catherwood, B. D., Interferon-
'/ inhibits I ,25-dihydroxyvitamin 0 3-stimulated synthesis of bone Gla protein in rat os-
teosarcoma cells by a pretranslational mechanism, Endocrinology, 127, 588, 1990.
123. Stein, G. S., Lian, J. B., Dworetzky, S. I., Owen, T. A., Bortell, R., Bidwell, J, P.,
and van Wijnen, A. J,, Regulation of transcription-factor activity during growth and
differentiation: involvement of the nuclear matrix in concentration and localization of
promoter binding proteins, J. Cell. Biochern., 47, 300, 1991.
Lian and Stein 427
124. Bortell, R., Owen, T. A., van Wijnen, A. J., Bidwell, J, P., Gavazzo, P., Breen,
E., DeLuca, H., Stein, G. S., and Lian, J, B., Vitamin D responsive protein/DNA
interactions at multiple promoter regulatory elements contribute to the level of osteocalcin
gene expression, Proc. Natl. Acad. Sci. U.S.A., 1992.
125. Terpening, C. M., Haussler, C. A., Jurutka, P. W., Galligan, M.A., Komm, B. S.,
and Haussler, M. R., The vitamin D-responsive element in the rat bone Gla protein
gene is an imperfect direct repeat that cooperates with other cis-elements in 1,25-dihy-
droxyvitamin D,-mediated transcriptional activation, Mol. Endocrinol., 5, 373, 1991.
126. Schiile, R., Rangarajan, P., Kliewer, S., Ransome, L. J,, Bolado, J,, Yang, N.,
Verma, I. M., and Evans, R. M., Functional antagonism between oncoprotein c-Jun
and the glucocorticoid receptor, Cell, 62, 1217, 1990b.
127. Owen, T. A., Bortell, R., Yocum, S. A., Smock, S. L., Zhang, M., Abate, C.,
Shalhoub, V., Aronin, N., Wright, K. L., van Wijnen, A. J,, Stein, J, L., Curran,
T., Lian, J. B., and Stein, G. S., Coordinate occupancy of AP-1 sites in the vitamin
D responsive and CCAAT box elements by fos-jun in the osteocalcin gene: a model for
phenotype suppression of transcription, Proc. Natl. Acad. Sci. U.S.A., 87, 9990, 1990.
128. Ozono, K., Liao, J., Kerner, S. A., Scott, R. A., and Pike, J, W., The vitamin D-
responsive element in the human osteocalcin gene, J. Bioi. Chern., 265, 21881, 1990.
129. Ross, T. K., Moss, V. E., Prahl, J. M., and DeLuca, H. F., A nuclear protein essential
for binding of rat 1,25-dihydroxyvitamin D3 receptor to its response elements, Proc.
Natl. Acad. Sci. U.S.A., in press.
130. Brown, T. A. and DeLuca, H. F., Phosphorylation of the I ,25-dihydroxyvitamin D3
receptor. A primary event in I ,25-dihydroxyvitamin D3 action, J. Bioi. Chern., 265,
10025, 1990.
131. Hsieh, J.-C., Jurutka, P. W., Galligan, M.A., Terpening, C. M., Haussler, C. A.,
Samuels, D. S., Shimizu, Y., Shimizu, N., and Haussler, M. R., Human vitamin D
receptor is selectively phosphorylated by protein kinase C on serine 51, a residue crucial
to its trans-activation function, Proc. Nat{. Acad. Sci. U.S.A., 88, 9315, 1991.
132. Lian, J, B., Stein, G. S., Bortell, R., and Owen, T. A., Phenotype suppression: a
postulated molecular mechanism for mediating the relationship of proliferation and dif-
ferentiation by fos/jun interactions at AP-I sites in steroid responsive promoter elements
of tissue-specific genes, J. Cell. Biochem., 45, 9, 1991.
133. Candeliere, G. A., Prud'homme, J,, and St-Arnaud, R., Differential stimulation of
fos and jun family members by calcitriol in osteoblastic cells, Mol. Endocrinol., 12,
1780, 1991.
134. Diamond, M. I., Miner, J, N., Yoshinaga, S. K., and Yamamoto, K. R., Transcription
factor interactions: selectors of positive or negative regulation from a single DNA element,
Science, 249, 1266, 1990.
135. Lian, J, B. and Stein, G. S., Transcriptional control of vitamin D regulated proteins,
J. Cell. Biochem., 49, 1992.
136. Feldman, D. and Malloy, P. J., Hereditary 1,25-dihydroxyvitamin D resistant rickets:
molecular basis and implications for the role of I ,25(0H),D 3 in normal physiology, Mol.
Cell. Endocrinol., 72, C57, 1990.
137. Marx, S. J,, Liberman, U. A., Eil, C., Gamblin, G. T., Degrange, D. A., and
Balsan, S., Hereditary resistance to I ,25-dihydroxyvitamin D3 suppresses parathyroid
hormone secretion from bovine parathyroid cells in tissue culture, Endocrinology, 117,
2214, 1984.
138. Liberman, U. and Marx, S. J,, Vitamin D dependent rickets, in Primer on Metabolic
Bone Diseases and Disorders of Mineral Metabolism, Fauvus, M. J., Ed., American
Society of Bone and Mineral Research, Kelseyville, CA, 1990, 178.
139. Glorieux, F. H., Rickets, the continuing challenge, New Engl. J. Med., 325, 1875,
1991.
140. Marks, S. J., Jr., Osteopetrosis-multiple pathways for the interruption of osteoclast
functions, Appl. Pathol., 5, 172, 1987.
141. Shalhoub, V., Jackson, M. E., Lian, J. B., Stein, G. S., and Marks, S. C., Jr.,
Gene expression during skeletal development in three osteopetrotic rat mutations: evidence
for osteoblast abnormalities, J. Bioi. Chern., 266, 9847, 1991.
142. DeLuca, H. F., Newly discovered actions of 1,25-dihydroxyvitamin 0 3 and osteoporosis:
basis and treatment, Proc. Soc. Exp. Bioi. Med., 191, 1989.
143. Avioli, L. V., Significance of osteoporosis: a growing international health care problem,
Calcif. Tissue Int., 49, S5, 1991.
144. Eastell, R., Yergey, A. L., Vieira, N. E., Cedel, S. L., Kumar, R., and Riggs,
B. L., Interrelationship among vitamin D metabolism, true calcium absorption, parathy-
roid function, and age in women: evidence of an age-related intestinal resistance to I ,25-
dihydroxyvitamin A action, J. Bone Min. Res., 6, 125, 1991.
145. Franz, K. B., Abnormalities in parathyroid hormone secretion and I ,25-dihydroxyvi-
tamin 0 3 formation in women with osteoporosis, New Engl. J. Med., 320, 1697, 1989.
146. Silverberg, S. J., Shane, E., de Ia Cruz, L., Segre, G. V., Clemens, T. L., and
Bilezikian, J. P., Abnormalities in parathyroid hormone secretion and l ,25-dihydroxy-
vitamin 0 3 formation in women with osteoporosis, New Engl. J. Med., 320, 277, 1989.
147. Gallagher, J. C., Jerpbak, C. M., Lee, W. S. S., Johnson, K. A., DeLuca, H. F.,
and Riggs, B. L., 1,25-dihydroxyvitamin 0 3 : short- and long-term effects on bone and
calcium metabolism in patients with postmenopausal osteoporosis, Proc. Nat!. Acad. Sci.
U.S.A., 79, 3325, 1982.
148. Geusens, P., Vanderschueren, D., Verstraeten, A., Dequeker, J., Devos, P., and
Bouillon, R., Short-term course of I ,25(0H) 2 D, stimulates osteoblasts but not osteoclasis
in osteoporosis and osteoarthritis, Calcif. Tissue Int., 49, 168, 1991.
149. Tsai, K. S., Heath, H., Kumar, R., and Riggs, B. L., Impaired vitamin D metabolism
with aging in women. Possible role in pathogenesis of senile osteoporosis, J. Clin. Invest.,
73, 1668, 1984.
150. Gallagher, J. C. and Riggs, B. L., Action of 1,25-dihydroxyvitamin 0 3 on calcium
balance and bone turnover and its effect on vertebral fracture rate, Metabolism, 39, 30,
1990.
151. Aloia, J. F., Vaswani, A., Yeh, J. K., Ellis, K., Yasumura, S., and Cohn, S. H.,
Calcitriol in the treatment of postmenopausal osteoporosis, Am. J. Med., 84, 401, 1988.
152. Caniggia, A., Nuti, R., Galli, M., Lore, F., Turchetti, V., and Righi, G. A., Effect
of a long-term treatment with I ,25-dihydroxyvitamin 0 3 on osteocalcin in postmenopausal
osteoporosis, Calcif. Tissue Int., 38, 328, 1986.
153. Gallagher, J. C., Riggs, B. L., Recker, R. R., and Goldgar, D., The effect of calcitriol
on patients with postmenopausal osteoporosis with special reference to fracture frequency,
Proc. Soc. Exp. Bioi. Med., 191, 287, 1989.
154. Falch, J. A., Odegaard, 0. R., Finnanger, A.M., and Matheson, I., Postmenopausal
osteoporosis: no effect of three years treatment with 1,25-dihydroxycholecalciferol, Acta
Med. Scand., 221, 199, 1987.
155. Ott, S.M. and Chestnut, C. H., III, Calcitriol treatment is not effective in postmeno-
pausal osteoporosis, Ann. Int. Med., 110, 267, 1989.
156. Manolagas, S.C., Hustmyer, F. G., and Yu, Z.-P., Immunomodulating properties of
1,25-dihydroxyvitamin 0 3 , Kidney Int., 38, S9, 1990.
157. Lemire, J. M. and Archer, D. C., 1,25-dihydroxyvitamin D, prevents the in vivo
induction of murine experimental autoimmune encephalomyelitis, J. Clin. Invest., 87,
1103, 1991.
158. Kragballe, K., Gjertsen, B. T., De Hoop, D., Karlsmark, T., van de Kerkbof, P. C.,
Larko, 0., Nieboer, C., Roed-Peterson, J., Strand, A., and Tikjb, G., Double-blind,
right/left comparison of calcipotriol (MC903) and betamethasone valerate in treatment of
psoriasis vulgaris, Lancet, 337, 193, 1991.
Lian and Stein 429
!59. Okano, M., I a!pha,25-(0H) 2D 3 use on psoriasis and ichthyosis, Int. J. Dermatol., 30,
62, 1991.
160. Huckins, D., Felson, D. T., and Holick, M., Treatment of psoriatic arthritis with oral
1,25-dihydroxyvitamin 0 3 : a pilot study, Arthritis Rheum.. 33:1723-1727.
161. Ostrem, V. K., Tanaka, Y., Prahl, J., DeLuca, H. F., and Ikekawa, N., 24- and 26-
homo-1 ,25-dihydroxyvitarnin 0 3 : preferential activity in inducing differentiation of human
leukemic cells HL-60 in vitro, Proc. Nat/. Acad. Sci. U.S.A., 84, 2610, !987.
162. Paulson, S. K., Perlman, K., DeLuca, H. F., and Stern, P. H., 24- and 26-homo-
1,25-dihydroxyvitamin 0 3 analogs: potencies on in vitro bone resorption differ from those
reported for cell differentiation, J. Bone Min. Res., 5, 201, 1990.
163. Zhou, J.-Y., Norman, A. W., Lubbert, M., Collins, E. D., Uskokovik, M. R., and
Koeffler, H. P., Novel vitamin D analogs that modulate leukemic cell growth and
differentiation with little effect on either intestinal calcium absorption or bone calcium
mobilization, Blood, 74, 82, 1989.
164. Binderup, L. and Bramm, E., Effects of a novel vitamin D analog MC903 on cell
proliferation and differentiation in vitro and on calcium metabolism in vivo, Biochem.
Pharmacal., 37, 889, 1988.
165. Bouillon, R., Allewaert, K., Tan, B. K., van Baelen, H., de Clercq, P., and Vandewalle,
M., Biological activity of I ,25-dihydroxyvitamin D analogs: influence of relative affin-
ities for DBP and vitamin D receptor, in Vitamin D Gene Regulation Structure-Function
Analysis and Clinical Application, Norman, A. W., Bouillon, R., and Thomasset, M.,
Eds., Walter de Gruyter, Berlin, 1991, 131.
166. Evans, D. B., Thavarajah, M., Uskokovic, M. R., and Kanis, J. A., Increased potency
of I ,25-dihydroxyvitamin 0 3 on human osteoblast-like cells following structural side-
chain modification, Bone, II, 439, 1990.
167. Marie, P. J. Coones, D., Hott, M., and Miravet, L., Comparative effects of a novel
vitamin D analogue MC-903 and I ,25-dihydroxyvitamin 0 3 on alkaline phosphatase
activity, osteocalcin and DNA synthesis by human osteoblastic cells in culture, Bone,
II, 171, !990.
168. Valaja, T., Mahonen, A., Pirskanen, A., and Maenpaa, P. H., Affinity of 22-oxa-
1,25(0H) 2D3 for I ,25-dihydroxyvitarnin D receptor and its effects on the synthesis of
osteocalcin in human osteosarcoma cells, Biochem. Biophys. Res. Commun .. 169, 629,
1990.
169. Pernalete, N., Mori, T., Nishii, Y., Slatopolsky, E., and Brown, A. J., The activity
of22-oxacalcitriol in osteoblast-like (ROS 17/2.8) cells, Endocrinology, 129,778, 1991.
170. Tanaka, Y. and DeLuca, H. F., 26,26,26,27,27,27 -hexafluoro-1 ,25-dihydroxyvitamin
0 3 : a highly potent, long-lasting analog of 1,25-dihydroxyvitamin 0 3 , Arch. Biochem.
Biophys., 229, 348, 1984.
171. Kiriyama, T., Sumiaki, 0., Ejima, E., Kurihara, N., Hakeda, Y., Ito, N., Izumi,
M., Kumegawa, M., and Nagataki, S., Effect of a highly potent fluoro analog of I ,25-
dihydroxyvitamin 0 3 on human bone-derived cells, Endocrinology, 128, 81, 1991.
172. lnaba, M., Okuno, S., Inoue, A., Nishizawa, Y., Morii, H., and DeLuca, H. F.,
DNA binding property of vitamin 0 3 receptors associated with 26,26,26,27,27,27-hex-
afluoro-I,25(0H)2D3, Arch. Biochem. Biophys.. 268, 35, 1989.
173. Heinrichs, A., Stein, J., Lian, J. B. and Stein, G. S., unpublished observations.
174. Lian, J. B. and Stein, G. S., unpublished observations.
175. Lian, J. B., Stein, G. S., and Uskokovic, M., unpublished observation.
176. Tilyard, M. W., Spears, G. F. S., Thomson, J., and Dovey, S., Treatment of post-
menopausal osteoporosis with calcitriol or calcium, N. Engl. J. Med., 326, 357, 1992.
Chapter 19
RETINOIC ACID REGULATION OF GENE

EXPRESSION IN ADIPOCYTES
Donald B. Jump, Gerald J. Lepar, and Ormond A. MacDougald
TABLE OF CONTENTS
I. Introduction ...................................................... 432
II. Regulation of Expression of the S 14 Protein in Cultured

Adipocytes ....................................................... 434
III. Effects of RA on Adipocyte Gene Expression ................... 437
IV. RAR Expression in Cultured Adipocytes ......................... 438
V. Identification of Prospective RA and Glucocorticoid

Response Cis-Acting Elements ................................... 439
VI. Role of Tissue-Specific Factors in RA and DEX Control

of Sl4 Gene Expression ......................................... 440
VII. Model for the Tissue-Specific, Hormone, and Nutrient

Controlled Expression of the S 14 Gene in Cultured
Adipocytes ....................................................... 444
VIII. Summary: Does RA Function in Intermediary

Metabolism? ..................................................... 447
References ............................................................... 448
0·8493·6961-4/93/$0.00 +$.50
I. INTRODUCTION
Retinol and retinoic acid have profound effects on a wide range of phys-
iological and biochemical processes. 1-11 For example, the aldehyde derivative
of retinol (11-cis-retinal) functions in vision. 5 Retinol is converted to retinoic
acid (RA) in multiple tissues, 2-4 and RA has profound effects on pattern
formation during early development, skeletal growth, cellular growth, and
differentiation. 3.4. 6-11 The diverse effects of RA are due to its regulation of
the abundance and/or activity of specific proteins. As shown in Table 1, these
proteins include hormones and growth factors, 12-14 cell surface and nuclear
receptors, 15 -17·19 ·22 -26 and nuclear trans-acting factors, 18 ·20 ·21 specific en-
zymes/8·27-33 plasma proteins, 38 ·39 ·42 and proteins involved in cellular archi-
tecture and other cellular functions. 34-36 .4°·43 -47
RA regulates cellular functions by binding to intracellular retinoic acid
receptors (RR). 48 -57 Two distinct families of RR have been identified, i.e.,
RAR and RXR, each with multiple isoforms. The RAR family consists of
RARa and RAR[3, which are found in many tissues, and RAR')', which occurs
predominantly in skin and lung. 53 The RXR family consists of an a, [3, and
')' isoforms and RXR')' is found mainly in liver, lung, and muscle. 54 Even
though there is low sequence homology between RARs and RXRs, each binds
RA and DNA. Tissues like liver and white adipose tissue express multiple
forms of the RRs. 4H,4 9·55 ·56 The physiological significance of multiple receptor
isoform expression in tissues has not been explained.
Both RARs and RXRs are structurally similar to steroid and thyroid
hormone receptors (TR). 48 -54 Both function as ligand-activated transcription
factors in a manner similar to the regulation of gene expression described for
steroid and thyroid hormones. 51 RRs contain Zn-finger protein sequence mo-
tifs which function in receptor binding to DNA, at retinoic acid response
elements (RARE). The carboxyl terminal end of the molecule functions in
ligand binding. In contrast to steroid receptors, RR do not interact with heat
shock proteins, like HSP90, 57 proteins which prevent steroid receptor binding
to DNA. These observations suggest that RR may function more like TRs.
If analogous to TRs, RRs may be associated with the nuclear compartment
in the absence of ligand. RA binding to nuclear receptors sets in motion a
sequence of events that culminate in a change in transcription of the cis-
linked gene. Such transcriptional changes lead to corresponding changes in
mRNA coding for specific proteins and eventual changes in cell function.
The DNA targets for RR binding, i.e., RARE, have been the subject of
intense investigation. Umesono et al. 58 first reported that the thyroid hormone
response element (TRE) located upstream from the rat pituitary growth hor-
mone (GH) gene was a target for RAR binding and action. The GH TRE
used in these studies was palindromic and represented a modified version of
the TRE identified in the GH gene. The notion that RAR might utilize TREs
was reinforced by the finding that RA treatment of cultured pituitary cells
Jump et al. 433
TABLE 1
Cellular Proteins Regulated by Retinoic Acid
Induced (I) or
Protein Cellitissue repressed (R) Ref.
Hormones/Growth Factors
Growth Hormone Pituitary cells 12, 13

TGF[32 Keratinocytes 14
Receptors/Nuclear Trans-Acting Factors
I ,25-(0H),D, Receptors Rat osteosarcoma Ros 17/2 15

RAR[3 HepG2 Hepatoma 16, 17
c-Fos Rheumatoid synoviocytes R 18
Progesterone receptors T-470 Breast cancer R 19
Zif268 transcription factor Rat fetal calvaria 20
AP-2 Transcription factor NT2 Teratocarcinoma 21
MSH Receptors Cloudman S91 melanoma 22
lnterleukin-6 receptors Myeloma R 23
lnterleukin-2 receptors Thymocyte 24
EGF receptors Corneal endothelium 25
EGF receptors Corneal epithelium R 26
Enzymes
Trans glutaminase Macrophages 27

Alkaline phosphatase Rat osteosarcoma Ros 17/2.8 R 28
Collagenase Rheumatoid synoviocytes R 18
Phosphoenolpyruvate carboxykinase H411E Hepatoma 29
G,a F9 Teratocarcinoma 30
Alcohol dehydrogenase Hep3B hepatoma 31
!-Plasminogen activator F9 Teratocarcinoma 32
Glycerophosphate 3T3-F442A Adipocytes 33
Dehydrogenase
Other Proteins
[3, Laminin F9 Teratocarcinoma 34-36

CRBP and CRABP Pl6 Embryonal carcinoma 37
Complement factor H L929 Fibroblast 38
Adipsin 3T3-F442A Adipocytes R 39
Keratins, K5,K6, Kl4 Keratinocytes R 40
Loricrin Keratinocytes R 41
Apolipoprotein AI HepG2 hepatoma I 42
Neurofilament (NF-M) PCI2 Neuroblastoma 43
GAP-43 PCI2 Neuroblast 43
Collagen type X Chick chondrocytes 44
Sl4 Protein 3T3-F442A Adipocytes 33
Collagen F9 Teratocarcinoma 45
Fibronectin F9 Teratocarcinoma 46
Vimentin F9 Teratocarcinoma 47
elevated GH protein production, mRNAGH levels and GH gene transcrip-

tion.12·13 Subsequent reports showed that TR and RR formed heterodimers
and bound to the GH TRE to induce GH gene transcription in an additive
fashion. 13.62,63
The finding that RA, as well as triiodothyronine (T3), regulated the GH
TRE suggested that other T 3-regulated genes might be targets for RA control
or that RA-regulated genes might be affected by T 3. When similar studies
were carried out using the myosin heavy chain TRE, the formation of het-
erodimers inhibited the T3 activation of myosin heavy chain gene transcrip-
tion.62 Conversely, RA-regulated genes, like RARj31 6·17 and f31-laminin, 36
were not regulated by T3. These results showed that structural features within
response elements determine receptor binding.
The structure of response elements for TR, RAR, RXR, and vitamin D3
receptors might be direct repeats rather than palindromic. 59-61 Each of the four
receptors could recognize the same sequence, i.e., AGGTCA, when it was
present in DNA as a direct repeat. However, receptor binding specificity was
determined by the nucleotide spacing between the repeats. For example,
RXR would bind to AGGTCA when one nucleotide separated the repeats, 60
while vitamin D3 receptors, T 3 receptors, and RARs bound to DNA when the
spacing between repeats was 3, 4, and 5 nucleotides, respectively. 59-60
The basic components of the RA regulatory system are now being defined.
However, recent studies in our laboratory suggest that there are additional
cellular factors which contribute to the RA regulation of gene expression. 33
This review will show the evolution of thought that led us to investigate the
role RA plays in adipocyte gene expression. We will begin with an exami-
nation of the deficient T 3 control of S 14 gene expression in 3T3-F442A cells
and progress to an analysis of RA action on S14 gene transcription in these
cells. We will show that glucocorticoids and tissue-specific factors contribute
significantly to the RA-mediated control of S 14 gene transcription in cultured
adipocytes.
II. REGULATION OF EXPRESSION OF THE S14

PROTEIN IN CULTURED ADIPOCYTES
The Sl4 protein (17 kDa; 4.9 pi) is expressed predominantly in tissues
which synthesize triglycerides such as liver, lactating mammary gland, white
and brown adipose tissues, 64- 72 and cultured adipocytes. 73 In rat liver, tran-
scription of the S14 gene is regulated by hormones, 67 -70 dietary, 69 ·74-76 devel-
opmental, 77 and tissue-specific factors. 64 •66 ·67 Hepatic S 14 gene transcription
was rapidly stimulated by T3,68 insulin, 70 dietary carbohydrate69 ·70 and sup-
pressed by dietary fat 74-76 and factors which elevated cellular levels of cAMP. 70
The overall pattern of control of S 14 gene expression is typical of many
proteins involved in lipid and carbohydrate metabolism. 78 For these reasons,
the S14 gene can be used to study the tissue-specific, hormone, and nutrient
control of gene transcription.
Jump et al. 435
TABLE 2
T 3 Binding Capacity in Rat Liver,
3T3-F442A, and 3T3-Ll Adipocytes
Tissue/cells T 3 Binding capacity•

(#TR/nucleus)
Rat liver 3273 ± 299

3T3-F442A Adipocytes 420 ± 150
3T3-LI Adipocytes 2773 ± 680
' T 3 binding capacity was measured in isolated rat

liver or adipocyte nuclei as described. 73 •83 Results
are expressed as the number ofT 3 receptors (TR)/
nucleus; mean ± SD. Number of determinations
ranged from 3 to 5. Affinity of T 3 binding varied
<twofold.
We were interested in developing a cell culture model to study both the

hormonal and tissue-specific control of S14 gene transcription. The 3T3-
F442A cell line was selected because of the purported role of the S 14 protein
in lipid metabolism. 67·71 When F442A cells differentiate from preadipocytes
(fibroblast) to adipocytes, 79 •80 there is a distinct change in morphology ac-
companied by: (1) an engorgement of triglyceride; (2) induction of enzymes
involved in fatty acid and triglyceride synthesis; and (3) acquisition of re-
sponsiveness to both lipogenic and lipolytic hormones. 79- 82 Thus, the 3T3-
F442A cell line appeared to be well suited to study both the cellular and
molecular mechanisms of fat cell differentiation and hormonal regulation of
genes involved in lipid metabolism.
Our initial studies on S 14 gene expression in cultured adipocytes revealed
two surprising findings. First, mRNA 514 was expressed in F442A adipocytes
at vastly reduced levels (:S0.005 units) when compared to mouse liver or
white adipose tissue (WAT) (2:0.3 units). 73 We would have expected mRNA 514
to be much higher if the S 14 protein played a critical role in triglyceride
synthesis. Second, while glucocorticoids rapidly induced mRNA 514 both in
vivo and in F442A adipocytes, T3 had no effect on S14 gene expression in
F442A preadipocytes or adipocytes. 73 T 3 rapidly induced S14 gene expression
in both mouse liver and WAT. 73
Measurement ofT 3 receptor binding capacity in isolated F442A adipocyte
nuclei showed that TR were expressed at 420 ± 150 TR per nucleus (Table
2), representing an 85% reduction from the T 3 binding capacity when com-
pared to rat liver. Since levels of TR expression in WAT are comparable to
liver, 84 the low T 3 binding capacity indicated that F442A cells do not share
the same phenotypic features as adipocytes in adult WAT.
Hausdorf et al. 85 and Moustaid and Sul 86 reported that S14 and fatty acid
synthase (FAS) were induced by T3 in 3T3-Ll adipocytes, respectively. 3T3-
Liver WAT F442A L1

TR alpha 1 -.
erbA alpha 2 -.
TR Beta 1 _.
FIGURE I. T, receptor isofonn analysis in mouse liver, white adipose tissue, 3T3-F442A,
and 3T3-LI adipocytes. Total RNA (25 f.Lg) from mouse liver, epididymal fat (WAT), 3T3-
F442A adipocytes (F442A), and 3T3-LI adipocytes (LI) was electrophoretically separated on
I% agarose-2.2 M formaldehyde gels and transferred to nitrocellulose. 33 Northern blots were
hybridized with eDNA inserts derived from rc-erbAa or rc-erbAI3 plasmids (generously provided
by H. Towle, Univ. of Minnesota). The rc-erbAa hybridized to two mRNAs: TRal, 5 kb;
c-erbA-2, 2.6 kb, while the rc-erbAI3 probe hybridizes to I mRNA of 6.2 kb 87 Duplicate
samples were electrophoresed for WAT, F442A, and Ll adipocytes. Rat liver ISS and 28S
ribosomal RNAs served as standards and were visualized by ethidium bromide staining.
Ll cells were also cloned from mouse embryonic fibroblast. 79 We have found
that following a 72 h treatment of Ll adipocytes with T 3 , mRNA 514 was
induced fivefold. Analysis of T 3 binding capacity in these cells showed that
TR were expressed at 70% of the level observed in rat liver (Table 2). Thus,
different adipocyte cell lines display different T 3 regulatory profiles.
Molecular cloning studies have shown that at least three ligand-binding
isoforms of TR, i.e., c-erbAai (TRal), c-erbAJ31 (TRJ31), and c-erbAJ32
(TRJ32), are expressed in various rodent tissues. 87 - 92 While TRJ32 is pituitary
specific, TRal and TRJ31 are expressed in several tissues. 87•88 In addition,
three variants of these proteins (i.e., c-erba2, c-erba3, and rev-erbAa) have
been described which do not bind T 3 . 88 -93 Northern analysis was used to
examine the TR isoforms expressed in mouse liver and WAT, and 3T3-F442A
and 3T3-Ll adipocytes (Figure 1). A rat liver c-erbAa eDNA hybridized to
two mRNAs in liver, WAT, and both adipocyte cell lines. The 5-kb mRNA
codes for TRal, while the 2.5-kb mRNA codes for c-erbAa2. The hybrid-
ization of rc-erbAa to liver RNA is weak when compared to the hybridization
signal in WAT, F442A, and Ll. TRJ31 (6.2 kb) was expressed in both liver
and WAT, but not in either adipocyte cell lines. The relative abundance of
TRa I, c-erbAa2, and TRJ31 in liver and Ll cells agrees favorably with
previous reports. 87
The differential T 3 responsiveness of F442A and Ll is due, at least in
part, to differences in T 3 binding capacity. Ifthese cells are "true" progenitors
to adipocytes in WAT, then the absence of TRJ31 expression in F442A and
Jump et al. 437
TABLE 3
Dexamethasone and Retinoic Acid Regulation of mRNA814 Levels in
3T3-F442A and 3T3-Ll Cells•
3T3-F442A 3T3-Ll
Treatment Units Fold Units Fold
Vehicle 0.005 ± 0.001 0.004 ± 0.002

Retinoic acid 0.070 ± 0.013 14 0.160 ± 0.120 40
Dexamethasone 0.740 ± 0.013 148 0.240 ± 0.060 60
Dexamethasone + retinoic acid 1.800 ± 0.220 360 0.798 ± 0.079 200
' Adipocytes were exposed to I f.LM dexamethasone, retinoic acid, or a combination of treat-
ments for 72 h. 33 Control cultures were exposed to vehicle (0.1% ethanol) for 72 h. RNA
was extracted and analyzed for expression of mRNA 514 . Results are expressed as mean +
SD. Number of determinations ranged from 6 to 12 for each treatment.
Ll, and the low T 3 binding capacity in F442A cells, suggests that the T3
regulatory system in adipocyte precursors might be under developmental con-
trol. The F442A and Ll cells might represent early to intermediate stages in
development, and thus they may not be satisfactory models for T3 action.
III. EFFECTS OF RA ON ADIPOCYTE GENE

EXPRESSION
The studies on T3 regulation of S 14 gene expression in cultured adipocytes
show that embryo-derived cells do not reflect the T3 regulatory systems seen
in WAT derived from adult animals. Based on the role of RAin development
and differentiation 3·4·6 · 11 and the fact that RA regulates some T 3 responsive
genes, we speculated that RA might regulate genes involved in lipid metab-
olism in these cells.
RA blocks the differentiation of both F442A and Ll preadipocytes (fi-
broblasts) to adipocytes, 94-97 indicating that RA was functional in these cells.
We then determined whether RA regulated S14 gene expression in cultured
F442A and Ll adipocytes. Expression of the S14 gene in F442A cells was
dependent on the adipocyte phenotype and treatment of cells with glucocor-
ticoids.73 Thus, effects of RA on adipocyte S14 gene expression was deter-
mined in the presence and absence of the glucocorticoid agonist, dexameth-
asone (DEX) (Table 3). Treatment of F442A and Ll adipocytes with RA for
72 h induced mRNA 814 14- and 40-fold, respectively. DEX induced a 148-
and 60-fold increase in mRNA 814 abundance in F442A and Ll adipocytes,
respectively. The combination of both treatments induced mRNA 514 360- and
200-fold in F442A and Ll adipocytes, respectively. While RA and DEX
promoted S14 gene expression to different levels in the F442A and Ll cells,
the induction pattern showed that both factors regulated S 14 gene expression
through independent pathways in each cell type.
Further investigation showed that RA and DEX interacted at the tran-

scriptional level to regulate S 14 gene expression in F442A adipocytes _33
Treatment of F442A adipocytes with 0.1 nM DEX or 1 JJM RA for 72 h
induced a fivefold increase in both mRNA514 and S14 gene transcription.
Following treatment of adipocytes with both DEX (0.1 nM) and RA (1 fLM),
S14 gene transcription and mRNA 514 levels increased by 27- and 86-fold,
respectively. When DEX was elevated to 1 jJM in the presence of RA, Sl4
gene transcription and mRNA 514 levels rose by 87- and 262-fold, respectively.
Thus, RA and DEX synergistically regulate Sl4 gene expression. In the
presence of 0.1 nM DEX, RA effects were dose-dependent (ED 50 = 5 nM)
and rapid (50-fold within 4 h). This induction pattern indicates that the prin-
cipal site for both RA and DEX action on S14 gene expression is at the level
of S 14 gene transcription. The synergistic regulation of S 14 gene transcription
suggests that RA may play an important modulatory role in DEX-mediated
gene expression.
The specificity of RA action on adipocyte gene expression was determined
by measuring the effects of RA on the expression of other genes. Changes
in a glycerophosphate dehydrogenase (GPD), FAS, and 13-actin mRNA levels
and run-on activity were measured following a 72-h treatment of adipocytes
with RA. 33 RA had no effect on FAS or 13-actin gene expression, but sup-
pressed GPD gene transcription and mRNA 0 p 0 levels by 75%. 73 In addition,
others reported that RA suppressed adipsin gene expression through post-
transcriptional regulatory mechanisms. 37 The lack of any significant effect of
RA on 13-actin or FAS indicated that RA does not induce generalized changes
in adipocyte gene expression. More importantly, this limited analysis hints
at a possible role for RA in modulating triglyceride synthesis in fat cells.
IV. RAR EXPRESSION IN CULTURED ADIPOCYTES
RA clearly regulates gene expression in both 3T3-F442A and 3T3-Ll

adipocytes. In order to identify which RAR might mediate these changes in
gene expression, a and 13 RAR cDNAs were used as hybridization probes
for electrophoretically separated RNA. RARa was expressed as 2 mRNAs
of 2.8 and 3.8 kb in both the L1 and F442A cells (Figure 2). No apparent
difference in the level of RARa mRNA was detected between Ll or F442A
cell lines or between the preadipocyte and adipocyte phenotypes. RARa was
expressed at higher levels in cultured adipocytes than in liver.
In contrast to the RARa mRNA expression, analysis of RARI3 mRNA
levels show no detectable RNA in any of the L1 or F442A cells, but there
were high levels of RARI3 mRNA in liver. Zelent et a!. 53 reported that RARI3
was more abundant than RARa in liver. We have not determined whether
other RAR isoforms, like RARy5 3 or RXR, 54 were expressed in F442A or
Ll cells. However, Haq and ChytiP6 recently reported that RARa, RARI3,
and RAR-y were all expressed in rat WAT. We have also found that both
Jump et al. 439
RAR Alpha
RAR Beta
FIGURE 2. Retinoic acid receptor isofonn analysis in mouse liver, 3T3-F442A, and 3T3-Ll
cells. Total mRNA (25 flog) isolated from mouse liver (lane I) and 3T3-F442A preadipocytes
(lanes 2,3), 3T3-F442A adipocytes (lanes 4,5), 3T3-Ll preadipocytes (lanes 6, 7), and 3T3-Ll
adipocytes (lanes 8,9) was electrophoretically separated in I% agarose-2.2 M formaldehyde
gels." Following electrophoresis RNA was transferred to nitrocellulose and hybridized with
either 32 P-labeled inserts derived from pRARaO (RAR alpha) or pRAR[3 (RAR beta) plasmids.
RARa hybridizes to two mRNA of 2.8 and 3.8 kb, while RAR[3 hybridizes to two mRNAs of
3.2 and 3.6 kb." Rat liver 18S and 28S ribosomal RNAs served as standards and were visualized
by ethidium bromide staining.
RARa and RARI3 were expressed in rat white epididymal fat (not shown).
If the mRNA levels coding for these various RAR isoforms reflect the level
of active receptor in these cells, then RARa is the predominant RAR isoform
expressed in both F442A and Ll cells. In addition, RARI3 is upregulated by
RA in some cell types. 16 • 17 We found no evidence for an effect of RA on
RARI3 expression in either preadipocytes or adipocytes (not shown).
V. IDENTIFICATION OF PROSPECTIVE RA
AND GLUCOCORTICOID RESPONSIVE
CIS-ACTING ELEMENTS
Both RA and DEX induced rapid and dose-dependent increases in adi-

pocyte mRNA 514 by activating S14 gene transcription. 33 •73 RARa and glu-
cocorticoid receptors (GR) might regulate S14 gene transcription through
interaction with cis-linked response elements. In order to localize these ele-
ments, we transfected cells with plasmids containing 5' flanking regions
isolated from the rat liver S14 gene. These regions were fused upstream from
the chloramphenicol acetyltransferase (CAT) gene (i.e., S14-CAT) and used
to measure changes in CAT activity induced by RA and DEX. Under these
conditions, changes in CAT activity reflect changes in transcription of the
CAT gene which is regulated by the S14 5' flanking sequences.
Adipocytes could not be transiently transfected, so stably transfected 3T3-
F442A adipocyte cell lines were developed to examine S 14-CAT gene expres-
sion. Fibroblasts were transfected with Sl4-CAT fusion genes in the presence
of pSV2-Neo using the lipofection technique. 98 Stably transfected colonies
were pooled and differentiated to adipocytes. Cells were treated with DEX
or RA only after >80% of the cells acquired an adipocyte phenotype. Fol-

lowing hormone treatment, cell extracts were prepared and used to measure
CAT activity. 99 Under these transfection conditions, endogenous receptors
must be recruited to regulate integrated S14-CAT fusion genes.
Figure 3 illustrates a set of 5' deletion mutations of the S 14 promoter
that were used to localize the boundaries of the DEX and RA response regions.
Plasmids with 5' end points at -4315, -2110, and -1588 bp were equally
responsive to DEX (fourfold induction). However, plasmids with 5' end points
at -1075 and -290 bp were not responsive to DEX treatment. When an
element extending from -1588 to - 1075 bp was fused to the -290 bp S14
promoter, DEX control was conferred to the S14 promoter. Thus, the bound-
aries for the DEX regulatory unit controlling S 14 gene transcription in adi-
pocytes is located between -1588 and -1075 bp upstream from the 5' end
of the S14 gene.
Since the GH TRE also functioned as a RARE, 58 •62 •100 we examined the
possibility that RA activated S 14 gene transcription through the S 14 TRE at
-2.55 to 2. 75 kb. 68 •69 · 100 Using a similar strategy to locate RA-responsive
regions, successive 5' deletions from the S14 promoter showed that RA
induction of CAT activity increased from 1.7- to 4-fold using plasmids with
5' end points at -4315 and -1588 bp, respectively (Figure 3). The increased
RA responsiveness with successive deletions may indicate the presence of
elements located between -1588 and -4315 bp which modulate RA action.
However, deletions with 5' end points at -1075 and -290 showed no
response to RA. Full RA activity was restored by ligating the - 1588 to
- 1075 bp region to the - 290 bp S 14 promoter element. These results show
that the boundaries for both the RA and glucocorticoid response units are
within the same 513-bp fragment extending from -1588 to -1075 bp. These
studies rule out the S14 TRE as a target for RA action.
VI. ROLE OF TISSUE-SPECIFIC FACTORS IN RA

AND DEX CONTROL OF S14 GENE EXPRESSION
In preadipocytes, the S14 gene is quiescent and unresponsive to both

DEX and RA treatment. 33 •73 In contrast to genes like FAS and GPD, 79 - 82 S14
gene expression does not rise when cells differentiate from preadipocytes to
adipocytes. However, once cells express the adipocyte phenotype, S14 gene
transcription can be activated by both RA and DEX. Thus, S14 gene expres-
sion is dependent on both adipocyte-specific factors and hormone regulated
factors. 33 •73 This pattern of control can be due to: ( 1) tissue-specific regulation
of RA and glucocorticoid regulatory networks; (2) regulation of key tissue-
specific accessory proteins or transcription factors; or (3) a combination of
these factors.
In order to define the molecular basis for the tissue-specific hormonal
regulation of S 14 gene transcription, we determined whether the glucocor-
DELETION ANALYSIS OF THE S 14 PROMOTER <....
:;::
FOLD INDUCTION
~
~
3 5 7
f2-
5 14-4.3-CAT 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1~1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ( c.T I
-4315 -27S -2:5,0 +19
514-2.1-CAT
...
lllllllllllllllllllllllllllllllllllllllllllllllll CAT I
-2110 +1'9
514-1 6-CAT
...
OJIIIIIII!liiiiiiUIIIIIIIIIIIIIII CAT I
-1588 +u - Dexamethasone
... CSJ Retinoic Acid
514-1 I-CAT Jllllllfflll\111111111 CA~
-1075 +19
514- 29-CAT
...
1111111 CAT I
-2'90 +U
075 290
514-PA-CAT
-1588 ~
FIGURE 3. Analysis of retinoic acid and glucocorticoid induction of chloramphenicol acetyltransferase (CAT) expression in various deletion mutations of the
Sl4 promoter. 3T3-F442A fibroblasts were stably transfected with plasmids containing various 5' deletions of the SJ4 promoter. The location of the TRE in
pSl4-4.3-CAT is between -2750 and -2550 bp. Stably transfected cells were differentiated to adipocytes and treated with I ].1M RA or DEX for 72 h. CAT
activity was assayed and results are expressed as fold induction, mean ± SD, N = 3 to 6.
t
.....
100 - MMTV-CAT
90 ISSSI TK14M-CAT
80 CJ RSV -CAT
70 ~ S14 -PA-CAT
rn
·c:
+'
;:)
60
50
:f;.
·;; 40
:,J
()
< 30
~ 20
u
10
0 ~~
Veh Dex RA
is
Veh
i Dex RA
Precdipocytes Adipocytes
FIGURE 4. Tissue-specific expression of SI4-CAT in 3T3-F442A cells. 3T3-F442A pre-

adipocytes were stably transfected with either MAM-Neo-CAT (Clontech), TK14AA-CAT (gen-
erously provided by R. Koenig, University of Michigan), RSV-CAT (generously provided by
S. Conrad, Michigan State Univ.), or S14-PA-CAT. Stably transfected preadipocytes and adi-
pocytes were treated with vehicle, I fLM DEX, or I fLM RA for 72 h. Cells were harvested and
assayed for CAT activity. Results are expressed as mean ± SE (N = 4) CAT units (pmoles of
acetylated chloramphenicol formed per 100 f.Lg protein per h).
ticoid regulatory network was operative in F442A preadipocytes. Accordingly,

preadipocytes were transfected with pMAM-neo-CAT, a plasmid that contains
the murine mammary tumor virus long terminal repeat (MMTV-LTR) ligated
upstream from CAT. The MMTV-LTR contains glucocorticoid response ele-
ments (GRE) which bind ligand activated GRs and augment transcription
from the cis-linked CAT gene. 101 Both preadipocytes and adipocytes were
treated with vehicle or DEX (1 ~ for 72 h. The basal levels of expression,
i.e., 10 units CAT, was not affected by adipocyte differentiation (Figure 4).
DEX treatment increased CAT activity sevenfold in each phenotype indicating
that GRs were functional in both preadipocytes and adipocytes.
In order to determine whether the RA regulatory network was operative
in both phenotypes, F442A preadipocytes were stably transfected with the
TK14AA-CAT plasmid. The TK14AA-CAT plasmid contains a modified
version of the rat pituitary GH TRE ligated upstream from the thymidine
kinase promoter and the CAT reporter gene. 102 The GH TRE binds RRs and
functions as a RAR£. 13 •58 •62 •63 Basal expression of TK14AA-CAT (i.e., 8
units), was not affected by differentiation. However, RA treatment for 72 h
induced CAT activity fivefold in both preadipocytes and adipocytes (Figure
4). Thus, the RA regulatory network is functional in both phenotypes.
As a final control study, RSV-CAT was stably transfected into F442A
preadipocytes and CAT activity measured in both preadipocyte and adipocyte
Jump et al. 443
phenotypes in the presence and absence of DEX. Expression of CAT from

RSV-CAT remained unaffected by differentiation status and hormone treat-
ment (Figure 4). Thus, the effects of DEX and RA on CAT expression from
the MAM-neo-CAT and TK14AA-CAT plasmids were due to transcriptional
activation of the CAT gene and not to effects on enzyme stability or other
generalized mechanisms.
In F442A adipocytes, the Sl4-PA-CAT plasmid is responsive to both RA
and DEX (Figure 3). This plasmid contains a hormone responsive region
(- 1588 to - 1075 bp) ligated upstream from the S 14 promoter (- 290 to
+ 19 bp). Expression of Sl4-PA-CAT in preadipocytes was low (0.5 units)
and remained unaffected by either DEX or RA treatment (Figure 4). Differ-
entiation of F442A cells to adipocytes augmented the basal level of CAT
activity tenfold and treatment with either RA or DEX enhanced CAT activity
fourfold. Essentially the same results were obtained when Ll cells transfected
with the various plasmids described above. The lack of hormonal control of
S 14 gene expression in F442A and Ll preadipocytes cannot be attributed to
the absence of intact RA and glucocorticoid regulatory networks. Acquisition
of hormone responsiveness of the Sl4 gene in both F442A and Ll cells
requires the expression of adipocyte-specific factors that accompany differ-
entiation of cells to adipocytes.
Analysis of the S 14 promoter ( - 290 to + 19 bp) shows that it functions
as a weak promoter (i.e., <0.01 units) in both phenotypes and in either the
absence or presence of hormones. While additional studies are still required,
we do not view the S 14 promoter as a primary target for tissue-specific control
of Sl4 gene expression. The region extending from -1588 to -1075 bp is
the most likely target for both hormonal and tissue-specific control. To ex-
amine this possibility, DNase I footprint analysis was used to determine
whether changes in DNA-protein interaction occurred within this region during
adipocyte differentiation.
Figure 5 illustrates the pattern of DNase I footprints in the - 1588 to
- 1435 bp region. Two protected regions were detected at - 1554 to -1530
bp (region 1) and - 1477 and - 1457 bp (region 2). Conversion of preadi-
pocytes to adipocytes leads to greater protection of region 1 and less protection
of region 2. This indicates that proteins which bind to these regions are under
control during adipocyte differentiation. This analysis was expanded to other
regions and is summarized in Figure 5. Four protected regions were detected
within the 513 bp tissue-specific enhancer. Proteins binding to regions 1 and
4 are induced following differentiation of preadipocytes to adipocytes, while
proteins interacting with region 2 are suppressed. Proteins interacting with
region 3 remain unaffected by the differentiation status of the cells and are
referred to as constitutive factors (CFl). These preliminary studies suggest
that the adipocyte-specific hormonal control of S 14 gene transcription is due
to a change in nuclear proteins which interact with the S 14 tissue-specific
enhancer.
-1554
Reg ion 1
-1 530
-1 477
Reg io n 2
-1 457
FIGURE 5. DNase I footprint analysis of the S 14 enhancer region. A DNA fragment extending
from - 1380 to -1588 bp relative to the Sl4 gene transcription start site was isolated. The
noncoding strand was ·12 P-labeled with T 4 polynucleotide kinase and used for DNase I footprint
analysis. 103 In lanes labeled "D" , the 32 P-labeled DNA was treated with dimethyl sulfate to
generate a G-ladder. In lanes labeled " P" and "A", the 32 P-DNA was mixed with nuclear
proteins (30 JLg) extracted from Ll preadipocyte and Ll adipocyte nuclei , respectively. Samples
were digested with DNase I; the 32 P-DNA was extracted and electrophoresed on a 8% acrylamide
sequencing gel, dried , and exposed to X-ray film.
VII. MODEL FOR THE TISSUE-SPECIFIC,

HORMONE, AND NUTRIENT CONTROLLED
EXPRESSION OF THE S14 GENE IN
CULTURED ADIPOCYTES
Figure 6 illustrates our current view of the arrangement of factors which
regulate Sl4 gene transcription in cultured adipocytes. Both in vivo and in
vitro studies show that transcription of the S 14 gene is subject to complex,
multifactorial control. Our studies have defined three regulatory regions lo-
cated upstream from the 5' end of the Sl4 gene. They include: (1) the
promoter, extending from - 290 to - I bp which functions in the organization
of the transcription initiation complex; 103 (2) the tissue-specific enhancer (TSE)
Jump et al. 445
'- pre-mRNA
Promoter
FIGURE 6. Organization of functional elements controlling Sl4 gene transcription in adipo-

cytes. The diagram illustrates the Sl4 gene, transcription start site (horizontal arrow), the promoter
region ( -290 to -1 bp), the tissue specific enhancer (TSE) located between -1588 to -1075
bp, and the Sl4 thyroid hormone response element (TRE) between -2750 to -2550 bp. Ligand
activated RR (RAR-RA) and glucocorticoid receptors (GR-DEX), T 3 • and the c-erbAa.2 are
illustrated. Tissue-specific factors (TSFl, TSF2. TSF3) and constitutive factor (GFI) are shown
bound to the TSE.
extending from - 1588 to - 1075 bp, which also functions as a ligand-

activated enhancer; and (3) the TRE extending from -2750 to -2550 bp
which functions as a target for T3 control in rat liver, 68 •69 · 100 but not cultured
adipocytes.
In the absence of enhancers, the S 14 promoter is a weak promoter ( < 0. 0 1
units CAT) that is not regulated by tissue-specific factors or hormones. The
TSE functions both to increase the basal rate of S 14 gene transcription in
adipocytes and to confer responsiveness to both RA and DEX. Thus, the TSE
appears to be a major functional unit controlling both tissue-specific and
hormonal regulation of S 14 gene transcription in adipocytes.
The TSE maps at the same location as a DNase I hypersensitive site (at
-1525 to -1335 bp) identified as Hss-2 in rat liver. 67 - 70 DNase I hypersen-
sitive sites typically mark the location of prospective cis-acting elements. 104
Chromatin structure studies showed that the organization of this region was
regulated by tissue-specific factors 66 and by factors regulated by dietary car-
bohydrate. 69 This dietary carbohydrate control of S 14 gene transcription re-
quires insulin, but is inhibited by A-kinase regulated mechanisms. 70 Trans-
fection analysis of cultured hepatocytes indicated that this region was a target
for insulin-dependent, carbohydrate-regulated transcription factors. 105
TABLE 4
Changes in DNA-Protein Interaction within the S14
Tissue-Specific Enhancer during Adipocyte
Differentiation
Region Location• Binding activity Identification
- 1554 to - 1530 Increase TSFl

2 - 1477 to - 1457 Decrease TSF2
3 - 1372 to - 1352 No change CFI
4 -1117 to -1094 Increase TSF3
Note: TSF, Tissue-specific factors; CFI, Constitutive factor.
' Location of protected regions relative to the 5' end of the S 14 gene.
These diverse approaches indicate that in different cells, such as hepa-

tocytes and adipocytes, the TSE region may be regulated by different regu-
latory networks. For example, we have found that the pattern of DNase I
protection using liver and adipocyte nuclear extracts was not identical (not
shown). Moreover, the TSE region functioned in the insulin/carbohydrate
control of S14 gene expression in liver. 105 However, no such control has been
detected in adipocytes (not shown). Alternatively, certain core factors which
interact with the TSE may be required for this element to function in all
tissues in which the S14 gene is expressed. S14 gene expression is regulated
by glucocorticoids in liver, WAT, and F442A and L1 adipocytes. 73 Therefore,
the glucocorticoid-regulated element(s) within the TSE may be a common
functional target in all tissues where the S14 gene is expressed.
Perhaps the most interesting finding of this analysis was that neither DEX
nor RA regulate S14 gene transcription in preadipocytes, despite the presence
of functional RA and glucocorticoid regulatory networks. We attribute this
to the differentiation-dependent expression of trans-acting factors which in-
teract with the TSE. Tissue-specific factors which interact with regions 1, 2,
and 4 and constitutive factors which interact with region 3 are likely to be
involved in the tissue-specific control of S 14 gene transcription in adipocytes.
However, we do not know the identity of the proteins which interact with
these regions or the precise role they play in regulating S 14 gene transcription.
Others have reported that AP-1 binding proteins, i.e., c-Fos/c-Junul6,JO? and
C/EBP function in adipocyte-specific control of gene transcription. 108 - 110 We
have no evidence for a direct interaction of these proteins with elements within
the S 14 TSE. Moreover, we have no evidence for direct interaction of RARa
or GR with the TSE. DNA sequence analysis has not revealed canonical
response elements for either RARE, RXRE, or GRE within the -1588 to
- 1075 bp region. It remains possible that neither RR or GR bind directly to
the S 14 DNA and that these regulatory networks control S14 gene transcription
through the modulation of other transcription factors which interact with the
Jump et al. 447
TSE. Clearly, additional studies will be required to decipher the molecular

basis for the multifactorial control of S 14 gene transcription in cultured adi-
pocytes.
In contrast to rat liver, the S 14 TRE does not function as a target for T 3 -
activated TR in F442A cells. Moreover, this region is not a target for RA
control in F442A or L1 cells. However, in a tissue like liver or WAT, where
both RR and TR receptors are present and the S14 gene is active, both T 3
and RA might regulate S14 gene transcription. This may be mediated by
independent interaction of TR and RR with their corresponding response
regions. Alternatively, TR and RR might form heterodimers and interact with
one or both of the specific response elements. Under these conditions, T 3 and
RA might antagonize or augment each other's action.
However, in F442A adipocytes, the low T 3 binding capacity would ex-
clude T 3 as a potent inducer of S14 gene transcription. On the other hand,
TRs are expressed at high levels in L1 adipocytes and these cells are responsive
to T3 • Preliminary transfection analysis of L1 cells with S14-CAT fusion
genes shows that the S14 TRE is not the target for T3 action (not shown).
This raises the possibility that high levels of c-erbAa2 expression in both
F442A and L1 cells might block TRa1 action at the S14 TRE. Alternatively,
TRf31 is not expressed in either F442A or L1 cell types, but is expressed in
liver. The S14 TRE might be a target for TRf31 and not TRa1 or c-erbAa2.
These issues remain unresolved and require further investigation.
What is clear from these studies is that the complex multifactorial control
of S14 gene transcription can be explained, at least in part, by interaction of
adipocyte specific trans-acting factors with defined cis-linked elements. How-
ever, these elements are not clustered near the 5' end of the S14 gene, but
are dispersed over a 2. 7 kb region. In fact, two spatially distinct ligand
activated enhancers, i.e., TSE and TRE, have been defined. For these distal
elements to function in the initiation of S 14 gene transcription they must
interact with the promoter. Regulating the interaction of these distal elements
with promoter elements may represent another important step by which hor-
mones, nutrients, and tissue-specific factors control the rate of initiation of
S 14 gene transcription.
VIII. SUMMARY: DOES RA FUNCTION IN

INTERMEDIARY METABOLISM?
While RA action on cell growth and differentiation 34- 36 •54- 60 has been well
characterized, several recent reports show that RA regulates the level of
mRNAs coding for proteins involved in intermediary metabolism. These pro-
teins include phosphoenolpyruvate carboxykinase, 29 the S 14 protein, 33 trans-
glutaminase,27 GPD, 33 alcohol dehydrogenase, 31 and G,a. 30 Although these
findings are based on cell culture studies, they show that RA is a regulator
factor in intermediary metabolism which must also function in vivo. The role
of RA in intermediary metabolism is supported further by the presence of RA

(as both 13-cis-RA and all-trans-RA) in plasma at 1 to 3 ng/ml or 3 to 10
nM 111 • 112 and by the wide tissue distribution of RR, particularly in lipogenic
tissues such as liver and WAT In addition to providing new insight into the
interaction of the RA regulatory system with tissue-specific and glucocorti-
coid-regulated factors, the aclipocyte cell culture model provides a means to
investigate the metabolic consequences of RA action on adipocyte metabo-
lism.
ACKNOWLEDGMENTS
The authors wish to thank Drs. Maija Zile and Henry Bayley for many
helpful suggestions during the preparation of this manuscript. We also wish
to thank Dr. P. Chambon, B. Spiegelman, H. Towle, L. Kedes, H-S. Sui,
R. Koenig, and S. Conrad for the various cDNAs, expression vectors, and
adipocyte cell lines used in these studies. This research was supported by
grants from the National Institutes of Health (GM36851, DK43220), Amer-
ican Diabetes Association, The Upjohn Company, and Michigan State Uni-
versity.
REFERENCES
I. Wolbach, S. B., Effects of vitamin A deficiency and hypervitaminosis A in animals, in
The Vitamins, Vol. I, Sebrell, W. B., Jr. and Harris, R. S., Eds., Academic Press, New
York, 1954, 106.
2. Ganguly, J., Rao, M. R. S., Murthy, S. K., and Sarada, K., Systemic mode of action
of vitamin A, in Vitamins and Hormones, Vol. 38, Munson, P. L., Diczfalusy, E.,
Glover, J., and Olson, R. E., Eds., 1980, 1.
3. Roels, 0. A., Biochemical systems, in The Vitamins, 2nd ed., Vol. I, Sebrell, W. H.,
Jr. and Harris, R. S., Eds., Academic Press, New York, !967, 245.
4. Wolf, G., Multiple functions of Vitamin A, Physiol. Rev., 64, 873, 1984.
5. Wald, G., The visual function of the Vitamins A, Vitam. Horm., 18, 417, 1960.
6. Napoli, J. L., The biogenesis of retinoic acid: a physiologically significant promoter of
differentiation, in Chemistry and Biology of Synthetic Retinoids, Dawson, M. I. and
Okamura, W. H., Eds., CRC Press, Boca Raton, FL, 229, 1990.
7. Sani, B. B., Cellular retinoic acid binding proteins and action of retinoic acid, in Chem-
istry and Biology of Synthetic Retinoids, Dawson, M. I. and Okamura, W. H., Eds.,
CRC Press, Boca Raton, FL, 1990, 365.
8. Wolf, G., Recent progress in vitamin A research: nuclear retinoic acid receptors and
their interaction with gene elements, J. Nutr. Biochem., 1, 284, 1991.
9. Chytil, F., Retinoic acid: biochemistry and metabolism, J. Am. Acad. Dermatol .. IS,
741, 1986.
10. Melton, D. A., Pattern formation during animal development, Science, 252, 234, 1991.
11. Blomhoff, R., Green, M. H., Berg, T., and Norum, K. R., Transport and storage of
vitamin A, Science, 250, 399, 1990.
Jump eta!. 449
12. Morita, S., Fernadez-Mejia, C., and Melmed, S., Retinoic acid selectively stimulates
growth hormone secretion and messenger ribonucleic acid levels in rat pituitary cells,
13. Bedo, G., Santisteban, P., and Aranda, A., Retinoic acid regulates growth hormone
gene expression, Nature (London), 339, 231, 1989.
14. Glick, A. B., Flanders, K. C., Danielpour, D., Yuspa, S. H., and Sporn, M. B.,
Retinoic acid induces transforming growth factor ~2 in cultured keratinocytes and mouse
epidermis, Cell Regul., I, 87, 1989.
15. Petkovich, P. M., Heersche, J. N. M., Tinker, D. 0., and Jones, G., Retinoic acid
stimulates I ,25-dihydroxyvitamin D, binding in rat osteosarcoma cells, J. Bioi. Chem.,
259, 8274, 1984.
16. Sucov, H. M., Murakami, K. K., and Evans, R. M., Characterization of an auto-
regulated response element in the mouse RAR~ gene, Proc. Nat/. Acad. Sci. U.S.A.,
87, 5392, 1990.
17. de The, H., Marchio, A., Tiollais, P., and Dejean, A., Differential expression and
ligand regulation of the RAR a and ~ genes, EMBO J., 8, 429, 1989.
18. Lafyatis, R., Kim, S-J., Angel, P., Roberts, A. B., Sporn, M. B., Karin, M., and
Wilder, R., lnterleukin-1 stimulates and all-trans retinoic acid inhibits collagenase gene
expression through its 5' activator protein-1-binding site, Mol. Endocrinol., 4, 973, 1990.
19. Clarke, C. L., Roman, S. D., Graham, J., Koga, M., and Sutherland, R. L.,
Progesterone receptors regulated by retinoic acid in human breast cancer cell line T-4 7D,
J. Bioi. Chem., 265, 12694, 1990.
20. Suva, L. J., Ernst, M., and Rodan, G. A., Retinoic acid induces zif268 early gene
expression in rat preosteoblastic cells, Mol. Cell. Bioi., 11, 2503, 1991.
21. Luscher, B., Mitchell, P., Williams, T., and Tjian, R., Regulation of transcription
factor AP-2 by the morphogen retinoic acid and second messengers, Genes Dev., 3,
1507, 1989.
22. Chakraborty, A. K., Orlo, S. J., and Pawelek, J. M., Stimulation of the receptor for
melanocyte stimulating hormone by retinoic acid, FEBS, 276, 205, 1989.
23. Side!, N., Taga, T., Hirano, T., Kishimoto, T., and Saxon, A., Retinoic acid induced
growth inhibition of a human myeloma cell line via down-regulation of IL-6 receptors,
J. Immunol., 146, 3809, 1991.
24. Sidell, N. and Ramsdell, F. R., Retinoic acid upregulates interleukin-2 receptors on
activated human thymocytes, Cell. Immunol., 115, 299, 1988.
25. Junquero, D., Modat, G., Coquelet, C., and Bonne, C., Retinoids enhance the number
ofEGF receptors in corneal endothelial cells, Exp. Eye Res., 51, 49, 1990.
26. Ubels, J. L., Iorfino, A., and O'Brien, W. J., Retinoic acid decreases the number of
EGF receptors in corneal epithelium and Chang conjunctival cells, Exp. Eye Res., 52,
763, 1991.
27. Chiocca, E. A., Davies, P. J, A., and Stein, J, P., The molecular basis of retinoic
acid action, J. Bioi. Chem., 263, 11584, 1988.
28. Imai, Y., Rodan, S. B., and Rodan, G. A., Effects of retinoic acid on alkaline phos-
phatase mRNA, catecholamine receptors and G-proteins in ROS 17/2.8 cells, Endocri-
nology, 122, 456, 1988.
29. Lucas, P. C., O'Brien, R. M., Mitchell, J, A., Davis, C. M., lmai, E., Forman,
B. M., Samuels, H. H., and Granner, D. K., A retinoic acid response element is part
of a pleiotropic domain in the PepCK gene, Proc. Nat/. Acad. Sci. U.S.A., 88, 2184,
1991.
30. Chan, S.D. H., Strewler, G. J., and Nissensa, R. A., Transcription activation of Gsa
gene expression by retinoic acid and parathyroid hormone-related protein in F9 terato-
carcinoma cells, J. Bioi. Chem., 265, 20081, 1990.
31. Duester, G., Shear, M. L., McBride, M.S., and Stewart, M. J., Retinoic acid response
element in human alcohol dehydrogenase ADH3: Implication for regulation of retinoic
acid synthesis, Mol. Cell. Bioi., 11, 1638, 1991.
32. Darrow, L., Riches, R. j., Pecorino, L. T., and Strickland, S., Transcription factor
SPl is important for retinoic acid induced expression of tissue plasminogen activator gene
during F9 teratocarcinoma cell differentiation, Mol. Cell. Bioi., 10, 5883, 1990.
33. Lepar, G. j. and Jump, D. B., Retinoic acid and dexamethasone interact to regulate
S 14 gene transcription in 3T3-F442A adipocytes, Mol. Cell. Endocrinology, 77, 84, 65,
1992.
34. Strickland, S., Smith, K. K., and Marotti, K. K., Hormonal induction of differentiation
in teratocarcinoma stem cells: generation of parietal endoderm by retinoic acid and db-
cAMP, Cell, 21, 347, 1980.
35. Wang, S. Y. and Gudas, L. J,, Isolation of eDNA clones specific for collagen IV and
laminin from mouse teratocarcinoma, Proc. Nat/. Acad. Sci. U.S.A.. 80, 5880, 1983.
36. Vasios, G. W., Gold, J. D., Petkovich, M., Chambon, P., and Gudas, L. J,, A
retinoic acid response element is present in the 5' flanking region of the laminin 131 gene,
Proc. Nat!. Acad. Sci. U.S.A .. 86, 9099, 1989.
37. Wei, L-N., Blaner, W. S., Goodman, D. S., and Nguyen-Huu, C. M., Regulation of
cellular retinoid-binding proteins and their mRNAs during embryonal carcinoma cell
differentiation induced by retinoic acid, Mol. Endocrinol., 3, 454, 1989.
38. Munoz-Canoves, P., Vik, D. P., and Tack, B. F., Mapping a retinoic acid responsive
element in the promoter region of the complement factor H gene, J. Bioi. Chern., 265,
20065, 1990.
39. Antras, j., Lasnier, F., and Pairault, J,, Adipsin gene expression in 3T3-F442A
adipocytes is post-transcriptionally down-regulated by retinoic acid, J. Bioi. Chern .. 266,
1157, 1991.
40. Stellmach, V., Leask, A., and Fuchs, E., Retinoid mediated transcriptional regulation
of keratin genes in human epidermal and squamous carcinoma cells, Proc. Nat/. Acad.
Sci. U.S.A.. 88, 4582, 1991.
41. Hohl, D., Lichti, U., Breitkreutz, D., Steinert, P.M., and Roop, D. R., Transcription
of the human loricrin gene in vitro is induced by calcium and cell density and suppressed
by retinoic acid, J. Invest. Derrnatol., 96, 414, 1991.
42. Rottman, J. N., Widom, R. L., Nadal-Ginard, B., Mahdavi, V., and Karathanasis,
S. K., A retinoic acid response element in the apolipoprotein AI gene distinguishes
between two different retinoic acid pathways, Mol. Cell Bioi., II, 3814, 1991.
43. Scheibe, R. J., Ginty, D. D., and Wagner, J. A., Retinoic acid stimulates differentiation
of PC12 cell that are deficient in cAMP-dependent protein kinase A, J. Cell Bioi., 113,
1173, 1991.
44. Pacifici, M., Golden, E. B., Iwamoto, M., and Adams, S. L., Retinoic acid treatment
induces type X collagen gene expression in cultured chick chondrocytes, Exp. Cell Res.,
195, 38, 1991.
45. Glover, A., Oshima, R. G., and Adamson, E. D., Epithelial layer formation in dif-
ferentiation aggregates of F9 embryonal carcinoma cells, J. Cell Bioi., 96, 1690, 1983.
46. Linder, S. K., Krondahl, U., Sennerstrom, R., and Ringerty, N. R., Retinoic acid
induced differentiation of F9 embryonal carcinoma cells, Exp. Cell Res.. 132, 453, 1981.
47. Lehtonen, E., Lehto, V. P., Badley, R. A., and Virtanen, 1., Formation of vinculin
plaques precedes other cytoskeletal changes during retinoic acid induced teratocarcinoma
cells differentiation, Exp. Cell Res., 144, 191, 1983.
48. Petkovich, M., Brand, N.J., Krust, A., and Chambon, P., A human retinoic acid
receptor which belongs to the family of nuclear receptors, Nature (London), 330, 444,
1987.
49. Giguere, V., Ong, E. S., Segui, P., and Evans, R. M., Identification of a receptor for
the morphogen retinoic acid. Nature (London), 330, 624, 1987.
50. Brand, N., Petkovich, M., Krust, A., Chambon, P., de The, H., Marchio, A.,
Tiollais, P., and Dejean, A., Identification of a second human retinoic acid receptor,
Nature (London), 332, 850, 1988.
Jump et al. 451
51. Evans, R. M., The steroid and thyroid hormone receptor superfamily, Science, 240,
889, 1988.
52. Benbrook, D., Lenhardt, E., and Pfahl, M., A new retinoic acid receptor identified
from a hepatocellular carcinoma. Nature (London), 333, 669, 1988.
53. Zelent, A., Krust, A., Petkovich, M., Kastner, P., and Chambon, P., Cloning of
murine a and J3 retinoic acid receptors and a novel 'Y receptor predominantly expressed
in skin, Nature (London), 339, 714, 1989.
54. Mangelsdorf, D. J., Ong, E. S., Dyck, J, A., and Evans, R. M., Nuclear receptor
that identifies a novel retinoic acid response pathway, Nature (London), 345, 224, 1990.
55. DeLuca, L. M., Retinoids and their receptors in differentiation, embryogenesis and
differentiation, FASEB J., 5, 2924, 1991.
56. Haq, R. U. and Chytil, R., Expression of nuclear retinoic acid receptors in rat adipose
tissue, Biochem. Biophys. Res. Comm., 176, 1539, 1991.
57. Dalman, F. C., Sturzenbecker, L. J., Levin, A. A., Lucas, D. A., Perdew, G. W.,
Petkovich, M., Chambon, P., Grippo, J, F., and Pratt, W. B., Retinoic acid receptor
belongs to a subclass of nuclear receptors that do not form docking complexes with hsp90,
Biochemistry, 30, 5605, 1991.
58. Umesono, K., Giguere, V., Glass, C. K., Rosenfeld, M. G., and Evans, R. M.,
Retinoic acid and thyroid hormone induce gene expression through a common responsive
element, Nature (London), 336, 262, 1988.
59. Umesono, K., Murakami, K. K., Thompson, C. C., and Evans, R. M., Direct repeats
as selective response elements for the thyroid hormone, retinoic acid and vitamin 0 3
receptors, Cell, 65, 1255, 1991.
60. Mangelsdorf, D. J., Umesono, K., Kliewer, S. A., Borgmeyer, U., Ong, E. S., and
Evans, R. M., A direct repeat in the cellular retinol-binding protein type II gene confers
differential regulation by RXR and RAR, Cell, 66, 555, 1991.
61. Naar, A. M., Boutin, J.-M., Lipkin, S. M., Yu, V. C., Holloway, J. M., Glass,
C. K., and Rosenfeld, M.G., The orientation and spacing of core DNA-binding motifs
dictate selective transcriptional responses to three nuclear receptors, Cell, 65, 1267, 1991.
62. Glass, C. K., Lipkin, S. M., Devary, 0. V., and Rosenfeld, M. G., Positive and
negative regulation of gene transcription by a retinoic acid-thyroid hormone receptor
heterodimer, Cell, 59, 697, 1989.
63. Forman, B. M., Yang, C.-R., Au, M., Casanova, J., Ghysdael, J., and Samuels,
H. H., A domain containing leucine-zipper-like motifs mediate novel in vivo interaction
between the thyroid hormone and retinoic acid receptors, Mol. Endocrinol., 3, 1610,
1989.
64. Jump, D. B., Narayan, P., Towle, H. C., and Oppenheimer, J, H., Rapid effects of
triiodothyronine on hepatic gene expression, J. Bioi. Chem .. 259. 2789, 1984.
65. Jump, D. B. and Oppenheimer, J. H., High basal expression and 3.5. 3' -triiodothy-
ronine regulation of mRNA" 4 in lipogenic tissues. Endocrinology, 117, 2259, 1985.
66. Jump, D. B., Wong, N. C. W., and Oppenheimer, J, H., Chromatin structure and
methylation state of a thyroid hormone responsive gene in rat liver, J. Bioi. Chem., 262,
778, 1985.
67. Jump, D. B., Thyroid hormone regulation of rat liver Sl4 gene expression, in Gene
Regulation by Steroids Hormones IV, Roy. A. K. and Clarke. J. H., Eds., Springer-
Verlag, New York, 1989, 144.
68. Jump, D. B., Rapid induction of rat liver S14 gene transcription by thyroid hormone,
J. Bioi. Chem., 264, 4698, 1989.
69. Jump, D. B., Bell, A., and Santiago, V., Thyroid hormone and dietary carbohydrate
interact to regulate rat liver S 14 gene transcription and chromatin structure, J. Bioi.
Chem., 265, 3474, 1990.
70. Jump, D. B., Bell, A., Lepar, G., and Hu, D., Insulin rapidly induces rat liver S14
gene transcription, Mol. Endocrinol., 4, 1655, 1990.
71. Oppenheimer, J, H., Schwartz, H. L., Mariash, C. N., Kinlaw, W. B., Wong,
N. C. W., and Freake, H. C., Advances in our understanding of thyroid hormone action
at the cellular level, Endocrine Rev., 8, 288, 1987.
72. Kinlaw, W. B., Ling, N.C., and Oppenheimer, J. H., Identification ofrat Sl4 protein
and comparison of its regulation with that of mRNA 514 employing synthetic peptide
antisera, J. Bioi. Chern., 264, 19779, 1989.
73. Lepar, G. J. and Jump, D. B., Hormonal regulation of the Sl4 gene in 3T3-F442A
cells, Mol. Endocrinol., 3, 1207, 1989.
74. Clarke, S. D., Armstrong, M. K., and Jump, D. B., Nutritional control of rat liver
fatty acid synthase and Sl4 mRNA abundance, J. Nutr., 120, 218, 1990.
75. Clarke, S. D., Armstrong, M. K., and Jump, D. B., Dietary polyunsaturated fats
uniquely suppress rat liver fatty acid synthase and Sl4 mRNA content, J. Nutr., 120,
225, 1990.
76. Blake, W. L. and Clarke, S.D., Suppression of rat hepatic fatty acid synthase and Sl4
gene transcription by dietary polyunsaturated fat, J. Nutr., 120, 1727, 1990.
77. Jump, D. B., Veit, A. M., Santiago, V. S., Lepar, G., and Herberholz, L., Tran-
scriptional activation of the rat liver S 14 gene during post-natal development, J. Bioi.
Chern., 263, 7254, 1988.
hydrate and lipid metabolism, Annu. Rev. Nutr., 7, l 57, 1987.
79. Green, H. and Kehinde, 0., Subhnes of mouse 3T3 cells that accumulate lipid, Cell,
I, 113, 1974.
80. Spiegelman, B. M., Frank, M., and Green, H., Molecular cloning of mRNA from
3T3 adipocytes, J. Bioi. Chern., 258, 10083, 1983.
81. Bernlohr, D. A., BoJanowski, M.A., Kelly, T. J., Jr., and Lane, M.D., Evidence
for an increase in transcription of specific mRNA during differentiation of 3T3-Ll pre-
adipocytes, J. Bioi. Chern., 260, 5563, 1985.
82. Wise, L. S., Sui, H. S., and Rubin, C. S., Coordinate regulation of the biosynthesis
of ATP-citrate lyase and malic enzyme during adipocyte differentiation, J. Bioi. Chern.,
259, 4827, 1984.
83. Oppenheimer, J, H., Schwartz, H. L., and Surks, M. I., Tissue differences in the
concentration ofT, nuclear binding sites in the rat liver, kidney, pituitary, heart, brain,
spleen and testes, Endocrinology, 95. 897, 1974.
84. Buergi, U. and Abuehl, U., Nuclear triiodothyronine binding sites in rat adipocytes,
Metabolism, 33, 326, 1984.
85. Hausdorf, S., Clement, J., and Loos, U., Expression of Sl4 gene during differentiation
of 3T3-Ll cells, Horm. Metab. Res., 20, 723, 1988.
86. Moustaid, N. and Sui, H. S., Regulation of expression of the fatty acid synthase gene
in 3T3-Ll cells by differentiation and triiodothyronine, J. Bioi. Chern., 266, 18550,
1991.
87. Murray, M. B., Zilz, N., McCreary, N. L., MacDonald, M. J., and Towle, H. C.,
Isolation and characterization of rat eDNA clones for two distinct thyroid hormone re-
ceptors, J. Bioi. Chern., 263, 12770, 1988.
88. Hodin, R. A., Lazar, M.A., Wintman, B. I., Darling, D. S., Koenig, R. J,, Larsen,
P. R., Moore, D. D., and Chin, W. W., Identification of a thyroid hormone receptor
that is pituitary specific, Science, 244, 76, 1989.
89. lzumo, S. and Madhavi, V., Thyroid hormone receptor isoforms generated by alternative
splicing differentially activate myosin HC gene transcription, Nature (London), 334, 539,
1988.
90. Lazar, M. A., Hodin, R. A., Darling, D. S., and Chin, W. W., Identification of a
rat c-erbA-related protein which binds deoxyribonucleic acid, but does not bind thyroid
hormone, Mol. Endocrinol., 2, 893, 1988.
Jump et al. 453
91. Mitsuhashi, T., Tennyson, G. E., and Nikodem, V. M., Alternative splicing generates
messages encoding rat c-erbA proteins that do not bind thyroid hormone, Proc. Nat/.
Acad. Sci. U.S.A.. 85. 5804, 1988.
92. Koenig, R. J., Lazar, M. A., Hodin, R. A., Brent, G. A., Larsen, P. R., Chin,
W. W., and Moore, D. D., Inhibition of thyroid hormone action by a non-hormone
binding c-erbA protein generated by alternative mRNA splicing, Nature, 337, 659, 1989.
93. Cook, C. B. and Koenig, R. J., Expression of erbAo: and [3 mRNAs in regions of adult
rat brain, Mol. Cell. Endocrinol., 70. 13, 1990.
94. Kuri-Harcuch, W., Differentiation of 3T3-F442A cells into adipocytes is inhibited by
retinoic acid, Differentiation, 23, 164, 1982.
95. Castro-Munozledo, F., Marsch-Moreno, M., Beltran-Langarica, A., and Kuri-
Harcuch, W., Commitment of adipocyte differentiation in 3T3 cells is inhibited by
retinoic acid and the expression of lipogenic enzymes in modulated through cytoskeleton
stabilization, Differentiation, 36, 211, 1987.
96. Pairault, J. and Lasnier, F., Control of adipogenic differentiation of 3T3-F442A cells
by retinoic acid, dexamethasone and insulin: a topographic analysis, J. Cell Physiol.,
132, 279, 1987.
97. Stone, R. L. and Bernlohr, D. A., The molecular basis for inhibition of adipose
conversion of murine 3T3-Ll cells by retinoic acid, Differentiation, 45, 119, 1990.
98. Feigner, P. L., Gadek, T. R., Holm, M., Roman, R., Chan, H. W., Wenz, M.,
Northrop, J. P., Ringold, G. M., and Danielsen, M., Lipofection: a highly efficient,
lipid-mediated DNA-transfection procedure, Proc. Nat/. Acad. Sci. U.S.A., 84, 7413,
1987.
99. Gorman, C. M., Moffat, L. F., and Howard, B. H., Recombinant genomes which
express chloramphenicol acetyltransferase in mammalian cells, Mol. Cell Bioi., 2, 1044,
1982.
100. Zilz, N. D., Murray, M. B., and Towle, H. C., Identification of multiple thyroid
hormone response elements located far upstream from the rat S14 promoter, J. Bioi.
Chern., 265, 8136, 1990.
101. Chandler, V. L., Maler, B. A., and Yamamoto, K. R., DNA sequences bound spe-
cifically by glucocorticoid receptor in vitro render a heterologous promoter hormone
responsive in vivo, Cell, 33, 489, 1983.
102. Brent, G., Larsen, P.R., Harney, J. W., Koenig, R. J., and Moore, D. D., Functional
characterization of the rat growth hormone promoter elements required for induction by
thyroid hormone with and without a co-tranfected [3-type thyroid hormone receptor, J.
Bioi. Chern., 264, 178, 1989.
103. MacDougald, 0. A. and Jump, D. B., Identification of functional cis-acting elements
in the S14 promoter, Biochern. J., 280, 761, 1991.
104. Gross, D. S. and Garrard, W. T., Nuclease hypersensitive sites in chromatin, Annu.
Rev. Biochern., 57, 159, 1988.
105. Jacoby, D. B., Zilz, N.D., and Towle, H. C., Sequences within the 5' flanking region
of the S14 gene confer responsiveness to glucose in primary hepatocytes, J. Bioi. Chern.,
264, 17623, 1989.
106. Distel, R. J., Ro, H.-S., Rosen, B. S., Groves, D. L., and Spiegelman, B. M.,
Nucleoprotein complexes that regulate gene expression in adipocyte differentiation: direct
participation of c-fos, Cell, 49, 835, 1987.
107. Herrera, R. 0., Robinson, H. S., Xanthopoulos, K. G., and Spiegelman, B. M., A
direct role for c/EBP and AP-I binding site in gene expression linked to adipocyte
differentiation, Mol. Cell Bioi., 9, 5331, 1989.
108. Christy, R. J., Yang, V. W., Ntambi, J. M., Geiman, D. E., Landschulz, W. H.,
Friedman, A. D., Nakabeppu, Y., Kelly, T. J., and Lane, M. D., Differentiation-
induced gene expression in 3T3-Ll preadipocytes: CCAAT/enhancer binding protein
interacts with and activates the promoters of adipocyte-specific genes, Genes Dev., 3,
1323, 1989.
109. Friedman, A. D., Landschulz, W. H., and McKnight, S. L., C/EBP activates the
promoter of the serum albumin gene in cultured hepatoma cells, Genes Dev., 3, 1314,
1989.
110. McKnight, S. L., Lane, M. D., and Gluecksohn-Waelsch, S., Is CCAAT/enhancer
binding protein a central regulatory of energy metabolism?, Genes Dev., 3, 2021, 1989.
Ill. Tang, G. and Russel, R. M., 13-cis-retinoic acid is an endogenous compound in human
serum, J. Lipids Res., 31, 175, 1990.
112. Cullum, M. E. and Zile, M. H., Metabolism of all-trans-retinoic acid and all-trans-
retinyl acetate. Demonstration of common physiological metabolites in rat small intestinal
muscosa and circulation, J. Bioi. Chern., 260, 10590, 1985.
Chapter 20
THE ROLE OF VITAMIN K IN CLOTTING

FACTOR BIOSYNTHESIS
J. W. Suttie
TABLE OF CONTENTS
I. Introduction ...................................................... 456
II. -y-Carboxyglutamic Acid ......................................... 456

A. The Vitamin K-Dependent Carboxylase .................. 459
1. Mechanism of the Carboxylase Reaction .......... 460
2. Cell Biology of the Carboxylase .................. 462
3. Relationship of Carboxylation to Other
Vitamin K-Metabolizing Enzymes ................ 463
III. Structure of Vitamin K-Dependent Plasma Proteins ............. .464

A. Genomic Structure of Vitamin K-Dependent
Proteins .................................................. 468
B. Congenital Deficiencies of Vitamin K-Dependent
Clotting Factors .......................................... 468
IV. The Propeptide Recognition Site ................................. 470

A. Cellular Expression Systems and Propeptide
Function .................................................. 470
B. Carboxylase Substrates and Propeptide Function ........ .472
C. Dual Role of the Propeptide .............................. 473
V. Gene Expression and Vitamin K-Dependent Proteins ........... .474
References .............................................................. .47 6
0-8493-6961-4/93/$0.00 +$.50
I. INTRODUCTION
Vitamin K was not discovered until the early 1930s when Dam 1 noted a
hemorrhagic syndrome in chicks fed a lipid-free diet. This response could be
cured by the addition of alfalfa meal to the diet or by the administration of
a lipid extract of green plants. By 1939, a series of investigations led by Dam
and his collaborators in Denmark, Almquist at Berkeley, and Doisy at St.
Louis University had established that the form of the vitamin which is found
in alfalfa, and now called vitamin K 1 or phylloquinone, was 2-methyl-3-
phytyl-1 ,4-naphthoquinone. A large number of bacteria, including numerous
faculative and obligatory anaerobes of the human lower bowel, synthesize a
series of vitamers with unsaturated multiprenyl side chains at the 3-position.
These were characterized soon after phylloquinone and were originally called
vitamin K 2 , but are now more correctly designated as menaquinones (Figure
1). This early history of the discovery and characterization of vitamin K has
been adequately reviewed. zA
The hemorrhagic condition that resulted from the dietary lack of vitamin
K was shown by Dam et al. 5 to be associated with a decrease in plasma
prothrombin activity, and, for some time, was thought to be due solely to a
lowered concentration of prothrombin (factor II). Later, as they were dis-
covered, it was shown that the synthesis of clotting factors VII, IX, and X,
which are involved in the activation of prothrombin to thrombin, was also
depressed in the deficient state. 4 These four plasma proteins were collectively
called the ''vitamin K-dependent clotting factors'', and for years were thought
to be the only proteins whose synthesis required the vitamin. It was not until
the 1970s, after the biochemical role of the vitamin had been established,
that two proteins now known to play an anticoagulant, rather than a procoa-
gulant, role in hemostasis, protein C6 , and protein S7 were discovered. The
general role of the vitamin K -dependent procoagulants and anticoagulants in
normal hemostasis is shown in Figure 2. Another vitamin K-dependent protein
(protein Z) of unknown function is also found in plasma. A limited number
of extrahepatic, vitamin K-dependent proteins, including osteocalcin or Bone
Gla Protein (BGP) and Matrix Gla Protein (MGP), have been identified. 8
However, efforts to understand the enzymology of the vitamin K-dependent
modification of proteins have concentrated on the activity expressed in hepatic
tissue.
II. y-CARBOXYGLUTAMIC ACID
Although vitamin K was known by the late 1930s to regulate the activity
of a few specific proteins, the lack of a general understanding of the mech-
anism of protein biosynthesis prevented serious experimental approaches to
the cellular and molecular mechanisms involved until the mid-1960s. By the
early 1940s, the 4-hydroxy-coumarins were identified as indirect anticoag-
ulants that functioned by antagonizing the action of vitamin K. 9 Studies
Suttie 457
c<r
0 0
0 0
Menadione Phylloquinone
0
Menaqui none-7
FIGURE 1. Biologically active forms of vitamin K. Phylloquinone, produced by plants, and
menaquinone-7 (MK-7), an example of the long chain menaquinone produced by bacteria, are
natural forms of the vitamin. Menadione is a synthetic compound which has biological activity
after it is alkylated to MK-4 in the liver.
directed toward an understanding of the role of vitamin K concentrated on

an understanding of the biosynthesis of prothrombin. These studies were aided
by the ability to easily regulate the plasma activity of prothrombin by oral
anticoagulant (usually warfarin) administration. Although an early hypothesis
suggested that vitamin K regulated the expression of the prothrombin gene, 10
by the mid-1960s a number of animal model studies strongly suggested that
vitamin K was involved in converting an intracellular inactive precursor of
prothrombin to biologically active prothrombin. 11
This hypothesis was strengthened by the observation of immunochemi-
cally similar, but biologically inactive, prothrombin molecules that increased
in the plasma of anticoagulant-treated patients. Characterization of this "ab-
normal" or undercarboxylated prothrombin obtained from anticoagulant-treated
bovine plasma revealed that its amino acid composition, amino and carboxyl
terminal amino acids, and molecular weight appeared to be identical to those
of prothrombin. This protein could be electrophoretically separated from nor-
mal prothrombin in the presence, but not the absence, of Ca 2 +-containing
buffers, and in contrast to normal vitamin K-dependent proteins, it was not
adsorbed by insoluble barium salts. Calcium ions are required for normal
prothrombin activation, and the abnormal prothrombin lacked the Ca2 + bind-
ing properties of normal prothrombin and did not demonstrate a Ca2 +-de-
pendent association with phospholipid surfaces. This inability of the plasma-
abnormal prothrombin to bind Ca 2 + in solution suggested to investigators in
"Intrinsic Pathway" "Extrinsic Pathway"
Factor XI, XII

·contact Activation·
! !
lla
x.
~~-TM n.
~-~~~·~\1'111"
_ ...//····---- .~
~··· inact.
VIlla
0
''
---------<---.••••••,,, _n_•~L--' "-

',, ___ ................• ~
j
Thrombin
Fibrinogen - - - - - • Fibrin
FIGURE 2. The role of vitamin K-dependent proteins in blood coagulation. The classical
vitamin K-dependent clotting factors, II (prothrombin), VII, IX, and X. have procoagulant
hemostatic roles and circulate as inactive zymogens which are converted to active serine proteases
by specific peptide cleavage (e.g .• X to Xa). Protein C and proteinS are involved in the process
by which the rapid generation of thrombin from prothrombin is brought under control by inac-
tivation of accessory factors Va and VIlla of the hemostatic system. The other clotting factors
shown do not require vitamin K for their biosynthesis.
this field that the function of vitamin K was to attach a calcium-binding

prosthetic group to a liver precursor of prothrombin.
Low molecular weight, calcium-binding peptides could be isolated from
enzymatic digests of normal prothrombin, and these acidic peptides were
subsequently shown by Stenflo 12 and Nelsestuen 13 to contain 'Y-carboxyglu-
tamic acid (Gla), a previously unrecognized acidic amino acid. Further in-
vestigation of the chemistry of prothrombin revealed 14 that all 10 of what had
once been thought to be glutamic acid residues in the first 42 residues of
bovine prothrombin were modified in this fashion (Figure 3). These residues
are required for the calcium-mediated interaction of the vitamin K-dependent
proteins with a negatively charged phospholipid surface and are essential for
the formation of the protein conformation that imparts this property. These
studies, which led to the characterization of the vitamin K-dependent modi-
fication of prothrombin, have been reviewed. 15
Suttie 459
VITAMIN K-DEPENDENT PLASMA PROTEINS
-40 Residues with 10-13 •cia• residues COOH

I
7-Carboxyglutamic acid NH 2 -C-H
I
CH 2
I
HC-COOH
I
COOH
FIGURE 3. "(-Carboxyglutamic acid. The vitamin K-dependent proteins are typical glycopro-
teins that are unique in that they contain a number of 'Y-carboxyglutamyl (Gla) residues in a very
homologous (see Table I) amino terminal domain.
I Vit KH 2 Vit KO I
CH2
I
CH 2
I
)>-/'·
/
CHz
I
H-C-COOH
I
COOH 02 C0 2 COOH
Vitamin K-dependent carboxylase

y -glutamyl carboxylase
Vitamin K 2,3-epoxidase
FIGURE 4. The liver microsomal vitamin K-dependent carboxylase. The enzyme utilizes the
hydronaphthoquinone form of the vitamin (Vit KH 2 ) to drive the carboxylation of peptidyl bound
glutamyl residues to "(-carboxyglutamyl residues. The coproduct of the reaction is vitamin K
2,3-epoxide (Vit KO).
A. THE VITAMIN K-DEPENDENT CARBOXYLASE

The enzyme responsible for the posttranslational conversion of glutamyl
residues in newly synthesized proteins to -y-carboxyglutamyl residues in ma-
ture proteins (Figure 4) was first demonstrated 16 in a crude rat liver microsomal
preparation. The activity is usually called the vitamin K-dependent carbox-
ylase or -y-glutamyl carboxylase, and as seen in Figure 4, it could also be
called a vitamin K-2,3-epoxidase. Early studies of the properties of the en-
zyme, which have been reviewed in detail, 17- 21 demonstrated that the reaction
required 0 2 , C0 2 , and the reduced (hydronaphthoquinone) form of vitamin
K. These early studies established that the only source of energy needed to
drive this unique carboxylation reaction comes from the reoxidation of the
reduced vitamin.
Most studies of the enzyme hve utilized crude microsomes solubilized in
various detergents and short peptides containing Glu-Glu sequences as sub-
strates. The enzyme is usually studied at 25°C or lower, and most studies
have been carried out in very crude preparations. Reported measurements of
kinetic contents have been variable but have been in the range of 75 fLM for
reduced vitamin K, 2 mM for COzfHC03 - , and 100 fLM for 0 2.22 Most studies
have utilized short peptides such as Phe-Leu-Glu-Glu-Leu or Boc-Glu-Glu-
Leu-OMe as Glu site substrates. These have a Km of a few mM in contrast
to the low Km values measured for the natural protein substrates of the en-
zyme.22
There have been few attempts to understand the mechanism of action of
the enzyme by the use of steady-state kinetic measurements. It is difficult to
saturate the enzyme with the nonvariable substrates, but Uotila 23 has studied
the calf liver enzyme at a number of substrate concentrations. He concluded
that the carboxylase has a sequential reaction mechanism which includes a
quintemary complex of the enzyme with all of the four substrates. The data
are consistent with an ordered steady-state addition of the substrates to the
enzyme and suggest that C0 2 is either the first or last substrate to bind.
1. Mechanism of the Carboxylase Reaction

The vitamin K-dependent carboxylase appears to represent a unique bio-
chemical mechanism that is not yet understood in any detail. Early studies
considered two possibilities: that the vitamin was a cofactor utilized to abstract
the hydrogen on the -y-position of the glutamyl residue to allow attack of C02
at this position, or that it was involved as a C0 2 carrier. Little evidence to
support the latter hypothesis was ever obtained. Synthesis of the pentapeptide
Phe-Leu-Glu-Glu-Leu, tritiated at the -y-carbon of each Glu residue, 24 was
used to demonstrate that the enzyme catalyzed a vitamin KH 2 - and 0 2-de-
pendent, but C0 2 -independent release of tritium from this substrate. These
data appeared to offer conclusive evidence that the role of the vitamin was
not to activate or carry the C0 2 but to remove the -y-hydrogen of the Glu
substrate, and to concurrently be converted to the epoxide. Epoxide formation
is stimulated by the presence of a Glu site substrate of the carboxylase, and
at saturating concentrations of C0 2 there is an apparent equivalent stoichi-
ometry between epoxide formation and Gla formation. 25
The enzyme will also catalyze a vitamin KH 2 and oxygen-dependent
exchange of 3 H2 0 into the -y-position of a Glu residue in the substrate Boc-
Glu-Glu-Leu-OMe. 26 ·27 Exchange of 3 H with the -y-carbon hydrogen is de-
creased as the concentration of HC0 3 - in the media is increased. Studies 28
Suttie 461
OH H+
c¢: R ~(-)
Glu Glu
C02
\.., Gla
~lOH
~
0
[ KH-OOH] (Qo R
H2 0 0
FIGURE 5. Proposed mechanism for the vitamin K-dependent carboxylase/epoxidase system.

The available data strongly support the vitamin K-dependent formation of a carbanion at the
-y-position of the Glu residue in a reaction coupled to epoxide formation. This event is followed
by carboxylation in a step not involving the vitamin. Neither the chemical nature of the proposed
oxygenated intermediate nor the mechanism by which hydrogen abstract is linked to epoxide
formation can be determined from the available data. Recent data (see text) suggests that the
key intermediate may be a I ,2-dioxetane which forms after the initial 0 2 attack. In the absence
of a Glu substrate, the enzyme will also carry out a low rate of epoxidation not linked to hydrogen
abstraction.
utilizing 'Y- 3 H-labeled Glu substrates have demonstrated a close association

between epoxide formation, Gla formation, and 'Y-C-H bond cleavage. The
efficiency of the carboxylation reaction, Gla/-y-C-H bond cleaved, was found
to be independent of Glu substrate concentrations, and the data suggest that
this ratio approaches unity at high C0 2 concentrations. These studies have
also clearly demonstrated that the enzyme will carry out the half-reaction of
epoxide formation at a diminished, but significant, rate in the absence of a
Glu site substrate. The available data are consistent with the model shown in
Figure 5, which indicates that the role of vitamin K is to abstract the x-
hydrogen to form a carbanion which can be attacked by C0 2 . The available
data do not offer conclusive proof that the abstraction event initially leaves
a carbanion. It is possible that the initial event is a radical cleavage and that
the carbon radical formed is subsequently reduced to a carbanion.
A chemical reaction, which results in the 0 2 -dependent intermolecular
condensation of diethyl adipate to ethyl cyclopentanone-2-carboxylate and
the formation of an a-keto epoxide from 2,4-dimethyl-1-naphthol has recently
been proposed as an organic model for this enzymatic carboxylation. 29 The
key step postulated is the formation of a dioxetane ring from a peroxy anion
adduct which results from the initial attack of 0 2 • The dioxetane then de-
composes to yield a strong base, the epoxy alkoxide, needed for proton
abstraction. There is good support for the proposed pathway of the model
reaction, and the available enzymological data are also consistent with this
model. However, no direct evidence for the dioxetane intermediate is available
for the enzymatic reaction.
A knowledge of the functional groups participating in the reaction would

aid studies of mechanism, but these data will depend on the availability of
significant amounts of pure enzyme. Progress in purifying the carboxylase
activity was minimal for some time, but recent reports suggest that purification
may have now been accomplished. A relatively simple procedure that yields
a roughly 1000-fold purification of the enzyme has been described, 30 and it
provides a convenient preparation for a number of studies of the enzyme. The
initial report 31 that the enzyme had been purified to homogeneity has now
been shown to represent the purification of the heat shock protein BiP, con-
taminated with the carboxylase. A second reported purification 32 of 7 x 103 -
fold has yielded a 95-kDa protein that has been cloned 33 and found not to be
related in any significant degree to other known proteins. A third report 34 of
a purification of 3.5 X 105 -fold by a significantly different strategy has
yielded a 98-kDa protein of apparent homogeneity. Which of these two pro-
teins is the carboxylase enzyme will not be known until additional studies
are completed.
2. Cell Biology of the Carboxylase

The -y-glutamyl carboxylase appears to be located on the luminal face of
the rough endoplasmic reticulum and has been shown to utilize a high mannose
form of prothrombin as a substrate. 22 The action of the enzyme can be shown
to be definitely posttranslational under some conditions, but it is difficult to
rule out the possibility that some cotranslational carboxylation may occur.
Studies of the intracellular transport of prothrombin in the rat 35 have shown
that prothrombin secretion occurs at a slower rate than that of albumin, and
that the extent of carboxylation of prothrombin has some influence on secretion
rate.
It has usually been assumed that the carboxylation of all vitamin K-
dependent proteins proceeds in the same fashion, but this may not be the
case. Wallin and Martin 36 have pointed out apparent differences between the
processing of prothrombin and factor X by rat liver microsomes and have
described some differences in the manner in which vitamin K-dependent
proteins are processed by human and rat microsomes. The significance of
these observations is not yet understood. An acquired vitamin K deficiency
produced by anticoagulant administration results in the secretion into the
plasma of large quantities of uncarboxylated and partially carboxylated forms
of vitamin K-dependent clotting factors in both the human and bovine, but
only low levels in the rat. 37 The specific alterations in the cell biology of
protein secretion responsible for this interesting species difference have not
been identified. However, the most likely explanation is that it is related to
differences in the degradation rate of prothrombin precursors within the lumen
of the endoplasmic reticulum of the two species.
Suttie 463
"r
HC-COOH
I
COOH
SH SH
t...,_,._,J
FIGURE 6. Vitamin K metabolism in liver. The vitamin K epoxide produced as a coproduct

of the carboxylation reaction is reduced by dithiol-dependent epoxide and quinone reductases to
the enzyme substrate form, the hydronaphthoquinone. These two reductive activities are the sites
of action of the clinically useful 4-hydroxycoumarin anticoagulants, and it is possible that both
activities are catalyzed by the same enzyme. The quinone form of the vitamin can also be reduced
by pyridine nucleotide-dependent reductases that are much less coumarin sensitive.
3. Relationship of Carboxylation to Other Vitamin

K-Metabolizing Enzymes
The reactions shown in Figure 6 summarize the known major metabolic
transformation of vitamin K in rat liver microsomes and point out the rela-
tionship of the carboxylase to these other activities. At least in in vitro in-
cubations, three forms of vitamin K, the quinone (vitamin K), the hydro-
quinone (vitamin KH 2 ), and the 2,3-epoxide (vitamin KO) can feed into a
hepatic vitamin K cycle. Vitamin K can be reduced to vitamin KH 2 by a
number of NAD(P)H-linked reductases including one that appears to be a
microsomal-bound form of the extensively studied liver DT-diaphorase
activitt 8 •39 and also by a dithiol-dependent reductase. 40 Vitamin KH 2 can
serve as a substrate for the microsomal carboxylase/epoxidase activity that
forms -y-carboxyglutamyl residues and converts the vitamin to vitamin KO.
This epoxide is the substrate for another microsomal enzyme, the 2,3-epoxide
reductase which reduces it to vitamin K. The epoxide reductase utilizes a
sulfhydryl compound (dithiothreitol in in vitro incubations) as a reductant and
appears to be 41 the enzyme that is the site of the physiological action of the
4-hydroxycoumarins as anticoagulants. In crude in vitro systems, dithiothrei-
tol will also reduce vitamin K to vitamin KH 2 in a 4-hydroxycoumarin-
sensitive reaction. It is likely that this reduction is catalyzed by the same
enzyme that reduces vitamin KO to vitamin K, 42 but final proof of this
relationship will depend on purification of the two activities. Recent evidence
suggests that the physiological reductant for the enzyme(s) that catalyzes the
reduction of vitamin K to vitamin KH 2 may be thioredoxin. 43 •44
III. STRUCTURE OF VITAMIN K-DEPENDENT

PLASMA PROTEINS
The vitamin K-dependent plasma proteins are a family of structurally
related glycoproteins that represent less than 0.5% of the total plasma protein
pool. Their distinguishing feature is the presence of from 9 to 13 Gla residues
in an amino terminal domain. The majority of these proteins are zymogen
forms of a serine protease, and a number of them contain a second post-
translational modification in the form of 13-hydroxyaspartate or 13-hydrox-
yasparagine.
Early data, based on partial amino acid sequences, indicated that there
was a great deal of homology in the amino terminal, Gla-containing region
of the vitamin K-dependent proteins. The complete amino acid sequence of
prothrombin was established45 by Magnusson in 1975, but because of the
small quantities of material available, determination of the primary sequence
of the other vitamin K-dependent proteins, particularly factor VII, represented
a formidable task in protein chemistry. By the late 1970s it was clear that the
primary sequence of these proteins could be more easily obtained by utilizing
the tools of molecular biology than through direct protein sequencing. Our
current understanding of the amino acid sequence of the vitamin K-dependent
proteins is, therefore, based on both direct peptide sequencing and, particularly
in the case of the human forms, an analysis of the eDNA structure. 46 The
data presented in Table 1 indicate the close structural homology of the amino
terminal Gla-containing region of some of the vitamin K-dependent proteins.
In addition to a homologous Gla domain, prothrombin (factor II), factor VII,
factor IX, factor X, and protein C share a serine protease domain containing
the reactive serine and neighboring histidine and aspartate residues first iden-
tified in pancreatic digestive enzymes such as trypsin, chymotrypsin, and
elastase. This serine protease homology is shared with the nonvitamin K-
dependent proteins of the plasma coagulation system, Kallikrein and factors
XI and XII, as well as plasmin and plasminogen activators of the fibrin lytic
systems. 46
The availability of the complete sequence of the primary gene product of
all of the vitamin K-dependent proteins has provided information on sequence
homologies which were not previously apparent. Prothrombin contains two
Suttie 465
"Kringle" domains. These looped sequences. named for their resemblance

to the Danish pastry, are also found in the zymogen of the major enzyme of
fibrinolysis, plasminogen, and its activators, urokinase, and tissue plasmin-
ogen activator, but not in the other vitamin K-dependent clotting factors. With
the exception of prothrombin, the plasma vitamin K-dependent clotting factors
also contain one or more epidermal growth factor (EGF) precursor-like do-
mains. The second important posttranslational modification of the vitamin K-
dependent protein, ~-hydroxylation of aspartyl residues, is located in this
region. Although it appears that this residue confers Ca 2 + -binding properties
to the vitamin K-dependent proteins that are independent of the Gla mediated
interactions, the details of this role have not been determined.
The circulating forms of the vitamin K-dependent plasma proteins are,
therefore, composed of a number of common domains. 46 . 49 A homologous
region containing the Gla residues that gives these proteins their ability to
form Ca 2 +-dependent interactions with negatively charged phospholipids is
present at the amino terminus. With the exception of proteinS. these proteins
circulate as zymogens and only in the case of prothrombin does their proteo-
lytic activation release the Gla domain. In the case of prothrombin, the amino
terminal Gla domain is followed by two Kringle domains, and the other
proteins by two or more EGF domains. The remainder of these proteins consist
of regions of unidentified function and the carboxy terminal serine protease
domain. These relationships are illustrated in Figure 7.
The vitamin K-dependent proteins are relatively large, contain a signif-
icant amount of carbohydrate, and are readily degraded by protease action.
Because of these factors, strongly defracting crystals of the intact proteins
have been difficult to obtain, and progress in obtaining their tertiary structure
by X-ray crystallography has moved forward slowly. Most of the effort has
centered on prothrombin and has been directed toward obtaining the structure
of the fragment 1 portion of bovine prothrombin. Early efforts yielded models
of low resolution, and detailed locations of individual amino acids could not
be determined. Park and Tulinsky 50 then obtained a 2.8 A resolution structure
of bovine fragment 1. This structure, and subsequent refinements, provided
a clear picture of the Kringle region of fragment I , but the first 35 residues
of this peptide, which contain all of the Gla residues, were disordered. A
second structure 51 eventually at 2.2 A resolution 52 containing 7 Ca2 + atoms
is now available. This solution has provided a clear model of the manner in
which the charges of the Gla residues are ordered to provide two internal
carboxylate surfaces of high negative charge density which hold the Ca2 +
ions. A detailed explanation of the interactions between the remaining Gla
residues, additional Ca 2 + ions, and a negatively charged phospholipid surface
is, however, still lacking. Molecular dynamics simulations based on the avail-
able crystallographic data have also provided insights into the conformational
changes resulting from interactions with Ca2 + ionsY
TABLE 1 0\
0\
""'
Amino Acid Sequence of the "Gia" Region of Selected Vitamin
K-Dependent Plasma Proteins
+I +10
F-11 Human Ala Asn Thr Phe Leu Gla Gla Val Arg Lys Gly Asn Leu Gla Arg
Bovine Ala Asn Lys Gly Phe Leu Gla Gla Val Arg Lys Gly Asn Leu Gla Arg
Chicken Ala Asn Lys Gly Phe Leu Gla Gla Met lie Lys Gly Asn Leu Gla Arg
Rat Ala Asn Ser Gly Phe Leu Gla Gla Leu Arg Lys Gly Asn Leu Gla A
F-X Human Ala Asn Ser Phe Leu Gla Gla Met Lys Lys Gly His Leu Gla Arg
F-IX Human Tyr Asn Ser Gly Lys Leu Gla Gla Phe Val Gin Gly Asn Leu Gla Arg
F-VII Human Ala Asn Ala Phe Leu Gla Gla Leu Arg Pro Gly Ser Leu Gla Arg
P-C Human Ala Asn Ser Phe Leu Gla Gla Leu Arg His Ser Ser Leu Gla Arg
P-S Human Ala Asn Ser Leu Leu Gla Gla Thr Lys Gin Gly Asn Leu Gla Arg
+20 +30
F-11 Human Gla Cys Val Gla Gla Thr Cy> Ser Tyr Gla Gla Ala Phe Gla Ala Leu
Bovine Gla Cys Leu Gla Gla Pro Cys Ser Arg Gla Gla Ala Phe Gla Ala Leu
Chicken Gla Cys Leu Gla Gla Thr Cys Asn Tyr Gla Gla Ala Phe Gla Ala Leu
Rat Gla Cys Val Gla Gla Gin Cys Ser Tyr Gla Gla Ala Phe Gla Ala Leu :;;..:
;;:::
F-X Human Gla Cys Met Gla Gla Thr Cys Ser Tyr Gla Gla Ala Arg Gla Val Phe
F-IX Human Gla Cys Me Gla Gla Lys Cys Ser Phe Gla Gla Ala Arg Gla Val Phe "~:
F-VII Human Gla Cys Lys Gla Gla Gin Cys Ser Phe Gla Gla Ala Arg Gla lie Phe ;::
$:l
P-C Human Gla Cys lie Gla Gla lie Cys Asp Phe Gla Gla Ala Lys Gla lie Phe ;::
P-S Human Gla Cys lie Gla Gla Leu Cys Asn Lys Gla Gla Ala Arg Gla Val Phe i:l...
+40 C'l
;::
F-11 Human Gla Ser Ser Thr Ala Asp Val Phe Trp Ala
I Lys Tyr "'
Bovine Gla Ser Leu Ser Ala hr Asp Ala Phe Trp Ala Lys Tyr "'~
Chicken Gla Ser Thr Val Asp hr Asp Ala Phe Trp ':::i
Rat Gla Ser Pro Gin Asp hr Asp Val Phe Lys Tyr ~
Trp Ala
I
F-X Human Gla Asp Ser Asp Lys hr Asn Gla Phe Trp Asn Lys Tyr """"c:;·
F-IX Human Gla Asn Thr Gla Arg hr Thr Gla Phe Trp Lys Gin Tyr ;::
~1
F-VII Human Lys Asp Ala Gla Lys Leu Ser Tyr ~
Arg 'Thr Phe Trp Ile
I s:::
P-C Human Gln Asn Val Asp Asp Thr Leu Ala Phe Trp Ser Lys His ::::
Gla Asn Asp Pro Asp Tyr Phe Tyr Pro Lys Tyr
;;;·
P-S Human Gla Thr
Note: References to the primary literature can be obtained in reviews46 A 9 with the exception of the rat sequence 111
which is identical to that of the mouse 111 Although the sequence of the Gla region of most of these proteins
was determined by peptide sequences, complete structures for most were obtained only when eDNA structures
were available.
.,..
0\
-...l
prothromb'n l K K
factor VII l E '

factor IX l EE ' '
factor X I l E I E II I '
protein C E E '
proteinS l E EE E
100 amino acids
FIGURE 7. Structural domains of the vitamin K-dependent plasma clotting factors. The cleav-
age site(s) that give rise to the active proteases are indicated by arrows. Abbreviations used:
y, y-carboxyglutamic acid-containing region; K, kringle; E, epidermal growth factor-like region.
The serine protease catalytic region is shown as a solid bar. The structures are drawn to scale,
and the 100-residue length is indicated. Adapted from Fung and MacGillivray.••
A. GENOMIC STRUCTURE OF VITAMIN

K-DEPENDENT PROTEINS
The genes for the vitamin K-dependent proteins have been characterized
by DNA sequence analysis, and the similarities of their structures have been
reviewed. 46 .49 The relationship between factors IX, X, and protein Care much
closer than the relationship of this group to prothrombin. Although they all
share strong similarities in the exon coding for the Gla and preproleader
regions, they definitely diverge to two families downstream of the Gla exon.
Not only does prothrombin contain three exons coding for the two Kringle
regions rather than the EGF coding exons, but differences exist in the re-
mainder of the gene. The serine protease (thrombin) region of prothrombin
is coded for by five different exons which do not share homology in terms
of intron sites with the factor IX family. The basis for the divergent evolu-
tionary pathway of these closely related proteins is not yet defined, but a
clear definition of the gene structure of all of the vitamin K-dependent proteins
is nearly complete. 54
B. CONGENITAL DEFICIENCIES OF VITAMIN K-DEPENDENT

CLOTTING FACTORS
Classical hemophilia, a congenital factor VIII deficiency, is a sex-linked
recessive characteristic that is the most common hereditary clotting defect in
the human population. Severe hemophilia is found in about 1 per 25,000 of
the population, and, when mildly affected patients are considered, the inci-
dence of the disease might be as high as 1 in every 3000 to 4000 live births. 55
Factor VIII is not a vitamin K-dependent factor, but in many cases the bleeding
Suttie 469
tendency resulting from a factor VIII deficiency can be cured by increasing

the concentrations of factor VII •. The disease state has been known for many
years to include a large number of variants, and these are rapidly being defined
by modem methods of gene analysis.
Three of the classical vitamin K-dependent clotting factors, factors VII,
IX, and X, were first discovered because of a congenital abnormality that
resulted in a coagulation disorder. Christmas Disease, or hemophilia B, is a
congenital deficiency of factor IX that is present in the human population at
about 15% the incidence of hemophilia A. It is a sex-linked recessive bleeding
disorder that is inherited in the same manner as hemophilia A. The disease
encompasses a heterogeneous group of molecular disorders, and about half
of the affected individuals have severely reduced circulating antigen levels
while the remainder have circulating antigenically similar, but biologically
less active forms of the protein. 56 Congenital deficiencies of prothrombin,
factor X, and factor VII are much less common, are autosomally transmitted,
and exist in a number of forms that may or may not be associated with
circulating biologically inactive, but immunochemically related forms of the
proteins.
As protein C and protein S function as natural anticoagulants rather than
procoagulants, a congenital defect in these proteins might be expected to lead
to a thrombotic rather than a hemorrhagic state. Griffin et a!.5 9 first described
a patient with recurrent thrombophlebitis who was shown to have a decreased
protein C concentration in plasma, and numerous cases of thrombin disorders
associated with a heterozygous congenital protein C deficiency have been
reported. It is now clear that structural variants or decreased concentrations
of either protein C or protein S can result in thrombotic episodes. The extent
of these congenital abnormalities in the population is not yet known, but it
is likely that a significant portion of the population with recurrent thrombotic
disorders of undiagnosed etiology may have an abnormality in one of these
proteins. 58 ·59
In addition to congenital defects of vitamin K-dependent proteins arising
from a mutation in the gene directing the synthesis of a specific clotting
factor, it is possible that an enzyme involved in the posttranslational vitamin
K-dependent carboxylation of these proteins or in another biosynthetic pro-
cessing step could be altered. Such an event might result in a multiple de-
ficiency disease, and a limited number of cases of hemostatic disorders as-
sociated with altered concentrations or activities of all of the vitamin K-
dependent proteins have been reported. One of the best studied60 has been a
female with severe functional abnormalities of all of the vitamin K-dependent
coagulation factors. Immunologic and nonphysiologic measures indicated the
presence of abnormal, dysfunctional proteins. This individual showed partial
correction following administration of vitamin K and recurrence of clinical
symptoms following discontinuation of therapy.
A mutation in a protein essential to intracellular protein processing, a

defect in the carboxylase itself, or an interference with vitamin K metabolism
are the most logical explanations for a functional abnormality of all of the
vitamin K-dependent proteins. A child who maintains about 40% of the normal
levels of prothrombin, factor VII, factor IX, and factor X on a daily intake
of 10 mg of vitamin K has been described by Pauli et al. 61 This patient also
exhibited a high circulating concentration of vitamin K epoxide, suggesting
that the defect is an inability to metabolize the epoxide to vitamin K. This
metabolic alteration is the same as that produced as the result of oral anti-
coagulant therapy and would be consistent with a partial reversal of the
congenital defect by high doses of vitamin K. Whether or not the same
abnormality is present in other individuals with combined defects of vitamin
K-dependent proteins, has not been established.
IV. THE PROPEPTIDE RECOGNITION SITE
The vitamin K-dependent proteins comprise approximately 1% of the

total proteins synthesized by the liver and secreted into the plasma. This minor
intracellular pool is, however, efficiently recognized, and specific Glu resi-
dues are selectively carboxylated by the vitamin K-dependent carboxylase
located at the lumenal surface of the endoplasmic reticulum. Some indication
of the basis for the selectivity of carboxylation was obtained when it was
shown that the intracellular precursor form of prothrombin was larger and
more basic than the secreted plasma form 62 and that isolated intracellular
forms of prothrombin were much better substrates for the in vitro carboxylase
than des--y-carboxy plasma prothrombin. 63 . 65 These observations were con-
sistent with structural information obtained through eDNA sequencing 46 which
revealed that the primary gene product of all of the vitamin K-dependent
plasma proteins contained a basic amino acid hydrophobic residue rich ''pro-
peptide'' between the signal peptide region of the primary gene product and
the amino terminal end of the mature plasma form of the protein. These
observations suggest that this homologous amino terminal extension (Table
2) might play an important role in the recognition of these proteins by the
microsomal carboxylase. This hypothesis was strengthened by the observation66
that an amino terminal extension homologous to this region was also present
in the precursor form of the vitamin K-dependent bone Gla protein.
A. CELLULAR EXPRESSION SYSTEMS AND

PROPEPTIDE FUNCTION
The possibility that recombinant clotting factors, particularly factor VII,
factor IX, and protein C, might have a role as hemostatic or antithrombotic
drugs made them an attractive target for attempts to obtain biologically active
proteins through the use of mammalian expression systems. Early studies
demonstrated rF-IX production in hepatoma, 67•68 baby hamster kidney (BHK), 69
TABLE 2 Vl
Amino Acid Sequences in the Propeptide Region of Human Vitamin §

(i;•
K-Dependent Plasma Proteins
Protein -20 -10

Prothrombin Leu Val His Ser Gln His Val Phe
Leu Ala Pro Gln Gin Ala
Factor X Leu Leu Leu Gly Glu Ser Leu Phe
lle Arg Arg Glu Gin Ala
Factor IX Leu Ser Ala Glu Cys Thr Val Phe
Leu Asp His Glu Asn Ala
Factor VII Gln Gly Cys Leu Ala Ala Val Phe
Val Thr Gln Glu Glu Ala
Protein C Pro Ala Pro Leu Asp Ser Val Phe
Ser Ser Ser Glu Arg Ala
ProteinS Leu Pro Val Ser Glu Ala Asn Phe
Leu Ser Lys Gln Gin Ala
-I +I
Prothrombin Arg Ser Leu Leu Gln Arg Val Arg Arg Ala
Factor X Asn Asn lie Leu Ala Arg Val Thr Arg Ala
Factor IX Asn Lys lle Leu Asn Arg Pro Lys Arg Tyr
Factor VII His Gly Val Leu His Arg Arg Arg Arg Ala
Protein C His Gln Val Leu Arg Ile Arg Lys Arg Ala
ProteinS Ser Gln Val Leu Val Arg Lys Arg Arg Ala
Note: The amino terminal residue of the mature form of the plasma protein is designated as
residue + I. The site of signal peptide cleavage for factor IX is between residues - 19
and - IS"I.I" and between -24 and - 25 for protein C. 77 The position of cleavage
has not been determined with certainty for all of the other factors.
....~
or Chinese hamster ovary (CH0)70 cells, and rF-VII production in BHK

cells. 71 The vitamin K-dependent proteins are complex and for full activity
require -y-glutamyl carboxylation, glycosylation, disulfide bond formation,
proteolytic processing, and often 13-aspartyl hydroxylation. These problems
and their relationship to the low levels of expression of biologically active
recombinant clotting factors that were obtained in most of the early studies
have been discussed. 72 Early attempts at expression of vitamin K-dependent
proteins suggested that -y-glutamyl carboxylation often became limiting as
cellular expression levels were increased. All of the vitamin K-dependent
plasma proteins have now been expressed in cultured mammalian cell lines,
and by selection of appropriate cell lines and expression vectors, increased
quantities of biologically active recombinant clotting factors can now be
produced. Of these efforts, the product closest to commercial use is probably
rhF-VII which has been tested and found effective as a hemostatic agent in
both animal models 73 •74 and in limited clinical situations. 75 •76
The availability of these well-characterized cellular expression systems
has been utilized to test the hypothesis that the homologous amino terminal
propeptide of the vitamin K-dependent proteins is an important signal for )'-
carboxylation. Both rPC 77 and rF-IX 78 secreted by cells transfected with con-
structs which had been modified so that the propeptide region was deleted
were found to lack Gla residues. A second posttranslational modification of
these proteins, 13-aspartyl hydroxylation, was, however, not dependent on
this region of the primary gene product. 79 Further studies78 ·80 demonstrated
that the well conserved - 16 Phe, - 10 Ala, and -6 Leu residues of the
propeptide region (see Table 2) are of particular importance in directing
efficient carboxylation. Residues -1 to -4 are presumably more important
in directing the action of a propeptidase that is required for cleavage of the
amino terminal extension prior to secretion. A natural F-IX variant with a
mutation of the -4 Arg residue to a Gin has been shown81 to secrete F-IX
with the propeptide still attached. This variant lacks much of the normal
biological activity, but does retain most of the Gla content of normal F-IX. 80
Further studies of the cellular expression of rF-11 with specific mutations in
the propeptide region 25 confirmed the importance of the - 10 Ala and - 16
Phe residues and demonstrated that the entire amino terminal region of the
propeptide (residues - 14 to - 18) were important for efficient carboxylation.
Although mutations in all residues was not investigated, it was demonstrated
that mutations at the - 14 Ala and -8 Val residues did not impair carbox-
ylation. Much of the work in this area has recently been summarized. 83
B. CARBOXYLASE SUBSTRATES AND PROPEPTIDE FUNCTION

The importance of specific regions or residues of the propeptide for
carboxylase recognition has also been assessed through the synthesis of peptide
substrates for the carboxylase. The enzyme is routinely assayed with a tri or
penta peptide containing a Glu-Glu sequence such as Phe-Leu-Glu-Glu-Leu
Suttie 473
(FLEEL) as a substrate. 22 These substrates have Km values in the range of a

few mM. Ulrich et a!. 84 demonstrated that a 28-amino acid peptide consisting
of residue - 18 to + 10 of human prothrombin and containing the + 6 and
+ 7 Glu sites was carboxylated with a Km of 3.6 fLM, about 500- to 1000-
fold lower than FLEEL. Peptides comprising the first 10 residues of pro-
thrombin ( + I to + 10) of this region and a portion of the propeptide region
(- 10 to + 10) were no better than FLEEL as a substrate. The importance of
the - 16 Phe residue of the propeptide region was demonstrated by the syn-
thesis of the 28-residue peptide with Ala at the -16 position. Carboxylation
of this substrate was significantly impaired when compared to the native Phe-
containing peptide. An extension of these initial studies 85 demonstrated that,
at least in vitro, the hF-11 (prothrombin) propeptide was more efficient at
directing carboxylation than the hF-IX propeptide, and that the carboxylation
of peptides containing a Ser rather than an Arg at residues - 1 or -4 is not
drastically impaired. These are the residues altered in two characterized ge-
netic variants of F-IX, 80 and the lack of involvement of these two residues
in directing carboxylation has also been demonstrated through the use of
substrates representing the -18 to +41 region of hF-IX. 86 This substrate,
and variants of it, were produced as recombinant fusion proteins in E. coli.
Price et a!. 87 have identified a consensus sequence, Glu-X-X-X-Glu-X-Cys,
within all known vitamin K-dependent proteins, and have suggested that it,
as well as the propeptide region, serve as a recognition site for the substrate/
carboxylase interaction. Comparison of the in vitro carboxylation of a peptide
comprising residues - 18 to + 36 of prothrombin, which contains the con-
sensus sequence, with the carboxylation of a -18 to + 10 prothrombin hom-
olog, which lacks this region, suggests 82 that this suggested consensus se-
quence is not important in directing in vitro carboxylation.
C. DUAL ROLE OF THE PROPEPTIDE

Both the in vitro and cellular expression studies are consistent with the
hypothesis that the carboxy terminal region of the propeptide contains a
recognition site for binding to a propeptidase, and that residue - 10 and the
amino terminal region of the propeptide serves to recognize the carboxylase
and to serve a "docking" function. Knobloch and Suttie88 have shown that
a 20-residue peptide containing the octadeca propeptide of human factor X
strongly stimulates the activity of the rat liver vitamin K-dependent carbox-
ylase when assayed with small peptide substrates such as FLEEL. This in-
creased activity brought about by the noncovalently bound propeptide is mainly
the result of a lowering of the Km at the Glu binding site of the carboxylase. 89
These data suggest that this region of the vitamin K-dependent protein serves
both a regulatory and a recognition function. When these studies were ex-
tended,90 it was demonstrated that the - 10 Ala and -16 Phe that had been
shown to be important in directing carboxylation in cellular expression systems
and in covalently linked substrates were also important in influencing the
interaction of enzyme and substrate at the Glu site. The ability of the free
propeptide to stimulate the carboxylase is not limited to weak binding high
Km substrates. Des-'Y-carboxy Bone Gla Protein (dBGP) is not a particularly
good carboxylase substrate, but in the presence of free propeptide it is effi-
ciently carboxylated91 with a Km of only 11 !J.M. These data indicating a role
of the propeptide in modulating the Glu binding site of the enzyme do not
suggest that the presumed docking function of the propeptide is not important.
A second protein originally found in bone, MGP, does not have an amino
terminal propeptide cleaved during protein processing. Rather, it has an in-
ternal sequence in the mature protein which is homologous to this region. 87
The Km of dMGP as a substrate is only 0.2 !J.M, and its carboxylation is not
stimulated, but is effectively inhibited by free propeptide, 91 presumably by
displacing it from the enzyme. These data make it clear that the propeptide
region of the few proteins that are substrates for the carboxylase serves a dual
role, it anchors or docks the substrate with the enzyme, and also alters the
affinity of the enzyme for the Glu residue which will be carboxylated.
Why both a docking and regulatory role for the propeptide should be
required for normal carboxylation is not readily apparent. The manner by
which the enzyme carboxylates its multisite substrates is not known, and it
is possible that some Glu (potential Gla) sites may have a high affinity for
the enzyme, or that it is an ordered series of events. The location of the Gla
residues in partially carboxylated forms of prothrombin92 suggests that the
most amino terminal Gla site is most readily carboxylated but that there may
also be high affinity sites elsewhere in the Glu region. The published studies
of the in vitro carboxylation of numerous peptides with multiple Glu sites
have not addressed the question of which of the possible sites are carboxylated.
This question is not easily addressed in a substrate with ten potential Gla
sites, but is more easily approached by studies of the carboxylation of dBGP
which has only three carboxylated Glu residues. Preliminary data from this
system93 suggests that the most carboxy terminal Glu is preferentially car-
boxylated, but the influence of the propeptide region on the pattern of car-
boxylation has not yet been determined.
V. GENE EXPRESSION AND VITAMIN

K-DEPENDENT PROTEINS
Both rat (H-35) and human (HepG2) hepatoma cell lines 94. 96 have been
shown to secrete prothrombin and other vitamin K-dependent proteins. 95 •97
The general observation has been that warfarin treatment increases the intra-
cellular pool of prothrombin and decreases secretion into the culture medium,
while vitamin K administration increases secretion. It is not clear from the
available data if these responses are related in any way to changes in the rate
of expression of the prothrombin gene, or if they merely represent a response
to the altered ability of the cells to effectively carry out the 'Y-glutamyl
Suttie 475
carboxylation step. The H-35 rat hepatoma synthesis systems have also been
used 98 to demonstrate an increase of prothrombin synthesis upon the addition
of the amino terminal, fragment-! portion (residues 1 through 156) of pro-
thrombin to the media. Indirect evidence suggested that this response was
related to an effect on prothrombin gene expression, but no direct evidence
was presented, and the observation has not been repeated.
Plasma concentrations of vitamin K-dependent clotting factors are nor-
mally maintained in a narrow range, and there is some evidence for a humoral
factor that is involved in this control. Karpatkin and his co-workers 90 · 100 have
demonstrated that the intravenous injection of a small amount of plasma
obtained from an anticoagulant-treated rabbit with a low level of vitamin K-
dependent clotting factors into a normal rabbit will result in an increase in
the activities of these factors in the recipient rabbit. The humoral factor
responsible for this response has been called a coagulopoietin(s). This response
has also been observed in the rat, 101 and there are indications 101 . 103 that this
treatment increases biological activity of prothrombin more than prothrombin
antigen. It has been demonstrated 104 that this factor appears early during the
development of a deficiency state and that the activity is variable between
different animals. The chemical nature of the proposed coagulopoietin is not
known, and the current data do not clearly differentiate an effect on expression
of the vitamin K-dependent proteins, or of components of the carboxylase
system.
The concentration of undercarboxylated prothrombin in normal human
plasma is very low and increases substantially with oral anticoagulation or
vitamin K deficiency, and slightly in many patients with liver disease. 105 It
has been shown 106 · 107 that this protein is increased more dramatically in cases
of primary hepatocellular carcinoma than in other liver disorders, and its
presence can be used to screen for this disease. This increase is independent
of vitamin K status, and the molecular basis of the observation has been
studied in a rat model. 108 A series of Morris hepatoma tumors were trans-
planted into recipient Buffalo strain rats. Transplantation of some but not all
tumor lines resulted in increased levels of circulating undercarboxylated pro-
thrombin, and the tumor mass had low levels of carboxylase. The tumor mass
of those lines not secreting this marker had much higher levels of the car-
boxylase. The data suggest that the secretion of abnormal prothrombin by
hepatocellular tumors is the result of normal expression of the prothrombin
gene and a failure of the tumor to fully express the carboxylase gene.
These examples indicate that neither the role of vitamin K, if any, in
regulating gene expression, nor the role of other factors, which influence
gene expression on vitamin K metabolism, is yet understood. Progress in this
area should, however, be rapid. The vitamin K-dependent proteins themselves
have been cloned, and mRNA probes have been successfully used to follow
the regulation of prothrombin synthesis during fetal and neonatal develop-
ment. 109 • 110 The currently confused status of the possible regulation of clotting
factor genes by vitamin K status should therefore soon be clarified. The

apparent purification and cloning of the carboxylase will also make available
the tools with which to probe the factors which regulate the expression of
this unique and important enzyme.
REFERENCES
1. Darn, H., Hemorrhages in chicks reared on artificial diets: a new deficiency disease,
Nature (London), 133, 909, 1934.
2. Suttie, J, W., Vitamin K, inHandbookofVitamins, 2nded., Machlin, L. J., Ed., Marcel
Dekker, New York, 1991.
3. Bentley, R. and Meganathan, R., Biosynthesis of vitamin K (menaquinone) in bacteria,
Microbial. Rev., 46, 241, 1982.
4. Suttie, J, W., Vitamin K, in The Fat-Soluble Vitamins, Diplock, A. T., Ed., William
Heinemann, London, 1985, 225.
5. Darn, H., Schonheyder, F., and Tage-Hansen, E., Studies on the mode of action of
vitamin K, Biochem. J., 30, 1075, 1936.
6. Stenflo, J,, A new vitamin K-dependent protein. Purification from bovine plasma and
preliminary characterization, J. Bioi. Chern., 251, 355, 1976.
7. Di Scipio, R. G., Herrnodson, M.A., Yates, S. G., and Davie, E. W., A comparison
of human prothrombin, factor IX (Christmas factor), factor X (Stuart factor), and protein
S, Biochemistry, 16, 698, 1977.
8. Price, P. A., Role of vitamin K-dependent proteins in bone metabolism, Annu. Rev.
Nutr., 8, 565, 1988.
9. Link, K. P., The discovery of dicumarol and its sequels, Circulation, 19, 97, 1959.
10. Olson, R. E., Vitamin K induced prothrombin formation: antagonism by actinomycin
D, Science, 145, 926, 1964.
11. Suttie, J. W., Metabolism and properties of a liver precursor to prothrombin, Vitam.
Horm., 32, 463, 1974.
12. Stenflo, J., Fernlund, P., Egan, W., and Roepstorff, P., Vitamin K dependent mod-
ifications of glutamic acid residues in prothrombin, Proc. Nat/. Acad. Sci. U.S.A., 71,
2730, 1974.
13. Nelsestuen, G. L., Zytkovicz, T. H., and Howard, J, B., The mode of action of
vitamin K. Identification of -y-carboxyglutamic acid as a component of prothrombin. J.
Bioi. Chern., 249, 6347, 1974.
14. Magnusson, S., Sottrup-Jensen, L., Petersen, T. E., Morris, H. R., and Dell, A.,
Primary structure of the vitamin K-dependent part of prothrombin, FEBS Lett., 44, 189,
1974.
15. Stenflo, J, and Suttie, J, W., Vitamin K-dependent formation of -y-carboxyglutamic
acid, Annu. Rev. Biochem., 46, 157, 1977.
16. Esrnon, C. T., Sadowski, J, A., and Suttie, J. W., A new carboxylation reaction. The
vitamin K-dependent incorporation of H14C03 into prothrombin, J. Bioi. Chern., 250,
4744, 1975.
17. Suttie, J, W., Mechanism of action of vitamin K: synthesis of Y-carboxyglutamic acid,
CRC Crit. Rev. Biochem., 8, 191, 1980.
18. Johnson, B. C., Post-translational carboxylation of preprothrombin, Mol. Cell Biochem.,
38, 77, 1981.
Suttie 477
19. Vermeer, C. and de Boer-van den Berg, M.A. G., Vitamin K-dependent carboxylase,
Haematologia, 18, 71, 1985.
20. Olson, R. E., The function and metabolism of vitamin K, Annu. Rev. Nutr., 4, 281,
1984.
21. Friedman, P. A. and Przysiecki, C. T., Vitamin K-dependent carboxylation, Int. J.
Biochem., 19, I, 1987.
22. Suttie, J, W., Vitamin K-dependent carboxylation of glutamyl residues in proteins,
BioFactors, 1, 55, 1988.
23. Uotila, L., Vitamin K-dependent carboxylase from calf liver: studies on the steady-state
kinetic mechanism, Arch. Biochem. Biophys., 264, 135, 1988.
24. Friedman, P. A., Shia, M.A., Gallop, P.M., and Griep, A. E., Vitamin K-dependent
-y-carbon-hydrogen bond cleavage and the non-mandatory concurrent carboxylation of
peptide bound glutamic acid residues, Proc. Nat/. Acad. Sci. U.S.A., 76, 3126, 1979.
25. Larson, A. E., Friedman, P. A., and Suttie, J. W., Vitamin K-dependent carboxylase:
stoichiometry of carboxylation and vitamin K 2,3-epoxide formation, J. Bioi. Chern.,
256, 11032, 1981.
26. McTigue, J. J. and Suttie, J. W., Vitamin K-dependent carboxylase: demonstration of
a vitamin K- and 0 2-dependent exchange of 'H from 3 H2 0 into glutamic acid residues,
J. Bioi. Chern., 258, 12129, 1983.
27. Anton, D. L. and Friedman, P. A., Fate of the activated -y-carbon-hydrogen bond in
the uncoupled vitamin K-dependent -y-glutamyl carboxylation reaction, J. Bioi. Chern.,
258, 14084, 1983.
28. Wood, G. M. and Suttie, J, W., Vitamin K-dependent carboxylase. Stoichiometry of
vitamin K epoxide formation, -y-carboxyglutamyl formation, and -y-glutamyl-3H cleavage,
J. Bioi. Chern., 263, 3234, 1988.
29. Dowd, P., Ham, S. W., and Geib, S. J,, Mechanism of action of vitamin K, J. Am.
Chern. Soc., 113, 7734, 1991.
30. Harbeck, M. C., Cheung, A. Y., and Suttie, J, W., Vitamin K-dependent carboxylase:
partial purification of the enzyme by antibody affinity techniques, Thrombosis Res.. 56,
317, 1989.
31. Hubbard, B. R., Ulrich, M. M. W., Jacobs, M., Vermeer, C., Walsh, C., Furie,
B., and Furie, B. C., Vitamin K-dependent carboxylase: affinity purification from bovine
liver by using a synthetic propeptide containing the -y-carboxylation recognition site,
32. Wu, S-M., Morris, D. P., and Stafford, D. W., Identification and purification to near
homogeneity of the vitamin K-dependent carboxylase, Proc. Nat/. Acad. Sci. U.S.A.,
88, 2236, 1991.
33. Wu, S-M., Cheung, W-F., Frazier, D., and Stafford, D., Cloning and expression of
the eDNA for human -y-glutamyl carboxylase, Science, 254, 1634, 1991.
34. Berkner, K. L., Harbeck, M., Lingenfelter, S., Bailey, C., Sanders-Hinck, C. M.,
and Suttie, J, W., Purification and identification of the bovine liver-y-carboxylase, Proc.
Nat/. Acad. Sci. U.S.A., 89, 6242, 1992.
35. Kvalvaag, A. H., Tollersrud, 0. K., and Helgeland, L., A study on the intracellular
transport of prothrombin, albumin and transferrin in rat, Biochim. Biophys. Acta, 937,
319, 1988.
36. Wallin, R. and Martin, L. F., Early processing of prothrombin and factor X by the
vitamin K-dependent carboxylase, J. Bioi. Chern., 263, 9994, 1988.
37. Shah, D. V., Swanson, J, C., and Suttie, J, W., Abnormal prothrombin in the vitamin
K-deficient rat, Thrombosis Res., 35, 451, 1984.
38. Wallin, R., Gebhardt, 0., and Prydz, H., NAD(P)H dehydrogenase and its role in
the vitamin K (2-methyl-3-phytyl-1 ,4-naphthoquinone)-dependent carboxylation reaction,
Biochem. J., 169, 95, 1978.
39. Fasco, M. J. and Principe, L. M., Vitamin K 1 hydroquinone formation catalyzed by

DT-diaphorase, Biochem. Biophys. Res. Commun., 104, 187, 1982.
40. Fasco, M. J. and Principe, L. M., Vitamin K 1 hydroquinone formation catalyzed by
a microsomal reductase system, Biochem. Biophys. Res. Commun., 97, 1487, 1980.
41. Suttie, J. W., Vitamin K antagonists, in Hemostasis and Thrombosis: Basic Principles
and Clinical Practice, 3rd ed., Colman, R. W. and Hirsh, J., Eds., J. B. Lippincott,
Philadelphia, in press.
42. Gardill, S. L. and Suttie, J. W., Vitamin K epoxide and quinone reductase activities:
evidence for reduction by a common enzyme, Biochem. Pharmacal., 40, 1055, 1990.
43. van Haarlem, L. J. M., Soute, B. A. M., and Vermeer, C., Vitamin K-dependent
carboxylase. Possible role for thioredoxin in the reduction of vitamin K metabolites in
liver, FEBS Lett., 222, 353, 1987.
44. Silverman, R. B. and Nandi, D. L., Reduced thioredoxin: a possible physiological
cofactor for vitamin K epoxide reductase. Further support for an active site disulfide,
45. Magnusson, S., Petersen, T. E., Sottrup-Jensen, L., and Claeys, H., Complete pri-
mary structure of prothrombin: Isolation, structure and reactivity of ten carboxylated
glutamic acid residues and regulation of prothrombin activation by thrombin, in Proteases
and Biological Control, Vol. 2, Reich, E., Rifkin, D. B., and Shaw, E., Eds., Cold
Spring Harbor Conf. on Cell Proliferation, 1975, 123.
46. Davie, E. W., The blood coagulation factors: their cDNAs, genes, and expression, in
Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 2nd ed., Colman,
R. W., Hirsh, J., Mardar, V. J., and Salzman, E. W., Eds., Lippincott, Philadelphia,
1987, 242.
47. Walz, D. A., Hewett-Emmett, D., and Guillin, M-C., Amino acid sequences and
molecular homology of the vitamin K-dependent clotting factors, in Prothrombin and
Other Vitamin K Proteins, vol. I, Seegers, W. H. and Walz, D. A., Eds., CRC Press,
Boca Raton, FL, 1986, 125.
48. Forie, B. and Forie, B. C., The molecular basis of blood coagulation, Cell, 53, 505,
1988.
49. Fung, M. R. and MacGillivray, R. T. A., Organization of the genes coding for the
vitamin K-dependent clotting factors, in Current Advances in Vitamin K Research, Suttie,
J. W., Ed., Elsevier, New York, 1988, 143.
50. Park, C. H. and Tulinsky, A., Three dimensional structure of the Kringle sequence:
structure of prothrombin fragment I, Biochemistry, 25, 3977, 1986.
51. Soriano-Garda, M., Park, C. H., Tulinsky, A., Ravichandran, K. G., and Skrzyp-
czak-Jankun, E., Structure of Ca2 + prothrombin fragment I including the conformation
of the Gla domain, Biochemistry, 28, 6805, 1989.
52. Soriano-Garda, M., Padmanabhan, K., deVos, A.M., and Tulinsky, A., TheCa'+
ion and membrane binding structure of the Gla-domain of Ca-prothrombin fragment I,
Biochemistry, in press.
53. Charifson, P. S., Darden, T., Tulinsky, A., Hughey, J. L., Hiskey, R. G., and
Pedersen, L. G., Solution conformations of the -y-carboxyglutamic acid domain of bovine
prothrombin fragment I, residues 1-65, Proc. Nat/. Acad. Sci. U.S.A., 88, 424, 1991.
54. Patthy, L., Evolutionary assembly of blood coagulation proteins, Sem. Thrombos. Hae-
mostas., 16, 245, 1990.
55. Rizza, C. R., The clinical features of clotting factor deficiencies, in Human Blood
Coagulation, Haemostasis and Thrombosis, Biggs, R., Ed., Blackwell Scientific, Oxford,
1972, 210.
56. Shapiro, S. S. and Martinez, J., Molecular variations of coagulation proteins, in He-
mostasis and Thrombosis, Bowie, E. J. W. and Sharp, A. A., Eds., Butterworths, London,
1985, 197.
Suttie 479
57. Griffin, J, H., Evatt, B., Zimmerman, T. S., Kleiss, A. J., and Wideman, C.,
Deficiency of protein C in congenital thrombotic disease, J. Clin. Invest., 68, 1370,
1981.
58. Clouse, L. H. and Comp, P. C., The regulation of hemostasis: the protein C system,
New Engl. J. Med., 314, 1298, 1986.
59. Esmon, C. T., The regulation of natural anticoagulant pathways, Science, 235, 1348,
1987.
60. Chung, K-S, Bezeaud, A., Goldsmith, J, C., McMillan, C. W., Menache, D., and
Roberts, H. R., Congenital deficiency of blood clotting factors II, VII, IX, and X,
Blood, 53, 776, 1979.
61. Pauli, R. M., Lian, J, B., Mosher, D. F., and Suttie, J, W., Association of congenital
deficiency of multiple vitamin K-dependent coagulation factors and the phenotype of the
warfarin embryopathy: clues to the mechanism of teratogenicity of coumarin derivatives,
Am. J. Hum. Genet., 41, 566, 1987.
62. Swanson, J, C. and Suttie, J, W., Prothrombin biosynthesis: characterization of pro-
cessing events in rat liver microsomes, Biochemistry, 24, 3890, 1985.
63. Soute, B. A. M., Vermeer, C., De Metz, M., Hemker, H. C., and Lijnen, H. R.,
In vitro prothrombin synthesis from a purified precursor protein. Ill. Preparation of an
acid-soluble substrate for vitamin K-dependent carboxylase by limited proteolysis of
bovine descarboxyprothrombin, Biochim. Biophys. Acta, 676, 101, 1981.
64. Shah, D. V., Swanson, J, C., and Suttie, J, W., Vitamin K-dependent carboxylase:
effect of detergent concentrations, vitamin K status, and added protein precursors on
activity, Arch. Biochem. Biophys., 222, 216, 1983.
65. Evans, M. R., Sung, M. W. P., and Esnouf, M. P., The early stages in the biosynthesis
of prothrombin, Biochem. Soc. Trans., 12, 1051, 1984.
66. Pan, L. C. and Price, P. A., The propeptide of rat bone -y-carboxyglutamic acid protein
shares homology with other vitamin K-dependent protein precursors, Proc. Nat/. Acad.
Sci. U.S.A., 82, 6109, 1985.
67. de Ia Salle, H., Altenburger, W., Elkaim, R., Dott, K., Dieterle, A., Drillien, R.,
Cazenave, J-P., Tolstoshev, P., and Lecocq, J-P., Active -y-carboxylated human factor
IX expressed using recombinant DNA techniques, Nature (London), 316, 268, 1985.
68. Anson, D. S., Austen, D. E. G., and Brownlee, G. G., Expression of active human
clotting factor IX from recombinant DNA clones in mammalian cells, Nature (London),
315, 683, 1985.
69. Busby, S., Kumar, A., Joseph, M., Halfpap, L., Insley, M., Berkner, K., Kurachi,
K., and Woodbury, R., Expression of active human factor IX in transfected cells, Nature
(London), 316, 271, 1985.
70. Kaufman, R. J., Wasley, L. C., Forie, B. C., Forie, B., and Shoemaker, C. B.,
Expression, purification, and characterization of recombinant -y-carboxylated factor IX
synthesized in Chinese hamster ovary cells, J. Bioi. Chern., 261, 9622, 1986.
71. Thim, L., Bjoern, S., Christensen, M., Nicolaisen, E. M., Lund-Hansen, T., Pedersen,
A. W., and Redner, U., Amino acid sequence and posttranslational modifications of
human factor VIla from plasma and transfected baby hamster kidney cells, Biochemistry,
27, 7785, 1988.
72. Yan, S. C. B., Grinnell, B. W., and Wold, F., Post-translational modifications of
proteins: some problems left to solve, Trends Biochem. Sci., 14, 264, 1989.
73. Brinkhous, K. M., Hedner, U., Garris, J. B., Diness, V., and Read, M. S., Effect
of recombinant factor VIla on the hemostatic defect in dogs with hemophilia A, hemophilia
B, and von Willebrand Disease, Proc. Nat/. Acad. Sci. U.S.A., 86, 1382, 1989.
74. Diness, V., Lund-Hansen, T., and Hedner, U., Effect of recombinant human FVIIA
on warfarin-induced bleeding in rats, Thrombosis Res., 59, 921, 1990.
75. Redner, U., Glazer, S., Pingel, K., Alberts, K. A., Blomback, M., Schulman, S.
and Johnsson, H., Successful use of recombinant factor VIla in patient with severe
haemophilia A during synovectomy, Lancet, 2, 1193, 1988.
76. Macik, B. G., Hohneker, J., Roberts, H. R., and Griffin, A.M., Use of recombinant
activated factor VII for treatment of a retropharyngeal hemorrhage in a hemophilic patient
with a high titer inhibitor, Am. J. Hematol., 32, 232, 1989.
77. Foster, D. C., Rudinski, M. S., Schach, B. G., Berkner, K. L., Kumar, A. A.,
Hagen, F. S., Sprecher, C., Insley, M., and Davie, E. W., Propeptide of human protein
Cis necessary for gamma-carboxylation, Biochemistry, 26, 7003, 1987.
78. Jorgensen, M. J,, Cantor, A. B., Furie, B. C., Brown, C. L., Shoemaker, C. B.,
and Furie, B., Recognition site directing vitamin K-dependent -y-carboxylation residues
on the propeptide of factor IX, Cell, 48, 185, 1987.
79. Rabiet, M-J., Jorgensen, M. J,, Furie, B., and Furie, B. C., Effect of propeptide
mutations on post-translational processing of factor IX. Evidence that [3-hydroxylation
and-y-carboxylation are independent events, J. Bioi. Chem., 262, 14895, 1987.
80. Handford, P. A., Winship, P. R., and Brownlee, G. G., Protein engineering of the
propeptide of human factor IX, Protein Eng., 4, 319, 1991.
81. Bentley, A. K., Rees, D. J, G., Rizza, C., and Brownlee, G. G., Defective propeptide
processing of blood clotting factor IX caused by mutation of arginine to glutamine at
position -4, Cell, 45, 343, 1986.
82. Huber, P., Schmitz, T., Griffin, J,, Jacobs, M., Walsh, C., Furie, B., and Furie,
B. C., Identification of amino acids in the -y-carboxylation recognition site on the pro-
peptide of prothrombin, J. Bioi. Chem., 265, 12467, 1990.
83. Furie, B. and Furie, B. C., Molecular basis of vitamin K-dependent -y-carboxylation,
Blood, 75, 1753, 1990.
84. Ulrich, M. M. W., Furie, B., Jacobs, M. R., Vermeer, C., and Furie, B. C., Vitamin
K-dependent carboxylation. A synthetic peptide based upon the -y-carboxylation recog-
nition site sequence of the prothrombin propeptide is an active substrate for the carboxylase
in vitro, J. Bioi. Chem., 263, 9697, 1988.
85. Hubbard, B. R., Jacobs, M., Ulrich, M. M. W., Walsh, C., Furie, B., and Furie,
B. C., Vitamin K-dependent carboxylation. In vitro modification of synthetic peptides
containing the-y-carboxylation recognition site, J. Bioi. Chem., 264, 14145, 1989.
86. Wu, S-M., Soule, B. A. M., Vermeer, C., and Stafford, D. W., In vitro -y-carbox-
ylation of a 59-residue recombinant peptide including the propeptide and the -y-carbox-
yglutamic acid domain of coagulation factor IX. Effect of mutations near the propeptide
cleavage site, J. Bioi. Chem., 265, 13124, 1990.
87. Price, P. A., Fraser, J, D., and Metz-Virca, G., Molecular cloning of matrix Gla
protein: implications for substrate recognition by the vitamin K-dependent -y-carboxylase,
88. Knobloch, J. E. and Suttie, J. W., Vitamin K-dependent carboxylase. Control of enzyme
activity by the "propeptide" region of factor X, J. Bioi. Chem., 262, 15334, 1987.
89. Cheung, A., Engelke, J, A., Sanders, C., and Suttie, J, W., Vitamin K-dependent
carboxylase: influence of the "propeptide" region on enzyme activity, Arch. Biochem.
Biophys., 274, 574, 1989.
90. Cheung, A., Suttie, J, W., and Bernatowicz, M., Vitamin K-dependent carboxylase:
structural requirements for propeptide activation, Biochim. Biophys. Acta, 1039, 90, 1990.
91. Engelke, J, A., Hale, J, E., Suttie, J, W., and Price, P. A., Vitamin K-dependent
carboxylase: utilization of decarboxylated bone Gla protein and matrix Gla protein as
substrates, Biochim. Biophys. Acta, 1078, 31, 1991.
92. Liska, D. J, and Suttie, J, W., Location of -y-carboxyglutamyl residues in partially
carboxylated prothrombin preparations, Biochemistry, 27, 8636, 1988.
93. Benton, M. E., Price, P. A., and Suttie, J, W., The processivity of the vitamin K-
dependent carboxylase, FASEB J., abstr., in press.
94. Munns, T. W., Johnston, M. F. M., Liszewski, M. K., and Olson, R. E., Vitamin
K-dependent synthesis and modification of precursor prothrombin in cultured H-35 hep-
atoma cells, Proc. Nat!. Acad. Sci. U.S.A., 73, 2803, 1976.
Suttie 481
95. Fair, D. S. and Bahnak, B. R., Human hepatoma cells secrete single chain factor X,
prothrombin, and antithrombin III, Blood, 64, 194, 1984.
96. Karpatkin, S., Finlay, T. H., Ballesteros, A. L., and Karpatkin, M., Effect of warfarin
on prothrombin synthesis and secretion in human HEP G2 cells, Blood, 70, 773, 1987.
97. Fair, D. S. and Marlar, R. A., Biosynthesis and secretion of factor VII, protein C,
protein S, and the protein C inhibitor from a human hepatoma cell line, Blood, 67, 64,
1986.
98. Graves, C. B., Munns, T. W., Carlisle, T. L., Grant, G. A., and Strauss, A. W.,
Induction of prothrombin synthesis by prothrombin fragments, Proc. Nat/. Acad. Sci.
U.S.A., 78, 4772, 1981.
99. Friedman, E. W., Karpatkin, M., and Karpatkin, S., Evidence suggesting the reg-
ulation of coagulation factor levels in rabbits by a transferable plasma agent, Blood, 48,
949, 1976.
I00. Karpatkin, M. and Karpatkin, S., Humoral factor that specifically regulates prothrom-
bin (factor II) levels (coagulopoietin-11), Proc. Nat/. Acad. Sci. U.S.A., 76, 491, 1979.
101. Shah, D. V., Nyari, L. J,, Swanson, J, C., and Suttie, J. W., Effect of a humoral
factor on plasma prothrombin and vitamin K-dependent liver carboxylase levels in the
rat, Thrombosis Res., 19, Ill, 1980.
102. Karpatkin, S., Chang, R-J., Pierce, W., and Karpatkin, M., Effect of coumadin-
induced coagulopoietin plasma on vitamin K-dependent carboxylation of liver micro-
somes, Br. J. Haematol., 55, 673, 1983.
103. Owens, M. R. and Cimino, C. D., A plasma factor enhances activity of vitamin K-
dependent coagulation proteins, Thromb. Haemostas., 50, 749, 1983.
I 04. Shah, D. V. and Suttie, J. W., Coagulopoietin activity in rat plasma: influence of degree
of hypoprothrombinemia and reproducibility of the response, Thromb. Res., 39, 43,
1985.
105. Blanchard, R. A., Forie, B. C., Jorgensen, M., Kruger, S. F., and Forie, B.,
Acquired vitamin K-dependent carboxylation deficiency in liver disease. New Engl. J.
Med., 305, 242, 1981.
106. Liebman, H. A., Forie, B. C., Tong, M. J., Blanchard, R. A., Lo, K-J., Lee, S-D.,
Coleman, M. S., and Forie, B., Des--y-carboxy (abnormal) prothrombin as a serum
marker of primary hepatocellular carcinoma. New Engl. J. Med., 310, 1427, 1984.
107. Naraki, T. and Watanabe, K., Quantitative measurement of PIVKA-II by the enzyme
immunoassay diagnostic kit Eitest Mono P-II, in Vitamin K-Dependent Proteins and Their
Metabolic Roles, Elsevier, New York, 1990, 83.
108. Shah, D. V., Engelke, J, A., and Suttie, J. W., Abnormal prothrombin in the plasma
of rats carrying hepatic tumors, Blood, 69. 850, 1987.
109. Kisker, C. T., Perlman, S., Bohlken, D., and Wicklund, B., Measurement of pro-
thrombin mRNA during gestation and early neonatal development. J. Lab. Clin. Med.,
112, 407, 1988.
110. Jamison, C. S. and Degen, S. J. F., Prenatal and postnatal expression of mRNA coding
for rat prothrombin, Biochim. Biophys. Acta, !088, 208, 1991.
Ill. Dihanich, M. and Monard, D., eDNA sequence of rat prothrombin, Nucleic Acids
Res., 18, 4251, 1990.
112. Degen, S. J, F., Shaefer, L. A., Jamison, C. S., Grant, S. G., FitzGibbon, J. J.,
Pai, J-A., Chapman, V. M., and Elliott, R. W., Characterization of the eDNA coding
for mouse prothrombin and localization of the gene on mouse chromosome 2. DNA Cell
Bioi., 9, 487, 1990.
113. Diuguid, D. L., Rabiet, M. J,, Forie, B. C., Liebman, H. A., and Forie, B., Molecular
basis of hemophilia B: a defective enzyme due to an unprocessed propeptide is caused
by a point mutation in the factor IX precursor, Proc. Nat/. Acad. Sci. U.S.A., 83, 5803,
1986.
Chapter 21
ROLE OF ASCORBATE IN REGULATING THE

COLLAGEN PATHWAY
Richard I. Schwarz
TABLE OF CONTENTS
I. Introduction ...................................................... 484
II. Historical (Importance of Cell Culture Conditions and

Developmental Programming) ................................... 485
III. Ascorbate Interaction with the Collagen Pathway-

Secretion ......................................................... 488
IV. Ascorbate Interaction with the Collagen Pathway -

Translation ....................................................... 492
V. Ascorbate Interaction with the Collagen Pathway -

mRNA Levels ................................................... 497
VI. Ascorbate Interaction with the Collagen Pathway -

Other Collagen Types ............................................ 498
VII. Ascorbate Induction of the Collagen Pathway -

Overview ........................................................ 499
VIII. Indirect Roles of Ascorbate in Regulating Collagen

Production ....................................................... 500
IX. Conclusions ...................................................... 501
References ............................................................... 502
0-8493-6961-41931$0.00 + $.50
I. INTRODUCTION
Ascorbate is a small molecule with unusual biological activities. It has

long been recognized as a reducing agent and for its ability to cure scurvy,
yielding its alternate name, Vitamin C. 18 •22 Even though these are important
attributes, they have overshadowed its other properties. For instance, with
the advent of cell culture, it became clear that while humans required ascorbate
for survival, cultures of human cells proliferate well in its absence. Conse-
quently, ascorbate is the only required vitamin needed solely for the main-
tenance of differentiated functions that are necessary at the tissue level. One
such function requiring ascorbate is collagen biosynthesis, and it is this in-
teraction of ascorbate with the collagen pathway that is the primary subject
of this chapter. The focus will be to resolve where and when changes take
place in the collagen pathway after the addition of ascorbate that lead to an
increase in collagen biosynthesis:
type-specific procollagen transcription ~ procollagen mRNA

pool size ~ translation rates ~ internal procollagen pool size
~ secretion rates ~ extracellular cleavage of propeptides ~
collagen fiber formation and extracellular modifications
But before focusing in more detail on the collagen pathway, one needs
a better overview of how ascorbate itself is regulated and how other functions
are affected by this molecule. While it is generally realized that humans,
other primates, and guinea pigs are incapable of making ascorbate because
of a mutation in a required enzyme (L-gulonolactone oxidase), 18 it is less well
known that most cells, even in a competent organism, lack the capacity to
make ascorbate as well. Instead, the vitamin is made in either the liver or
the kidney and transported to other tissues via the blood. 22 One consequence
is that the ascorbate concentration in each tissue varies depending on the
ability to actively transport ascorbate. In this regard ascorbate acts more like
a hormone than a vitamin. So it is not difficult to see why in a competent
organism, where most cells do not produce ascorbate, that a mutation that
deletes the synthetic capacity would have little effect, as long as the diet had
sufficient vitamin C. The reason behind this complex regulation is not known.
However, it may relate to the postulated role of ascorbate as a signaling
molecule between cell types. 28 ·36 ·8o
In addition to its hormone and signaling potential, one finds that ascorbate
has many critical biological activities in different tissues in the body. As-
corbate acts as a hydrophilic protector of biological molecules from oxidative
damage. Lipids in the blood are shielded from oxidation by the presence of
ascorbate. 21 •42 It also acts as a biological reducing agent with the ability to
reduce ferric to ferrous and cupric to cuprous. Other reducing agents like
NADH and glutathione are ineffective. This accounts for its role as a cofactor
Schwarz 485
in many oxygen-utilizing reactions that are catalyzed by enzymes containing

these ions: dopamine hydroxylation, 15 camitine hydroxylation, 62 alpha-ami-
dation of neuroendocrine peptides, 17 and proline hydroxylation. 57·58 The latter
has linked ascorbate to the collagen pathway because in collagen half the
proline residues are hydroxylated. Since the presence of hydroxyproline in
the collagen molecule stabilizes the normal triple helical conformation, the
action of ascorbate in altering the collagen pathway would appear to be
straightforward. As will be discussed below, however, this is not the case.
The collagen pathway is under complex regulation including feedback control.
Add to this that collagen production is regulated by other parameters, and
the resolution of specific interactions can be difficult unless the experimental
conditions are well understood.
To simplify the discussion, this review will concentrate on the direct
interaction of ascorbate with type I collagen expression in fibroblasts. This
is the collagen type, the cell type, and the mode of action of ascorbate that
has been studied in enough detail to yield a coherent picture. This information
will then be used as the background for the less developed examples of how
ascorbate can affect the expression of other collagen types. This review will
also touch on ascorbate's role in altering collagen expression indirectly. In
this case, a noncollagen pathway is affected by ascorbate and its product goes
on to trigger collagen production.
What will become clear is that ascorbate is a simple molecule, but defining
its role as the cure for scurvy has overlooked the potential of even a simple
molecule to alter the differentiated state of a cell. As details of the cell biology
of ascorbate action have accumulated, it has forced a new perspective on how
the cell regulates the synthesis of critical protein-like collagen.
II. HISTORICAL (IMPORTANCE OF

CELL CULTURE CONDITIONS AND
DEVELOPMENTAL PROGRAMMING)
Ascorbate has been known to play a role in collagen regulation because

of its association with the nutritional disease scurvy. Poor wound healing,
loose teeth, and other symptoms of the disease can be traced back to an
impairment in collagen expression. In the late 1920s Szent-Gorgyi recognized
that ascorbate could be vitamin C, 74 •75 and in the early 1960s Green and
Goldberg showed that cells in culture produced more collagen in the presence
of ascorbate (actually, their assays could only show an increase in the number
of hydroxyproline residues in collagen; a point discussed more extensively
below). 25 •26 In 1965, Peterkofsky and Udenfriend found that ascorbate was
an essential cofactor in the hydroxylation reaction. 58 This fact finally linked
ascorbate to a step in the collagen pathway. Several groups then went on to
show that proline hydroxylation stabilizes the triple helical conformation of
the collagen molecule (through increased hydrogen bonding) which gave
further importance to this step. 4 • 16 •27
Attempts to further clarify how ascorbate interacted with the collagen

pathway focused on cells in culture. But as more detailed questions were
asked, more inconsistencies were found in the ability of ascorbate to induce
collagen production. Part of the problem could be traced to the collagen assay.
In early studies collagen production was measured by the ratio of hydroxy-
proline to proline. This could be used as an indicator because in fibroblasts
hydroxyproline is found almost exclusively in collagen. However, this assay
can be misleading because a cell after addition of ascorbate could change the
level of hydroxylation while still producing the same number of collagen
molecules. Peterkofsky and Diegelmann in 1971 resolved this problem by
developing an assay based on the use of a bacterial collagenase to degrade
collagen which is independent of the level of hydroxylation. 56 Even with this
improved assay, cell cultures continued to give inconsistent results: some cell
types increased their collagen production, some showed no effect, and one
cell type was even reported to lower its collagen production. 52 •54 •65 •68
At this point, if anything was clear, it was that the interaction of ascorbate
with the cell was complex. Part of the problem was that other cell culture
parameters could affect collagen expression and negate the positive role of
ascorbate. This was most clearly shown by Peterkofsky when she took calvaria
fibroblasts from the chick embryo and put them in culture. 55 In the embryo
these cells produced 60% of their total protein production as procollagen, but
in culture, even in the presence of ascorbate, procollagen synthesis dropped
to 3%. This experiment raised the question of what could be learned about
the mode of action of a positive modulator in the presence of stronger and
unknown negative regulators. Another cell culture phenomenon that affected
the results was that as cells were passaged for many generations, collagen
production decreased. The cell strains first analyzed by Green and Goldberg
in the 1960s made -15% collagen; 26 cell lines, such as 3T3 cells, derived
from these strains made 8%. 26 And 10 years later, these lines were only
making a couple of percent collagen. 54 So again, even under the same cell
culture conditions, cells can lower their collagen production in a way that is
not reversible by ascorbate.
In addition to cell culture artifacts, the developmental programming of
the cell can alter the responsiveness of a cell to ascorbate. For type I collagen,
the major collagen-producing fibroblasts are in tissues that are the structural
organizers of the body: bone, tendon, skin, etc. The size of these tissues has
to be closely regulated to the overall size of the whole organism. 71 Therefore
in early development, when growth is rapid, these tissues grow rapidly; and
consequently in the adult, when growth of the whole organism has ceased,
so does the growth of these tissues. One would predict that net collagen
production would follow a similar pattern being very high in the embryo and
falling to a low maintenance level in the adult (Figure 1). Here again, the
developmental programming of a cell to lower collagen production restricts
the stimulatory effects that one can expect due to ascorbate addition.
Schwarz 487
FIGURE I. Stained frozen sections o f chick leg tendons from 16-d e mbryos (top) and from
a 4-month old chicken (bottom: 2.5 x lower magnification). Two critical changes occur over
thi s relatively short-time span as the o rgani sm goes from a s tage of rapid growth to attainment
of full size. First, the tissue goes from being hypercellular to being hypocellular. Second, the
level of collagen synthesis drops (see Reference 71). The consequence is that the capacity of
the tissue to produce collagen is dramat ically reduced.
In hindsight, one could predict that the strongest ascorbate effect would
be seen in embryonic cells in a cell culture environment permissive for main-
taining differentiated function. Indeed, one finds that primary avian tendon
(PAT) cells from 16-d chick embryos grown in low serum medium increase
their procollagen production from ~ 10 to ~50% of total protein synthesis
on addition of ascorbate. 68 Other cells producing type I procollagen, where
a relatively high procollagen rate has been maintained in culture, also show
a strong ascorbate effect on collagen production. What has made this so
interesting is that there is only evidence for the direct action of ascorbate on
the posttranslational modification step of prolyl hydroxylation. The impli-
cation is that the general increase in the collagen pathway occurs because
most steps are tightly coupled allowing feedback regulation. 66 The strong
evidence for this type of control is presented below.
III. ASCORBATE INTERACTION WITH THE

COLLAGEN PATHWAY-SECRETION
For 25 years, researchers have been adding ascorbate to cell cultures and
analyzing its effects on the collagen pathway. While many changes have been
observed, the most consistent effect has been that ascorbate causes cells to
maximally hydroxylate their proline residues-between 40 and 50% of the
prolines are hydroxylated in the collagen molecule. It had been suggested in
the 1950s based on the stability of the collagen helix in different species 16
and then shown in the 1970s based on resistance to protease digestion4 •27 that
hydroxylation of proline residues stabilizes the triple helical conformation, 19
a hallmark of the collagen molecule. It is now known that the enzyme, prolyl
hydroxylase, prefers the nonhelical form and the reaction is therefore self-
limiting as increasing hydroxylation triggers the change to a helical form.
The question can then be asked whether this conformation change has any
additional effects on the collagen pathway. zAo
From early studies using electron microscopy to visualize the changes
occurring during wound healing in the scorbutic guinea pig, there were in-
dications that collagen secretion was retarded. 63 This was deduced from swell-
ing inside cells of what was presumed to be unsecreted collagen. Further
evidence came from experiments where the hydroxylation reaction was in-
hibited by using the ferrous ion chelator, a,a'-dipyridyl. Ferrous ion is a
necessary cofactor in the hydroxylation reaction; so, removing iron from the
reaction would be equivalent to depriving the cells of ascorbate (the role of
ascorbate in the hydroxylation reaction is to keep the iron in the reduced
state). In these experiments, a,a'-dipyridyl depressed collagen secretion. 40
In cell culture with 3T3 and chick embryo fibroblasts, 5 5 Peterkofsky could
show that the presence of ascorbate stimulated the secretion of procollagen.
In this case, newly synthesized procollagen was tracked by using a short pulse
of radioactive proline. The fraction of labeled procollagen in the medium was
found to be greater when cells were treated with ascorbate.
Schwarz 489
Secretion, however, is a complicated parameter that can be easily mis-

interpreted. Secretion, unlike the hydroxylation reaction, does not perma-
nently alter the molecule, only its location. Therefore, it is the rate of the
reaction that is critical. The slow step in procollagen secretion follows first
order kinetics (see below). Therefore, the rate is dependent on a rate constant,
which changes depending on the presence or absence of ascorbate, and on
the procollagen pool size within the cell. The pool size, itself, is dependent
on translation, secretion, and internal degradation rates. One consequence is
that a cell cannot secrete faster than it can translate except for short transition
periods. In the case of 3T3 cells above, where collagen translation rates do
not change with ascorbate, the flux of radioactive procollagen to the outside
depends on when ascorbate was added and the length of the labeling period
used to measure secreted procollagen. 2 •55
This is probably best demonstrated with PAT cells where there are more
extensive experimental data. When ascorbate was given to these cells, there
was no change in translation rates for several hours, 65 but in an hour or less,
the secretion rates increased dramatically (Figure 2). 69 This increase in the
secretion rate, but not in the translation rate, caused the internal procollagen
pool to fall. As the internal procollagen pool dropped, this reduced the overall
rate of secretion so that it again matched the translation rate. With these
changes taking place one can see that the length of the radiolabeling period
can make a difference in the fraction of the label that is seen inside vs. outside
the cell. Except for the short transition period after addition of ascorbate, the
flux of procollagen molecules to the outside remains unchanged until the cell
can increase its rate of procollagen translation. This dependency on both a
secretion rate constant and the procollagen pool size in determining the flux
out of the cell has led to confusion in the literature about whether ascorbate
does affect the procollagen secretion process.
A significant improvement in the methodology of measuring procollagen
secretion rates was made by Kao et alY by adapting a pulse-chase protocol
to this problem. Because during the chase the pool of radioactive molecules
does not increase, one can focus directly on the rate of loss of radioactive
procollagen from the cell. Their data, using freshly isolated tendon cells in
short-term suspension culture, showed that procollagen was secreted from the
cell with first order kinetics. What was surprising about their experiments
was that procollagen secretion is described by two independent rate constants:
a fast rate where the half-life of a procollagen molecule in the internal pro-
collagen pool is ~ 15 min (about 60% of the molecules used this route), and
a slow rate where the half-life is ~ 115 min. A variety of interpretations are
possible to explain this data, but a simple one will suffice for this discussion.
There is a common pool for newly synthesized procollagen inside the en-
doplasmic reticulum (ER). From this pool, procollagen molecules are selected
for secretion by either a fast or slow route. What made this finding of two
routes even more fascinating was the observation that when proline hydroxy-
lation was inhibited using a,a'-dipyridyl, the consequence was to completely
100
"' "'
""
50
I
""
\
\
#: \
""'
\
Q3
() 20 \
\
c
"'
\
Ol \
c
·c: \
\
"iti
10 \
E
~ \
c \
(j)
Ol
\
~ \
0() 5 \
0
'\.
ct '"~
0 2 3 4 5 6 7
Hours
FIGURE 2. Secretion of 3 H-labeled procollagen from PAT cells during the chase period when
ascorbate was added only at the time of the chase (the open and closed triangles are independent
determinations). The short dashed line shows the secretion profile of cells always in the presence
of ascorbate and the long dashed line shows the secretion profile of cells in the absence of
ascorbate (for more details of these controls see Reference 69). After -1 h, the cells are able
to hydroxylate the already synthesized procollagen and secrete it at the fast rate.
shift procollagen secretion to the slow route. 34 The implication of this result
is that a step in the fast route, possibly a receptor, requires the triple helical
conformation of the procollagen molecule in order to proceed.
These pulse-chase experiments raised the question of what the positive
effect of ascorbate addition would be on the process of procollagen secretion
and whether it would be directly opposite to those obtained using the inhibitor,
Schwarz 491
a,a' -dipyridyl. In the experiments above, freshly isolated chick embryo ten-
don cells were used, and one cannot ask the question directly. In the chick
embryo, ascorbate is made in the kidney and liver, and freshly isolated tendon
cells are, therefore, already maximally stimulated with ascorbate. One way
to circumvent this problem is to grow the cells in culture without ascorbate
for more than 36 h. 65 When pulse-chase experiments were performed using
tendon cells in culture, the data obtained by ascorbate addition to scorbutic
cultures was completely compatible with the inhibitor studies. 69 With ascor-
bate, the cells show a biphasic secretion profile with half-lives of ~20 min
and~ 120 min. In confluent cultures almost all the procollagen ( ~90%) leaves
the cell by the fast route. This may mean that in freshly isolated cells, assayed
in suspension culture, that the round shape caused by being unattached to a
substratum may partially inhibit the fast route. In any case, scorbutic tendon
cells secrete all of their procollagen by the slow route. Because ascorbate is
actively and rapidly transported into the cell, one can gauge how quickly
secretion rates can change by addition of ascorbate only at the time of the
chase (Figure 2). In an hour or less the cells stopped secreting at the slow
rate and started secreting the already synthesized procollagen molecules at
the high rate. Because the prolyl hydroxylase enzyme is located in the ER, 13 • 14
this meant that the accumulated pool of procollagen must also be in the ER
and that the rate-controlling step in the secretion process is leaving the ER.
These experiments also showed that the underhydroxylated procollagen mol-
ecule shares the same pool and is capable of switching to the fast route if it
becomes hydroxylated and triple helical.
The reason why a triple helical molecule should be secreted six times
faster than its nonhelical counterpart is not understood but is critical to a
major action of ascorbate on the collagen pathway. The mechanism of pro-
collagen secretion has been debated for 25 years; yet it remains unclear
whether procollagen is secreted through the Golgi or more directly out of the
cell. Early experiments by Ross et a!. argued that the kinetics of labeling
during a pulse-chase experiment did not support transport through the Golgi. 63
Marchi and LeBlond tracked procollagen and other proteins using 3 H-proline
labeling in fibroblast cultures, and they presented evidence that the Golgi
was involved. 39 Antibody studies have localized procollagen in the Golgi.
Olsen et a!. showed by antibody staining that procollagen molecules could
be seen in the Golgi in tendon cells under conditions that sufficiently disrupted
the membranes and allowed antibody penetration. 43 -45 •53 On the other hand,
Pierce et a!. 59 using antibodies could not find type IV procollagen in the Golgi
of parietal yolk sack cells. Since EM immunostaining is subject to many
potential artifacts depending on the cell system and the antibody, definitive
conclusions are hard to draw using this technique alone. Moreover, the kinetic
data presented above would argue that procollagen can be secreted in more
than one way, making the interpretation of this earlier data even more difficult.
In addition, many other questions can be raised. Why is the Golgi apparatus
TABLE 1
Temperature Sensitivity of Secretion from
PAT Cells
Half-life, - asc Half-life, + asc Half-life

oc procollagen procollagen noncollagen
41 120' 20b 70
39 120' 20b 75
32 120' 20b 140
29 * 36 *
25 * * *
Note: * Not secreted with first order kinetics.
•.b Average value presented; differences between temper-
atures were not statistically significant.
so small in tendon cells when half their total protein production is secreted?69
Why are secretory vesicles in tendon cells difficult to see by electron mi-
croscopy? Why does secreted procollagen remain sensitive to endoglycosidase
H, 11 an indication that sugar rearrangements that should have occurred in the
Golgi have not taken place in the procollagen molecule? Why is the fast rate
of procollagen secretion so insensitive to temperature (Table 1), something
one would not expect if secretion were occurring by an energy intensive
process such as vesicle transport? Until more is understood about the process
of procollagen secretion, it is difficult to give a physical meaning to the
sixfold differences in secretion rates that appear to be critically dependent on
whether the procollagen molecule is triple helical.
IV. ASCORBATE INTERACTION WITH THE

COLLAGEN PATHWAY-TRANSLATION
The addition of ascorbate to fibroblast cultures has been reported to
increase, decrease, and not alter the amount of procollagen synthe-
sized. 52 •54 •65 •68 Needless to say, such diversity of action has not led to a
consensus for the role of ascorbate in regulating this step. For reasons that
will become clear, only cells devoting a significant portion of their total
protein synthesis to procollagen are capable of a dramatic response. Over the
last 15 years as cell culture techniques have improved, ascorbate has been
shown to be a strong inducer in chick tendon 64 - 68 and bone cells, 1 "47 and human
skin fibroblast. 8 •41 •60 •76 When these cell types were observed more closely, at
least part, and sometimes all, of the increase in procollagen production could
be attributed to an increase in procollagen mRNA translation rates.
One of the first indications for actual translational regulation came from
the inhibitor studies using a,a'-dipyridyl as described above. One would
Schwarz 493
expect that when secretion rates were reduced by addition of the inhibitor
that the pool of procollagen inside the cell would build up dramatically.
Instead, Kao et al. 34 observed that the internal procollagen pool remained
constant. They concluded that the inhibitor decreased procollagen translation
rates (procollagen degradation 5 could have played a role but was not dis-
cussed). This possibility for translational regulation could be tested directly
using this same cell type in primary culture (PAT cells). 64 In this case a,a'-
dipyridyl was given to cells fully induced with ascorbate, making half their
total protein production as procollagen, and to scorbutic cells, as a control.
The data was plotted as the ratio between counts incorporated into procollagen
and those into noncollagen protein (Figure 3). This controlled for the non-
specific decrease in protein synthesis ( ~40% )34 •64 resulting from a toxic effect
of chelation of ferrous ion by the inhibitor. As might be predicted, scorbutic
cells showed no specific change in procollagen production, confirming that
a cell inhibited at the proline hydroxylation step by a lack of ascorbate cannot
be further inhibited by a lack of ferrous ion. In contrast, fully induced cells
showed a specific and rapid reduction in procollagen translation rates. By 4
h the level had dropped by 2/3 and then leveled off with increasing time. In
the short time period of this experiment, there was no significant change in
procollagen mRNA levels.
The question can be raised that if there is a feedback linkage between
translation and secretion rates why is the drop in collagen production on
addition of a,a' -dipyridyl only 2/3 of the scorbutic rate. The answer is that
the linkage between these steps is better than it first appears. When one takes
into account that the inhibitor nonspecifically inhibits 1/3 of all protein syn-
thesis, then the procollagen translation rates need only drop by 2/3 to bring
them into synchrony with the ability of the cell to secrete what is made.
Translation regulation is also observed when one induces procollagen
production with ascorbate (Figure 4). In this case, the efficiency of translating
procollagen mRNA increases by ~2-fold and this continues until the cells
reach a fully induced state where upon it returns to the original rate. 64 In the
context of the kinetics of the induction process, secretion rates are rapidly
altered by ascorbate (< 1 h), but full induction of translation rates requires a
much longer time ( ~48 h) in large part because of the slow increase in
procollagen mRNA levels (starting at 12 hand requiring ~72 h to complete). 64
Until translation rates can equal secretion rates the pool of procollagen within
the ER is depressed, and a consequence is that the cell raises its translational
efficiency.
The kinetics of the induction process are incompatible with changes in
degradation rates playing a significant role. 5 One might assume that scorbutic
cells producing nonhelical collagen would have more degradation. This fol-
lows from the fact that nonhelical collagen (gelatin) is susceptible to most
proteases, but the triple helical collagen molecule requires specific proteases,
called collagenases, to denote their unusual ability to attack the triple helical
1.2
.....
z
w
(!'
c(
...J
...J
0
(.)
z
0
z......
z
w
(!'
c(
...J
...J
0
(.)
0
0::
....,
0..
0
i=
c(
0::
-----------~-----------~-----------
HOURS
FIGURE 3. Effect of adding cx,cx' -dipyridyl (0.3 mM) to ascorbate-induced (•) and uninduced
("')PAT cells. 64 The rates of procollagen synthesis have been normalized to that of noncollagen
protein synthesis to correct for the nonspecific inhibition of protein synthesis ( -40% in 6 h).
Each point in the graph reflects a 2-h pulse ending at the time indicated. The fully induced and
uninduced controls are shown as dashed lines ± SD on the right ordinate.
conformation. As logical as this hypothesis appears to be, the relevant data

do not support it. Ascorbate acts quickly to fully hydroxylate procollagen and
thereby stabilize the triple helical conformation. This change, if the model
were correct, would reflect itself in the ascorbate induction process as an
early jump in the levels of net procollagen translation. No such jump is
observed, and, instead , net translation rates are not significantly increased
in the first few hours after addition of ascorbate. 65
Pinnell et a!., 60 using human skin fibroblast, have also observed strong
evidence for translational control without evidence for a change in degradation
Schwarz 495
•
...J
w
~
ffi
~
w
~
ii::
w
;;::
!;(
...J
w
a:
I I I I I
48 12 16 20 24 28 32 36 40 44 48 52 56 60
HOURS AFTER ASCORBATE ADDITION
FIGURE 4. Kinetics of the relative increase in procollagen synthesis (ratio of procollagen

counts incorporated to noncollagen counts; solid line) and the accumulation ofprocollagen mRNA
after addition of ascorbate to PAT cells (dashed line). 64 Horizontal lines delineate the average
maximum values observed (6.8 ± 0.5 [SD] for the protein; 5.9 ± 0.4 for the RNA). During
the transition to the induced state, about twice as much procollagen is synthesized for a given
amount ofprocollagen mRNA. The initial drop in procollagen mRNA levels is only occasionally
observed and is probably an artifact of the extraction procedure for cytoplasmic RNA using
Triton X-100. This weak detergent only extracts a part of the procollagen mRNA that is mem-
brane-bound. Ascorbate addition appears to strengthen the interaction between the procollagen
mRNA and the membrane-bound ribosome reducing its extractability. Consistent with this view
is the observation that when the detergent SDS is used to extract all the RNA no drop is observed.
rates. On addition of ascorbate, these cells increased their levels ofprocollagen

synthesis by fourfold, but procollagen mRNA levels only increased twofold.
Using lysine hydroxylation as a marker of collagen, which was insensitive
to the presence of ascorbate, they could show directly that procollagen deg-
radation was not a factor in the translational induction by ascorbate.
Again, using human skin fibroblasts, Chan et a!. 8 observed similar changes
where collagen synthesis increased six- to sevenfold with ascorbate but pro-
collagen mRNA only increased two- to threefold. Interestingly, after addition
of ascorbate, full induction of procollagen synthesis occurred in 10 h and this
period was too short for an increase in procollagen mRNA levels to be
observed. While this appeared to be much more rapid than the induction in
PAT cells, it is not inconsistent with that data. The reason is that procollagen
mRNA levels in the human fibroblasts never dropped more than approximately
twofold. Since translational efficiency can increase approximately twofold
after ascorbate addition, there is no need for mRNA levels to increase in order
to obtain full induction. It is the slow increase in mRNA levels that retards
the induction process in PAT cells.
Owen et at. 47 have shown that for rat osteoblasts there is a strong ascorbate
induction of collagen synthesis from ~2.5 to~ 30% after~ 36 h. Even though
detailed kinetics were not done and no comparisons were made with procol-
lagen mRNA levels, this study again shows that ascorbate can have a dramatic
effect on the collagen pathway and the amount of procollagen synthesized.
One can ask why the earlier results with ascorbate were so variable when
the more recent studies show a common response to ascorbate. One important
difference is that earlier studies used fibroblastic cell lines making low levels
of collagen. As discussed before, this could be reflecting cell culture con-
ditions that are dominantly negative for collagen expression. 77 Or it could be
reflecting the fact that long-term cultures lower their collagen production by
a mechanism that is not reversible by ascorbate addition. In addition, a cell
making 1% procollagen no longer taxes the secretory machinery of the cell.
At that level of synthesis, changes in the secretion rate may still alter the
pool size, but the initial pool would be so small that these changes may not
feed back to alter translation rates. Therefore, even though a cell making low
levels of collagen could respond to ascorbate, its lack of increase in collagen
production could be due to multiple factors.
Finally, a discussion of translational control and collagen regulation would
not be complete without reference to some of the unusual features of the
translational machinery of the cell. As might be expected, embryonic tendon
cells have a high proportion of membrane-bound ribosomes to support the
high level of procollagen translation rates. 31 •66 This only drops slowly when
cells are grown without ascorbate in contrast to the more rapid drop in pro-
collagen synthesis. This leads to a large disparity between the relative number
of membrane-bound ribosomes and the level of secreted proteins. A finding
that promotes the concept that the number of membrane-bound ribosomes is
a characteristic trait of the differentiated state at the time when the cells were
isolated. Similarly, the pool of proline and glycine tRNA is elevated in PAT
cells and this too is unaffected by the absence of ascorbate but does drop as
cells age developmentally. 46 Also, there is evidence to support the model that
there are specific sites on membrane-bound ribosomes for each procollagen
chain helping to organize the interaction between collagen chains that results
in a type I collagen molecule with two a 1 and one a 2 chains. 78 •79 Needless
to say, much more needs to be known about procollagen translation and its
regulation before these observations can be understood in molecular terms.
Schwarz 497
V. ASCORBATE INTERACTION WITH THE

COLLAGEN PATHWAY-mRNA LEVELS
Despite the common assumption that mRNA levels are the rate-controlling
step in a protein induction process, its position at the end of this discussion
reflects the specific kinetics of the ascorbate induction process. In this case,
the last pool to increase after addition of ascorbate is procollagen mRNA
levels. 8 •64 After ascorbate addition to PAT cells, a sixfold increase in secretion
rates is followed by a sixfold increase in translation rates which is followed
and overlapped by a sixfold increase in procollagen mRNA levels. As seen
in Figure 4 procollagen mRNA levels in PAT cells begin to increase only
after 12 hand linearly increase over the next 72 h. 64 As was discussed earlier
in human fibroblasts (Chen et al.) 8 where there is only a twofold change in
mRNA levels between induced and uninduced states, then this can be com-
pensated for by translational efficiency. The consequence is that the induction
process appears to be more rapid. With PAT cells where there is a sixfold
change, mRNA levels have to increase in order to achieve full induction. The
limitations of having only a single copy gene restricts the amount of tran-
scription and accounts for the slow increase in the procollagen mRNA pool.
Transcription rates are increased threefold in fully il\duced PAT cells as com-
pared to those cells maintained without ascorbate. 38 The pool increases ~6-
fold because the half-life of the procollagen mRNA also increases 2-fold from
approximately 11 to 20 h. 38
This raises the important question of how the procollagen mRNA pool
is controlled. One must consider that there are in reality two pools-one is
a free (a relative term) pool and the other is membrane-bound as part of the
translation complex. If one assumes the latter has a long half-life (>20 h)
and the former a short half-life ( < 11 h), then the increased stability of the
mRNA after ascorbate addition may only be reflecting the increased trans-
lational activity of the procollagen mRNA. Conceptually, this ties mRNA
half-life to increases in translational activity and would be a direct link between
ascorbate's participation in the hydroxylation step and part of the increase in
procollagen mRNA levels. However, this does not resolve how ascorbate
increases transcriptional activity of the procollagen genes (a, and a 2 ).
Classically, one would postulate an interaction between ascorbate and a
factor that would bind to the promoter region of the collagen gene and stimulate
transcription. 61 Alternatively, one can use the lessons learned from other steps
in the pathway and postulate a feedback mechanism revolving around the free
pool of procollagen mRNA. In the latter case, as the free pool of procollagen
mRNA gets depleted by participation in the translational complex, this could
trigger more transcription of procollagen mRNA. Even though there is no
direct evidence for this feedback mechanism, it has two indirect lines of
evidence supporting it. First is the 12-h lag before procollagen mRNA levels
start to increase. 64 In the feedback model, the lag is predicted because the
other steps would have to be induced before the transcriptional activity could
increase. In the case of a transcription factor model, a fairly rapid increase
in transcriptional activity would be expected. Ascorbate is rapidly taken up
by the cell; 6 •69 so, there is no simple explanation for the delay in mRNA
induction using this model. Second is the fact that isoascorbate, an isomer
whose uptake by the cell is strongly selected against, once inside the cell is
as active as ascorbate at inducing procollagen production. 35 These data support
the idea that the proline hydroxylation step can not discriminate between the
isomers. Since other reducing agents can participate in the reaction in vitro
it is questionable whether a specific recognition site exists. Since the isomers
have the same reducing potential, they are interchangeable at this step. How-
ever, one can question whether a transcription factor that specifically bound
ascorbate would also bind its isomer with equal efficiency. Again, these
questions only make it clear that more needs to be known about the controls
regulating procollagen mRNA levels and how these controls are specifically
changed during the ascorbate induction process.
VI. ASCORBATE INTERACTION WITH THE

COLLAGEN PATHWAY-OTHER COLLAGEN TYPES
The role of ascorbate in the induction of type I collagen production is

only now becoming clear and this is the most studied system. The effect of
ascorbate addition on other collagen types shows both similarities and dif-
ferences to the type I example. Some of the differences can be explained by
the uniqueness of that particular collagen molecule, other differences may
reflect deviations between the specific collagen pathways, and some differ-
ences may only reflect an inability to optimize the cell culture conditions for
that cell type. With few studies to refer to, the chances for misinterpretation
increase. Nevertheless, an analysis of the effect of ascorbate on another
collagen type gives an opportunity to categorize the changes in terms of their
generality for all collagens from those changes that are more type specific.
The information not only reveals differences in the ascorbate induction process
but also uniquenesses in the regulation of specific collagen types.
The one change after ascorbate addition that appears to be universal across
collagen types is the induction of full hydroxylation of proline residues. 32
The question of whether this change goes on to stimulate procollagen secretion
rates is not easily answered for all collagen types due to a lack of data. For
instance, types II and X secrete at a higher rate when their chains are fully
hydroxylated, 50 but type IV appears to be unaffected. 32 The change in secretion
rates with ascorbate addition for types II and X is around sixfold. 50 However,
the fast rate has a half-life for secretion of ~ 120 min which is comparable
to the slow rate for type J.3 2 So the mechanism for secretion appears to be
different for each collagen type. Type IV collagen behaves uniquely but is
consistent with its unusual characteristics. Secretion rates are postulated to
Schwarz 499
change for type I collagen because the more hydroxylated chains stabilize the
triple helical conformation. With type IV collagen, even fully hydroxylated
molecules exist as single chains and may be secreted as single chains. 12 •50
Therefore, using the model that the increase in secretion rates is a result of
the conformation change of the procollagen molecule, the observation that
the secretion of typed IV collagen is not significantly changed by ascorbate
is consistent with that model (although more conclusive data is needed to
prove the point).
The effect of ascorbate on procollagen translation rates is even more
difficult to establish with other cell types. For instance, indirect evidence has
shown that ascorbate causes a remodeling of the ER in chondrocytes making
collagen types II and X. 5 1 In type X more procollagen is made after ascorbate,
but the kinetics have not been studied in enough detail to show a linkage
between this step and secretion rates. 37 With type II the analysis is quite
complex. Ascorbate has not been shown to increase synthesis and in some
cases there is a decrease. 37 Part of the problem is a consequence of chondro-
cytes being able to produce both type I and type II collagen. 20 •24 The presence
of ascorbate, especially for cells grown on plastic, can encourage the transition
to a more "fibroblastic" phenotype-reducing type II production. 3 It remains
to be seen if a chondrocyte, in suspension culture or in other conditions that
stabilize the chondrocyte phenotype, will respond to ascorbate by increasing
their type II procollagen translation rates.
The effect of ascorbate on mRNA levels remains to be understood in
more detail. In the case of chondrocytes, type X mRNA increases 14-fold,
7 days after ascorbate addition. In the same cells, type II mRNA drops 50%
but this occurs independently of ascorbate. 37
What conclusions can be drawn about the role of ascorbate in regulating
other collagen types? Even though there are strong analogies in some situations
to the type I example, more studies will have to be done especially with cell
types in cell culture conditions permissive for high procollagen expression.
As was seen with secretion and type IV collagen, the differences as well as
the similarities to the type I situation will be informative about how specific
cell types regulate each collagen type.
VII. ASCORBATE INDUCTION OF THE COLLAGEN

PATHWAY-OVERVIEW
Ascorbate can induce a major increase in type I collagen production.

Conversely, deprivation of ascorbate in humans causes scurvy with a major
symptom being an inability to produce type I collagen. It seems strange that
a major protein would be regulated by a molecule whose primary function is
that of a reducing agent. Moreover, why should an animal such as a chicken
that can make its own ascorbate, and presumably never be in a deprived state,
still show such a complex inductive response when this artificial situation
occurs in culture?
There are no definitive answers to these questions. Nevertheless, a partial

explanation comes from the tight interaction between steps in the collagen
pathway allowing feedback regulation, 60 •66 and from the fact that collagen is
the major structural protein of the body, and hence will be under intense
regulation. Having a pathway which is tightly coupled allows a tremendous
variety in regulatory controls and an ability to match small variations outside
the cell to changes inside the cell. In contrast, regulation at the transcriptional
level is much cruder when constrained by having only a single copy gene
and the requirement for high procollagen mRNA translation rates. 38 1t requires
long times (3 d) 64 to fully induce mRNA levels (low transcription rates) or
to reduce mRNA levels (long half-life of the mRNA, ~I d) 38 in PAT cells.
During development of the chick, changes in mRNA levels that required days
to become effective would overlap major alterations in tissue structure. There-
fore, in order to adequately regulate collagen production in this time frame
requires that collagen production be regulated at additional steps. The regu-
lation by ascorbate is an example of ways collagen regulation can occur at
other points in the pathway to fine-tune the process.
The hydroxylation step may have become an important step in regulating
the collagen pathway not because of the interaction of ascorbate, but because
cell density exerts its control at the same point. 29 •30 •7° Cultured PAT cells will
only make a high percentage of procollagen at a high cell density. 64 •68 One
can show that one subunit of the prolyl hydroxylase enzyme is regulated by
cell density. 30 Moreover, in a tissue, such as a tendon that goes through large
variations in cell density over the course of development (Figure I), regulation
at the proline hydroxylation step could be an important part of the develop-
mental control of collagen production. Because ascorbate interacts at the same
point, it may only be accidental that the collagen pathway is so dramatically
affected by this molecule.
Nevertheless, because the potential exists for ascorbate to regulate the
pathway and because ascorbate levels are tightly controlled in the blood
stream, ascorbate could act as a messenger to induce collagen production in
some cells. There is some support for this 28 •36 •80 and as more is understood
about development regulation, this possibility may become more significant.
VIII. INDIRECT ROLES OF ASCORBATE IN

REGULATING COLLAGEN PRODUCTION
Because ascorbate affects many pathways and because the collagen path-
way can be altered at many levels, it is not surprising that ascorbate can affect
collagen production indirectly. In these cases, the primary action of ascorbate
is not on the collagen pathway. As a consequence of that change, however,
there is an alteration in collagen synthesis. There are probably many such
interactions, but two well-documented cases will be discussed below.
The first involves the complete removal of ascorbate from the diet of the
guinea pig. The consequence is that the guinea pig stops eating after two
Schwarz 501
weeks and literally starves. 9 This unusual behavior may result from an im-
balance caused by a lack of an essential vitamin. In any case, collagen
production drops rapidly and can be shown to be equally affected by food
deprivation in the presence of ascorbate. 9 This is an unusual case. Lack of
appetite in the guinea pig is associated only with complete removal of the
vitamin from the diet and a comparable effect is not observed in humans.
Nevertheless, it is a good example of how in the whole organism the ascorbate
effect can become quite complex. Peterkofsky and her co-workers have ana-
lyzed this problem in detail and they have found that the lack of vitamin C
causes a change in the blood. 49 The serum from fasted animals can be shown
to reduce the collagen production in cells in culture. The active factor appears
to bind to and inhibit the activity of insulin-like growth factor I in the fasted
serum. 4 •8 How this factor alters the collagen pathway is still under investi-
gation.
The second case involves the oxidation of lipids to aldehydes. 7• 10 ·23 Chojk-
ier et a!. 7• 10 have shown that aldehydes can induce collagen production in
certain cells in culture. The human fibroblasts used were making about 3%
collagen, and this increased approximately twofold by either ascorbate or
aldehyde addition. In this case, the role of ascorbate is difficult to interpret
since in the highly oxygenated environment of cell culture, ascorbate breaks
down and can promote the oxidation of the lipids to aldehydes. This mech-
anism of collagen stimulation may be important in wound repair and in the
fibrosis due to injury seen in the lung and other tissues. It is probably not a
major route of ascorbate stimulation of collagen synthesis. Ascorbate is thought
to prevent lipid oxidation under normal circumstances. 21 In addition, ascorbate
and isoascorbate would be expected to have equal activities in stimulating
collagen if oxidation was the critical factor since the isomers are oxidized
outside the cell at about the same rate. 35 In fact, isoascorbate is about three-
to fivefold less effective than ascorbate at stimulating PAT cells because of
its inability to be taken up effectively by the cell. 35 Nevertheless, one has to
keep in mind when dealing with multifaceted molecules and multiregulated
pathways, alternative mechanisms have to be expected and can play important
roles in specific cases.
IX. CONCLUSIONS
Ascorbate acts as a specific inducer of collagen production in many cells
producing type I collagen. If it had worked directly on transcriptional regu-
lation of the collagen gene, it would have been a mainstream research topic,
but it would have revealed little new about the regulation of the collagen
pathway except perhaps a new promotor binding site. The fact that a molecule
with little more than a good reducing potential can act to specifically stimulate
procollagen synthesis requires a new look at protein regulation. The critical
concept that this molecule keeps teaching us about the collagen pathway (type
I) is that it is tightly coupled through feedback regulation. Therefore, ascorbate
levels (or cell density) by controlling the activity of the proline hydroxylation
step can control the flow through the whole pathway. Hydroxylation levels
control whether the collagen molecule is triple helical. This conformation
changes the secretion rates by sixfold. This in tum can alter the pool inside
the ER; the pool size alters translational efficiency; and translational efficiency
alters the half-life of the procollagen mRNA.
It is also clear that not all collagen types respond to ascorbate in the same
way. But this may be an advantage in that the differences may reveal even
more about how cells can control tissue-specific functions. The action of
ascorbate on other collagen types promises to be an area that will yield
important new information about specific points of control within the collagen
pathway.
At the same time one has to be aware that in the whole organism ascorbate
can affect many biological pathways in nonmesenchymal tissues, and these
changes can feedback to alter collagen production. Likewise, the ability of
the molecule to form free radicals in an oxygenated environment can modulate
the collagen pathway indirectly. In the end, as we better understand the
complexities of ascorbate action on the cell, this will all add to our under-
standing of how the differentiated state can be controlled.
After 70 years since the discovery of ascorbate, we can finally give a
molecular model for why scurvy is a collagen disease. But to its credit, this
specific inducer of collagen synthesis has been a key to unlocking the complex
controls that regulate the differentiated state of the cell.
REFERENCES
I. Aronow, M. A., Gerstenfeld, L. C., Owen, T. A., Tassinari, M. S., Stein, G. S.,
and Lian, J. B., Factors that promote progressive development of the osteoblast phe-
notype in cultured fetal rate calvaria cells, J. Cell. Physiol., 143, 213, 1990.
2. Bates, C. J., Bailey, A. J., Prynne, C. J., and Levine, C. I., The effect of ascorbic
acid on the synthesis of collagen precursor by 3T6 mouse fibroblasts in culture, Biochim.
Biophys. Acta, 278, 372, 1972.
3. Benya, P. D. and Shaffer, J. D., Dedifferentiated chondrocytes reexpress the differ-
entiated collagen phenotype when cultured in agarose gels, Cell, 30, 215, 1982.
4. Berg, R. A. and Prockop, D. J., The thermal transition of a nonhydroxylated form of
collagen. Evidence for a role for hydroxyproline in stabilizing the triple-helix of collagen,
S. Bienkowski, R. S., Curran, S. F., and Berg, R. A., Kinetics of intracellular degradation
of newly synthesized collagen, Biochemistry, 25, 2455, 1986.
6. Blanck, T. J. J. and Peterkofsky, B., The stimulation of collagen secretion by ascorbate
as a result of increased proline hydroxylation in chick embryo fibroblasts, Arch. Biochem.
Biophys., 171, 259, 1975.
7. Brenner, D. A. and Chojkier, M., Acetaldehyde increases collagen gene transcription
in cultured human fibroblasts, J. Bioi. Chem., 262, 17690, 1987.
Schwarz 503
8. Chan, D. Lamande, S. R., Cole, W. G., and Bateman, J. F., Regulation ofprocollagen
synthesis and processing during ascorbate-induced extracellular matrix accumulation in
vitro, Biochem. 1., 269, 175, 1990.
9. Chojkier, M., Spanheimer, R., and Peterkofsky, B., Specifically decreased collagen
biosynthesis in scurvy dissociated from an effect on proline hydroxylation and correlated
with body weight loss, J. Clin. Invest., 72, 826, 1983.
10. Chojkier, M., Houglum, K., Solis-Herruzo, J., and Brenner, D. A., Stimulation of
collagen gene expression by ascorbic acid in cultured human fibroblasts: a role for lipid
peroxidation, J. Bioi. Chern., 264, 16957, 1989.
II. Clark, C. C., The distribution and initial characterization of oligosaccharide unit on the
COOH-terminal propeptide extensions of the pro- I and pro- 2 chains of type I pro-
collagen, J. Bioi. Chern., 254, 10798, 1979.
12. Crouch, E. and Bornstein, P., Characterization of a type IV procollagen synthesized
by human amniotic fluid cells in culture, J. Bioi. Chern., 254, 4197, 1979.
13. Cutroneo, K. R., Guzman, N. A., and Sharawy, M. M., Evidence for a subcellular
vesicular site of collagen prolyl hydroxylation, J. Bioi. Chern., 249, 5989, 1974.
14. Diegelmann, R. F., Bernstein, L., and Peterkofsky, B., Cell-free collagen synthesis
on membrane-bound polysomes of chick embryo connective tissue and the localization
of prolyl hydroxylation on the polysome-membrane complex, J. Bioi. Chern., 248, 6514,
1973.
15. Diliberto, E. J., Jr., Daniels, A. J., and Viveros, 0. H., Multicompartmental secretion
of ascorbic acid and its dual role in dopamine-'Y-hydroxylation, Am. J. Clin. Nutr., 54,
1163S, 1991.
16. Doty, P. and Nishihara, T., The molecular properties and thermal stability of soluble
collagen, in Recent Advances in Gelatin and Glue Research, Stainsberg, G., Ed., Per-
gamon Press, London, 1957, 92.
17. Eipper, B. A. and Mains, R. E., The role of ascorbate in the biosynthesis of neuroen-
docrine peptides, Am. J. Clin. Nutr., 54, 1153S, 1991.
18. Englard, S. and Seifter, S., The biochemical functions of ascorbic acid, Annu. Rev.
Nutr., 6, 365, 1986.
19. Fessler, L. I. and Fessler, J. H., Protein assembly of procollagen and effects of hy-
droxylation, J. Bioi. Chern., 249, 7637, 1974.
20. Finer, M. H., Gerstenfeld, L. C., Young, D., Doty, P., and Boedtker, H., Collagen
expression in embryonic chicken chondrocytes treated with phorbol myristate acetate,
Mol. Cell. Bioi., 5, 1415, 1985.
21. Frei, B., Ascorbic acid protects lipids in human plasma and low-density lipoprotein
against oxidative damage, Am. J. Clin. Nutr., 54, 1113S, 1991.
22. Friedrich, W., in Vitamins, Walter de Gruyter, Berlin, 1988, 929.
23. Geesin, J. C., Gordon, J. S., and Berg, R. A., Retinoids affect collagen synthesis
through inhibition of ascorbate-induced lipid peroxidation in cultured human dermal
fibroblasts, Arch. Biochem. Biophys., 278. 350, 1990.
24. Gerstenfeld, L. C., Finer, M. H., and Boedtker, H., Quantitative analysis of collagen
expression in embryonic chick chondrocytes having different developmental fates, J.
Bioi. Chern., 264,5112, 1989.
25. Green, H. and Goldberg, B., Collagen synthesis by human fibroblast strains, Proc.
Soc. Exp. Bioi. Med., 117, 258, 1964.
26. Green, H., Goldberg, B., and Todaro, G., Differentiated cell types and the regulation
of collagen synthesis, Nature (London), 212, 631, 1966.
27. Jimenez, S. A. and Yankowski, R., Role of molecular conformation on secretion of
chick tendon procollagen, J. Bioi. Chern., 253, 1420, 1978.
28. Kalcheim, C. and Levie!, V., Stimulation of collagen production in vitro by ascorbic
acid released from explants of migrating avian neural crest, Cell Diff, 22, 107, 1988.
29. Kao, W.-Y.W., Kao, W.-C. C., and Schwarz, R. 1., Prolyl hydroxylase production
can be uncoupled from the regulation of procollagen synthesis, Exp. Cell Res., 157, 265,
1985.
30. Kao, W. W-Y. and Berg, R. A., Cell density-dependent increase in prolyl hydroxylase
activity in cultured L-929 cells requires de novo protein synthesis, Biochim. Biophys.
Acta, 586, 528, 1979.
31. Kao, W. W-Y., Flaks, J. G., and Prockop, D. J., Primary and secondary effects of
ascorbate on procollagen synthesis and protein synthesis by primary cultures of tendon
fibroblasts, Arch. Biochem. Biophys., 173, 638, 1976.
32. Kao, W. W-Y., Mai, S. H., Chou, K-L., and Ebert, J., Mechanism for the regulation
of post-translational modifications of procollagens synthesized by matrix-free cells from
chick embryos, J. Bioi. Chern., 258, 7779, 1983.
33. Kao, W. W-Y., Berg, R. A., and Prockop, D. J., Kinetics for the secretion of pro-
collagen by freshly isolated tendon cells, J. Bioi. Chern., 252, 8391, 1977.
34. Kao, W. W-Y., Prockop, D. J., and Berg, R. A., Kinetics for the secretion ofnonhelical
procollagen by freshly isolated tendon cells, J. Bioi. Chern., 254, 2234, 1979.
35. Kipp, D. and Schwarz, R. I., Effectiveness of isoascorbate versus ascorbate as an
inducer of collagen synthesis in primary avian tendon cells, J. Nutr., 120, 185, 1990.
36. Knaack, D., Podleski, T. R., and Salpeter, M. M., Ascorbic acid and acetylcholine
receptor expression, Ann. N.Y. Acad. Sci., 498, 77, 1987.
37. Leboy, P. S., Vaias, L., Uschmann, B., Golub, E., and Adams, S. L., Ascorbic acid
induces alkaline phosphatase, type X collagen, and calcium deposition in cultured chick
chondrocytes, J. Bioi. Chern., 264, 17281, 1989.
38. Lyons, B. L. and Schwarz, R. I., Ascorbate stimulation of PAT cells causes an increase
in transcription rates and a decrease in degradation rates of procollagen mRNA, Nucleic
Acids Res., 12, 2569, 1984.
39. Marchi, F. and Leblond, C. P., Radioautographic characterization of successive com-
partments along the rough endoplasmic reticulum-golgi pathway of collagen precursors
in foot pad fibroblasts of [3H]proline-injected rats, J. Cell Bioi., 98, 1705, 1984.
40. Margolis, R. L. and Lukens, L. N., The role of hydroxylation in the secretion of
collagen by mouse fibroblasts in culture, Arch. Biochem. Biophys., 147, 612, 1971.
41. Murad, S., Grove, D., Lindberg, K. A., Reynolds, G., Sivarajah, A., and Pinnell,
S. R., Regulation of collagen synthesis by ascorbic acid, Proc. Nat/. Acad. Sci. U.S.A.,
78, 2879, 1981.
42. Niki, E., Action of ascorbic acid as a scavenger of active and stable oxygen radicals,
Am. J. Clin. Nutr., 54, 1119S, 1991.
43. Nist, C., von der Mark, K., Hay, E. D., Olsen, B. R., Bornstein, P., Ross, R., and
Dehm, P., Localization of procollagen in chick corneal and tendon fibroblasts with
ferritin-conjugated antibodies, J. Cell Bioi., 65, 75, 1975.
44. Olsen, B. R., Berg, R. A., Kishida, Y., and Prockop, D. J., Collagen synthesis:
localization of prolyl hydroxylase in tendon cells detected with ferritin-labeled antibodies,
Science, 182, 825, 1973.
45. Olsen, B. R., Berg, R. A., Kishida, Y., and Prockop, D. A., Further characterization
of embryonic fibroblasts and the use of immunoferritin techniques to study collagen
biosynthesis. J. Cell Bioi., 64, 340, 1975.
46. Ouenzar, B., Weill, D., Agoutin, B., Keith, G., and Heyman, T., Relative content
of isoaccepting tRNAs for glycine and proline in avian tendon cells with different rates
of procollagen synthesis, Biochim. Biophys. Acta, 950, 429, 1988.
47. Owen, T. A., Aronow, M., Shalhoub, V., Barone, L. M., Wilming, L., Tassinari,
M. S., Kennedy, M. B., Pockwinse, S., Lian, J. B., and Stein, G. S., Progressive
development of the rate osteoblast phenotype in vitro: reciprocal relationships in expres-
sion of genes associated with osteoblast proliferation and differentiation during formation
of the bone extracellular matrix, J. Cell. Physioi., 143, 420, 1990.
Schwarz 505
48. Oyamada, I., Palka, J., Schalk, E. M., Takeda, K., and Peterkofsky, B., Scorbutic
and fasted guinea pig sera contain an insulin-like growth factor !-reversible inhibitor of
proteoglycan and collagen synthesis in chick embryo chondrocytes and adult human skin
fibroblasts, Arch. Biochem. Biophys., 276, 85, 1990.
49. Oyamada, I., Bird, T. A., and Peterkofsky, B., Decreased extracellular matrix pro-
duction in scurvy involves a humoral factor other than ascorbate, Biochem. Biophys. Res.
Commun., 152, 1490, 1988.
50. Pacifici, M., Independent secretion of proteoglycans and collagens in chick chondrocyte
cultures during acute ascorbic acid treatment, Biochem. J., 272, 193, 1990.
51. Pacifici, M. and Iozzo, R. V., Remodeling of the rough endoplasmic reticulum during
stimulation of procollagens secretion by ascorbic acid in cultured chondrocytes, J. Bioi.
Chern., 263, 2483, 1988.
52. Paz, M. A. and Gallop, P. M., Collagen synthesized and modified by aging fibroblasts
in culture, In Vitro, II, 302, 1975.
53. Pesciotta, D. M. and Olsen, B. R., The cell biology of collagen secretion, in Immu-
nochemistry of the Extracellular Matrix, Vol. II, Furthmayr, H., Ed., CRC Press, Boca
Raton, FL, 1982, I.
54. Peterkofsky, B., The effect of ascorbic acid on collagen polypeptide synthesis and proline
hydroxylation during the growth of cultured fibroblasts, Arch. Biochem. Biophys., 152,
318, 1972.
55. Peterkofsky, B., Regulation of collagen secretion by ascorbic acid in 3T3 and chick
embryo fibroblasts, Biochem. Biophys. Res. Commun., 49, 1343, 1972.
56. Peterkofsky, B. and Diegelmann, R., Use of mixture of protease-free collagenase for
specific assay of radioactive collage in the presence of other proteins, Biochemistry, 10,
988, 1971.
57. Peterkofsky, B. and Udenfriend, S., Conversion of proline to collagen hydroxyproline
in a cell-free system from chick embryo, J. Bioi. Chern., 238, 3966, 1963.
58. Peterkofsky, B. and Udenfriend, S., Enzymatic hydroxylation of proline in microsomal
polypeptide leading to formation of collagen, Proc. Nat/. Acad. Sci. U.S.A., 53, 335,
1965.
59. Pierce, G. B., Jones, A., Orfanakis, N. G., Nakane, P. K., and Lustig, L., Biosyn-
thesis of basement membrane by parietal yolk sac cells, Differentiation, 23, 60, 1982.
60. Pinnell, S. R., Murad, S., and Darr, D., Induction of collagen synthesis by ascorbic
acid, Arch. Dermatol., 123, 1684, 1987.
61. Quinones, S. R., Neblock, D. S., and Berg, R. A., Regulation of collagen production
and collagen mRNA amounts in fibroblast in response to culture conditions, Biochem.
J., 239, 179, 1986.
62. Rebouche, C. J,, Ascorbic acid and carnitine biosynthesis, Am. J. Clin. Nutr., 54,
1147S, 1991.
63. Ross, R. and Benditt, E. P., Wound healing and collagen formation, J. Cell Bioi.,
27, 83, 1965.
64. Rowe, L. B. and Schwarz, R. I., Role of procollagen mRNA levels in controlling the
rate of procollagen synthesis, Mol. Cell. Bioi., 3, 241, 1983.
65. Schwarz, R. 1., Mandell, R. B., and Bissell, M. J., Ascorbate induction of collagen
synthesis as a means for elucidating a mechanism for quantitative control of tissue-specific
function, Mol. Cell. Bioi., 1, 843, 1981.
66. Schwarz, R. I., Kleinman, P., and Owens, N., Ascorbate can act as an inducer of the
collagen pathway because most steps are tightly coupled, N. Y. Acad. Sci., 498, 172,
1987.
67. Schwarz, R. 1., Colarusso, L., and Doty, P., Maintenance of differentiation in primary
avian tendon cells, Exp. Cell Res., 102, 63, 1976.
68. Schwarz, R. I. and Bissell, M. J., Dependence of the differentiated state on the cellular
environment: modulation of collagen synthesis in tendon cells, Proc. Nat/. Acad. Sci.
U.S.A., 74, 4453, 1976.
69. Schwarz, R. I., Procollagen secretion meets the minimum requirements for the rate-
controlling step in the ascorbate induction of procollagen synthesis, J. Bioi. Chern .. 260,
3045, 1985.
70. Schwarz, R. I., Cell-to-cell signaling in the regulation of procollagen expression in
primary avian tendon cells, In Vitro Cell. Dev. Bioi., 27a, 698, 1991.
71. Schwarz, R. I., Farson, D. A., and Bissell, M. J., Requirements for maintaining the
embryonic state of avian tendon cells in culture, In Vitro, 15, 941, 1979.
72. Spanheimer, R. G., Bird, T. A., and Peterkofsky, B., Regulation of collagen synthesis
and mRNA levels in articular cartilage of scorbutic guinea pigs, Arch. Biochern. Biophys.,
246, 33, 1986.
73. Spanheimer, R. C. and Peterkofsky, B., A specific decrease in collagen synthesis in
acutely fasted, vitamin C-supplemented, guinea pigs, J. Bioi. Chern., 260, 3955, 1985.
74. Szent-Gyorgyi, A., Observations on the function of peroxidase systems and the chemistry
of the adrenal cortex. Description of a new carbohydrate derivative, Biochern. J .. 22,
1387, 1928.
75. Szent-Gyorgyi, A. and Haworth, W. N., 'Hexuronic acid' (ascorbic acid) as the an-
tiscorbutic factor, Nature, 131, 24, 1933.
76. Tajima, S. and Pinnell, S. R., Regulation of collagen synthesis by ascorbic acid.
Ascorbic acid increases type I procollagen mRNA, Biochern. Biophys. Res. Cornrn .. 106,
632, 1982.
77. Valmassoi, j. and Schwarz, R.I., High serum levels interfere with the normal differ-
entiated state of avian tendon cells by altering translational regulation, Exp. Cell Res.,
176, 268, 1988.
78. Veis, A. and Kirk, T. Z., The coordinate synthesis and cotranslational assembly of type
I procollagen, J. Bioi. Chern., 264, 3884, 1989.
79. Veis, A., Leibovich, S. J., Evans, J., and Kirk, T. Z., Supramolecular assemblies of
mRNA direct the coordinated synthesis of type I procollagen chains, Proc. Natl. Acad.
Sci. U.S.A., 82, 3693, 1985.
80. Vogel, Z., Daniels, M.P., Chen, T., Xi, Z-Y., Bachar, E., Ben-David, L., Rosenberg,
N., Krause, M., Duksin, D., and Kalcheim, C., Ascorbate-like factor from embryonic
brain. Role in collagen formation, basement membrane deposition, and acetylcholine
receptor aggregation by muscle cells, Ann. N. Y. Acad. Sci., 498, 13, 1987.
Chapter 22
DIETARY REGULATION OF METALLOTHIONEIN

EXPRESSION
Neil F. Shay and Robert J. Cousins
TABLE OF CONTENTS
I. Introduction ...................................................... 508
II. Metallothionein .................................................. 509

A. Roles of Metallothionein ................................. 509
B. Transcriptional Regulation of Metallothionein ............ 510
C. Turnover of Metallothionein .............................. 510
III. Dietary Regulation of Metallothionein Expression ............... 510

A. Metallothionein Levels and Zinc Uptake ................. 511
B. Metallothionein mRNA Studies .......................... 512
C. Other Studies ............................................. 51 5
IV. Other Genes Regulated by Dietary Zinc ......................... 517
V. Summary ........................................................ 520
References ............................................................... 521
0:oc_8;:-c49;;:o3--;c;69:-;c61c--:-4=/93/$0.00 +-$.5_0
I. INTRODUCTION
The divalent cation zinc is a required component of plant and animal

cells. The full spectrum of biological roles for this nutrient is only now
becoming appreciated. Zinc has been demonstrated to be an integral part of
at least 60 to 80 different enzymes. 1 It functions within enzymes providing
either a structural or catalytic role. Carbonic anhydrase, for example, has a
zinc ion coordinated within the active site of the enzyme, which has the ability
to catalyze over 36 million hydrations of carbon dioxide per second. 2 Zinc
"fingers" are polypeptide sequences within proteins containing cysteine and
histidine residues which allow for coordination of a zinc ion. 3 These structures
have been shown to interact with certain DNA sequences and function as part
of the DNA binding domain of transcription factors including TFIIIA of
Xenopus, 4 segmentation genes KruppeP and Hunchback6 of Drosophila, and
the human gene ZFY. 7 It has been proposed recently that the availability of
zinc for these fingers may be one way that the DNA-binding affinity of these
transcription factors, such as Spl, may be regulated, thereby regulating tran-
scription itself. 8 The zinc binding properties of the LIM motif, 9 which is
present in putative transcription factors, 10 is an example of another potential
zinc function at the gene level.
Decreased availability of zinc for cells can lead to dysfunction of one or
more of these zinc-requiring systems, which then leads to evidence of zinc
deprivation. Laboratory animals show the characteristic features of zinc de-
ficiency when supplied with a defined diet reduced in zinc content (e.g., 1
mg Zn per kg diet). Alopecia, dermatitis, and loss of appetite are typically
seen as animals become zinc deficient. 11 The murine "lethal milk" (lm)
syndrome is a genetic model for zinc deficiency. 12 Lethal milk dams are
unable to provide adequate amounts of zinc via milk for their offspring, despite
having adequate levels of zinc themselves. Without intervention, lm pups
die. With zinc supplementation, the pups grow in a normal fashion. This
zinc-responsive genetic lesion is most likely to reside in the mammary gland.
In humans, impaired growth and development, skin lesions, and altered im-
mune function are the most common signs of zinc deficiency. 11 The autosomal
recessive disease acrodermatitis enteropathica (AE) is the result of an inability
to efficiently absorb zinc from the diet. 13 AE is characterized by alopecia and
orificial skin lesions along with other signs of zinc deficiency. The condition
is successfully treated with an oral zinc supplement, apparently allowing
absorption of sufficient zinc to overcome an impaired intestinal uptake. A
similar condition is bovine Adema disease, which is characterized by de-
creased zinc absorption and immune dysfunction. 14 Collectively, these ex-
amples of congenital abnormalities are given to demonstrate that a variety of
genes have effects on the metabolic handling and/or utilization of zinc.
The focus of this chapter addresses the effects of dietary zinc on gene
expression. The genetic defects described above are examples of altered zinc
Shay and Cousins 509
metabolism and/or alteration of a zinc-requiring cellular function. Although

yet undetermined, this presumably involves altered protein structure or levels.
In contrast, there is substantial data on the regulation of metallothionein by
dietary zinc under practical conditions to suggest zinc has regulatory control
over some genes at the transcriptional level. There is recent evidence that the
response elements are not unique to this gene. Therefore, the mechanism of
transcriptional regulation of metallothionein expression by this micronutrient
may, in fact, parallel that for other zinc-responsive genes.
II. METALLOTHIONEIN
Metallothionein (MT) typically exists simultaneously as two distinct yet

very similar proteins, termed MT-1 and MT-2. In some species, for example,
humans, there may be several subclasses of different MT isoforms. 15 MTs
are hydrophilic, low molecular weight proteins (6 to 7 kDa) with a high
content (23 to 33 mol%) of cysteine residues. The high cysteine content allows
for the binding of heavy metals (usually 7 to 12 atoms/mole) via clusters of
thiolate bonds. The hydrophilic nature of the MT protein along with the
absence of disulfide bonds, despite the high cysteine content, makes MT a
heat stable protein.
A. ROLES OF METALLOTHIONEIN
Since the discovery of metallothionein, the function of this protein has
been debated and a number of potential roles have been postulated and tested.
The ability to serve as a cellular detoxification system has been well docu-
mented. 15 When a cell or organism is challenged with increased amounts of
a potentially damaging heavy metal, such as cadmium, metallothionein syn-
thesis is increased. The increased MT may then bind the cadmium, thus
reducing the toxic effect of the heavy metal. What is unclear is if this mech-
anism evolved as a protective mechanism against heavy metals or if it con-
veniently serves this role in addition to another role, such as the regulation
of free zinc levels within cells or protection against free-radical damage. 16
Zinc, a micronutrient required in the diet, 11 also activates the increase in MT
synthesis. Increases in MT protein allows for binding of excess zinc, thereby
serving as a homeostatic regulatory system for zinc. 16 It remains a largely
unanswered question whether the regulation of the required metal zinc, or the
toxic metals, such as mercury or cadmium, has been the driving force during
the evolution of the metallothionein regulatory system. Antiquity of the me-
tallothionein gene suggests the protein is serving a required, essential role. 15
Others have proposed roles for MT serving as an intracellular free radical
scavenger 17 or part of a system protecting the cells from insults such as ionizing
radiation or alkylating agents. 18 - 20
B. TRANSCRIPTIONAL REGULATION OF METALLOTHIONEIN

The metallothionein eDNA and gene were identified by Palmiter and co-
workers.21 As the protein was already known to be regulated by zinc, the
identification of metallothionein promoters allowed for the regulatory mech-
anism of metal induction to be elucidated. By combining the results of DNA
sequence information with the results of deletion mapping studies, unique,
short DNA sequences were found that mediated the metal response. These
metal response elements, or MREs, were identified in the 5' regulatory region
just upstream from the transcription start site. They were identified in trans-
genic mice using MT gene constructs containing deletions within the 5' region
with thymidine kinase as the reporter gene, and identifying when a construct
had either maintained or lost the characteristic metal response upon addition
of zinc or cadmium. 22 •23 It was apparent that these MREs were sites for trans-
acting transcription factors to bind and enhance the basal rate of transcription
of the MT genes. The various MT isoforms, and even the individual MREs
within one MT gene, have been shown to have a varied response to induction
by different metals. 24
Putative protein factors that bind to the MREs have been identified by a
variety of groups using DNA band shift analyses, but it is presently unclear
if there are several different MRE binding proteins or if some of these groups
have perhaps identified proteolytic fragments of a larger MRE binding protein.
Published reports identifying MRE binding proteins as 108, 25 74, 26 or 39
kDa 27 emphasize the potential heterogeneity of the results. The DNA sequence
of a MRE binding protein has not been cloned, nor has an antibody to such
a protein been produced. Study of the purified MRE binding proteins will be
fundamental to the understanding of how minerals regulate gene expression
in mammalian systems.
C. TURNOVER OF METALLOTHIONEIN
The turnover of MT has not been studied as well as the induction process.
However, evidence has shown that the type of metal ions bound to MT has
a direct influence on the stability of the protein to proteolytic activity. In
general, the cellular content of MT protein is a composite of dietary and
hormonal inputs vs. metal release and other factors that influence degradation.
Readers are referred to more complete discussions of MT turnover than space
permits in this chapter. 15 •16
III. DIETARY REGULATION OF

METALLOTHIONEIN EXPRESSION
From the discovery of metallothionein as a renal cadmium-associated
protein, the relationship between metals and MT has been a primary focus
of research on this protein. Along with oral doses or injections of metals into
test animals, it has been found that physiologically relevant levels of metals
(e.g., zinc) in the diet regulate levels of metallothionein in the kidney, liver,
intestine, and other tissues. 15 · 16 Increases in the abundance of MT is much
more striking in kidney, liver, and intestine, most likely due to changes in
the metabolic compartment that regulates transcriptional control by metals.
Early experiments done to quantitate changes in metallothionein levels by
dietary zinc used 35 S and/or 65 Zn labeling of cellular protein followed by
chromatography of cellular extracts. The increase in MT synthesis has since
been shown to be the result of increased MT gene transcription, with maximal
levels of MT mRNA abundance occurring a few hours before the increased
levels of protein are detected. 28 ·29
A. METALLOTHIONEIN LEVELS AND ZINC UPTAKE

Early work tested the alteration in metallothionein as a result of acute
changes in dietary zinc concentrations. 31 •32 A 1-d zinc-deficient diet ( < 1 mg
zinc per kg) fed to rats resulted in a reduction of the metallothionein protein
levels from normal to nearly undetectable levels. Repletion of zinc with a
diet containing 150 mg zinc per kg for one additional day boosted metallo-
thionein levels approximately 10-fold over the initial values. Subsequently,
another 1 d of zinc depletion dropped MT levels to near trace levels. Also,
when rats were deprived of zinc for 7 d and subsequently repleted with 25,
75, or 150 mg zinc per kg diet, both hepatic and intestinal MT increased in
proportion to the dietary level. The interpretations made from these studies
were (1) MT is rapidly synthesized to bind excess zinc, (2) MT serves as a
coordinating ligand to introduce new zinc into the intracellular pool, or (3)
MT serves in both functions. In another study, diets containing 1 to 2 g zinc
per kg diet fed to rats for up to 8 weeks was necessary to see dramatic
increases in liver MT levels. These superinduced MT levels could then be
dropped to trace amounts after only 3 d of feeding on a zinc-deficient diet. 32
Actinomycin D would block changes in metallothionein caused by increasing
the dietary zinc supply. 30 •31
The temporal relationship between MT mRNA and protein induction with
realistic dietary levels of zinc was more closely examined by McCormick et
al. 28 Rats were depleted of zinc for 4 d, then, after an overnight fast, were
fed by gavage a diet containing 1 (- Zn) or 125 ( + Zn) mg zinc per kg. In
response to the + Zn feeding, serum zinc values increased dramatically, peak-
ing at 3 h. They dropped slowly thereafter, not returning to normal levels for
15 h. During this period, the liver cytosol zinc content increased steadily, a
portion being associated with increases in both MT-1 and MT-2. MT abun-
dance was threefold higher in the + Zn rats as compared to the - Zn controls.
The increase in MT synthesis, as determined by presence of the MT mRNA
in polyribosomes, preceded the accumulation of zinc in the liver by a few
hours. The rate of MT synthesis increased to a maximum of 4.5-fold over
the initial rate I 0 h after administration of the + Zn diet. This pattern of
events for MT synthesis and zinc accumulation was also seen in the intestine,
where MT synthesis was elevated 4-fold in rats between 6 and 9 h after

feeding the + Zn diet. 33 This coincided with maximal levels of translatable
intestinal MT mRNA.
Food restriction in rats has also been shown to increase liver metallothi-
onein levels. 33 •34 These were increased up to tenfold in rats fed zinc-adequate
or zinc-deficient diets prior to fasting. It has not been demonstrated, however,
if this response is due to the liberation of zinc into the systemic circulation
from other body stores of zinc, such as muscle, redistribution of liver zinc,
a hormonal response due to the fasting, or to a combination of these factors.
Immunological methods have received limited use to measure metallo-
thionein levels in response to dietary treatments. A notable example is the
RIA data generated by Bremner and co-workers. 35 Using polyclonal ovine
antisera against rat MT -1, they demonstrated that dietary zinc depletion ( < 1
mg zinc per kg) reduced plasma and liver MT-1 levels. Food restriction also
elevated liver MT-1 levels. Plasma MT-1 is probably a reflection of hepatic
synthesis in response to upregulation of expression. Recently, the introduction
of an ELISA method for quantitating human MT protein levels by Grider et
a!. 36 has given nutritionists a tool that allows the measurement of MT levels
in human subjects. The results show that dietary zinc restriction to <0.4 mg/
day reduces by 59% erythrocyte MT levels in male subjects. 37 Similarly, zinc
supplementation increases erythrocyte MT by more than sixfold. These data
are assumed to reflect increased MT gene expression in progenitor cells in
marrow in response to dietary zinc intake and subsequent release of the MT-
containing reticulocytes into the circulation. Erythrocyte MT levels remain
elevated as long as supplementation continues, then fall with a half-life of
about 10 d. Since this is shorter than the life span of human erythrocytes (ca.
120 d), proteolysis or another event rendering the protein nonantigenic must
occur. Preliminary data with rats fed levels of dietary zinc between (1 vs.
180 mg zinc per kg) support the regulation of MT expression in marrow
cells. 38
There is a large volume of data on the inducibility of the metallothionein
gene. Only a few papers, which chronicle the evolution of methodology
applied to dietary regulation, have been cited here. A number of reviews can
be examined for a more complete presentation of the topic. 15 · 16 ·24
B. METALLOTHIONEIN mRNA STUDIES

Blalock et a!. 29 reported on a comprehensive set of experiments in rats
fed a diet with either deficient, adequate, or supplemented levels of both
copper and zinc. Rats were fed one of the nine different combinations of
copper (1 , 6, or 36 mg/kg) and zinc (5, 30, or 180 mg/kg) using a factorial
design. Rats were fed the diet for 14 d, and RNA was purified from liver,
kidney, small intestine, and the brain. Oligonucleotide probes were synthe-
sized corresponding to the 5' region of both rat MT-1 and -2 mRNA sequences.
After labeling the oligonucleotides with 32P-ATP, they were used as probes
Liver Kidney Intestine

Dietary Zinc (mg/Kg) 5 JO 180 5 JO 180 5 JO 180
~-actin-
Metallothionein -
Spleen Thymus Heart Lung

Dietary Zinc (mg/Kg) 5 JO 180 5 JO 180 5 JO 180 5 JO 180
~-actin-
Metallothionein -
FIGURE I. Northern blot analysis of metallothionein mRNA (bottom bands) and ~-actin
mRNA (top bands). Total RNA was extracted from liver, kidney. intestine, thymus, spleen,
lung, and heart of rats that were fed diets supplying 5, 30, or 180 mg zinc per kg (as indicated)
for 2 weeks. Each dietary group was composed of five rats. Each lane had 7.4 j.Lg of RNA.
Hybridization was to a mixed 60-base probe for metallothionein-1 and -2 mRNA plus a 60-base
probe for ~-actin mRNA. (From Cousins, R. J. and Lee-Ambrose, L. M .. J . Nutr. , 122, 56,
1992. With permission.)
for MT mRNA in the RNA samples affixed to nitrocellulose filters . After

hybridization with a 21 -mer and washing, the filters were used for autora-
diography. Then 32 P in each spot was quantitated using liquid scintillation
counting. It was found that the kidney, liver, and intestine all demonstrated
increases in MT mRNA abundance due to increased zinc in the diet, with no
effect seen in brain MT levels . The abundance of MT mRNA as determined
by autoradiography was highest in kidney and lower in the other tissues.
However, estimations of numbers of MT mRNA molecules per cell due to
differing amounts of zinc intake yielded relatively similar differences in these
tissues . Data from both methods of presentation show that copper at these
realistic dietary intakes does not affect MT expression in rats .
More recent data using 60-mer individual probes for MT-1 and MT-2
mRNAs have shown the same general response to dietary zinc at the deficient,
adequate and supplemental intake level. 39 As shown in Figure I , the decreasing
order of abundance is kidney > liver > intestine > heart > spleen. These

Dietary Zinc (mg/Kg) 5 30 1110 5 30 1110 5 30 1110
~-actin-
t.Aetallothionein 1-

Dietary Zinc (mg/Kg) 30 1110 ~ 30 1110 ~ 30 1110
~-act i n-
t.Aetallothionein 2 -
FIGURE 2. Northern blot analysis of metallothionein-1 mRNA, metallothionein-2 mRNA,

and [3-actin. Top panels: Metallothionein-1 mRNA (bottom bands) and [3-actin (top bands).
Bottom panels: Metallothionein-2 mRNA (bottom bands) and [3-actin (top bands). Total RNA
was extracted from liver, kidney , and intestine of rats that were fed diets supplying 5. 30, or
180 mg zinc per kg (as indicated) for 2 weeks. Each dietary group was composed of five rats.
Each Jane had 7.4 j.tg of RNA. Hybridization was to individual 60-base probes for metallothionein-
1 or -2 mRNA plus a 60-base probe for [3-actin mRNA. (From Cousins, R. J. and Lee-Ambrose,
L. M., J. Nutr., 122, 56 , 1992. With permission .)
probes do not reveal differences in expression in thymus or lung. Of interest

is that it appears that MT -1 mRNA is the more abundant message (Figure 2).
The significance of that difference is unclear at this time .
Huber and Cousins 40 demonstrated that maternal dietary zinc deficiency
can influence fetal gene expression. Pregnant rats were fed purified diets
containing 1 , 5, 30, or 180 mg zinc per kg representing deficient, marginal,
adequate, and supplemented levels of zinc, respectively. At the 20th d of
gestation, metallothionein mRNA and protein levels were measured in the
maternal liver, kidney, and placenta as well as the fetal liver. Fetal liver MT
mRNA and protein levels were found to be depressed when dams were fed
the deficient and marginal zinc diets (Figure 3). Maternal levels ofMT mRNA
and protein were likewise regulated by zinc, with the lowest amounts found
in the rats fed I and 5 mg zinc per kg diets and highest in rats fed the 180
mg zinc per kg diet. The significance of this study is apparent considering
Northern Blot
Fetal Liver Metollothionein mRNA
1AL 5R 180R 30R 30AL
._550b
FIGURE 3. Northern blot analysis of fetal liver metallothionein mRNA . Pregnant rats were
fed dietary zinc at 1, 5, 30, or 180 mg/kg diet from days 13 through 20 of gestation. Intake was
either restricted (R) to that of the I mg zinc per kg group or ad libitum (AL). Total RNA from
fetal liver was fractionated in 1. 1% agarose gels. Hybridization of RNA transferred by Northern
blot was to a 32 P-labeled 60-base probe oligonucleotide DNA probe for metallothioneins I and
2. The 550b metallothionein mRNA band is indicated. (From Huber, K. L. and Cousins, R. J..
J . Nutr., 118 , 1570, 1988. With permission .)
the impairment of fetal development due to zinc deficiency.4 1 The metallo-

thionein levels in fetal liver were many fold higher than levels found in adult
male and female rats, correlating with previous reports from other laborato-
ries . 15 This increased level of MT may be serving to provide adequate amounts
of zinc to zinc-requiring metalloproteins and cellular sites necessary in the
developmental program of the fetus. Alternatively, the MT may be serving
as a "sink" for zinc and other metals to supply the neonates with adequate
mineral levels for an initial period after birth.
C. OTHER STUDIES
Although metallothionein has been consistently considered a cytosolic
protein, immunocytochemical localization 42 has confirmed the presence of
metallothionein in the nucleus. Whether this is real or an artifactual contam-
ination has not been definitively addressed. However, with the identification
recently of zinc-finger-containing transcription factors, the possibility of me-
tallothionein playing a role in nuclear regulation of zinc may exist. In fact,
Zeng et al. 8 have speculated that metallothionein may play a role in the
regulation of zinc available for these zinc-containing nuclear metalloproteins,
perhaps adding another level of control to the process of gene regulation.
Specifically, they have shown that apo-metallothionein can remove zinc from
the transcription factor Spl, decreasing the DNA-binding activity of the pro-
tein. Since the binding constant of MT for metals is considerably greater than
other proteins, they propose metal exchange from Spl to apo MT occurs.
Presumably, the nuclear level of MT protein could regulate the availability
of Zn for Sp 1 and influence regulation of specific genes.
Cousins and Lee-Ambrose 39 have investigated the relationship between
the dietary regulation of metallothionein and the uptake of zinc into the nucleus
in a variety of rat tissues to determine the changes in nuclear zinc brought
about by dietary change. Growing male rats were fed 5, 30, or 180 mg zinc
per kg for a 2-week period. Nuclei purified by sucrose gradient methods from
the liver, kidney, and spleen had approximately the same total zinc content
as measured by atomic absorption. Subsequently, a constant dietary zinc intake
of either 5, 30, or 180 mg zinc per kg was provided to rats for 2 d, followed
by gavage of liquified diet containing those same levels of zinc. By tagging
the gavaged zinc with 65 Zn, it was shown that substantial amounts of dietary
zinc entered cell nuclei shortly after the oral feeding. Two hours after this
oral dose, nuclei were isolated from the tissues. Intestine and liver accu-
mulated substantial amounts of zinc and directly reflected the dietary zinc
level as shown in Table 1. For example, nuclei from kidney accumulated up
to 6.2% of the zinc in the oral dose during the 2-h period. Additionally,
chromatography of extracts from liver nuclei showed that 65 Zn introduced
into the portal supply bound to specific fractions of nuclear proteins, including
one specific fraction that could contain the zinc-containing metal responsive
element binding protein (MRE-BP) that stimulates transcription of the me-
tallothionein genes. Clearly, this study shows the rapidity of the association
of dietary zinc with nuclear proteins, consistent with the known temporal
regulation of the metallothionein gene by zinc administration. Westin and
Schaffne~ 3
showed that a MREd binding protein from nuclear extracts of
HeLa cells binds DNA in a zinc-dependent fashion. Andersen et al. 27 and
Searle 23 have made similar observations. This supports the hypothesis that an
influx of zinc from the diet rapidly reaches the nucleus and influences tran-
scription. The interrelationships among dietary zinc, MREs, and MRE-BP
are presented in Figure 4.
At the interface between contemporary molecular biology and traditional
dietary studies is the elegant transgenic mouse work done by Palmiter and
others. 22 •23 Chimeric fusion genes were created using recombinant techniques.
TABLE 1
Nuclear Zinc Derived from
Dietary Zinca
Dietary (% of total nuclear zinc

Zn intake derived from dietary zinc
(mg Zn/kg) Liver Kidney Spleen
5 0.1 0.3 0.1

30 0.8 0.9 0.3
180 4.8 6.2 1.6
• Rats were fed 30 mg zinc per kg of diet and

total zinc concentration of nuclei purified by
sucrose gradient methods was measured. Other
rats were fed a purified diet with 5, 30, or 180
mg zinc per kg containing 65 Zn 2 h prior to
purification of nuclei. Dietary zinc uptake was
calculated from the specific activity of the la-
beled diet.
Derived from Cousins, R. J. and Lee-Ambrose,

L. M., J. Nutr., 122, 56, 1992. With permission.
These genes had the mouse MT -1 promoter fused to the coding sequences of
a variety of genes including rat and human growth hormones. The gene
constructs were injected into the pronucleus of fertilized mouse ova. The
microinjected ova were then reimplanted into the oviducts of surrogate moth-
ers. Upon screening the offspring, it was found that approximately 25% were
expressing the transgenes. The level of growth hormone assayed in the trans-
genic animals was found to be regulated by zinc levels in coordination with
the native MT genes. This technology has many applications where dietary
zinc could provide the signal for transcriptional activation of trans genes. Zinc
is relatively nontoxic, and we have shown that the native MREs provide
excellent regulation by this micronutrient at the physiological range of dietary
intakes.
Of equal importance to zinc regulation of transgenes is the effect dietary
zinc could potentially have in metallothionein-related drug resistance. There
is convincing evidence that resistance to many alkylating agents (e.g., cis-
platin) used in chemotherapy is directly related to MT expression. 18-20 If
methods could be developed to block the role of dietary zinc in this activation,
drug resistance may be decreased.
IV. OTHER GENES REGULATED BY DIETARY ZINC

Recently, Yiangou et a!. 44 determined that several acute-phase protein
genes are induced by heavy metals. The mRNAs for C-reactive protein and
alpha-1-acid glycoprotein were found to be induced severalfold in the livers
Zinc = •
Dietary
Zinc Supply
~--,...•
.... MRE-BP
k ·~
~pl~m ~~
Nucleus 7 ~
C,,. . ~
, w- -
--
MRE Zn-regulated
Gene
7
FIGURE 4. Zinc-regulated gene expression. Diagram illustrates zinc-dependent gene tran-

scription. Zinc (•) supplied directly by the diet and indirectly from the endogenous supply enters
cells and is most likely found bound to metallothionein or other zinc-binding proteins. Metal-
lothionein may buffer or regulate the available free zinc supply. The zinc may bind in the
cytoplasm to the metal responsive element-binding protein (MRE-BP) or may be carried into
the nucleus by metallothionein or other proteins before interacting with the MRE-BP. Once the
MRE-BP has bound one or more zinc ions, it presumably has a higher affinity for the MRE
sequences of specific promoters and enhances transcription of the Zn-regulated gene by RNA
polymerase. The MRE-BP is shown here directly interacting with RNA polymerase-evidence
for this interaction has not yet been conclusively demonstrated.
of Balb/c mice treated with a variety of heavy metals given by i.p. injection.
Mercury elicited the strongest response, followed in order of response by Cd
> Pb > Cu > Ni > Zn > Mg. Other acute-phase reactants, such as alpha-
1-antitrypsin and alpha-1-antichymotrypsin, were unaffected by the metal
treatment. Additionally, albumin, which is downregulated during acute-phase,
was not repressed by the metal treatment, which indicates that the induction
of these mRNAs was not due to a metal-mediated inflammatory response.
Adrenalectomized animals were also challenged with metals, and the response
was still present, indicating that the induction by metals was not mediated
by the glucocorticoid induction pathway. They also noted that some DNA
sequence homologies exist between the MREs of the MT genes and sequences
in the promoter region of the alpha-1-acid glycoprotein and C-reactive protein
genes. Metal induction potential of these acute phase genes were tested in a
transiently tranfected cell system. Approximately 600 bp of the alpha-1-acid
glycoprotein promoter sequence was linked to the bacterial chloramphenicol

(CAT) gene in a pPS-CAT transfection vector. This plasmid was then used
to transiently transfect human HepG2 cells growing in culture. Addition of
mercury to the culture medium ( 1 J.LM to 1 nM) was sufficient to stimulate
CAT activity three- to fourfold in the HepG2 cells. This cell system, free of
the extraneous influences of hormones that would be produced at other sites
in vivo, strongly suggests that these genes are being directly regulated by the
presence of metals, and the site of interaction is probably the MRE-like
sequences found in the promoter region of these genes. Certainly, analysis
of deletion mutants by transfection and CAT assay will determine which sites
within the promoter are responsible for the metal induction. Dietary regulation
of these genes has not been examined, but would appear to be a promising
area to study.
Shay and Cousins 45 have conducted a "plus-minus" or differential hy-
bridization experiment with the express intent of finding mRNAs in the in-
testine and skin that are regulated by zinc levels in the diet. Rats were fed a
purified diet for 14 d containing either a deficient (1 mg zinc per kg) or
adequate (30 mg zinc per kg) level of zinc, after which mRNA was purified
from the intestine and skin. eDNA plasmid libraries were made and screened
with 32 P-labeled eDNA probes derived from the zinc-deficient and -adequate
pools of mRNA. We have identified several mRNAs that are regulated by
the zinc levels in the diet and are currently characterizing these mRNAs by
measuring their relative abundance under a variety of dietary conditions,
determining their tissue specificities, and sequencing the respective cDNAs
to determine the identity of these mRNAs or other homologous sequences
identified by DNA sequence data base search.
Finally, Hempe and Cousins46 discovered a low molecular weight zinc-
binding protein present in rat intestinal mucosa cells that could be a candidate
protein for an intracellular carrier of zinc. The protein could potentially be
functioning as a transporter of zinc from the luminal brush border to the
circulatory system. 47 The protein was identified using 65 Zn ligand blotting of
SDS-PAGE/size-fractionated protein as well as dot blotting of cytosolic pro-
tein fractions size-fractionated by Superose-HPLC. The protein was blotted
onto a PVDF membrane following SDS-PAGE, excised, and used for protein
microsequencing. Analysis of the N terminal amino acid sequence allowed
the known protein data bases to be searched, and it was found that this protein
was the "Cysteine-Rich Intestinal Protein" (CRIP) identified as a eDNA
sequence by Birkenmeier and Gordon48 in 1986. They screened a rat intestine
eDNA library to identify eDNA clones induced during weaning. The CRIP
eDNA was isolated, determined that it was nearly tissue-specific to the in-
testine, and that the mRNA was induced from low basal levels during suckling
to a higher, moderately abundant concentration after weaning.
Birkenmeier and Gordon 48 noted some homology between CRIP and
certain bacterial ferredoxins, whereas the similarity between CRIP and the
histidine and cysteine residues of the "LIM" motif was recognized by Freyd
et al. 10 The latter is found in some transcription factors and has some features
in common with the consensus sequence of zinc-finger proteins. Certainly,
the structural features of this protein coupled with the nearly absolute local-
ization to the intestine and induction at weaning give us strong circumstantial
evidence that this zinc-binding protein is crucially important to nutrient ab-
sorption. It will be both interesting and a challenge to design experiments
that define the nutritional regulation of CRIP.
V. SUMMARY
Metallothionein has proved to be enigmatic to those who have chosen to

study it. The mechanism of MT gene regulation has been clearly defined. It
is transcriptionally activated not only by numerous hormones and other cellular
factors, particularly in a tissue-specific mode by various cytokines, but uniquely
by dietary zinc. Data suggest that dietary copper does not have a major effect
on MT gene expression at practical intake levels. This metal regulation is
conferred by MRE sequences in the promoter region that apparently bind the
putative MRE-BP that "senses" the metal status within the cell and elevates
transcription of MT genes when bound to the MRE sequences. Several groups
have identified candidate MRE-binding proteins. It is not clear at the present
time if there is a single or multiple species of this protein. Metallothionein
gene regulation has been so well understood that it has served not only as a
model for nutritionists interested in gene regulation but for all biologists
interested in gene expression. The metal regulation of the MT promoter has
been used in a variety of chimeric constructs allowing genes of interest to be
put under the control of zinc concentrations in transfected cell systems. These
systems have been so widely accepted that transfection vectors are now com-
mercially available using the MT promoter to control inserted eDNA coding
sequences.
In contrast to the mechanism of expression, the function of metallothionein
has proved to be a frustration. Clearly, the protein is involved in toxicity
protection, 21 and recent genetic studies done in yeast suggest that the major
function of metallothionein may be as a free radical scavenger, 20 a function
that has been investigated for a number of years by various groups. The metal
regulation of MT also confers on it a role in metal homeostasis in vivo, and
we have tried to present here some results from investigations involving dietary
regulation of metallothionein by zinc and the implications of that regulation
on cellular zinc availability, in particular.
Recently, several genes have been identified that are either regulated by
metals or are serving important functions in the absorption or homeostasis of
minerals. With the current pace of advances in molecular cell biology, it is
hoped that the roles of these genes in nutrition will be rapidly determined.
REFERENCES
I. Vallee, B. L. and Galdes, A., The metallobiochemistry of zinc enzmes, Adv. Enzymol.,
57, 283, 1984.
2. Lehninger, A. L., Principles ofBiochemistry, Worth Publishers. New York, 1982, 220.
3. Miller, J., McLachlan, A. D., and Klug, A., Repetitive zinc-binding domains in the
protein transcription factor IliA from Xenopus oocytes, EMBO J .. 4, 1609, 1985.
4. Engelke, D. R., Ng, S.-Y., Shastry, B. S., and Roeder, R. G., Specific interaction
of a purified transcription factor with an internal control region of 5S RNA genes, Cell,
19, 7l7, 1980.
5. Rosenberg, U. B., Schroder, C., Preiss, A., Kienlen, A., Cote, S., Riede, I., and
Jackie, H., Structural homology of the product of the Drosophila Kruppel gene with
Xenopus transcription factor IliA, Nature, 319, 336, 1986.
6. Tautz, D., Lehman, R., Schnurch, H., Schuh, R., Seifert, E., Kienlin, A., Jones,
K., and Jackie, H., Finger protein of novel structure encoded by hunchback, a second
member of the gap class of Drosophila segmentation genes, Nature (London), 327, 383,
1987.
7. Page, D. C., Mosher, R., Simpson, E. M., Fisher, E. M. C., Mardon, G., Pollack,
J., McGillivray, B., de Ia Chapelle, A., and Brown, L. G., The sex-determining region
of the human Y chromosome encodes a finger protein, Cell, 51, 1091, 1987.
8. Zeng, J., Heuchel, R., Schaffner, W., and Kagi, J. H. R., Thionein (apometallothi-
onein) can modulate DNA binding and transcription activation by zinc finger containing
factor Sp1, FEBS Lett., 279, 310, 1991.
9. Li, P. M., Reichert, J., Freyd, G., Horvitz, H. R., and Walsh, C. T., The LIM
region of a presumptive Caenorhabditis elegans transcription factor is an iron-sulfur and
zinc-containing metallodomain, Proc. Nat/. Acad. Sci. U.S.A., 88, 9210, 1991.
10. Freyd, G., Kim, S. K., and Horvitz, H. R., Novel cysteine-rich motif and homeodomain
in the product of the Caenorhabditis elegans cell lineage gene lin-11, Nature (London),
344, 876, !990.
11. Hambidge, K. M., Casey, C. E., and Krebs, N. F., Zinc, in Trace Elements in Human
and Animal Nutrition II, Mertz, W., Ed., Academic Press, Orlando, 1986, I.
12. Piletz, J. E. and Ganschow, R. E., Zinc deficiency in murine milk underlies expression
of the lethal-milk (lm) mutation, Science, 199, 181, 1978.
13. Lombeck, I., Schnippering, H. G., Ritz!, F., Feinendegen, L. H., and Bremner,
H. J., Absorption of zinc in acrodermatitis enteropathica, Lancet, I, 855, 1975.
14. Flagstad, T., Zinc absorption in cattle with a dietary picolinic acid supplement, J. Nutr.,
111' 1996, 1981.
15. Kagi, J. H. R. and Kojima, Y., Chemistry and biochemistry of metallothionein, in
Metallothionein ll, Experientia Supplementum, Vol. 52, Kagi, J. H. R. and Kojima, Y.,
Eds., Birkhauser Verlag, Basel, 1987, 25.
16. Dunn, M. A., Blalock, T. L., and Cousins, R. J., Metallothionein, Proc. Soc. Exp.
Bioi. Med., 185, 107, 1987.
17. Thornalley, P. J. and Vasak, M., Possible role for metallothionein in protection against
radiation-induced oxidative stress. Kinetic and mechanism of its reaction with superoxide
and hydroxyl radicals, Biochim. Biophys. Acta, 827, 36, !985.
18. Lohrer, H. and Robson, T., Overexpression of metallothionein in CHO cells and its
effect on cell killing by ionizing radiation and alkylating agents, Carcinogenesis, 10,
2279, 1989.
19. Kelley, S. L., Basu, A., Teicher, B. A., Hacker, M. P., Hamer, D. H., and Lazo,
J_ S., Overexpression of metallothionein confers resistance to anticancer drugs, Science,
241, 1813, 1988.
20. Kaina, B., Lohrer, H., Karin, M., and Herrlich, P., Overexpressed human metal-
lothionein IIA gene protects Chinese hamster ovary cells from killing by alkylating agents,
Proc. Nat/. Acad. Sci. U.S. A., 87, 2710, 1990.
21. Durnam, D. M. and Palmiter, R., Transcriptional regulation of the mouse metallothi-
onein-1 gene by heavy metals, J. Bioi. Chern., 256, 6712, 1981.
22. Stuart, G. W., Searle, P. F., Cen, H. Y., Brinster, R. L. and Palmiter, R. D., A
12-base-pair DNA motif that is repeated several times in metallothionein gene promoters
confers metal regulation to a heterologous gene, Proc. Nat/. Acad. Sci. U.S.A., 81, 7318,
1984.
23. Searle, P. F., Zinc dependent binding of a liver nuclear factor to metal response element
MRE-a of the mouse MT-1 gene and variant sequences. Nucleic Acids Res., 18, 4683,
1990.
24. Hamer, D., Metallothionein, Annu. Rev. Biochem., 55, 913, 1986.
25. Seguin, C. and Prevost, J., Detection of a nuclear protein that interacts with a metal
regulatory element of the mouse MT-1 gene, Nucleic Acids Res., 16, 10547, 1988.
26. lmbert, J., Zafarullah, M., Culotta, V. C., Gedamu, L., and Hamer, D., Transcrip-
tion factor MBF-1 interacts with metal regulatory elements of higher eucaryotic MT genes,
Mol. Cell. Bioi., 9, 5315, 1989.
27. Andersen, R. D., Taplitz, S. J., Oberbauer, A.M., Calame, K. L., and Herschman,
R., Metal-dependent binding of a nuclear factor to the rat MT-1 promoter, Nucleic Acids
Res., 18, 6049, 1990.
28. McCormick, C. C., Menard, M.P., and Cousins, R. J., Induction of hepatic MT by
feeding zinc to rats of depleted zinc status, Am. J. Physiol., 240, E414, 1981.
29. Blalock, T. L., Dunn, M. A., and Cousins, R. J., Metallothionein gene expression in
rats: tissue-specific regulation by dietary copper and zinc, J. Nutr., 118, 222, 1988.
30. Richards, M. P. and Cousins, R. J., Metallothionein and its relationship to the me-
tabolism of dietary zinc in rats, J. Nutr., 106, 1591, 1976.
31. Richards, M. P. and Cousins, R. J., Zinc-binding protein: relationship to short term
changes in zinc metabolism, Proc. Soc. Exp. Bioi. Med., 153, 52, 1976.
32. Whanger, P. D. and Weswig, P. H., Effect of some copper antagonists on induction
of ceruloplasmin in the rat, J. Nutr., 100, 341, 1970.
33. Menard, M. P., McCormick, C. C., and Cousins, R. J., Regulation of intestinal
metallothionein biosynthesis in rats by dietary zinc, J. Nutr., 111, 1353, 1981.
34. Bremner, I. and Davies, N. T., The induction of metallothionein in rat liver injection
and restriction of food intake, Biochem. J., 149, 733, 1975.
35. Sato, M., Mehra, R. K., and Bremner, I., Measurement of plasma metallothionein-
1 in the assessment of the zinc status of zinc-deficient and stressed rats, J. Nutr., 114,
1683, 1984.
36. Grider, A., Kao, K. J., Klein, P. A., and Cousins, R. J., Enzyme-linked immuno-
sorbent assay for human metallothionein: correlation of induction with infection, J. Lab.
Clin. Med., 113, 221, 1989.
37. Grider, A., Bailey, L. B., and Cousins, R. J., Erythrocyte metallothionein as an index
of zinc status in humans, Proc. Natl. Acad. Sci. U.S.A., 87, 1259, 1990.
38. Huber, K. L. and Cousins, R. J., K562 cells respond to zinc and alter metallothionein
gene expression during differentiation, FASEB J., 6, 873A, 1992.
39. Cousins, R. J. and Lee-Ambrose, L. M., Nuclear zinc uptake and interactions and
metallothionein gene expression are influenced by dietary zinc in rats, J. Nutr., 122, 56,
1992.
40. Huber, K. L. and Cousins, R. J., Maternal zinc deprivation and interleukin-1 influence
metallothionein gene expression and zinc metabolism of rats, J. Nutr., 118, 1570, 1988.
41. Hurley, L. S., Teratogenic aspects of manganese, zinc and copper nutrition, Physiol.
Rev., 61, 249, 1981.
42. Cherian, M. G., Templeton, D. M., Gallant, K. R., and Banerjee, D., Biosynthesis
and metabolism of metallothionein in rat during perinatal development, in Metallothionein
II, Experientia Supplementum, Vol. 52, Kagi, 1. H. R. and Kojima, Y., Eds., Birkhauser
Verlag, Basel, 1987, 499.
43. Westin, G. and Schaffner, W., A zinc-responsive factor interacts with a metal-regulated
enhancer element (MRE) of the mouse metallothionein-1 gene, EMBO J., 7, 3763, 1988.
44. Yiangou, M., Ge, X., Carter, K. C., and Papaconstantinou, J., Induction of several
acute-phase protein genes by heavy metals: a new class of metal-responsive genes, Bio-
chemistry, 30, 3798, 1991.
45. Shay, N. F. and Cousins, R. J., Identification and cloning of zinc-regulated mRNAs
from rat intestine, FASEB J., 6, A483, 1992.
46. Hempe, J. M. and Cousins, R. J., Cysteine-rich intestinal protein binds zinc during
transmucosal zinc transport, Proc. Natl. Acad. Sci. U.S.A., 88,9671, 1991.
47. Hempe, J. M. and Cousins, R. J., Cysteine-rich intestinal protein and intestinal me-
tallothionein: an inverse relationship as a conceptual model for zinc absorption in the rat,
J. Nutr., 122, 89, 1992.
48. Birkenmeier, E. H. and Gordon, J. I., Developmental regulation of a gene that encodes
a cysteine-rich intestinal protein and maps near the murine immunoglobulin heavy chain
locus, Proc. Nat/. Acad. Sci. U.S.A., 83, 2516, 1986.
Chapter 23
IRON-DEPENDENT REGULATION OF
FERRITIN SYNTHESIS
H. N. Munro, Z. Kikinis, and R. S. Eisenstein
TABLE OF CONTENTS
I. Overview ........................................................ 526
II. Ferritin Subunits: Amino Acid Sequences, Messenger

RNAs, Genes .................................................... 526
III. Regulation of Ferritin Synthesis by Iron ......................... 528

A. Translation ............................................... 528
1. The Ferritin Model: The Iron-Responsive
Element ........................................... 528
2. Other Systems Involving the Iron-
Responsive Element .............................. 538
B. Transcription ............................................. 539
I. Response of H and L Genes to Iron .............. 539
2. Response of H and L Genes to Other
Agents ............................................ 539
IV. Conclusion ....................................................... 540
References ............................................................... 541
0-8493-6961-4/93/$0.00 +$.50
I. OVERVIEW
Since the middle of the last century, it has been known that tissues contain
iron-positive granules called hemosiderin. 1 In 1937, Laufberger isolated and
crystallized another iron-rich compound from horse spleen and named it fer-
ritin to reflect its high iron content. 2 It has since been shown that ferritin is
a ubiquitous constituent of animal, plant, and microorganismal cells. Ferritin
consists of a protein shell made from subunits, usually 24 in number, sur-
rounding a cavity in which iron can be stored. Harrison et al. 3 provide an
excellent account of the physical features of ferritin and its function in iron
storage. In 1974 to 1975, two groups4 ·5 recognized heterogeneity in the sub-
units of the protein shell and eventually these were classified either as light
(L) or heavy (H) based on their slightly different molecular weights.
Drysdale and Munro 6 had shown in 1966 that iron induces ferritin syn-
thesis in the liver of the rat by a mechanism that is not quenched by acti-
nomycin D and therefore represents a posttranscriptional event. In 1976,
Zahringer et al. 7 confirmed and extended this observation by showing that in
rats not receiving extra iron a large number of ferritin mRNAs lie dormant
in their liver cell sap. When the rats were given iron supplements, these
reserve mRNAs became attached to ribosomes in the liver and started to
translate ferritin mRNA actively. They proposed that, at low levels of intra-
cellular iron, a specific protein binds to the anterior end (5' untranslated
region, 5'UTR) of ferritin mRNA, thus blocking initiation of ferritin protein
synthesis. When more iron enters the cell, the regulatory protein loosens so
that ferritin protein synthesis can proceed to meet the need for more ferritin
molecules to trap the excess iron. These and other studies of ferritin up to
1978 are reviewed elsewhere. 8 The present chapter deals in detail with research
performed during the past decade which has considerably increased our un-
derstanding of the mechanisms involved in the translational regulation of
ferritin mRNA availability in order to achieve iron homeostasis.
II. FERRITIN SUBUNITS: AMINO ACID SEQUENCES,

MESSENGER RNAs, GENES
Using sequenator analysis or more commonly deducing amino acids from
nucleotide sequences involving eDNA and occasionally ferritin genes, anum-
ber of investigators have reported the amino acid sequence of the L and H
subunits of a series of mammals and the chicken as summarized elsewhere, 9
to which can be added the more recently analyzed L subunit chain of the
mouse 10 and also the three chains of the bullfrog" and the H chain of the
snail. 12 If we assign the human L subunit a value of 100, the L chain of
human liver and spleen are essentially identical (98% and 97%, respectively),
whereas the L chain of horse spleen shows only 88% identity, rabbit liver
85%, rat liver 84%, and mouse liver 82% similarity. In contrast, taking human
Munro et al. 527
5'UTR CODING REGION 3' UTR
l-MRNA POLY A
95% LOOP 55% AA

HOMOLoGY HOMOLOGY
H-MRNA POLY A
153 B
RAT FERRITIN L-GENE

ATG
>lkb -so
RAT FERRITIN H-GENE
FIGURE 1. Upper diagram: Comparison of the number of nucleotide bases (B) and amino
acids (AA) of the coding regions of rat ferritin L and H subunit mRNAs. Also shown are the
features of the 5' and 3' untranslated regions with loop and stem structure at the same position
in the 5' regions of both mRNAs. Lower diagram: Comparison of the genomic structures of the
expressed rat ferritin L and H subunit genes. The exons are indicated by thick bars and the
introns by thin lines. The location of the TATA box, initiating methionine codon (ATG), ter-
mination codons (TAG,TAA), and polyA polymerase (AATAA) are indicated. (From Munro,
H. N. eta!., Ann. N.Y. Acad. Sci., 526, 117, 1988. With permission.)
liver H subunit as 100, rat liver shows 95% identity of sequence, mouse 94%,
and chicken 93%, thus suggesting more conservation.
The major features of the mRNAs of rat ferritin L and H subunits are
shown in Figure I. Despite closeness in size, the reading frames for rat L
and H are only 55% homologous; a similar divergence is seen when human
Land H subunits are compared, namely 58%. However, when the Chou and
Fasman rules 13 are applied, both subunits show five alpha-helices, thus sup-
porting a common ancestral gene. This suggests that L and H subunits may
interchange in the construction of the protein shell and thus give rise to
numerous isoferritins.
Figure also shows that L and H subunit mRNAs have long 5' UTRs
which, for the most part, show little homology except for a sequence of 28
nucleotides close to the cap region. This region is highly conserved between
rat H and L and also between species as widely divergent as man and the
bullfrog. 14 On the basis of the reported sequences of this region for the L
chain of the human, rabbit, rat, and mouse, the consensus becomes:
UCUUGCUUCAACAGUGUUUGGACGGAAC L chain
UC~UGCUUCAACAGUGCUUGGACGGAAC H chain
The H ferritin of human, rat, mouse, and chicken are similar except for
substitution of C for U in two places in all four H sequences. On computer
analysis, the conserved sequence is able to form a double-stranded stem
supporting a loop (underlined) of six nucleotides. This sequence will be further
discussed under the section on regulation of ferritin mRNAs by iron.
Regarding the L and H ferritin genes of the rat, Southern blots show more
than 20 copies of the L gene 15 and 5 of the H gene. 16 Figure 1 displays the
exon and intron patterns of the single expressed L and H genes of the rat,
the others being considered as pseudogenes. The latter were evident when a
selection of L genes of the rat were sequenced and were found to contain no
introns, but to have polyadenylic acid tails and direct repeats in the adjacent
DNA, all of which is compatible with the random insertion of processed
pseudogenes into DNA by a retrovirus. Figure 1 shows that 3 introns (thin
lines) separate the 4 exons (thick lines) which encompass the 5 alpha-helices
of the L and H subunits, the small fifth alpha helix being present with the
fourth alpha helix on the last exon. Note also that the first and the last exons
also carry the 5' and 3' UTRs, respectively, of the mRNA. Irrespective of
whether the genes are from Lor H subunits, or are from man, rat, or chicken,
the middle two exons have identical lengths of 147 and 126 nucleotides,
respectively, representing 49 and 42 amino acids in the completed protein.
This is compelling evidence that the L and H genes share a common ancestry.
III. REGULATION OF FERRITIN SYNTHESIS

BY IRON
The levels of control of ferritin gene expression are illustrated in Figure
2. This involves regulation of transcription and also of translation in response
to iron and possibly to other factors.
A. TRANSLATION
1. The Ferritin Model: The Iron-Responsive Element
As described in the introduction, injection of iron salts into the peritoneal
cavity of the rat is quickly followed by the synthesis of extra ferritin subunits
by a mechanism that is not quenched by prior injection of actinomycin D. 8
Munro et al. 529
DNA
4- Actinomycin
RNA
- Cordycepin
MRNA
~AAA
~~~r
/ PDLYRIBOSOMES
~ 5' IReading I 3' AAA ~

frame
,!
& ~ ~-------
__
repressor
molecule 1....
Transferrin Ferric Hemin

NH 4 c it rate Fe
Fe~, ~Fe
DESFERR I-
OXAMINE '/
DPTA
FIGURE 2. Factors involved in regulation of ferritin subunit mRNAs. These include the supply
of mRNA from the nucleus, segregation of inactive ferritin mRNA, and its release for polyri-
bosome function when excess iron enters the cell. This is determined by the effect of the entering
iron on the free iron pool which can be restricted by chelation with desferioxamine which enters
the cell, but not with DPTA, which cannot enter. Note that iron can come from transferrin,
inorganic salts, or hemin, the latter through the action of heme oxygenase. (From Munro.
H. N. et al.. Ann. N.Y. Acad. Sci., 526. 117, 1988. With permission.)
This mechanism is therefore unlikely to involve transcription but rather trans-

lation. The interpretation was strengthened when a large stock of dormant
mRNAs of ferritin H and L subunits were identified in the cell sap of rats
untreated with iron. 7 The stored mRN A represents a reserve needed for rapid
availability of ferritin when the cell receives excessive amounts of iron. When
more sophisticated techniques became available, the dormant and the active
ferritin mRNA could be separated into free ribonucleoprotein (mRNP) par-
ticles and polysome-attached mRNA. Density gradients were used to separate
ferritin mRNAs at different levels on the gradient ranging from RNPs particles
at the top to large polyribosomes at the bottom. The ferritin mRNA was then
quantitated with Northern blots. Figure 3 illustrates 17 such a gradient for
separating L-mRNA from rat liver cell sap according to whether the animal
had received no extra iron (A) or had been injected with iron (B). As measured
-2.8
B
-0.97
2 3 4 ~ 6 78 9 10 II 12 13 14 I~ 16 17 18
FJGURE 3. Cytoplasmic distribution of the rat ferritin L subunit mRNA in control (above)
and iron-treated rats (below). Rats were injected intraperitoneally with ferric ammonium citrate
or saline (controls) and killed 3.5 h later. Sucrose gradients of liver homogenates were run. The
gradients were divided into 18 fractions from the top ("free'' mRNP) through monosomes to
the bottom (large polysomes). Each fraction was extracted from RNA, which was then run on
the Northern blot gel to show the quantity of L subunit RNA in each fraction using hybridisation
with labeled L subunit eDNA. The gel was then autoradiographed. (From Aziz , N. and Munro,
H. N., Nucleic Acid Res., 14,9 15 . 1986. With permission .)
by Northern blotting of different fractions of the gradient, it will be seen that,

in the case of the uninjected rats, most of the L-mRNA is in the free mRNP
fraction where it lies dormant. On the other hand, rats killed 3 h after injection
of iron show extensive translocation of L-mRNA to the polyribosome frac-
tions. This procedure was then repeated for ferritin H-mRNA which showed
a similar response to iron administration. Thus, the basic picture presented
by Zahringer et al. 7 is substantially correct.
It can be accepted that cells accumulate large stocks of ferritin Land H-
mRNAs in order to respond rapidly to increases in intracellular iron levels
which would otherwise cause peroxidation of lipids and DNA and result in
other disruptive actions. For example, Dougherty et al. 18 describe experiments
in which rats received a single peritoneal dose of iron salts and responded by
Munro et al. 531
exhaling ethane made from peroxidation of membrane-associated omega-3-

fatty acids. The resulting generation of free radicals could well be a factor
in aging. 19
The mechanism linking intracellular iron level to synthesis of ferritin can
now be considered. The occurrence of a highly conserved sequence near the
cap region of the S' UTR (Figure I) was examined by computer for the
presence of a stem-loop structure which was indeed found by us in the first
79 nucleotides of the S' untranslated region of L ferritin. 20 This finding has
been confirmed by several investigators including ourselves, using newer
computer programs with which details emerge regarding the stem-loop struc-
ture. This motif, a candidate for being involved in the response to iron, is
usually referred to as the iron regulatory element or IRE. In order to function
as a regulator, it has to act as the cis element to which a specific protein acts
as the trans element.
First, it was necessary to confirm that this consensus sequence is involved
in responding to the iron in the cell.zt Accordingly a chimera was made in
which the S' and 3' UTRs from rat liver L-mRNA are joined to the reading
frame of the bacterial protein CAT (chloramphenicol acetyl transferase). Using
retrovirus-mediated gene transfer, this construct and modifications of it were
inserted into rat hepatoma cells in culture. The inserted sequences included
the intact S' UTRf (SCAT3), then a sequence with deletion of the first 67
nucleotides of the S' UTR (SsCAT3) and finally another with removal of the
3' UTR (SCAT). Cells with the complete insert (SCAT3) showed a response
to extra iron in the medium by synthesizing more CAT (Figure 4). Deletion
of the first 67 bases (SsCAT3) resulted in the disappearance of CAT respon-
siveness to iron, showing that the regulatory sequence occurs within these 67
nucleotides. The removal of the 3' UTR (SCAT) did not affect responsiveness
to iron. On the other hand, Dickey et al. 22 studied the loss of the 3'-region
from the ferritin mRNA of the tadpole and concluded that it played a part in
the response of ferritin synthesis to iron. Using human H-mRNA with CAT
and also growth hormone as reported proteins, Caughman et alY reached
conclusions similar to ours. In a more recent paper, 24 they varied the position
of the stem-loop structure along the S' UTR and concluded that its normal
location near the cap was essential to its regulatory function, possibly because
it interfered in that position with initiation of protein synthesis.
The trans element interacting with the stem-loop structure is a protein
present in the cell sap. Using rat L-mRNA, Leibold and Munro 25 prepared a
32
P-labeled transcript of the initial6S nucleotides of the S' UTR which included
the conserved 28 nucleotides. The transcripts were added to cell sap harvested
from rat liver and rat heart and from hepatoma cells in culture and were
allowed to react. The products were then separated on polyacrylamide gels.
Figure S illustrates an experiment in which cell sap was extracted from rat
hepatoma cells either before or after iron treatment and was then reacted with
[ P]-S'LRNA containing the 28-nucleotide sequence. As compared to the
32
migration of RNA alone on the gel, the binding of RNA to the protein retarded
...
,~
- ;>'
AT(i
CAT ~
,., ~
5 CAT 3
.. ATG •0#
• 5• CAT 3
...
,~
&TG
CAT
•0#
~'
... '~
.....
5 CAT
CAT ~'
,~ ~o'
• 0#
EC EC EC s
5CAT3 5sCAT 3 5CAT
FIGURE 4. Inducibility of the bacterial enzyme chloramphenicol acetyl transferase (CAT) by

iron using the intact ferritin-CAT chimera (5CAT3) and the chimera with deletion of the first
67 nucleotides of the 5' UTR (5sCAT3) or deletion of the 3' UTR (SCAT). These were per-
manently transfected into Ff0-28 cells, and iron in the form of hemin-treated (E) or control
cells (C) were incubated for 14 h followed by CAT assay. (From Aziz, N. and Munro, H. N.,
Proc. Nat/ Acad. Sci. U.S.A., 84, 8478, 1987. With permission.)
Munro et al. 533
<(
z
0::
_J FT0-28 1
l{) 0 248 16
81-
82-
Free-
234567
FIGURE 5. RNA-protein complex formation with SIOO extracts from hepatoma cells (Ff0-
2B) treated with ferric ammonium citrate for 0, I, 2, 4. 8, and 16 h. Positions of complexes
Bl and B2 and free RNA are indicated. (From Leibold, E. A. and Munro. H. N .. Proc. Nat/.
Acad. Sci. U.S.A., 85.2171, 1988. With permission.)
the migration of the labeled 5' UTR-protein complex which appeared as two
bands, B-1 and B-2. When the cells were incubated with extra iron, there
was extensive loss of B-2 binding and a gain in B-1 binding. When the
fractions B-1 and B-2 were digested with RNAse T1, about 40 nucleotides
were protected from digestion including the 28-nucleotide section. Cross-
linking using UV irradiation showed that proteins B-1 and B-2 were both
about 90 kDa. in molecular weight. Rouault et al., 26 using human H-mRNA,
also showed binding to human cell sap protein. In an extensive survey of the
binding proteins in 12 different species, Rothenberger et alY showed it to
be present in cytoplasmic extracts from human, mouse, chicken, frog, fish,
and fly, but was absent from bacterial, yeast, and plant sources. The proteins
found in animal sources varied from 95 to 100 kDa. Finally, like the rat/ 5
the mouse also displayed two binding proteins, thus confirming another de-
viation confined to these rodents, namely an insertion of eight amino acids
into the L subunit of the rae and the mouse 10 but not found in other species. 9
The next step was to purify the binding protein in order to characterize
and identify its structure. Rouault et al.2 6 •29 •30 describe a procedure in which
the regulatory RNA stem-loop structure (IRE) is used to trap the human
binding protein (IRE-BP). In order to do so, the segment of RNA containing
the IRE is reacted with biotin which makes the product capable of attaching
to avidin bound to an agarose column. After two cycles, an almost pure
product was obtained which migrated on an SDS/PAGE gel as a single band
with a molecular weight of 90 kDa. A peptide sequence from this purified
protein was used to prepare oligonucleotides to screen a human eDNA library.
A clone was identified which encoded a protein of 87 kDa. 29 We 31 also
employed biotinylation affinity chromatography to isolate the rat B-1 binding
protein, the yield being considerably increased by including six head-to-tail
IRE repeats. This protein had a molecular weight of 95 kDa and its N-terminal
amino acid sequence was close to that of the human. 29 •30 Walden et al. 28
employed a biochemical method using nucleotide binding dyes to isolate a
single polypeptide chain with a mol wt of about 90 kDa from rabbit liver.
Rouault et al. 29 noted in the amino acid sequence of their protein that it
included cysteine residues in motifs suggestive of those found in iron-sulfur
proteins. In particular, the mitochondrial enzyme aconitase shows 30% ho-
mology in its amino acid sequence compared with human binding protein. 32
This enzyme is active when all four Fe-S are present but can be inactivated
by loss of one iron atom. They therefore theorized that the gain or loss of
this iron atom may be the mechanism by which the IRE-BP senses iron levels
in the cell. The degree of homology of IRE-BP and aconitase is also em-
phasized by Hentze and Argos 34 who remark that another protein, isopro-
pylmalate isomerase, shows 23% homology with the binding protein se-
quence. It would be important to sequence the cytosolic form of aconitase
since this is the cell compartment in which the binding protein is located.
The activity of the binding protein may also vary with sulfhydryl levels in
the cytosoP 4 and indeed Barton et aJ.3 5 present a mechanism in response to
iron level involving multiple pools of IRE-BP, their characteristics, and ki-
netics including the important role of sulfhydryl groups (Figure 6).
As a step towards understanding how iron regulates the binding protein,
studies have focused on hemin and on chelatable cytosolic iron as likely
candidates. Figure 2 shows that chelatable free iron can be replenished from
Munro et al. 535
I
OXlDUD:I J'O»! REDUCED FORH I EI'WGEIIOUS CDI!nEl(
ln.cttve : low affinity I Active : high affinity I lllUtA bound
STUL£ I STAJLE I STA!LE
¥HS/
~Fe 1 HS'-R* lmRNA
~
F'\
0
mRN~
lll VIVO
i'Ri :
R*-
a m
RNA
s/ 'Fe :
-- ~ :~ -~- ~-:~-~- ~~~~--- --------------
IIVITIO
: HS/ 0 :
REDUCED FORM I
Active : •tah .tlinity I
I
UNSTABLE I
FIGURE 6. Scheme for interconversions of the pools of IRE-binding protein (IRE-BP) in

response to cytosolic levels of iron. When the iron content of the cytosol is high, IRE-BP occurs
as an inactive form (R) with low affinity for the IRE motif, thus allowing ferritin subunit
synthesis to proceed. The sulfhydryl groups of R, are in the oxidized state. In response to a
reduction in cytosolic iron level, the R, can be converted in vivo to high affinity active protein
(R:) with reduced sulfhydryl groups. Since this form binds tightly to the 28-nucleotide IRE, it
prevents translation of the ferritin mRNAs and is thus converted to the third pool, the endogenous
complex made up of R; and mRNA. Note that by adding thiol reductants the first pool (R,) can
also be converted in vitro to an active form (R.) which however is unstable. (Diagram kindly
provided by Dr. Hugh A. Barton.)
iron-loaded transferrin, from inorganic iron, from ferric pyridoxalisonicoti-

noyl hydrazone (FEPIH, not illustrated in the figure), and from hemin. Using
in vitro studies in which hemin was incubated with purified IRE-BP, Lin et
al. 36 were able to prevent the inhibitory action of binding protein on translation
of ferritin mRNA. This action of hemin was not affected by chelators of iron,
indicating that release of free iron was not involved. In a later publication,
they 37 were able to show direct binding of hemin to a specified site on the
binding protein surface. Haile eta!. 38 confirm that hemin added in vitro inhibits
the binding of the protein IRE-BP to the IRE motif in ferritin 5' UTR.
However, they show that hemin in vitro also binds nonspecifically to other
RNA and DNA binding proteins and to a variety of nucleases. Contrary to
the in vitro studies of Lin et al., 36 •37 Rogers and Munro 39 showed that, using
rat hepatoma cells in culture, addition of hemin to the medium increased
translation of ferritin mRNA, but that this action of hemin could be prevented
by adding the chelator desferal to the medium, thus suggesting that hemin
has to yield free iron in order to stimulate ferritin synthesis.
TABLE 1
Contrasting Effects of ALA on the Iron-
Dependent Induction of Ferritin and of Heme
Oxygenase Synthesis in Rat Fibroblasts
FEPIH ALA SA H-F HO

13 11M j.tM lrnM Synthesis cpm Synthesis cpm
140 170
+ 690 290
+ 100 270 560
+ 500 110 590
+ 1000 130 1010
+ + 520 460
+ 100 + 530 480
+ 500 + 400 480
+ 1000 + 480 450
Note: ALA, 5-amino levulinic acid; SA, the inhibitor succinyl

acetone; HF, H ferritin subunit; HO, heme oxygenase.
This thesis was explored in depth by Eisenstein et al. 40 who used rat
fibroblasts and rat hepatoma cells in culture to determine the effects of hemin,
other metalloporphyrins and nonheme iron (FEPIH) on the synthesis of ferritin
and on heme oxygenase, the latter being the enzyme that releases iron from
hemin. It was assumed that the level of activity of each enzyme reflects the
level of substrate, namely heme for heme oxygenase and chelatable iron for
ferritin. (This has been confirmed by later experiments [unpublished] in which
55
Fe was used to allow identification of the amounts of iron in different cell
components.) As shown in the published paper, 40 addition of hemin to the
medium induced both proteins, the oxygenase peaking first and releasing iron
for the subsequent increase in ferritin synthesis. Tin-mesoporphyrin, an in-
hibitor of heme breakdown by the oxygenase, delayed the induction of ferritin.
On the other hand, desferal, the chelator of free iron, inhibited ferritin syn-
thesis without affecting syntheses of heme oxygenase, thus suggesting that
free chelatable iron is the dominant factor in ferritin synthesis. This was
confirmed with free iron from added FEPIH, which selectively stimulated
ferritin synthesis but not heme oxygenase. Finally, addition to the cultured
cells of aminolevulinic acid (ALA), a regulator of the rate of heme synthesis,
increased the demands for iron by the heme synthesis pathway and thus
reduced the amount available for ferritin synthesis. These responses can be
prevented by addition to the medium of succinyl acetone (SA), an inhibitor
of ALA utilization (Table 1). Altogether these experiments show that ferritin
synthesis is regulated by chelatable free iron and not by hemin, making the
free cytosolic iron the preferred regulator of the IRE-BP.
Finally, optimum structure of the IRE for interaction with the IRE-BP
will be explored. Using the most recent Zuketl 1 computer-generated program
Munro eta!. 537
I
I
FIGURE 7. Suboptimal folding analysis of the nucleotides 26 through 68 of ferritin L-mRNA

using the Zucker program. 41 Single-strand cleavage sites (Tl and Sl) are denoted by the broken
arrows, double-strand cleavage sites (VI) by the solid arrows. The intensity of cleavage is
indicated by the length of the arrow. (From Bettany, A. J. E. eta!., in press. With permission.)
for predicting suboptional folding of RNA (Figure 7), we 42 have analyzed the
stem-loop structure of the IRE of the rat L subunit and have compared it with
the cleavage sites for single-stranded areas of the IRE using enzymes T-1 and
S-1, represented in the Figure 7 by dotted line arrows on the six-membered
loop and also the major bulge of three unpaired nucleotides (38 to 41) on one
side of the stem. Similarly, scission of double-stranded areas was accom-
plished by using enzyme V-1 and is represented by solid arrows. Wang et
al. 43 explored the structure of the bullfrog H subunit using the chemical
nuclease l, 10-phenanthroline-copper which cleaves single-stranded and bulged
areas. This supplemented the evidence from their application of T-1 while
V-1 was applied for identification of the double-stranded areas. In both of
these studies the predicted stem-loop structure has been confirmed.
In addition to studies of secondary structure described above, there have
been several reports of nucleotide replacement or deletion and the effect of
this manoever on IRE function. One example comes from a comparison of
the IRE of the Land the H subunits of mammals. As shown earlier, 14 all H
subunit IREs of mammals so far examined have a cytidylic acid replacing
the uridy lie acid of the L subunit IREs in two places (one in the sixth position
of the CAGUGU loop) without apparent Joss of function. Barton et al. 35
replaced two nucleotides in the loop without greatly altering the affinity of
the IRE to bind the specific protein as determined by in vitro competition
and saturation assays. Similarly, exchange of adenylic acid for cytidylic acid
in the unpaired bulge also left protein binding unaffected. In contrast, Hentze
et al. 45 deleted a cytidylic acid from the loop and from the stem of the human
H-IRE. The result with each deletion was Joss of response to iron, which
may be attributed to the alteration in secondary structure of the IRE rather
than to Joss of a specific nucleotide. An even more extensive loss of binding
occurred when one leg of the stem of rat L-mRNA was replaced by other
nucleotides that did not pair with the other leg of the stem, but was restored
by inserting a new leg of nucleotides compatible with the other new leg. 42
Thus one stem can be replaced with another of a different nucleotide com-
position and can function so long as the nucleotides in both legs are able to
pair. This leads to the conclusion that the secondary conformation of the IRE
stem is more important than the primary sequence of its nucleotides so long
as they form a structure of paired nucleotides of adequate length. 35 This is
also the conclusion drawn by Leibold et a!. 46 who applied competition assays
to obtain the 50% binding affinity between ferritin native IRE and mutant
IRE. These conclusions require testing in vivo.
2. Other Systems Involving the Iron-Responsive Element

The IRE motif performs a different function in regulating the stability of
the transferrin receptor (TfR) in response to changes in the intracellular iron
level. 47.48 In the 3' UTR of the receptor, five stem-loop structures resembling
the single IRE in the 5' UTR of ferritin mRNA have been identified. 49 •50
Three of these provide the function of regulating TfR mRNA breakdown in
response to changes in intracellular iron levels. 50 When cells are treated with
the iron-chelator desferal, the TfR mRNA becomes more stable and thus
favors the entry of more iron into the cell, whereas when iron uptake into
the cell increases, the TfR mRNA becomes unstable and diminishes in amount
so that entry of more iron is discouraged. These changes in stability of the
mRNA of TfR are due to binding of a specific protein which recognizes the
3' IREs and either increases (stabilizes) or relaxes (destabilizes) its binding
according to the iron level. Both the IREs 49 and the protein51 associated with
the 3' UTR are able to function if transferred to the 5' UTR where they then
regulate ferritin subunit translation in spite of having quite different nucleotide
pairs in their stems when compared with the stem of the 5' IRE regulating
ferritin subunit synthesis. Note that the same binding protein coordinates both
the synthesis of ferritin and the amount of transferrin receptor in relation to
changes in intracellular iron status.
An iron-regulated protein restricted to erythroid cells is the erythroid 5-
aminolevulinic acid synthase (eALAS), the rate-limiting enzyme in proto-
porphyrin synthesis. In addition to this erythroid-specific form, there is also
a ubiquitous ALAS which is activated by porphyrinogenic drugs and is reg-
ulated by hemin. The mRNA of the erythroid ALAS displays an IRE motif
in its 5' UTR, 52 •53 whereas the housekeeping form of this enzyme does not
have this motif. In order to test the importance of individual nucleotides in
the loop CAGUGC in the eALAS mRNA, Cox et a!. 52 replaced it with
GGAUCC and observed that the mutant IRE no longer competed with H
ferritin IRE. However, positions 1 and 6 of their mutant loop were occupied
by nucleotides G and C opposing one another at the top of the stem to which
the G-C pair created by the mutation contributed another nucleotide pair while
reducing the loop to four nucleotides, a possible cause of the loss of function
of the mutant. In the course of their exploration of the IRE in eALAS,
Munro et al. 539
Dandekar et al. 53 performed a computer-assisted identification of putative

IREs which turned up in large numbers from which these investigators offer
potentially interesting examples. One such is aconitase, a member of the Fe-
S family thought to include the IRE-binding protein. 32 Another example re-
corded by Tanzi and Hyman 54 identifies a stem-loop structure in the gene for
the amyloid protein precursor in Alzheimer's disease. This motif occurs at
the point where the much smaller amyloid protein is released from the pre-
cursor and thus may regulate either the amount of the precursor protein or
the amount of amyloid released.
B. TRANSCRIPTION
I. Response of H and L Genes to Iron
The regulatory stem-loop motif in the 5' UTR of both the L- and H-
mRNAs of ferritin implies that activation of translation of these two subunits
will create new ferritin molecules with H and L subunits in proportion to the
dormant H-and L-mRNAs. However, in an experiment in which new ferritin
subunits in rat liver were labeled with 14C-leucine, we 9 •55 obtained a ratio of
new L to H subunits which increased substantially after an intraperitoneal
dose of iron dextran, implying either preferential synthesis of L-mRNA or
decreased rate of breakdown of L-mRNA. Using HeLa cells in culture, Cairo
et al. 56 found that transcription of L-mRNA increased after adding more iron
to the medium, but for technical reasons were unable to monitor the response
of H-mRNA. Using transcription run-off assays for measuring Land H mRNA
synthesis, we 57 were able to show a preferential increase in the rate of tran-
scription of L compared with H-mRNA by rat liver nuclei. The pattern of
increased abundance of labeled L-mRNA in the nuclei then appeared in the
ribonucleoprotein pool (Figure 2) and later in the polyribosomal fraction of
the cell. This resulted in production of isoferritins richer in the L subunit,
and therefore better able to store iron. 55 The different effects of iron on
transcription and on translation thus allow flexibility in the creation of ferritin
spheres with various LIH ratios.
2. Response of H and L Genes to Other Agents

Factors other than iron can alter the expression of ferritin genes, usually
differentially favoring the H subunit. In 1977, Konijn and Hershko 58 produced
an acute phase reaction by injecting turpentine into rats. In addition to es-
tablished acute phase liver proteins, they observed an increase in liver ferritin
5 to 10 h after injection. Campbell et al. 59 attributed this to a translational
response to inflammation. Subsequent work has identified interleukin-1 (IL-
l) as the agent released by macrophages in the injured tissues that causes the
acute phase response. 60 Rogers et al. 61 added IL-l to human hepatoma cells
in culture and within 2 h obtained increases in the amounts of both H and L
ferritin subunits without changing the amount of L-mRNA (H-mRNA was
not reported). Consequently, the response to IL-l must have involved the
transfer of dormant mRNP to active polyribosomes as discussed earlier under

translation. Cell iron participated in the response to IL-l only when it was
reduced by desferal to a level that prevented translocation of the mRNA to
an active status. In studies on muscle cells in culture, Wei et al. 62 compared
the action of IL-l and tumor necrosis factor (TNF) on the abundance of H-
mRNA which increased in amount following addition of each cytokine. Since
IL-l and TNF bind to different cell receptors, it is not surprising that their
actions were additive. In a subsequent paper, 63 the same group found that
actinomycin inhibited the action of TNF on the abundance of H-mRNA,
implying an effect of TNF on transcription. In addition they 64 also analyzed
the regulation of transferrin receptors in response to TNF and to IL-l. Using
human fibroblasts exposed to TNF and to IL-l, they observed increases in
TfR mRNA and protein after a lag of 12 to 24 h. In the same experiment,
however, ferritin H-mRNA increased within 4 hand the H-protein peaked at
8 h, after which synthesis of both H and L proteins declined rapidly. Con-
sequently, the late increase in TfR levels cannot be credited with providing
more iron for the cell within the first 4 to 8 h. It can be concluded that TNF
and IL-l increase H subunit levels through transcription provided that ferritin
mRNA translation is not obstructed by too low an intracellular iron level.
Finally, changes in H and L subunit abundance have been linked to cell
differentiation. Chou et al. 65 used a human promyelocyte cell line that dif-
ferentiated in vitro when incubated with TPA to develop into macrophages
or DMSO with amiloride to yield neutrophils. In each case there was an
increase in both H- and L-mRNA, the increase in the H-mRNA being sev-
eralfold greater. Testa et a!. 66 also described a cell proliferation model of
ferritin involving human lymphocytes stimulated by phytohemagglutinin (PHA)
and also a cell differentiation model of monocytes maturing to macrophages.
In each case, they evaluated levels of TfR expression, ferritin (protein and
mRNA), and IRE-BP activity. It is of interest that the cell proliferation model
(lymphocytes) in an iron-supplemented medium responded to PHA with an
increase limited to H-mRNA, whereas the differentiation model (monocytes)
showed similar large changes in H and L ferritin mRNA when iron was added
to the medium. Liau et al. 67 used the vascular smooth vessel cell to show
that growth-arrest results in increases in the ferritin H chain which cannot be
attributed to a rise in the intracellular iron level. They present evidence
suggesting cAMP as a likely factor acting at the posttranscriptional level to
increase the H chain selectively.
IV. CONCLUSION
The translational control of ferritin subunit synthesis and the stability of

transferrin receptors provide a well-authenticated linking of factors regulating
intracellular iron content. This relationship emphasizes ferritin shells with L
subunit abundance as favoring iron storage which can be provided by in-
Munro eta/. 541
creasing L-mRNA transcription preferentially. However several publications

have shown that ferritin can preferentially increase mRNA after inflammation.
Two cautions regarding such work must be voiced. First, if there is insufficient
free iron in the target cell, transcribed H- or L-mRNA will not undergo
translation unless there is iron supply that allows the IRE-BP to dissociate
from the IRE, a limitation recognized by one group. 62 Second, it is not often
remembered that mRNA and protein can also be regulated by varying the rate
of their breakdown, which therefore should be measured. Third, a number
of authors has examined the importance of specific nucleotides in the con-
served 28 nucleotides of the IRE, using gel retardation, etc., as the criteria
in vitro. Based on such tests it has been concluded that nucleotides are mostly
replaceable in the stem and perhaps some in the loop of the IRE. It has
generally been found that this is true for the stem, but the six-membered loop
should be tested by in vivo criteria. The perfect retention of the sequence
CACUGN, stretching from snaiP 2 to man, 9 is impressive, whereas the stems
of snail and of man are only 40% concordant. This makes imperative in vivo
studies of loop structure using nucleotide replacement.
Finally, Dickey et al. 11 have recognized the cellular function may deter-
mine the proportion of H and L subunits in the ferritin shell depending on
whether the cells require a housekeeping ferritin or a cell-specific ferritin.
This projects us into the realm of different H and L subunit abundances in
different tissues. For example heart ferritin has a larger size of ferritin shell
in which the H subunit predominates. 5 5 Ex Afrika semper aliquid novi (Pliny
the Elder).
REFERENCES
I. Wixom, R. L., Prutkin, L., and Munro, H. N., Hemosiderin: nature, formation and
significance, in International Review Experimental Pathology. Richter, G. W. and Ep-
stein, M.A., Eds., Academic Press, New York, 1980, 22, 193.
2. Lautberger, M. V., Sur Ia cristallisation de Ia ferritine, Bull. Soc. Chim. Bioi., 19,
1575, 1937.
3. Harrison, P. M., Andrews, S. C., Artymiuk, P. J., Ford, G. C., Lawson, D. M.,
Smith, J. M.A., Treffry, A., and White, J. L., Ferritin, in Iron Transport and Storage,
Ponka, P., Schulman, H. M., and Woodworth, R. C., Eds., CRC Press, Boca Raton,
FL, 1990, 81.
4. Linder, M. C., Moor, J. 8., and Munro, H. N., Subunit heterogeneity in rat liver
apoferritin: differential responses of the subunits to iron administration, J. Bioi. Chem.,
249, 7707, 1974.
5. Adelman, T. G., Arosio, P., and Drysdale, J. W., Multiple subunits in human ferritin:
evidence for hybrid molecules, Biochem. Biophys. Res. Commun., 63, 1056, 1975.
6. Drysdale, J. W. and Munro, H. M., Regulation of synthesis and turnover of ferritin
in rat liver, J. Bioi. Chem., 241, 3630, 1966.
7. Zahringer, J., Baliga, B. S., and Munro, H. N., Novel mechanism for translational
control in regulation of ferritin synthesis by iron, Proc. Nat/. Acad. Sci. U.S.A., 73,
857, 1976.
8. Munro, H. N. and Linder, M. C., Ferritin: structure, biosynthesis and role in iron
metabolism, Physiol. Rev., 58, 317, 1978.
9. Munro, H. N., Leibold, E. A., Aziz, N., Murray, M. T., White, K., and Rogers,
J., Ferritin gene structure and expression, in Iron Transport and Storage, Ponka, P.,
Schulman, H. M., and Woodworth, R. C., Eds., CRC Press, Boca Raton, FL, 1990,
134.
10. Beaumont, C., Dugast, I., Renaudie, F., Souroujun, M., and Grandchamp, B.,
Transcriptional regulation of ferritin H and L subunits in adult erythroid and liver cells
from the mouse, J. Bioi. Chern., 264, 7498, 1989.
II. Dickey, L. F., Sreedharan, S., Theil, E. C., Didsbury, J. R., Wang, Y-H., and
Kaufman, R. E., Differences in the regulation of messenger RNA for housekeeping and
specialized cell ferritin: a comparison of three different ferritin complementary DNAs,
the corresponding subunits and identification of the first processed pseudogenes in am-
phibia, J. Bioi. Chern., 262, 7901, 1987.
12. Von Dar!, M. and Bottke, W., Vollstandige Basensequenz des langsten Somaferritin-
clons: Aminosauresequenz des Somaferritins. EMBL-accession Nr X 56778, 1991.
13. Chou, P. Y. and Fasman, G. D., Empirical predictions of protein conformation, Annu.
Rev. Biochem., 47, 251, 1978.
14. Eisenstein, R. S., Bettany, A. J. E., and Munro, H. N., Regulation of ferritin gene
expression, in Advances in Inorganic Biochemistry, Eichhorn, G. L. and Marzilli, L. G.,
Eds., Elsevier, New York, 1990, 8, 91.
IS. Leibold, E. A. and Munro, H. N., Characterization and evolution of the expressed rat
ferritin light subunit gene and its pseudogene family, J. Bioi. Chern., 262, 7335, 1987.
16. Murray, M. T., White, K., and Munro, H. N., Conservation of ferritin heavy subunit
gene structure: implications for the regulation of ferritin gene expression, Proc. Nat/.
Acad. Sci. U.S.A., 84, 7438, 1987.
17. Aziz, N. and Munro, H. N., Both subunits of rat liver ferritin are regulated at a
translational level by iron induction, Nucleic Acids Res., 14, 915, 1986.
18. Dougherty, J. J., Croft, W. A., and Hoekstra, W. G., Effects of ferrous chloride and
iron dextran on lipid peroxidation in vivo in vitamin E and selenium adequate and deficient
rats, J. Nutr., Ill, 1784, 1981.
19. Floyd, R. A., Role of oxygen free radicals in carcinogenesis and brain ischemia, FASEB
J., 4, 258, 1990.
20. Munro, H. N., Leibold, E. A., Vass, J. K., Aziz, N., Rogers, J., Murray, M., and
White, K., Ferritin gene structure and expression, in Proteins of Iron Storage and
Transport, Spik, G., Montreuil, J., Crichton, R. R., and Mazurier, J., Eds., Elsevier,
Amsterdam, 1985, 331.
21. Aziz, N. and Munro, H. N., Iron regulates ferritin mRNA translation through a segment
of its 5' untranslated region, Proc. Nat/. Acad. Sci. U.S.A., 84, 8478, 1987.
22. Dickey, L. F., Wang, Y-H, Schull, G. E., Wortman, I. A., and Theil, E. C., The
importance of the 3' untranslated region in the translational control of ferritin mRNA,
J. Bioi. Chern., 263, 3071, 1988.
23. Caughman, S. W., Hentze, M. W., Rouault, T. A., Harford, J. B., and Klausner,
R. D., The iron-responsive element is the single element responsible for iron-dependent
translational regulation of ferritin biosynthesis: evidence for function as the binding site
for a translational repressor, J. Bioi. Chern., 263, 19048, 1988.
24. Goessen, B., Caughman, S. W., Harford, J. B., Klausner, R. D., and Hentze, M. W.,
Translational repression by a complex between the iron-responsive element of ferritin
mRNA and its specific cytoplasmic binding protein is position-dependent in vivo, EMBO
J., 9, 4127, 1990.
Munro et al. 543
25. Leibold, E. A. and Munro, H. N., Cytoplasmic protein binds in vitro to a highly
conserved sequence of the 5' untranslated region of ferritin heavy and light subunit
mRNAs, Proc. Nat/. Acad. Sci. U.S.A., 85, 2171, 1988.
26. Rouault, T. A., Hentze, M. W., Caughman, S. W., Harford, J, B., and Klausner,
R. D., Binding of a cytosolic protein to the iron responsive element of human ferritin
messenger RNA, Science, 241, 1207, 1988.
27. Rothenberger, S., Mollner, E. W., and Kuhn, L. C., The mRNA-binding protein
which controls ferritin and transferrin receptor expression is conserved during evolution,
Nucleic Acids Res., 18, 1175, 1990.
28. Walden, W. E., Patino, M. M., and Gaffield, L., Purification of a specific repressor
of ferritin mRNA translation from rabbit liver, J. Bioi. Chern., 264, 13765, 1989.
29. Rouault, T. A., Hentze, M. W., Haile, D. J., Harford, J, B., and Klausner, R. D.,
The iron-responsive element binding protein: a method for the affinity purification of a
regulatory RNA-binding protein, Proc. Nat/. Acad. Sci. U.S.A., 86, 5768, 1989.
30. Rouault, T. A., Tang, C. K., Kaptain, S., Burgess, W. H., Haile, D. J., Samaniego,
F., McBride, 0. W., Harford, J, B., and Klausner, R. D., Cloning of the eDNA
encoding an RNA-regulatory protein: the human iron-response element binding protein,
31. Bomford, A. B., Barton, H. A., Pettingell, W. H., Munro, H. N., and Eisenstein,
R. S., The iron-responsive element-binding protein of the rat: biochemical and functional
studies of the purified protein. (Submitted).
32. Rouault, T. A., Stout, C. D., Kaptain, S., Harford, J, B., and Klansman, R. D.,
Structural relationship between an iron-regulated RNA-binding protein (IRE-BP) and
aconitase: functional implications, Cell, 64, 881, 1991.
33. Kaptain, S., Downey, W. E., Tang, C., Philpott, C., Haile, D., Orloff, D. G.,
Harford, J, B., Rouault, T. A., and Klausner, R. D., A regulated RNA binding protein
also possesses aconitase activity, Proc. Nat/. Acad. Sci. U.S.A., 88, 10109, 1991.
34. Hentze, M. W., Rouault, T. A., Harford, J, B., and Klausner, R. D., Oxidation-
reduction and the molecular mechanism of a regulatory RNA-protein interaction, Science,
244, 357, 1989.
35. Barton, H. A., Eisenstein, R. S., Bomford, A., and Munro, H. N., Determinants of
the interaction between the iron-responsive element-binding protein and its binding site
in rat L-ferritin mRNA, J. Bioi. Chern., 265, 7000, 1990.
36. Lin, J. J., Daniels-McQueen, S., Patino, M. M., Gaffield, L., Walden, W. E., and
Thach, R. E., Derepression of ferritin messenger RNA translation by hemin in vitro,
Science, 247, 74, 1990.
37. Lin, J. J., Patino, M. M., Garfield, L., Walden, W. E., Smith, A., and Thach,
R. E., Cross-linking of hemin to a specific site on the 90-kd repressor protein, Proc.
Nat/. Acad. Sci. U.S.A., 88, 6068, 1991.
38. Haile, D. J., Rouault, T. A., Harford, J. B., and Klausner, R. D., The inhibition of
the iron responsive element RNA-protein interaction by heme does not mimic in vivo
iron regulation, J. Bioi. Chern., 265, 12786, 1990.
39. Rogers, J, and Munro, H. N., Translation of light and heavy subunit mRNAs is regulated
by intracellular chelatable iron levels in the rat hepatoma cells, Proc. Nat/. Acad. Sci.
U.S.A., 84, 2277, 1987.
40. Eisenstein, R. S., Garcia-Mayol, D., Pettingell, W., and Munro, H. N., Regulation
of ferritin and heme oxygenase synthesis in rat fibroblasts by different forms of iron,
Proc. Nat/. Acad. Sci. U.S.A., 88, 688. 1991.
41. Jaeger, J, A., Turner, D. H. and Zuker, M., Predicting optimal and suboptimal
secondary structure for RNA, Methods Enzyrnol., 183, 281, 1990.
42. Bettany, A. J. E., Eisenstein, R. S., and Munro, H. N., Role of RNA secondary
structure in the interaction between the iron regulatory element of ferritin mRNA and its
binding protein, in press.
43. Wang, Y-H, Sczekan, S. R., and Theil, E. C., Structure of the 5' untranslated regulatory
region of ferritin mRNA studied in solution, Nucleic Acids Res., 18, 4463, 1990.
44. Harrell, C. M., McKenzie, A. R., Patino, M. M., Walden, W. E., and Theil, E. C.,
Ferritin mRNA: interactions of iron regulatory element with translational regulator protein
P-90 and the effect on base-paired flanking regions, Proc. Nat/. Acad. Sci. U.S.A., 88,
4166. 1991.
45. Hentze, M. W., Caughman, S. W., Casey, J. L., Koeller, D. M., Rouault, T. A.,
Harford, J. B., and Klausner, R. D., A model for the structure and functions of iron-
responsive elements, Gene, 72, 201. 1988.
46. Leibold, E. A., Laudano, A., and Yang, Y., Structural requirements of iron responsive
elements for binding of the protein involved in both transferrin receptor and ferritin mRNA
post-transcriptional regulation, Nucleic Acids Res., 18, 1819, 1990.
47. Mullner, E. W. and Kuhn, L. C., A stem-loop in the 3' untrans1ated region mediates
iron-dependent regulation of transferrin receptor mRNA stability in the cytoplasm, Cell,
53, 815, 1988.
48. Mullner, E. W., Neupert, B., and Kuhn, L. C., A specific mRNA binding factor
regulates the iron-dependent stability of cytoplasmic transferrin receptor mRNA, Cell,
58, 373, 1989.
49. Casey, J. L., Hentze, M. W., Koeller, D. M., Caughman, S. W., Rouault, T. A.,
Klausner, R. D., and Harford, J. B., Iron-responsive elements: regulatory RNA se-
quences that control mRNA levels and translation, Science, 240, 924, 1988.
50. Casey, J. L., Koeller, D. M., Ramin, V. C., Klausner, R. D., and Harford, J. B.,
Iron regulation of transferrin receptor mRNA levels requires iron-responsive elements
and a rapid turnover determinant in the 3' untranslated region of the mRNA, EMBO J.,
8, 3693, 1989.
51. Koeller, D. M., Casey, J. L., Hentze, M. W., Gerhardt, E. M., Chan, L-N. L.,
Klausner, R. D., and Harford, J. B., A cytosolic protein binds to structural elements
within the iron regulatory region of the transferrin receptor mRNA, Proc. Nat/. Acad.
Sci. U.S.A., 86, 3574, 1989.
52. Cox, T. C., Bawden, M. J., Martin, A., and May, B. K., Human erythroid 5-
aminolevulinate synthase: promoter analysis and identification of an iron-responsive ele-
ment in the mRNA, EMBO J., 10, 1891, 1991.
53. Dandekar, T., Stripecke, R., Gray, N. K., Goosen, B., Constable, A., Johansson,
H., and Hentze, M., Identification of a novel iron-responsive element in murine and
human 5-erythroid amino levulinic acid synthase mRNA, EMBO J., 10, 1903, 1991.
54. Tanzi, R. E. and Hyman, B. T., Alzheimer's mutation, Nature (London), 350, 564,
1991.
55. Bomford, A., Conlon-Hollingshead, C., and Munro, H., Adaptive responses of rat
tissue isoferritins to iron administration, J. Bioi. Chern., 256, 948, 1981.
56. Cairo, G., Bardella, L., Schiaffonati, L., Arosio, P., Levi, S., and Barnelli-Zazzera,
A., Multiple mechanisms of iron-induced ferritin synthesis in HeLa cells, Biochern.
Biophys. Res. Cornrnun., 133, 314, 1985.
57. White, K. and Munro, H. N., Induction of ferritin subunit synthesis by iron is regulated
at both the transcriptional and translational levels, J. Bioi. Chern., 263, 8938, 1988.
58. Konijn, A.M. and Hershko, C., Ferritin synthesis in inflammation. I. Pathogenesis of
impaired iron release, Br. J. Haernatol., 37, 7, 1977.
59. Campbell, C. H., Solgonick, R., and Linder, M. C., Translational regulation of ferritin
synthesis in rat spleen: effects of iron and inflammation, Biochern. Biophys. Res. Corn-
rnun., 160, 453, 1989.
60. Dinarello, C. A., Biology of interleukin-1, FASEB J., 2, 108, 1988.
61. Rogers, J. T., Bridges, K. R., Durmowicz, G. P., Glass, J., Auron, P. E., and
Munro, H. N., Translational control during the acute phase response; ferritin synthesis
in response to interleukin-1, J. Bioi. Chern., 265, 14572, 1990.
Munro et al. 545
62. Wei, Y., Miller, S. C., Tsuji, Y., Torti, S. V., and Torti, F. M., Interleukin 1 induces
ferritin heavy chain in human muscle cells, Biochern. Biophys. Res. Cornrnun .. 169, 289.
1990.
63. Miller, L. L., Miller, S.C., Torti, S. V., Tsuji, Y., and Torti, F. M., Iron-independent
induction of ferritin H chain by tumor necrosis factor, Proc. Nat/. Acad. Sci. U.S.A..
88. 4946. 1991.
64. Tsuji, Y., Miller, L. L., Miller, S.C., Torti, S. V., and Torti, F. M., Tumor necrosis
factor-alpha and interleukin !-alpha regulate transferrin receptor in human diploid fibro-
blasts; relationship to the induction of ferritin heavy chain, J. Bioi. Chern .. 266, 7257.
1991.
65. Chou, C-C., Gatti, R. A., Fuller, M. L., Concannon, P., Wong, A., Chada, S.,
Davis, R. C., and Salser, W. A., Structure and expression of ferritin genes in a human
promyelocyte cell line that differentiates in vitro. Mol. Cell. Bioi.. 6. 566, 1986.
66. Testa, U., Kuehn, L., Petrini, M., Quaranta, M. T., Pelosi, E., and Peschle, C.,
Differential regulation of iron regulatory element-binding protein(s) in cell extracts of
activated lymphocytes versus monocytes-macrophages, J. Bioi. Chern., 266, 13925, 1991.
67. Liau, G., Chan, L. M., and Feng, P., Increased ferritin gene expression is both promoted
by cAMP and a marker of growth arrest in rabbit vascular smooth muscle cells, J. Bioi.
Chern., 266, 18819, 1991.
Chapter 24
VITAMIN 8 6 MODULATION OF STEROID-INDUCED

GENE EXPRESSION
Douglas B. Tully, Victoria E. Allgood, and John A. Cidlowski
TABLE OF CONTENTS
I. Introduction ...................................................... 548
II. The Steroid Hormone Receptors and Their Mechanism

of Action ........................................................ 549
III. Classical Actions of Vitamin B6 ••.•••••••.••••••••••.••••••••••• 551
IV. Effects of the Vitamin on Physical Properties of Steroid

Receptors ........................................................ 552
V. Evidence for a Direct Interaction between Pyridoxal

Phosphate and the Glucocorticoid Receptor ...................... 555
VI. In Vitro Evidence for Vitamin B6 Effects on Steroid

Hormone Receptor-Mediated Gene Expression ................... 558
VII. Summary ........................................................ 561
References ............................................................... 563
0-8493-6961-4/93/$0.00 + $.50
I. INTRODUCTION
The steroid hormones, glucocorticoid, estrogen, progesterone, androgen,

and aldosterone, regulate many physiological processes in various tissues and
cell types throughout the body. These hormones are intimately involved in
numerous fundamental biological processes including growth, development,
reproduction, regulation of metabolic state, and maintenance of homeo-
stasis. 1"3 Steroid hormones elicit their biological effects at the single cell level
by regulating the expression of distinct target genes. Each type of steroid
hormone binds specifically to a unique intracellular receptor protein, and this
binding initiates transformation of the receptor into a form which is capable
of selective interaction with DNA regulatory sequences which are typically
contained within or near the promoter regions of hormone-responsive genes.
Interaction of the steroid-receptor-complex with these DNA regulatory sites
called hormone response elements results in modulation of target gene expres-
sion, and ultimately leads to alterations in the protein composition and phe-
notype of steroid-responsive cells.
Because steroid hormone receptors are critically involved in so many
physiological and biological processes, there has been considerable interest
in determining both the mechanisms through which steroid hormone action
is mediated and related regulatory mechanisms which serve to control the
degree of cellular responsiveness to steroid hormones. Numerous factors have
been shown to influence the capacity of cells to respond to steroid hormone
stimulation including the number of intracellular receptors, 4 •5 the state of
cellular differentiation, 6 - 9 the stage of the cell cycle, 10- 12 as well as nutritional
components such as butyric acid, 8 •9 and vitamin 8 6 • 13 - 15 Such interactions
between nutritional factors, such as vitamin B6 , and the endocrine system
will serve as the focus for this chapter. For example, early evidence from
whole animal studies suggests that glucocorticoid induction of tyrosine ami-
notransferase is decreased in response to the administration of pharmacological
doses of vitamin B6 and elevated under conditions of vitamin deficiency. 16
However, these studies measured enzymatic activity, and since tyrosine ami-
notransferase is itself a pyridoxal phosphate dependent enzyme, 17 these results
cannot distinguish between an effect of the vitamin on glucocorticoid receptor
function or an effect of the vitamin on tyrosine aminotransferase enzyme
activity. More recent evidence demonstrates that vitamin B6 alters the steroid
responsiveness of certain cell types by modulating the ability of steroid hor-
mone receptors to activate target gene expression. 18 These recent observations
coupled with in vitro studies showing that pyridoxal phosphate affects various
biophysical properties of steroid receptors serve as the basis for the idea that
intracellular vitamin B6 is a physiological regulator of steroid hormone action.
In this chapter, we will review the data on which this hypothesis is founded
and discuss possible mechanisms through which this essential vitamin may
influence steroid hormone action.
Tully eta!. 549
II. THE STEROID HORMONE RECEPTORS AND

THEIR MECHANISM OF ACTION
The steroid hormone receptors belong to a superfamily of ligand-activated

transcription factors which exert their biological effects through regulation of
target gene expression. 19 ·20 Despite the diverse array of physiological events
mediated by different steroid hormones, receptors for the glucocorticoid,
estrogen, progesterone, and androgen hormones exhibit many structural and
functional similarities. 3 In early studies which employed limited proteolytic
digestion of steroid receptor preparations, it was demonstrated that the steroid
hormone receptors all share a common domain structure. 21 -23 Within the past
6 years, cDNAs encoding the glucocorticoid, estrogen, progesterone, and
androgen hormone receptors from many different species have been cloned.
DNA sequence analysis has demonstrated that considerable DNA sequence
homology exists among the different members of this family of ligand-de-
pendent transcription factors. In addition, DNA and amino acid sequence
analyses have confirmed earlier suggestions that the steroid receptors share
a common modular organization. 3·24-27 Structure-function studies and muta-
tional analyses have subsequently verified that all of the steroid hormone
receptors are composed of three specific domains. The NH 2 -terminal region
of the receptor molecules is the most variable in size and character. This
domain of the receptor proteins contains antigenic determinants which permit
antibody discrimination among the different classes of steroid hormone re-
ceptor,28-30 and, not surprisingly, displays the lowest degree of DNA and
amino acid sequence homology. 3 Despite this variability, however, the NH 2-
terminal domain of the receptors appears to be required for maximal receptor
activity and may mediate some aspect of the modulatory effects receptors
have on transcriptional activation of target genes. 31 ·32 A region of somewhat
greater sequence homology exists within the COOH-terminal portion of the
receptor molecules. This region of the receptors has been shown to be ex-
tremely important for steroid hormone action, because it is this region of the
proteins which is responsible for high affinity, specific ligand binding. 3·33
Within this ligand binding domain there exist two short segments of 42 and
22 amino acids which are hydrophilic in nature and conserved to some extent
among all members of the receptor superfamily. The fact that these sequence
segments exhibit a relatively high degree of sequence homology among the
different steroid receptors suggests that they are of some functional signifi-
cance, and it has been proposed that these regions may be involved directly
in ligand binding, dimerization of receptor molecules, or interaction of re-
ceptor with other proteins. 33 ·34 However, establishing a definitive role for
these short sequences within the ligand binding domains has proven to be
difficult, since deletions, insertions, or mutations within the steroid binding
domain have generally interfered with steroid binding activity. The region of
greatest sequence homology, however, exists between the variable NH 2-ter-
minal and the ligand binding domains of the receptors. 3 •35 This highly con-
served region consists of 66 amino acids which permit the receptor molecules
to recognize and bind to specific DNA regulatory sequences, the steroid
response elements, located within or near to the promoter regions of steroid
responsive genes. Contained within this DNA binding domain are nine highly
conserved cysteine residues, eight of which are thought to form two zinc
fingers, with each finger coordinately binding one zinc atom. 36 It is believed
that one zinc finger determines target gene specificity by distinguishing be-
tween different steroid response elements, while the other zinc finger contacts
the sugar-phosphate backbone of the DNA helix to stabilize binding of the
receptor to the steroid response element. 37 -40
It is remarkable that despite such a high degree of conservation of amino
acid sequence and protein structure, different steroid receptors remain capable
of distinguishing between different hormonally regulated genes, activating
transcription only of the appropriate target genes in the appropriate cell or
tissue type. It is possible that other regions of the receptors may participate
in determining target gene specificity. Alternatively, cell- or tissue-specific
factors may participate in regulating receptor function, perhaps through direct
interaction with the receptor molecules. Although only a speculative notion,
the NH 2 -terminal region, which exhibits the most variability among different
receptors, may be responsible for promoting such receptor-specific functions.
This could explain variations in the cell- or tissue-specific expression of many
hormonally regulated genes. For example, expression of the tyrosine ami-
notransferase gene is selectively induced in parenchymal cells of liver, but
not in other tissues, following glucocorticoid treatment, 41 One could speculate
that a liver-specific factor may interact with the NH 2 -terminal portion of the
glucocorticoid receptor to allow glucocorticoid-specific induction of the ty-
rosine aminotransferase gene in the liver, while absence of an analogous
factor in other cell types would inhibit glucocorticoid induction of tyrosine
aminotransferase gene expression in nontarget tissues. Alternative explana-
tions may also apply, however, and in the case of the tyrosine aminotransferase
gene it has been reported that tissue specific methylation of CpG dinucleotides
in certain regions of DNA may inhibit expression of tyrosine aminotransferase
gene in nonexpressing cells. 42 Theories such as these are currently being
investigated in an effort to determine how specific gene expression is mediated
in the presence of similar steroid receptors.
Nonetheless, given the extensive sequence and structural homology that
exists among the different steroid hormone receptors, it is not surprising to
learn that the receptors all act through essentially the same pathway to exert
their diverse biological functions. Each of the steroid hormone receptors is
believed to occur in an oligomeric complex containing two molecules of heat
shock protein 90 and perhaps a smaller heat shock protein or other protein
in addition to receptor. 43 •44 While the receptor is associated with this oligo-
meric complex, it is maintained in an unactivated state, which is operationally
Tully eta!. 551
defined as being a non-DNA-binding state. When the receptor protein binds

its specific steroid ligand, the receptor undergoes a poorly understood process
termed transformation or activation. 45 -47 Although the precise nature of the
changes that occur in steroid receptor proteins following ligand binding and
activation have not been fully elucidated, it is likely that transformation
involves dissociation of the receptor from its oligomeric complex with heat
shock proteins and may also involve a conformational change in the receptor
protein itself. The primary point that is generally accepted is that the activated,
liganded receptor has acquired high affinity for the specific DNA regulatory
sites that typically reside within or near the promoter regions of hormone
responsive genes. Binding of activated steroid receptors to these DNA reg-
ulatory sites results in alteration of the level of expression of specific target
genes. 48 - 51 In most cases, expression of target genes is specifically induced
in response to hormone treatment, although it is clear that expression of some
genes is physiologically suppressed following hormone exposure. 52 •53 It is
through this regulation of gene expression that steroid hormones exert their
primary biological effects.
Efforts directed toward understanding the regulation of steroid hormone
action have revealed that several additional mechanisms contribute to the
overall degree of hormone response in many cell types. Alterations in the
ligand binding capacity of the receptor, 35 the number of functional receptors
in a particular tissue or cell type, 4 •5 the ability of the receptor to undergo
translocation to or exit from the nucleus, 54 and the affinity of the receptor
for binding to DNA 5 5 have all been shown to affect the level of gene expression
which occurs following exposure of cells or tissues to steroid hormones.
Additional factors including the state of differentiation or development of the
cell or tissue, 6 •7 the stage of cell cycle, 10 the concentration of serum com-
ponents, 56 intracellular second messenger levels, 56 •57 and the concentrations
of nutritional components such as butyric acid 8 •9 and vitamin B6 13 - 15 have also
been shown to influence the degree of hormone responsiveness of selected
tissue types.
III. CLASSICAL ACTIONS OF VITAMIN B6

Vitamin B6 is a water soluble vitamin essential for normal growth, de-
velopment, and metabolism. 14 •58 ·59 The physiologically active form of the
vitamin pyridoxal 5'-phosphate (pyridoxal phosphate) is synthesized in liver
from the dietary precursors pyridoxine, pyridoxal, and pyridoxamine, which
are biologically inactive. 60 -62 Active pyridoxal phosphate is released into cir-
culation in association with albumin, which is thought to protect the vitamin
from degradation while in circulation. 63 -65 However, since the physiologically
active, phosphorylated form of the vitamin is unable to cross cell membranes,
target tissues contain membrane-associated alkaline phosphatases which con-
vert the pyridoxal phosphate to pyridoxal. Pyridoxal readily passes through
the cell membrane, after which active pyridoxal phosphate is regenerated by

pyridoxal kinase enzymes resident within target cells. 66
Classically, pyridoxal phosphate is known for its role as an essential
cofactor for a diverse array of enzymes including aminotransferase (trans-
aminase), racemase, aldolase, dehydratase, cystathionase, and decarboxylase
enzymes involved in numerous aspects of intermediary metabolism_67·70 Pyr-
idoxal phosphate-dependent, amino acid-specific aminotransferase and de-
carboxylase enzymes carry out many of the reactions of amino acid
metabolism67 - 70 and neurotransmitter production. 71 - 74 The vitamin also plays
an important role as a cofactor for enzymes involved in lipid metabolism, 75
gluconeogenesis, 75 and immune function. 76 ·77 One of the most extensively
characterized pyridoxal phosphate-dependent enzymes is aspartate amino-
transferase, for which both cytoplasmic and mitochondrial isozymes have
been purified and crystallized. 78 - 79 X-ray crystallographic and other studies
with these isozymes have demonstrated that pyridoxal phosphate physically
associates with certain lysine amino acids of the enzymes. 78 - 81 Specifically,
the C4-carboxaldehyde group of pyridoxal phosphate interacts through Schiff's
base formation with the epsilon amino group of a specific lysine residue at
the active site of the enzyme. With the cofactor covalently attached to the
active site lysine, the pyridine ring of pyridoxal phosphate serves as an electron
sink providing a low-energy carbanionic path to catalyze reactions at the
alpha-, beta-, or gamma-carbons of the specific amino acid substrateY- 70
While the role pyridoxal phosphate plays as an enzyme cofactor is clearly its
most well characterized function, we have recently reported an additional role
for this important vitamin in modulating hormone-dependent gene expres-
sion. 14 • 18 These new studies suggest that vitamin B6 may also function as an
important physiological mediator of endocrine function.
IV. EFFECTS OF THE VITAMIN ON PHYSICAL

PROPERTIES OF STEROID RECEPTORS
Consideration of a potential physiologic interplay between vitamin B6

and steroid receptors is actually a relatively new field of investigation which
has its roots in classical biochemistry rather than traditional physiology or
nutrition. In much of this early work, pyridoxal phosphate was used as a
biochemical reagent rather than in physiologic quantities as a vitamin. None-
theless, these early experiments using pyridoxal phosphate as a chemical
reagent have contributed much to our understanding of the structure and
function of steroid hormone receptors and have laid a foundation for current
investigations of possible physiologic interactions between this important vi-
tamin and the endocrine system. In 1978, Nishigori et al. first demonstrated
that binding of partially purified progesterone receptor to ATP-Sepharose
could be inhibited by treatment of the receptor preparation with pyridoxal
phosphate. 82 Modak had previously reported that relatively low concentrations
Tully et al. 553
of pyridoxal 5' -phosphate on the order of 0.5 to 1.0 mM acted competitively

with nucleoside triphosphates to inhibit the activity of a variety of DNA
polymerases, 83 and it is probably this report that first prompted Nishigori and
Toft to examine the effects of pyridoxal phosphate on progesterone receptor.
Subsequently, Cake et al. reported that pyridoxal phosphate could inhibit
the in vitro association of partially purified rat liver glucocorticoid receptor
with DNA cellulose, phosphocellulose, and isolated liver nuclei. 84 However,
these studies did not distinguish between two possible effects of the vitamin:
inhibition of the receptor activation process or inhibition of receptor associ-
ation with these synthetic matrices. Early work from our laboratory addressed
these concerns using intact thymic lymphocytes, which contain specific, high
affinity, saturable glucocorticoid receptors. 85 Exposure of thymocytes to the
synthetic glucocorticoid hormone dexamethasone results in activation and
nuclear localization of glucocorticoid receptors. However, glucocorticoid re-
ceptors localized to the nucleus under physiological conditions of hormone
treatment were efficiently extracted from nuclei by treatment with 5 mM
pyridoxal phosphate. Only the physiologically active form of the vitamin,
pyridoxal 5' -phosphate, was effective; neither precursors (pyridoxine, pyri-
doxamine phosphate), metabolites (pyridoxal), nor nonphysiological analogs
(5-deoxypyridoxal) were capable of extracting nuclear receptors. These data
suggested that pyridoxal phosphate does not affect the process of receptor
activation, but instead acts to promote release of receptors previously con-
tained within or associated with the nucleus. This study also demonstrated
that the C4-carboxaldehyde group of pyridoxal phosphate is required for
efficient extraction of nuclear receptors. 85 As described above, it is through
this functional moiety that pyridoxal phosphate interacts with the enzymes
for which it serves as cofactor. Reduction of pyridoxal phosphate with either
hydroxylamine or semicarbazide produces derivatives of pyridoxal phosphate
which are incapable of extracting nuclear receptors, as shown in Figure I.
These data were the first to demonstrate a direct correlation between the
ability of pyridoxal phosphate to interact with proteins, via Schiff's base
formation, and its ability to extract glucocorticoid receptors previously lo-
calized to the nucleus. In subsequent studies, in vitro exposure to pyridoxal
phosphate has been shown to inhibit the association of progesterone, 86 estro-
gen,87 and androgen 88 receptors with DNA and isolated nuclei and to inhibit
both DNA- and steroid-binding by glucocorticoid receptor. 89 Further, whole
animals with elevated levels of vitamin B6 exhibit decreased levels of uterine
nuclear estrogen receptors, 90 while vitamin B6 deficiency has been shown to
result in enhanced nuclear localization of both uterine estrogen receptor91. 92
and prostatic androgen receptor. 93 Thus, with the caveat that the high con-
centrations of vitamin B6 used in these experiments really do not prove a
direct effect of the vitamin on receptor, pyridoxal phosphate treatment appears
to modulate at least two functions common to all the steroid receptors: DNA
binding and nuclear localization.
cpm
!
~
500
CD
E
~ 400
~
300
20
100
FIGURE 1. Inhibition of pyridoxal phosphate extraction of nuclear dexamethasone receptors

by hydroxylamine and semicarbazide. Nuclei from 20 fLI of cells previously exposed to ['H]
dexamethasone ± unlabeled dexamethasone were extracted with 200 fLI of the following solutions
for I h at 3°C. Controls represent initial binding values and were not extracted. All solutions
were 1.5 mM MgCI 2 (pH 7.0 ± 0.05). PPAL, 5 mM pyridoxal phosphate; SC, 10 mM semi-
carbazide; HA, 10 mM hydroxylamine; PPAL + SC, 5 mM pyridoxal phosphate and 7.5 mM
semicarbazide (pH 7.4); PPAL + HA, 5 mM pyridoxal phosphate and 7.5 mM hydroxylamine.
Each value represents the mean ± SD from four separate determinations. (From Cidlowski, J.
A. and Thanassi, J. W., Biochem. Biophys. Res. Commun., 82, 1140, 1978. With permission.)
The observation that pyridoxal phosphate affects nuclear binding of glu-

cocorticoid receptors prompted further examination into the mechanism through
which pyridoxal phosphate acts on steroid hormone receptors. Thus, we ex-
amined the effect of pyridoxal phosphate treatment on the conformation or
subunit composition of the glucocorticoid receptor isolated from HeLa S3
cells. 94 These cells are of human origin and contain high affinity, saturable
glucocorticoid receptors. The unactivated cytosolic receptor prepared from
these cells was known from previous studies to migrate on sucrose density
gradients with a sedimentation coefficient of 8S, while the activated or nuclear
form of receptor sediments as a 4S complex. Treatment of the unactivated,
cytosolic (8S) form of receptor with pyridoxal phosphate, followed by re-
duction with sodium borohydride, produces a receptor form with a sedimen-
tation profile similar to the activated or nuclear receptor, migrating with a
sedimentation coefficient of 4S. This effect on the sucrose density gradient
sedimentation profile of human glucocorticoid receptor was specific for the
physiologically active form of the vitamin and also required the C4-carbox-
aldehyde group; vitamin analogs lacking the C4-carboxaldehyde group, pyr-
idoxine and pyridoxamine phosphate, did not alter the sedimentation coef-
ficient of the receptor, supporting the idea that pyridoxal phosphate may
Tully et al. 555
interact directly with the glucocorticoid receptor through Schiff's base for-
mation, to alter its sedimentation profile. Other lysine-reactive compounds,
including dinitrofluorobenzene, picryl sulfonic acid, succinate, or p-nitro-
phenol acetate, had no effect on receptor sedimentation profiles. 94 Thus, the
glucocorticoid receptor exhibits an alteration in conformation or subunit com-
position as a result of exposure specifically to the physiologically active form
of vitamin 8 6 , pyridoxal phosphate. The requirement for the carboxaldehyde
moiety of pyridoxal phosphate further supports the notion of a direct inter-
action between the receptor and the vitamin. Pyridoxal phosphate treatment
has also been shown to have similar effects on other steroid receptors. Both
progesterone86 and estrogen receptors 87 exhibit a reduction in sedimentation
coefficient on sucrose density gradients following treatment with pyridoxal
phosphate. These observations suggest that pyridoxal phosphate may act through
a single, common mechanism to influence multiple members of the steroid
hormone receptor family of proteins. However, it still remained unclear whether
the vitamin was acting directly on the receptor or on another protein in the
oligomeric complex.
V. EVIDENCE FOR A DIRECT INTERACTION

BETWEEN PYRIDOXAL PHOSPHATE AND THE
GLUCOCORTICOID RECEPTOR
Subsequent work from our group was directed towards determining if,
in fact, a direct interaction between pyridoxal phosphate and the glucocorticoid
receptor exists. Initial experiments addressing this question examined the
effect of pyridoxal phosphate treatment on site-specific, limited proteolytic
digestion of the rat thymic glucocorticoid receptorY As shown in Figure 2,
the unactivated receptor elutes from S-200 Sephacryl as a 56 to 60 A form,
while treatment of the receptor extract with exogenous trypsin generates a
much smaller 18 to 20 A form of receptor. Treatment of unactivated receptor
preparations with pyridoxal phosphate generates an intermediately sized re-
ceptor, which elutes as a 42 A form. This shift in size of conformation or
subunit composition is consistent with the observed shift in sucrose density
gradient sedimentation following treatment of receptor preparations with pyr-
idoxal phosphate. Surprisingly, however, pretreatment of unactivated glu-
cocorticoid receptor preparations with pyridoxal phosphate prior to treatment
with trypsin prevented the generation of the smaller 18 to 20 A form of
receptor. This protection from trypsinolysis was specific for the active form
of the vitamin, since neither pyridoxine nor pyridoxal were capable of pre-
venting cleavage with trypsin. These observations suggest that pyridoxal phos-
phate may bind directly to the glucocorticoid receptor at or near a site which
is susceptible to cleavage with trypsin, thereby preventing proteolytic diges-
tion.
The idea of a direct interaction of pyridoxal phosphate with the gluco-
corticoid receptor was further examined using an antibody which specifically
UNACTIVATED RECEPTORS
• TRYPSIN • NABH 4
• TRYPSIN
o PPAL. NABH4• TRYPSIN
ov MY
' ' '

ALDO
vv
'
0 20 40 60 80 100 120
FRACTION
FIGURE 2. The influence of trypsin on the metabolism of rat thymocyte glucocorticoid re-
ceptor. Cytosols from thymocytes incubated with ['H] dexamethasone were treated with Wor-
thington trypsin 10 f.l-g/ml for 30 min at 4cC. The trypsin reaction was stopped by the addition
of I mg of soybean trypsin inhibitor (Sigma), and samples were treated with dextran-coated
charcoal as described under Methods prior to chromatography on Sephacryl. The dosed circles
were from untreated (unactivated) cytosol, and the open triangles were from samples to which
a small amount of solid sodium borohydride was added prior to trypsin. The open circles were
derived from samples of cytosol treated with 5 mM pyridoxal phosphate for 3 h at 0 to 4°C,
reduced by the addition of solid sodium borohydride, and then trypsin treated. The patterns are
representative of three similar experiments. (From Cidlowski. J. A., Biochemistry, 19, 6162,
1980. With permission.)
Tully et al. 557
recognizes pyridoxal phosphate. 95 In early experiments with this antibody,

the sucrose density gradient sedimentation profile of the pyridoxal phosphate-
treated glucocorticoid receptor incubated with antibody was altered, such that
the receptor sedimented as a much larger complex. In contrast, incubation
with the antibody failed to alter the sedimentation profile of glucocorticoid
receptors from extracts that were not treated with pyridoxal phosphate. These
results provided further evidence that pyridoxal phosphate may interact di-
rectly with the glucocorticoid receptor.
Subsequent two-dimensional gel analysis of glucocorticoid receptor ex-
tracts treated with pyridoxal phosphate demonstrated that the same protein
present in HeLa cells which binds radiolabeled glucocorticoid hormone is
also recognized specifically by the antibody against pyridoxal phosphate.
These observations made using the anti-pyridoxal phosphate antibody, coupled
with the data on pyridoxal phosphate protection of receptor against proteolysis
by trypsin, provide the most direct evidence to date that pyridoxal phosphate
can interact directly with the glucocorticoid receptor. It follows, then, that
the effects of pyridoxal phosphate on many activities and physical properties
of steroid hormone receptors (DNA binding, nuclear localization, and con-
formation or subunit composition) may be mediated through a direct inter-
action of the vitamin with each of the individual steroid receptors. These
observations also suggest that the vitamin effects are mediated through the
receptor molecules themselves and not through other receptor-associated fac-
tors.
Based on these findings, we decided to examine glucocorticoid receptors
treated in vitro with or without pyridoxal phosphate by isoelectric focusing.
If pyridoxal phosphate interacts directly with the glucocorticoid receptor, one
might expect that the isoelectric point of the receptor would shift due to
elimination of the positive charge on the side chain of the bound lysine and
£he concurrent addition of the negatively charged phosphate moiety of pyri-
doxal phosphate. The results obtained from isoelectric focusing of pyridoxal
phosphate-treated receptor extracts indicated that the isoelectric point of the
receptor was shifted to more acidic forms as a result of in vitro treatment
with pyridoxal phosphate. 15 This result is consistent with the hypothesis that
pyridoxal phosphate interacts directly with a lysine residue of the receptor,
and so alters its charge as anticipated, thus rendering the receptor protein
more acidic. This pyridoxal phosphate-induced alteration in the isoelectric
point of the receptor may have important biological ramifications. A more
acidic form of the receptor would probably tend to interact less favorably
with other acidic molecules, such as DNA. Thus, the specific alteration in
surface charge that would result upon pyridoxal phosphate binding may serve
as the mechanism by which DNA binding and nuclear localization of the
glucocorticoid receptor is inhibited by treatment with pyridoxal phosphate.
A caveat in this series of experiments is that these results are still based on
in vitro treatment of receptor extracts with the vitamin and may not reflect
how pyridoxal phosphate would react in vivo. To ascertain whether the is-
oelectric point of glucocorticoid receptor would be altered by pyridoxal phos-

phate in vivo, glucocorticoid receptor extracts were prepared from HeLa cells
cultured under conditions of elevated intracellular concentrations of pyridoxal
phosphate and subjected to isoelectric focusing. Results of this experiment
indicated that the in vivo treated glucocorticoid receptor also focuses as a
more acidic form.
This data was thus the first to demonstrate that the glucocorticoid receptor
may be influenced not only by in vitro treatment with pyridoxal phosphate,
but also by in vivo exposure. Although this observed effect of the vitamin
on the isoelectric point or surface charge of the glucocorticoid receptor mea-
sured a physical property of the receptor protein and did not directly examine
physiological functions of the receptor, it stimulated the provocative idea that
pyridoxal phosphate may be able to act in vivo to affect physiological functions
of the glucocorticoid receptor including DNA binding and modulation of target
gene expression.
VI. IN VIVO EVIDENCE FOR VITAMIN B6 EFFECTS

ON STEROID HORMONE RECEPTOR-MEDIATED
GENE EXPRESSION
To ascertain whether alterations in the in vivo concentrations of pyridoxal

phosphate could, in fact, affect the ability of the glucocorticoid receptor to
activate expression of a glucocorticoid-regulated target gene, we introduced
into HeLa cells a glucocorticoid responsive reporter plasmid, pGMCS. This
plasmid contains the entire coding sequence for the bacterial enzyme chlor-
amphenicol acetyltransferase (CAT). Expression of the CAT gene in this
plasmid is controlled by a promoter obtained from the long terminal repeat
of the mouse mammary tumor virus which is known to confer hormone
responsiveness upon heterologous genes. 96 In preliminary studies transfecting
this plasmid into HeLa cells, we determined that induction of CAT activity
occurs in direct response to glucocorticoid hormone treatment, is mediated
by the glucocorticoid receptor, and is not influenced by other steroid hor-
mones.19 In parallel studies, we determined conditions under which the in-
tracellular concentration of pyridoxal phosphate could be altered. 19 We found
that supplementation of the cell culture medium with the pyridoxal phosphate
synthesis precursor pyridoxine produced an increase in intracellular pyridoxal
phosphate concentration, while culture in medium supplemented with a pyr-
idoxal phosphate synthesis inhibitor, 4-deoxypyridoxine, resulted in a slight
intracellular pyridoxal phosphate deficiency. When HeLa cells transfected
with the glucocorticoid-responsive plasmid pGMCS were grown in medium
supplemented with pyridoxine, the level of glucocorticoid-induced CAT ac-
tivity was suppressed, as shown in the left panel of Figure 3; in contrast,
cells cultured under conditions of mild vitamin deficiency exhibited an en-
hanced level of glucocorticoid-induced CAT activity. These effects were
Tully et al. 559
3.0
~~ 2.0
5!!.
<';
1.0
COli - .... - COli pBLCAT2

- -•.- COli---.-
pGMCS pRSVCAT
FIGURE 3. Effect of alterations in intracellular pyridoxal phosphate concentration on CAT

activity derived from glucocorticoid-responsive and nonresponsive plasmids in transfected HeLa
cells. Following transfection with either the glucocorticoid-responsive plasmid pGMCS or a
glucocorticoid-insensitive plasmid, pBLCAT2 or pRSVCAT, HeLa cells were cultured in un-
altered medium (CON), pyridoxine-deficient medium (-B6), or medium supplemented with either
1 rnM pyridoxine (PYR) or 5 rnM 4-deoxypyridoxine (4-DEOXY) for 48 h. Cells were then
stimulated with 100 nM dexamethasone for 8 hand assayed for CAT activity. The values shown
are representative of the mean ± SE from two independent transfections, with CAT activity
from cells grown in unaltered medium assigned a value of 1.0, and CAT activity from cells
grown in other media expressed as a percentage of this value. The first 4 lanes are from cells
transfected with pGMCS, lanes 5 through 8 from cells transfected with pBLCAT2, and lanes 9
through 12 from cells transfected with pRSVCAT. (From Allgood, V. E., Powell-Oliver, F. E.,
and Cidlowski, J. A., J. Bioi. Chem., 265, 12424, 1990. With permission.)
observed under vitamin concentrations in which neither the cellular concen-

tration of glucocorticoid receptor nor its capacity to bind steroid were affected,
suggesting that changes in the amount of CAT activity were probably indic-
ative of alterations in the ability of the glucocorticoid receptor to induce CAT
gene expression. However, it was not clear if the effects of altered pyridoxal
phosphate concentration represented specific effects on glucocorticoid recep-
tor-dependent gene expression or perhaps reflected nonspecific actions on
other transcription factors involved in the induction of CAT gene expression
or CAT enzymatic activity. To distinguish between these possibilities, we
examined the effect of variations in intracellular pyridoxal phosphate con-
centration on the level of CAT activity derived from other transfected CAT
reporter plasmids. For this analysis, we measured CAT enzymatic activity
derived from CAT reporter plasmids regulated by constitutive, hormone-
independent promoters. 19 Data from these experiments, shown in the middle
and right panels of Figure 3, demonstrated that the same pyridoxal phosphate
concentrations which influence the level of glucocorticoid-induced gene
expression from the pGMCS reporter plasmid have no effect on the level of
CAT activity derived from reporter plasmids containing constitutive, hor-
mone-independent promoters. Thus, intracellular pyridoxal phosphate does
not act indiscriminately to influence gene transcription in a nonspecific man-
ner, but instead acts specifically to modulate the level of glucocorticoid
receptor-mediated gene expression.
Based on the sequence homology and structural similarities between the

glucocorticoid and other steroid hormone receptors, as well as the similarity
in in vitro effects on other steroid hormone receptors resulting from pyridoxal
phosphate treatment, we examined the possibility that alterations in intracel-
lular pyridoxal phosphate concentration might also affect the physiological
capacity of the progesterone, androgen, and estrogen receptors to activate
target gene expression.
We first investigated the effect of alterations in pyridoxal phosphate
concentration on gene expression mediated by the androgen receptor endog-
enously expressed in E8.2 cells, a derivative of the L929 cell line, and
progesterone receptor endogenously expressed in T47D cells. Extensive DNA
binding and transcriptional activation studies have demonstrated that the glu-
cocorticoid, androgen, and progesterone receptors can all act through the
same hormone response element present in the mouse mammary tumor virus
promoter to activate gene expression. 96- 98 Therefore, we transfected these cells
with the same hormone-inducible CAT reporter construct described above,
pGMCS, and examined the effect of altered pyridoxal phosphate concentra-
tions on the level of androgen- and progesterone-induced gene expression.
Results from these experiments demonstrated that gene expression mediated
through the androgen receptor in E8.2 cells and progesterone receptor-me-
diated gene expression in T47D cells were both affected in a manner analogous
to that which we had observed with the glucocorticoid receptor. 99 That is,
the level of androgen- or progesterone-induced CAT gene expression was
suppressed when cells were cultured under conditions which elevate intra-
cellular pyridoxal phosphate concentration, and hormone-induced gene
expression was enhanced under conditions which produce pyridoxal phosphate
deficiency. These observations are presented in Figure 4. Thus, the effects
of pyridoxal phosphate concentration on steroid receptor-mediated transcrip-
tional activation were not restricted to the glucocorticoid receptor, but ex-
tended to other members of the steroid hormone receptor family, including
at least the androgen and progesterone receptors.
Pyridoxal phosphate treatment had also been shown to exert some effect
on physical properties of the estrogen receptor, 87 and we examined the effect
of pyridoxal phosphate concentration on the level of estrogen receptor-me-
diated gene expression. For these studies, we cotransfected HeLa cells with
an expression vector directing production of the estrogen receptor along with
an estrogen receptor-responsive CAT reporter plasmid. Results from these
studies confirmed our hypothesis that alterations in intracellular pyridoxal
phosphate concentration would similarly affect the level of estrogen-induced
CAT gene expression. The level of estrogen receptor-mediated gene expres-
sion was reduced under conditions of elevated pyridoxal phosphate concen-
tration and enhanced under the opposite condition of pyridoxal phosphate
deficiency, 99 as shown in Figure 5. These observations demonstrate that pyr-
idoxal phosphate modulation of steroid receptor-mediated gene expression is
not restricted to the glucocorticoid receptor-regulated plasmid which contains
Tully et al. 561
E8.2 CELLS T47D CELLS
>-
I- 2.0 2.0
>
I-
u<{
I- 1.0 1.0
<{
u
CON PYR 4DXY CON PYR 4DXY

pGMCS pGMCS
ANDROGEN PROGESTERONE
RECEPTOR RECEPTOR
FIGURE 4. Vitamin B6 influences transcriptional activation of the androgen and progesterone
receptors. Left: 16 h after transfection with pGMCS, E8.2 cells were cultured for 48 h in either
control medium (CON), medium supplemented with 2 mM pyridoxine (PYR), or medium sup-
plemented with 3 mM 4-deoxypyridoxine (4DXY). Cells were then treated with 10 nM Rl881
or vehicle for 12 h prior to CAT activity determination. Right: 16 h after transfection with
pGMCS, T47D cells were cultured for 48 h in either control medium, medium supplemented
with 3 mM pyridoxine, or medium supplemented with 2 mM 4-deoxypyridoxine prior to 24-h
treatment with 1 nM progesterone or vehicle and CAT activity determination. The data shown
are representative of at least three individual experiments. (From Allgood, V. E. and Cidlowski,
J. A., J. Bioi. Chern., 267, 3819, 1992. With permission.)
sequences derived from the mouse mammary tumor virus promoter, and that
variation of intracellular pyridoxal phosphate concentration also affects tran-
scription mediated through other hormonally regulated promoters.
VII. SUMMARY
These studies on transcriptional activation by the glucocorticoid, andro-
gen, progesterone, and estrogen receptors provide the most direct evidence
to date that pyridoxal phosphate, the active form of vitamin B6 , acts not only
to affect some physical properties of steroid hormone receptors, but also
influences the physiological function of these molecules. The demonstration
that alterations in intracellular pyridoxal phosphate concentration have dra-
matic effects on hormonally regulated target gene expression suggests that
vitamin B6 may act as a physiological regulator of steroid hormone action.
Despite the efforts of many laboratories, surprisingly little is known about
how steroid hormone action may be modulated in different tissues and cell
types. Although the specific mechanisms through which receptor-mediated
HeLo CELLS
2.0 r
>- ..=-
1-
>
i=
r-
u 1.0 r-.--
<{
1-
<{
u
-
CON PYR4DXY
pERENF1CAT
+
pH EO
ESTROGEN
RECEPTOR
FIGURE 5. Estrogen receptor-mediated gene expression is modulated by alterations in intra-
cellular pyridoxal phosphate concentration. HeLa cells were transfected with the estrogen receptor
expression vector pHEO and the estrogen-regulated CAT reporter pERENFICAT. Cells were
exposed to control medium (CON) 16 h after transfection medium. medium supplemented with
I mM pyridoxine (PYR), or medium supplemented with 5 mM 4-deoxypyridoxine (4DXY) prior
to 24-h treatment with 5 nM 17 beta-estradiol or vehicle and CAT activity determination. The
data shown reflect results from at least three individual transfection experiments. (From Allgood,
V. E. and Cidlowski, J. A., J. Bioi. Chern., 267. 3819, 1992. With permission.)
gene expression is induced have been much more clearly identified, much
less is known about the mechanisms cells may employ to regulate the mag-
nitude of their response to steroid hormones. The observations which have
been described here are consistent with the hypothesis that pyridoxal phosphate
interacts with the steroid hormone receptors, and, in so doing, alters both
physical properties of the receptors and their transcriptional activation ca-
pacity. Thus by regulating the intracellular concentration of pyridoxal phos-
phate, differentiated cell types may be able to modify the magnitude of their
response to steroid hormone stimulation. Identification of a site on the steroid
receptors with which pyridoxal phosphate interacts will facilitate proof of this
hypothesis. However, it is clear that, regardless of mechanism, vitamin B6
acts to modulate the capacity of many different steroid hormone receptors to
induce target gene expression. Thus, vitamin B6 has an important role in the
regulation of steroid hormone action.
Tully et al. 563
REFERENCES
I. Beato, M., Transcriptional control by nuclear receptors. FASEB J., 5, 2044, 1991.
2. Munck, A., Mendel, D. B., Smith, L. I., and Orti, E., Glucocorticoid receptors and
actions, Am. Rev. Respir. Dis., 141, S2, 1990.
3. Evans, R. M., The steroid and thyroid hormone receptor superfamily, Science, 240,
889, 1988.
4. Vanderbilt, J. N., Miesfeld, R., Maler, B. A., and Yamamoto, K. R., Intracellular
receptor concentration limits glucocorticoid-dependent enhancer activity, Mol. Endocri-
nol .. 1, 68, 1987.
5. Gehring, U., Mugele, K., and Ulrich, Cellular receptor levels and glucocorticoid re-
sponsiveness of lymphoma cells, J. Mol. Cell. Endocrinol .. 36, 107. 1984.
6. Ballard, P. L., Glucocorticoids and differentiation, in Glucocorticoid Hormone Action.
Monographs in Endocrinology, Vol. 12. Baxter, J. D. and Rousseau, G. G., Eds.,
Springer-Verlag, Berlin, 1979, 493.
7. Kalinyak, J. E., Griffin, C. A., Hamilton, R. W., Bradshaw, J. G., Perlman, A. J.,
and Hoffman, A. R., Developmental and hormonal regulation of glucocorticoid receptor
messenger RNA in the rat, J. Clin. Invest., 84. 1843, 1989.
8. Littlefield, B. A., Cidlowski, N. B., and Cidlowski, J. A., Modulation of glucocorticoid
effects and steroid receptor binding in butyrate-treated HeLa S 3 cells, Arch. Biochem.
Biophys., 201, 174, 1980.
9. Littlefield, B. A. and Cidlowski, J. A., Increased steroid responsiveness during sodium
butyrate-induced "differentiation" of HeLa S, cells, Endocrinology, 114, 566, 1984.
10. Cidlowski, J. A. and Michaels, G. A., Alteration in glucocorticoid binding site number
during the cell cycle in HeLa cells, Nature (London), 266, 643, 1977.
11. Tomkins, G. M., Gelehrter, T., Granner, D., Martin, D., Samuels, H. H., and
Thompson, E. B., Control of gene expression in higher organisms, Science, 166, 1474,
1969.
12. Griffin, M. J. and Ber, R., Cell cycle events in the hydrocortisone regulation of alkaline
phosphatase in HeLa S, cells, J. Cell Bioi., 40, 297, 1969.
13. Allgood, V. E. and Cidlowski, J. A., Novel role for vitamin B6 in steroid hormone
action: a link between nutrition and the endocrine system, J. Nutr. Biochem., 2, 641,
1991.
14. Allgood, V. E., Powell-Oliver, F. E., and Cidlowski, J. A., Influence of vitamin B6
on the structure and function of the glucocorticoid receptor, Ann. N.Y. Acad. Sci., 585,
452, 1990.
15. Compton, M. M. and Cidlowski, J. A., Vitamin B6 and glucocorticoid action, Endocrine
Rev., 7, 140, 1986.
16. DiSorbo, D. M. and Litwack, G., Changes in the intracellular levels of pyridoxal 5'-
phosphate affect the induction of tyrosine aminotransferase by glucocorticoids, Biochem.
Biophys. Res. Commun., 99, 1203, 1981.
17. Valeriote, F. A., Auricchio, F., Tomkins, G. M., and Riley, D., Purification and
properties of rat tyrosine aminotransferase, J. Bioi. Chern., 244, 3618, 1969.
18. Allgood, V. E., Powell-Oliver, F. E., and Cidlowski, J. A., Vitamin B6 influences
glucocorticoid receptor-mediated gene expression, J. Bioi. Chern., 265, 12424, 1990.
19. Beato, M., Gene regulation by steroid hormones, Cell, 56, 355, 1989.
20. O'Malley, B. W., The steroid receptor superfamily: more excitement predicted for the
future, Mol. Endocrino/., 4, 363, 1990.
21. Reichman, M. E., Foster, C. M., Eisen, L. P., Torain, B. F., and Simons, S. S.,
Limited proteolysis of covalently labeled glucocorticoid receptors as a probe of receptor
structure, Biochemistry, 23, 5376, 1984.
22. Sherman, M. R., Moran, M. C., Tuazon, F. B., and Stevens, Y.-E., Structure,
dissociation, and proteolysis of mammalian steroid receptors, J. Bioi. Chern., 258, 10366,
1983.
23. Cidlowski, J. A., Pyridoxal phosphate induced alterations in glucocorticoid receptor

metabolism by proteases, Biochemistry, 19, 6162, 1980.
24. Bellingham, D. L. and Cidlowski, J. A., The glucocorticoid receptor: functional im-
plications of protein and gene structure, in Steroid Receptors and Disease, Sheridan,
P. J., Blum, K., and Trachtenberg, M. C., Eds., Marcei-Dekker, New York, 1988,97.
25. Gronemeyer, H., Green, S., Kumar, V., Jeltsch, J.-M. and Chambon, P., Structure
and function of the estrogen receptor and other members of the nuclear receptor family,
in Steroid Receptors and Disease, Sheridan, P. J., Blum, K., and Trachtenberg, M. C.,
Eds., Marcel-Dekker, New York, 1988, 153.
26. Beato, M., Gene regulation by steroid hormones, Cell, 56, 335, 1989.
27. Krust, A., Green, S., Argos, P., Kumar, V., Walter, P., Bomert, J. M., and Chambon,
P., The chicken oestrogen receptor sequence: homology with v-erba and the human
oestrogen and glucocorticoid receptors, EMBO J., 5, 891, 1986.
28. Carlstedt-Duke, J., Stromstedt, P.-E., Wrange, 0., Bergman, T., Gustafsson, J.-A.,
and Jornvall, H., Domain structure of the glucocorticoid receptor protein, Proc. Nat/.
Acad. Sci. U.S.A., 84, 4437, 1987.
29. Rusconi, S. and Yamamoto, K. R., Functional dissection of the hormone and DNA
binding activities of the glucocorticoid receptor, EMBO J., 6, 1309, 1987.
30. Carlstedt-Duke, J., Okret, S., Wrange, 0., and Gustafsson, J.-A., Immunochemical
analysis of the glucocorticoid receptor: identification of a third domain separate from the
steroid-binding and DNA-binding domains, Proc. Nat!. Acad. Sci. U.S.A., 79, 4260,
1982.
31. Tora, L., White, J., Brou, C., Tassel, D., Webster, N., Scheer, E., and Chambon,
P., The human estrogen receptor has two independent nonacidic transcriptional activation
functions, Cell, 59, 477, 1989.
32. Hollenberg, S.M. and Evans, R. M., Multiple and cooperative trans-activation domains
of the human glucocorticoid receptor, Cell, 55, 899, 1988.
33. Carson-Jurica, M.A., Schrader, W. T., and O'Malley, B. W., Steroid receptor family:
structure and functions, Endocrine Rev., 11, 201, 1990.
34. Forman, B. M. and Samuels, H. H., Interactions among a subfamily of nuclear hormone
receptors: the regulatory zipper model, Mol. Endocrinol., 4, 1293, 1990.
35. Green, S. and Chambon, P., Nuclear receptors enhance our understanding of transcrip-
tion regulation, Trends Genet., 4, 309, 1988.
36. Freedman, L. P., Luisi, B. F., Korszun, Z. R., Basavappa, R., Sigler, P., and
Yamamoto, K. R., The function and structure of the metal coordination sites within the
glucocorticoid receptor DNA binding domain, Nature (London), 334, 543, 1988.
37. Umesono, K. and Evans, R. M., Determinants of target gene specificity for steroid/
thyroid hormone receptors, Cell, 57, 1139, 1989.
38. Danielsen, M., Hinck, L., and Ringold, G. M., Two amino acids within the knuckle
of the first zinc finger specify DNA response element activation by the glucocorticoid
receptor, Cell, 57, 1131, 1989.
39. Mader, S., Kumar, V., Verneuil, H., and Chambon, P., Three amino acids of the
oestrogen receptor are essential to its ability to distinguish an oestrogen from a gluco-
corticoid responsive element, Nature (London), 338, 271, 1989.
40. Green, S., Kumar, V., Theulaz, I., Wahli, W., and Chambon, P., The N-terminal
DNA-binding 'zinc finger' of the oestrogen and glucocorticoid receptors determines target
gene specificity, EMBO J., 7, 3037, 1988.
41. Schmid, E., Schmid, W., Jantzen, M., Mayer, D., Jastorff, B., and Schutz, G.,
Transcription activation of the tyrosine aminotransferase gene by glucocorticoids and
cAMP in primary hepatocytes, Eur. J. Biochem., 165, 499, 1987.
42. Becker, P. B., Ruppert, S., and Schutz, G., Genomic footprinting reveals cell type-
specific DNA binding of ubiquitous factors, Cell, 51, 435, 1987.
Tully et al. 565
43. Mendel, D. B. and Orti, E., Isoform composition and stoichiometry of the -90-kDa
heat shock protein associated with glucocorticoid receptors, J. Bioi. Chern., 263, 6695,
1988.
44. Denis, M., Wikstrom, A.-C., Gustafsson, J.-A., the molybdate-stabilized nonactivated
glucocorticoid receptor contains a dimer of Mr 90,000 non-hormone-binding protein, J.
Bioi. Chern., 262, 11803, 1987.
45. Currie, R. A. and Cid1owski, J. A., Physicochemical properties of the cytoplasmic
glucocorticoid receptor complex in HeLa S3 cells, J. Steroid Biochem., 16, 419, 1982.
46. Holbrook, N. J., Bodwell, J, E., Jeffries, M., and Munck, A., Characterization of
nonactivated and activated glucocorticoid receptor complexes from intact rat thymus cells,
J. Bioi. Chern., 258, 6477, 1983.
47. Vedeckis, W. V., Subunit dissociation as a possible mechanism of glucocorticoid receptor
activation, Biochemistry, 22, 1983, 1983.
48. Burnstein, K. and Cidlowski, J. A., Regulation of gene expression by glucocorticoids,
Annu. Rev. Physiol., 51, 683, 1989.
49. Yamamoto, K. R. and Alberts, B. M., Steroid receptors: elements for modulation of
eukaryotic transcription, Annu. Rev. Biochem., 45, 721, 1975.
50. Rousseau, G. G., Control of gene expression by glucocorticoid hormones, Biochem. J.,
224, 1, 1984.
51. Ringold, G. M., Steroid hormone regulation of gene expression, Annu. Rev. Pharm.
Toxicol., 25, 529, 1985.
52. Drouin, J., Sun, Y. L., and Nemer, M., Glucocorticoid repression of proopiomelan-
ocortin gene transcription, J. Steroid Biochem., 34, 63, 1989.
53. Mordacq, J. C. and Linzer, D. I. H., Co-localization of elements required for phorbol
ester stimulation and glucocorticoid repression of proliferin gene expression, Genes De-
velop., 3, 760, 1989.
54. Picard, D. and Yamamoto, K. R., Two signals mediate hormone-dependent nuclear
localization of the glucocorticoid receptor, EMBO J., 6, 3333, 1987.
55. Nordeen, S. K., Sub, B. J., Kuhne!, B., and Hutchison, C. A., III, Structural
determinants of a glucocorticoid receptor recognition element, Mol. Endocrinol., 4, 1866,
1990.
56. Aronica, S.M. and Katzenellenbogen, B.S., Progesterone receptor regulation in uterine
cells: stimulation by estrogen, cyclic adenosine 3' ,5' -monophosphate, and insulin-like
growth factor I and suppression by antiestrogens and protein kinase inhibitors, Endocri-
nology, 128, 2045, 1991.
57. Groul, D. J., Harrigan, M. T., and Bourgeois, S., Modulation of glucocorticoid-
induced responses by cyclic AMP in lymphoid cell lines, in Gene Regulation by Steroid
Hormones, Vol. 4, Roy, A. K. and Clark, J. H., Eds., Springer-Verlag, Berlin, 1989,
41.
58. Tryfiates, G. P., Vitamin B6 Metabolism and Role in Growth, Food and Nutrition Press,
Westport, CT, 1980, I.
59. Merrill, A. H. and Henderson, J. M., Diseases associated with deficiencies in vitamin
8 6 metabolism or utilization, Annu. Rev. Nutr., 7, 137, 1987.
60. Merrill, A. H., Henderson, J. M., Wang, E., McDonald, B. W., and Millikan,
W. J., Metabolism of vitamin 8 6 by human liver, J. Nutr., 114, 1664, 1964.
61. McCormick, D. B., Two interconnected B vitamins: riboflavin and pyridoxine, Physiol.
Rev., 69, 1170, 1989.
62. Wada, H. and Snell, E. E., The enzymatic oxidation of pyridoxine and pyridoxamine
phosphates, J. Bioi. Chern., 236, 2089, 1961.
63. Dempsey, W. B. and Christensen, H. N., The specific binding ofpyridoxal5-phosphate
to bovine plasma albumin, J. Bioi. Chern., 237, 1113, 1962.
64. Anderson, B. B., Newmark, P. A., and Rawlins, M., Plasma binding of vitamin 8 6
compounds, Nature (London), 250, 502, 1974.
65. Lumeng, L. and Li, T. K., Mammalian vitamin 8 6 metabolism: regulatory role of
protein binding and the hydrolysis of pyridoxal 5 '-phosphate in storage and transport, in
Vitamin B6 Metabolism and Role in Growth, Tryfiates, G. P., Ed., Food and Nutrition
Press, Westport, CT, 1980, 27.
66. Brin, M., Vitamin B6 : Chemistry, absorption, metabolism, catabolism, and toxicity, in
Human Vitamin B6 Requirements, National Academy of Sciences, Washington, D.C.,
1976.
67. Leklem, J. E., Vitamin 8 6 , in Handbook of Vitamins, 2nd ed., Machlin, L. J., Ed.,
Marcel Dekker, New York, 1991, chap. 9.
68. Hayashi, H., Wada, H., Yoshimura, T., Esaki, N., and Soda, K., Recent topics in
pyridoxal 5' -phosphate enzyme studies, Annu. Rev. Biochem., 59, 87, 1990.
69. Sauberlich, H. E., Biochemical systems and biochemical detection of deficiency, in The
Vitamins, Sebrell, W. H. and Harris, R. S., Eds., Academic Press, New York, 1968,
31.
70. Walsh, C., Enzymatic reactions requiring pyridoxal phosphate, in Enzymatic Reaction
Mechanisms, W. H. Freeman, San Francisco, 1979, chap. 24.
71. Buffoni, F., Histamine and related amine oxidases, Pharmacal. Rev., 18, 1163, 1966.
72. Kahlson, G. and Rosengren, E., Histamine, Annu. Rev. Pharmacal., 5, 305, 1965.
73. Lovenberg, W., Weissbach, H., and Udenfriend, S., Aromatic L-amino acid decar-
boxylase, J. Bioi. Chern., 237, 89, 1962.
74. Roberts, E. and Frankel, S., Gamma-amino butyric acid in brain: its formation from
glutamic acid, J. Bioi. Chern., 187, 55, 1950.
75. Leklem, J. E., Vitamin 8 6 metabolism and function in humans, in Clinical and Phys-
iological Applications of Vitamin B6 , Leklem, J. E. and Reynolds, R. D., Eds., Alan
R. Liss, New york, 1988, 3.
76. Axelrod, A. E., Immune processes in vitamin deficiency states, Am. J. Clin. Nutr., 24,
265, 1971.
77. Chandra, R. K. and Purl, S., Vitamin 86 modulation of immune responses and in-
fection, in Vitamin 86: its role in health and disease, Reynolds, R. D. and Leklem, J.
E., Eds., Alan R. Liss, New York, 1984, 163.
78. Harutyunyan, E. G., Malashkevich, V.N., Kochkina, V.M., and Torchinsky, Y. M.,
Conformational changes in chicken heart cytosolic aspartate aminotransferase as revealed
by X-ray crystallography, in Chemical and Biological Aspects of Vitamin B6 , Evange-
lopoulos, A. E., Ed., Alan R. Liss, New York, 1984, 205.
79. Jansonius, J. N., Eichele, G., Ford, G. C., Kirsch, J. F., Picot, D., Thaller, C.,
and Vincent, M. G., Crystallographic studies on the mechanism of action of mitochon-
drial aspartate aminotansferase, in Chemical and Biological Aspects of Vitamin B6 , Evan-
gelopoulos, A. E., Ed., Alan R. Liss, New York, 1984, 195.
80. Gehring, H., Sandmeier, E., Kirsten, H., Carrassi, R., Christen, P., andJansonius,
J. N., Aspartate aminotransferase: studies on conformational adaptations and 'H-transfer
during transamination, in Chemical and Biological Aspects of Vitamin B6 Catalysis,
Evangelopoulos, A. E., Ed., Alan R. Liss, New York, 1984, 223.
81. Taylor, J. E., Metzler, D. E., and Arnone, A., Modeling inhibitors in the active site
of aspartate aminotransferase, Ann. N.Y. Acad. Sci., 585, 58, 1990.
82. Nishigori, H., Moudgil, V. K., and Toft, D., Inactivation of avian progesterone receptor
binding to ATP-Sepharose by pyridoxal 5 '-phosphate, Biochem. Biophys. Res. Commun.,
80, 112, 1978.
83. Modak, M. J., Observations on the pyridoxal 5'-phosphate inhibition of DNA poly-
merases, Biochemistry, 15, 3620, 1976.
84. Cake, M. H., DiSorbo, D. M., and Litwack, G., Effect of pyridoxal phosphate on
the DNA binding site of activated hepatic glucocorticoid receptor, J. Bioi. Chern., 253,
4886, 1978.
85. Cidlowski, J. A. and Thanassi, J. W., Extraction of nuclear glucocorticoid receptor
complexes with pyridoxal phosphate, Biochem. Biophys. Res. Commun., 82, 1140, 1978.
Tully et al. 567
86. Nishigori, H. and Toft, D., Chemical modification of the avian progesterone receptor
by pyridoxal 5'-phosphate. J. Bioi. Chern, 254. 9155, 1979.
87. Muldoon, T. G. and Cidlowski, J. A., Specific modification of rat uterine estrogen
receptor by pyridoxal 5'-phosphate, J. Bioi. Chem., 255, 3100, 1980.
88. Hiipakka, R. A. and Liao, S., Effect of pyridoxal phosphate on the androgen receptor
from rat prostate: inhibition of receptor aggregation and receptor binding to nuclei and
to DNA-cellulose, J. Steroid Biochem.. 13. 841. 1980.
89. Silva, C. M., Tully, D. B., Petch, L. A., Jewell, C. M., and Cidlowski, J, A.,
Application of a protein-blotting procedure to the study of human glucocorticoid receptor
interactions with DNA, Proc. Nat/. Acad. Sci. U.S.A., 84. 1744, 1987.
90. Holley, J,, Bender, D. A., Coulson, W. F., and Symes, E. K., Effects of vitamin B6
nutritional status on the uptake of ['H)-oestradiol into the uterus, liver, and hypothalamus
of the rat, J. Steroid Biochem.. 18. 161, 1983.
91. Bowden, J. F., Bender, D. A., Coulson, W. F., and Symes, E. K., Increased uterine
uptake and nuclear retention of ['H] oestradiol through the oestrous cycle and enhanced
end-organ sensitivity to oestrogen stimulation in vitamin 8 6 deficient rats, J. Steroid
Biochem., 25, 359, 1986.
92. Bunce, G. E. and Vessal, M., Effect of zinc and/or pyridoxine deficiency upon oestrogen
retention and oestrogen receptor distribution in the rat uterus, J. Steroid Biochem., 26,
303, 1987.
93. Symes, E. K., Bender, D. A., Bowden, J. F., and Coulson, W. F., Increased target
tissue uptake of, and sensitivity to, testosterone in the vitamin 8 6 deficient rat, J. Steroid
Biochem., 20, 1089, 1984.
94. O'Brien, J. M. and Cidlowski, j. A., Interaction of pyridoxal phosphate with gluco-
corticoid receptors from HeLa S, cells. J. Steroid Biochem., 14, 9. 1981.
95. Viceps-Madore, D., Cidlowski, J. A., Kittler, j. M., and Thanassi. J. W., Prepa-
ration, characterization, and use of monoclonal antibodies to vitamin 8 6 • J. Bioi. Chern.,
258, 2689, 1983.
96. Chandler, V. L., Maler, B. A., and Yamamoto, K. R., DNA sequences bound spe-
cifically by glucocorticoid receptor in vitro render a heterologous promoter hormone
responsive in vivo, Cell, 33, 489. 1983.
97. Cato, A. C. B., Skroch, P., Weinmann, J., Butkeraitis, P., and Ponta, H., DNA
sequences outside the receptor-binding sites differently modulate the responsiveness of
the mouse manunary tumour virus promoter to various steroid hormones, EMBO J., 7,
1403, 1988.
98. Ham, J., Thomson, A., Needham, M., Webb, P., and Parker, M., Characterization
of response elements for androgens, glucocorticoids and progestins in mouse mammary
tumour virus, Nucleic Acids Res., 16, 5263, 1988.
99. Allgood, V. E. and Cidlowski, J. A., Vitamin 86 modulates transcriptional activation
by multiple members of the steroid hormone receptor superfamily. J. Bioi. Chern., in
press.
569
INDEX
A Amino acids, physiological effects,

360-361
Absorption of lipids, 61 Amino acid sequences
Acetyl CoA decarboxylase Gla region of vitamin K-dependent
biotinylation, 290 proteins, 466
mRNA, 272-287 of intestinal immunolactase, 39-40
regulation lactase, 38-41
short -term, 290-291 propeptide region of vitamin K-
transcription, translation, and dependent plasma proteins, 4 71
posttrans1ational processing, Androgen receptors and vitamin B6
280-289 modulation, 547-562, see also
Aconitase, 539 Vitamin B6
Acrodermatitis enteropathica, 508, see Anterior pituitary, see Pituitary
also Metallothionein Anticoagulants, 456-457, 463, see also
Actinomycin and adipocyte Vitamin K
differentiation, 218 Apolipoprotein B mRNA editing,
Acute phase reaction, 539 298-313
Adaptation Arachidonic acid in Ca 2 + protein
of pancreatic lipase and colipase, regulation, 124-125
63-73, see also Lipases, Area postrema, 12
pancreatic Ascorbate (vitamin C)
preduodenal to triglycerides, 62-63 collagen pathway induction: overview,
Adema disease, 508 499-500
Adenylyl cyclase, 8 collagen pathway interactions
Adipocyte activity, 207-223, see also type I
Dietary fats; Lipogenesis mRNA levels, 497-498
Adipocytes secretion, 488-492
differentiation in 3T3-Ll cells, temperature and, 492
216-219 translation, 492-496
retinoic acid regulation, 437-438 types II through X, 498-499
cis-acting elements, 439-440 conclusions, 501-502
in cultured cells, 438-439 indirect regulation of collagen, 50 I
model for, 444-447 lipid oxidation and, 501
summary and conclusions, 447-448 methodological and historical
tissue-specific factors in, 440-444 considerations, 485-488
Sl4 protein regulation in, 434-437 overview of function, 484-485
Adipocyte-specific factors, 440-444
Adipose tissue, fatty acid synthase and B
p90 mRNAs in, 210-211
Age Bacteria, nutrient receptors in, 7-8
and G6PD activity, 189 Beta cells, see Pancreatic beta cells
lactase glycosylation and, 38 Bile acids and colon cancer, 256
Alkaline phosphatase, 397-398 Biological half-life, see Half-life
Alpha-1-acid glycoprotein, 517-518 Biotin
Alternate promoter regions in glucokinase and acetyl CoA decarboxylase activity,
gene, 162-163 290
and hepatic glucokinase gene in glucagon gene expression, II 0

transcription, 155-156 in lac operon activation, 8
BiP, Ca'' and induction of, 131-133 synthesis, 7
Blood glucose regulation, 144-145, see Cancer
also Glucokinase breast, 250-251, 255-256, 258
Bone colon, 250, 25 I, 253-256
calbindin-D gene expression, 382-383 liver, 252-253, 257-258
growth and differentiation. 392-419, lung, 255
see also Vitamin D ovarian, 251
Bone Gla protein, 474 pancreatic, 251
Brain calbindin-D gene regulation, prostatic, 251
383-384 skin, 249, 252
Brain cancer, 255 CAP, see Catabolite activator protein
Breast cancer, 250-251, 255-256, 258 Carbohydrate/fat conversion, 271-272
Brown adipose tissue, see Adipocytes Carbohydrates, dietary, see Dietary
carbohydrate
c Carboxylase reaction, 460-464, see also
Vitamin K
Caerulin, 75 Carcinogenesis
Calbindin-D gene dietary fat and
intestinal and renal gene expression, caloric effects, 257
379-382 eicosanoid metabolism, 256-257
nervous system expression, 383-384 induction of specific genes, 258
promoter analysis, 384-386 initiation effects, 257-258
membrane effects, 256
skeletal (bone) expression, 382-383
toxic effects, 256
types and functions, 378
epidemiological studies, 250-251
Calcium
experimental studies
absorption,lactase and, 29
breast cancer, 255
calbindin-D gene regulation, 381-382
colon cancer. 253-254
endoplasmic reticulum in homeostasis,
effect of omega-3 fatty acids,
119-121
255-256
metabolic stress and induction of
liver cancer, 252-253
GRP78/RIP, 131-133
nutritional considerations. 251
in secretory protein processing and pancreatic cancer, 254-255
export, 133-135 skin cancer, 252
supplementation in lactose intolerance, mechanisms
28 chemical, 248
translational activities genes important to, 250
initiation, 121-123 multistage. 249
peptide chain elongation, 128-131 Camitine palmitoyltransferase
phosphorylation/dephosphorylation, future directions, 332
127 general aspects, 320-321
sequestered Ca 2 + requirements, hepatic fatty acid oxidation and, 320
123-126 identification and variations, 323-324
in stress states, 127-128 problems of study. 321-323
Caloric intake and carcinogenesis, 25 I, promoter analysis, 329-331
257 regulation
cAMP of catalytic activity
and acetyl CoA decarboxylase activity, by lipid domain alterations,
290-291 326-327
Index 571
by malonyl CoA, 324-326 Cyclic adenosine 3' ,5' -monophosphate,

by phosphorylation, 326 see cAMP
of gene expression, 329 Cysteine-rich protein, 519-520
of mass
cloning and sequencing, 328-329 D
purification and antibody
generation, 327-328 Developmental programming in cell
Catabolite activator protein, 8 culture, 486-488
Catabolite repression, 11 Diabetes mellitus
CAT gene glutaminase activity and, 337, 343, 344
in adipocyte differentiation, 442-443 pyruvate kinase (L-type) expression in
and PK gene expression, 179-181 rats, 170-174
CCK (cholecystokinin), 12, 75 and renal function in rats, 171-172,
CCK gene, 98-102 174
Cell culture techniques, developmental urea cycle enzymes in, 345
programming, 486-488 Dietary carbohydrate, see also High
Chelatable and nonchelatable iron, carbohydrate diets
534-536 opposing relationship to proteins,
Chloramphenicol acetyltransferase gene, 35-355
see CAT gene Dietary fat
Cholecystokinin, see CCK and acetyl CoA decarboxylase activity,
286-287
Christmas disease (hemophilia B), 469
and carcinogenesis, 248-258, see also
Circumventricular organs, 14
Carcinogenesis
Cis-acting regions, 4-5
saturated vs. unsaturated, 254
in adipocyte differentation, 444-447
Dietary protein, see also High protein
calbindin-D gene, 384-386
diets
in ferritin synthesis, 530-534
and acetyl CoA decarboxylase activity,
L-type pyruvate kinase gene, 179-183
286
Clotting factors
enzyme synthesis and
biosynthesis, 455-476, see also
response and dietary composition,
Vitamin K
361-364
deficiency syndromes, 468-470
serine-threonine dehydratase,
recombinant, 472 367-369
Colipases tyrosine aminotransferase, 364-367
mRNA, 80 fasting/refeeding and hormonal
pancreatic, 63-73 secretion, 356-360
Collagen pathway opposing relationship to carbohydrates,
ascorbate regulation 354-355
indirect, 500-501 physiological effects of amino acids,
overview, 499-500 360-361
type I as stimulus to organ growth, 355-356
mRNA levels, 497-498 urea cycle and, 341-342, 345-346
secretion, 488-492 Digestion of lipids, 60-61
translation, 492-493 l ,25-Dihydroxyvitamin D,, see Vitamin D
types II through X, 498-499 o.,o.' -dipyrydyl, 488-491
Colon cancer, 250, 251, 253-256
4-hydroxy-Coumarins, 456-457, 464 E
C-reactive protein and heavy metals,
517-518 EF-2, see Elongation factor 2
Eicosanoids, 256-257, 356 hormonal regulation

Elimination constants, see Half-life in vivo, 211-214
Elongation factor l (EF-1), 128 in 3T3-Ll adipocytes, 219-221
Elongation factor 2 (EF-2), 128-131 nutritional regulation in vivo, 211-214
Endocrine regulation, see Hormonal transcription, 214-216
regulation Fatty acid synthesis, see also Lipogenesis
Endoplasmic reticulum in calcium hormonal regulation, 208
homeostasis, 119-121 Fecal pH and lactose intolerance, 27
Endothelin, 356 Feedback mechanisms, 3-4, 501-502
Enhancer elements, 5 Ferritin
Enteric nervous system, 11-12 subunits: amino acid sequences,
Enterocytes, 11-12, 41-42 mRNAs, genes, 526-528
Enteroendocrine cells, see Enterocytes synthesis and iron regulation
Epidermal growth factor conclusions, 540-541
and G6PD activity, 195 transcription, H and L gene
in lactase activity, 34 responses, 539-540
Erythroid 5-aminovulinic acid synthetase, translation
538-539 ferritin model: iron-responsive
Estrogen receptors and vitamin B6 element, 528-538
modulation, 547-562, see also other systems involving IRE,
Vitamin B 6 538-539
Eukaryotic gene expression, 76-78 Fibroblast growth factor, 14
Fos-Jun interactions in osteoclast
F differentiation, 412-415
Fructose, and L-type PK gene expression,
Fasting, see also Fasting/refeeding 170-178
and CCK gene, 99-100
and gastrin gene mRNA, 95-97 G
and somatostatin gene expression,
102-103 beta-Galactosidase, 7-8
Fasting/refeeding Gastric inhibitory peptide, 12, 357
collagen production and, 501 Gastrin gene
in diabetes mellitus, 211 fasting and, 95-97
hormonal secretion and model systems, 98
intestinal and pancreatic peptides, nutrients increasing mRNA, 97-98
357-349 G cells, 98
pituitary hormones, thyroxine, and Geographic patterns
glucocorticoids, 359-360 of lactase phenotypes, 30-3
and lipogenic enzyme expression, of milk consumption, 24-25
211-214 GH3 pituitary cells, 121-123
urea cycle and, 342, 345 Glomerular filtration rate, 355
Fat synthesis, see Lipogenesis Glucagon
Fatty acid oxidation, see also Lipid and fatty acid synthesis, 208
oxidation and lipogenic enzyme expression, 211
68-kDa camitine palmitoyltransferase starvation effects on, 357-358
and, 320-332 Glucagon gene, 105-106
Fatty acid synthase gene, 208-209 model systems, 109-110
cloning, 209-211 nutrient control of expression
differentiation-dependent expression in intestinal, 108
3T3-Ll cells, 216-219 pancreatic, I 06-108
Index 573
Glucocorticoid receptors glutaminase cDNAs. 343

DNA binding, 553 heterogeneity in liver. 343-344
in vivo vitamin 8 6 studies, 558-561 regulation, 344-345
isoelectric point, 557-558 in glutamine metabolism, 337-338
nuclear localization, 553-554 kidney-type, 337-338, 338, 341
sedimentation coefficient, 554-555 regulation. 340-343
trypsin cleavage of, 555-557 Glutamine
vitamin 8 6 modulation, 547-562, see metabolic pathways
also Vitamin 8 6 glutaminase, 337-338
Glucocorticoids glutamine synthetase. 337
calbindin-D gene inhibition, 381 hepatic and urea cycle, 338-340
and G6PD activity, 190, 193-195 sources and roles, 336-337
in lactase activity, 34 Glutamine cycle hypothesis, 344
and L-type PK gene expression, 178 Glutamine synthetase. 337
and osteoblast regulation by vitamin D. GLUT-2 transporter. 146, 147-149
403 Glycerol-3-phosphate acyltransferase, 209
starvation/refeeding and, 359-360 Glycogen phosphorylase. 14
Glucokinase Glycolysis. 211
cloning of G6PD
hepatic cDNAs, 150 dietary effects on, 188-190
islet cDNAs, 150-153 ethanol and, 196-197
function hepatocyte studies, 197-198
in beta cells, 147-148 hormonal regulation, 193-195
in liver, 149 molecular study techniques, 191-193
isoforms, 157-158 nutrient regulation. 195-196
aberrant pyruvate kinases and, 198
in liver and islets, 159-160 rat gene and pseudogene structure,
pituitary, 161-162 198-201
activity differences, 158-159 signals increasing, 190
human liver, 160
G proteins, 10
in starvation/refeeding, 362-364
Growth factors
Glucokinase gene, alternate promoter
epidem1al
regions, 153-157, 162-163
and G6PD activity. 195
Gluconeogenesis, 342, 358, 361
in lactase activity, 34
Glucose
fibroblast, 14
and beta cell glucokinase, 155-156
Growth hormone (somatotropin), 359-360
fatty acid biosynthesis from, 271-272
GRP78 and Ca 2 +, 118, 131-138
and L-type PK gene expression
in diabetic rats, 170-174
in rat hepatocytes, 174-178 H
Glucose-menhaden oil diet in rats,
188-189 ·Half-life of enzymes, 15-16
Glucose-6-phosphate dehydrogenase, see Hemophilia, 468-469
G6PD Hemosiderin, 526
Glucose receptors, 10-11 Hepatic glutaminase, see Glutaminase;
Glucose repression, 9, 354, 369 Glutamine
Glutaminase Hepatic 68kDa camitine
gene expression palmitoyltransferase, 319-332, see
coordinate with urea cycle genes, also Camitine palmitoyltransferase
345-346 HepG2 cells and sequestered Ca2 +, 125
Hexokinase genes, 145-146, see also and L-type PK gene expression

Glucokinase in diabetic rats, 170-174
Hexokinase type IV, see Glucokinase in rat hepatocytes, 174-178
H gene iron response, 539 starvation effects on, 357-359
High carbohydrate diet, 193-195 Insulin-like growth factor I
High protein diets (somatomedin), 360
hepatic effects, 358 Interleukin-1 (IL-l), 539-540
long-term effects, 364 Intestinal calbindin-D gene expression,
urea cycle and, 342 379-382
Hormonal regulation Ionophores, 123-124
of adipocyte differentiation, 219-221, IP3, see Inositol triphosphate
444-447 Iron, see also Ferritin
of apolipoprotein B mRNA editing, in ferritin synthesis
304-308 translation, 528-539,
of calbindin-D gene transcription, 539-540
in bone, 382-383 Iron responsive element-binding protein,
intestinal and renal, 379-381 534
nervous system, 383-384 Isoferritins, see Ferritin
of fatty acid synthesis, 208
of glutaminase activity, 340 K
of G6PD activity, 193-195
of hepatic glucokinase gene Kidneys, see also Renal
transcription, 155-156 calbindin-D gene expression
of lipases, 75 cancer of, 255
of lipogenic enzyme genes in 3T3-Ll L-type pyruvate kinase in, 171-172
cells, 219-221 Kinases
mechanisms, 11-14 gluco-, 145-163, see also Glucokinase
HSY-1 infection, 128 hexo-, 145-146
13-Hydroxyaspartate, 464 liver phosphofructokinase, 208-221,
4-Hydroxy-cournarins, 456-457, 464 see also Liver phosphofructokinase
Hypoglycemia, 359-360 L-type pyruvate kinase gene, 169-184,
see also Pyruvate kinase gene
I
L
Immunologic applications of vitamin D,
418 lac operon, 8
Induction defined, 5 Lactase, see also Lactose
Initiation factors and Ca2+, 127 active sites for, 35
Inositol triphosphate in sequestered Ca 2 + in calcium metabolism, 29
release, 118-121 cell surface processing, 35-38
Insulin cellular and molecular biology
and acetyl CoA decarboxylase activity, glycosylation, processing, and
284-286 transport, 36-38
and adipocyte differentiation, 219-221 molecular structure, 38-40
and fatty acid synthesis, 208 species comparison, 40
and G6PD activity, 190, 193-195 structure and function, 35-36
and hepatic glucokinase gene conclusions, 44-46
transcription, 155-156 genetics and development
and lipogenic enzyme expression, adaptation vs. inheritance, 31-32
211-214 development, 32-33
Index 575
distribution and hormonal effects, Lipogenic enzyme expression, 208-221,

33-34 see also individual enzymes and
regulation of decrease, 40-42 genes
molecular biology of decrease, 44 Lipolysis
molecular biology of persistence, 44 intraduodenal, 60-62
Lactose, see also Lactase intragastric, 60, 61-62
deficiency in humans, 29-31 Liver
intolerance in blood glucose regulation, 144
clinical symptoms, 26-28 cancer of, 252-253, 257-258
enzyme abnormalities in, 43 dietary fat effects, 252-253
genetic hypothesis, 32 fatty acid synthase and p90 mRNAs in,
milk drinking and, 24-25 210-211
mosaicism in, 43 glucokinase cDNAs, 150
nutritional effects, 29 glucokinase function in, 149
treatment, 28 glucokinase isoforms in, 159-160
Lactose absorption test, 27 glutamine utilization and synthesis,
Lactose breath hydrogen test, 27 336-346, see also Glutaminase;
Lethal milk syndrome, 508 Glutamine
Leukocyte differentiation and iron, 540 G6PD activity, 191-193
L gene iron response, 539 lipogenesis in, apo B mRNA editing
Lipases
and. 307-308
biosynthesis, 71-73, 80-81
L-type pyruvate kinase in, 170-183
gastric, 61-62
Liver phosphofructokinase gene, 208-209
mRNA, 79-80
cloning, 209-211
pancreatic, 59-62
differentiation-dependent expression in
adaptation. 63-64
3T3-Ll cells. 216-219
after intraduodenal perfusate or
hormonal regulation
meal intake, 64-65
quantitative factors, 68-71 in vivo. 211-214
regulators of. 73-80 in 3T3-Ll adipocytes, 219-221
triglyceride characteristics and. nutritional regulation in vivo, 211-214
65-68 transcription. 214-216
high fat diet and, 78-79 Low density lipoproteins, apo 8100 and.
transcriptional regulation, 80 298-313. see also Adipocytes
preduodenal, 57-59 L-type pyruvate kinase gene, 169-184,
prepyloric. 63 see also Pyruvate kinase gene
Lipid oxidation Lung cancer, 255
ascorbate and. 50 I
and cellular iron content. 530-531
Lipids, see also Lipases; Triglycerides
M
adaptation to. 62-79
dietary and carcinogenesis. 247-258. Matrix Gla protein. 398-399
see also Dietary fats Menaquinones. 456, 457
digestion, 57-61 Metabolic acidosis. glutaminase activity
Lipogenesis in, 343. 344
acetyl CoA decarboxylase and, Metallotheionein
270-291, see also Acetyl CoA roles of. 509
decarboyxlase transcriptional regulation, 510
and G6PD activity. 194 turnover, 51 0
hepatic and apo B mRNA editing. Metallothionein, dietary regulation,
307-308 510-517
Metal response element-binding protein, Osteocalcin gene, 406-414, see also

518 Vitamin D
Metal responsive elements, 510 Osteoclast formation and bone resorption,
Milk, see also Lactase; Lactose 404-406
consumption and lactose intolerance, Osteopontin, 398
24-25 Osteoporosis, 417-418
Mosaicism in lactose intolerance, 43 Ovarian cancer, 251
mRNAs
apolipoprotein B editing, 298-313 p
and ascorbate/collagen interactions,
497-498 Pancreatic beta cells
colipase, 80 in blood glucose regulation, 144-145
gastrin gene glucokinase cDNAs, 150-153
fasting and decrease, 96-97 glucokinase function in, 147-148
nutrients increasing, 97-98 glucokinase isoforms in, 159-160
lipase, 79-80 heterogeneity, 157
Multiple enzyme specificities of hormonal effects on transcription,
immunolactase, 35 156-157
Multistage model of carcinogenesis, 249 Pancreatic cancer, 251, 254-255
Pancreatic lipases, see Lipases, pancreatic
N Parabrachial nucleus, 12-14
p90 gene, 208-209
cloning, 209-211
Nervous system, calbindin-D gene differentiation-dependent expression in
expression, 383-384 3T3-Ll cells, 216-219
Neural regulatory mechanisms, 11-14 hormonal regulation
Nucleus of tractus solitarius, 12-14 in vivo, 211-214
Nutrient receptors in 3T3-Ll adipocytes, 219-221
conclusions and prospects, 18-19 nutritional regulation in vivo, 211-214
evolution of genetic control transcription, 214-216
mechanisms, 2-4 pH, fecal and lactose intolerance, 27
genetic control elements, 4-6 Phosphoenolpyruvate carboxykinase, 15,
importance, 6-7 342, 344
in mammals, 11-14 Phosphoenolpyruvate carboxylase, 361
temporal and quantitative aspects, Phosphoenolpyruvate decarboxylase,
14-18 447-448
in unicellular organisms Phosphofructase
bacteria, 7-8 liver, 208-221, see also Liver
yeasts, 8-11 phosphofructase
muscle, 219
Phospholipases, see Lipases
0 Phylloquinone, see Vitamin K
Pituitary gland, glucokinase isoforms in,
Omega-3 fatty acids and carcinogenesis, 159-160
255-256 Pituitary hormones, starvation/refeeding
Oncogenes, dietary fat and, 250, 258 and, 359-360
Ornithine decarboxylase, 356 Polypeptide chain elongation and Ca'+,
Osteoblast regulation, 395-404, see also 123, 128-131
Vitamin D Polyunsaturated fatty acids and acetyl
Osteocalcin, 398, 406-409, 419 CoA decarboxylase activity, 287
Index 577
Population studies of lactase deficiency, s

30-31
Preduodenal lipases, see Lipases, Saccharomyces cerevisiae, see also Yeasts
preduodenal glucose repression and, 9-10
Procollagen, 489-490, see also Collagen Saturated fat, see Dietary fat; Lipogenesis;
pathway Triglycerides
Progesterone receptor, 553, 560-561 Secretagogues, starvation effects on,
Proline hydroxylation, 488 357-358
Propeptide region of vitamin K, 470-474 Secretin in lipase regulation, 75
Prostatic cancer, 251 Serine-threonine dehydratase, 367-369
Protein kinase C, II 0 Sl4 gene
Proteins retinoic acid regulation, 437-438
c. 469 cis-acting elements in, 439-440
dietary, see Dietary proteins tissue-specific factors in, 440-444
pancreatic secretory, 78 tissue-specific enhancer, 444-447
processing transcription initiation complex,
Ca 2 + in, 118, 121-128 444-447
pancreatic secretory, 78 T3 regulation of, 434-437
S, 469 Silencer elements, 5
Proto-oncogenes, see Oncogenes 68-kDa hepatic carnitine
Pyridoxal phosphate, see Vitamin B6 palmitoyltransferase, 319-332, see
Pyridoxine, see Vitamin 8 6 also Carnitine palmitoyltransferase
Pyruvate kinase, 170 Skin cancer, 249, 252
L-type, 170, 198 Small intestine, L-type pyruvate kinase in,
R-type, 198 171-172, 174
Pyruvate kinase gene, L-type, 174-183 Somatomedin (insulin-like growth factor
1), 360
Q Somatostatin, 96-97, 102-105
fasting and gene expression, 102-103
Quantitative aspects of gene expression, model systems, 104-105
14-18 nutrient effects on gene expression,
103-104
R Somatotropin (growth hormone), 359-360
Sl4 protein, 369
RAR receptors, 432-433 Starvation, see Fasting; Fasting/refeeding
Receptors, see specific types Steroid hormone receptors, see also
Renal calbindin-D gene expression, Glucocorticoid receptors
379-382 mechanism of action, 549-551
Renal cancer, 255 vitamin 8 6 modulation, 547-562, see
Reticulocytes and Ca2 + , 126 also Vitamin 8 6
Retinoic acid Stress proteins and Ca 2 + , 127-128
calbindin-D gene regulation, 384
cellular proteins regulated by, 43S T
regulatory function overview, 432-434
Retinoic acid receptor types, 432-434, Temporal aspects
see also Retinoic acid of gene expression, 14-18
Retinoic acid response elements, 432-433 of lipase biosynthesis, 72-73
Retinol, see Retinoic acid Thapsigargin, 126
Rickets, 416-417 3T3-F442A cell line adipocyte/Sl4 protein
RXR receptors, 432-433 studies, 434-437
3T3-Ll cell studies, lipogenic enzyme and adipocyte differentiation, 220-221

genes, 216-221 apo B mRNA editing modulation,
Thyroid hormone receptors, isoforms 304-307
found in rodent tissues, 436-437 in fatty acid synthesis, 208
Thyroid hormone response element, and G6PD activity, 195
432-433 and hepatic glucokinase gene
Thyroid hormones, see also Thyroxine, transcription, 155-156
Triiodothyronine (T3) receptor binding characteristics,
and adipocyte differentiation, 220-221 436-437
in apo B mRNA editing modulation, regulatory functions, 433
304-307 in S14 gene expression, 434-437
and L-type PK gene expression, 178 Tryptophan, 361
Thyroxine Tumor necrosis factor, 540
in lactase development, 34 Tumor suppressor genes, dietary fat and,
starvation/refeeding and, 359-360 250, 258
Time, see Temporal aspects Tyrosine aminotransferase, 364-367,
Trans-acting elements, 4-5 548-552
in adipocyte differentation, 447
Transcriptional regulation, see also u
mRNAs
in lactase decline, 42 Urea cycle, 339-341
lipogenic enzyme genes, 214-216 Urea cycle enzyme genes, 345-346
pancreatic lipases, 80 Urea cycle enzymes, 342-343
Transducing mechanisms, 3-4
Transferrin receptor, 538 v
Transglumatinase, 447-448
Translation, Ca 2 + in, 121-128, see also Vagus nerve, 11
Calcium Viral infections and Ca 2 +, 128
Triglycerides, see also Lipases; Vitamin B6
Lipogenesis classical actions, 551-552
and acetyl CoA decarboxylase activity, direct interactions with glucocorticoid
286-287 receptor, 555-558
adaptation to effects on physical properties of steroid
pancreatic, 63-7 3 receptors, 552-555
preduodenal, 62-63 in vivo studies, 558-561
regulation of Vitamin C, 484-502, see also Ascorbate
hormonal, 75 Vitamin D
mechanisms of pancreatic lipase bone physiology and, 393-394
and colipase, 76-80 in vitro studies, 394
nutritional, 73-75 in vivo studies, 393-394
digestion in osteoblast regulation, 394
by pancreatic lipases, 59-60 growth and differentiation, 395,
by preduodenal lipases, 57-59 399-403
sequence of, 60-61 normal sequelae, 395-404
significance of gastric and pancreatic, osteocalcin gene regulation, 406-407
61-62 phenotype suppression hypothesis,
perspectives and general conclusion, 412-415
80-83 sequence organization and, 407-409
S 14 protein and synthesis, 434-437 vitamin D response element activity,
Triiodothyronine (T 3) 409-411
Index 579
in osteoclast differentiation and bone cellular expression systems and,

resorption, 404-406 470-472
structure and general function, 392-393 dual role of, 473-474
summary and conclusions, 419 gene expression and vitamin K-
therapeutic potential, 415-419 dependent proteins, 474-476
Vitamin D-dependent calcium binding types and nomenclature, 456, 457
protein, 378-386, see also
Calbindin-D gene w
Vitamin D responsive elements, 408-415
Vitamin K White adipose tissue, see Adipocytes
carboxylation, 456-459
cell biology of carboxylase, 462-463 y
mechanism, 460-462
relationship to other vitamin K- Yeasts, nutrient receptors in, 7-8
metabolizing enzymes, 459-464
plasma proteins dependent on z
congenital deficiencies, 468-470
genomic structure, 468 Zinc
structure, 464-468 metallothionein and, 511-517
propeptide recognition site other genes regulated by, 527-520
carboxylation substrates and, Zinc and zinc deficiency, 508-520, see
472-473 also Metallothionein

Nutrition and Gene Expression (PDFDrive)

Uploaded by

Copyright:

Available Formats

Nutrition and Gene Expression (PDFDrive)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nutrition and Gene Expression (PDFDrive)

Uploaded by

Copyright:

Available Formats

Nutrition

Carolyn D. Berdanier, Ph.D.

Boca Raton London New York

CRC Press is an imprint of the

Boca Raton Ann Arbor London Tokyo

Reissued 2018 by CRC Press

© 1993 by CRC Press, Inc.

No claim to original U.S. Government works

Library of Congress Cataloging-in-Publication Data

Nutrition and gene expression / editors, Carolyn D. Berdanier, James

A Library of Congress record exists under LC control number: 92016664

ISBN 13: 978-1-315-89601-4 (hbk)

University of Georgia Carolyn D. Berdanier

Carolyn D. Berdanier, Ph.D., is a Professor of Nutrition at The Uni-

James L. Hargrove, Ph.D., is an Associate Professor of Nutrition at

Victoria E. Allgood, Ph.D. Charles 0. Brostrom, Ph.D.

Naima Moustaid, Ph.D. Cynthia Smas, M.S.

J. W. Suttie, Ph.D. Malcolm Watford, Ph.D.

Takehiko Tanaka, M.D., Ph.D.

Index .................................................................... 569

NUTRIENT RECEPTORS AND GENE EXPRESSION

James L. Hargrove and Carolyn D. Berdanier

II. The Evolution of Genetic Control Mechanisms .................... 2

III. Elements of Genetic Control. ...................................... 4

IV. The Importance of Receptors in Nutrient-Gene

V. Nutritional Regulation of Genes in Unicellular

VI. Nutrient Sensors in Mammals ..................................... 11

VII. Temporal and Quantitative Aspects of Gene Expression .......... 14

VIII. Conclusions and Prospects ........................................ 18

The scope of potential nutrient-gene interactions is enormous. More than

II. THE EVOLUTION OF GENETIC

Evolutionary theory postulates that present-day genetic mechanisms are

metabolism, reproduction, motility, and sensation. Because the availability

III. ELEMENTS OF GENETIC CONTROL

Trans-Acting Factors RNA Polymerase II A

IV. THE IMPORTANCE OF RECEPTORS IN

In order to produce specific responses of any kind, nutrients or their

glucokinase. 15 In addition, binding of glucose or its metabolites to plasma

V. NUTRITIONAL REGULATION OF GENES IN

A. NUTRIENT-GENE INTERACTIONS IN BACTERIA

B. NUTRITIONAL REGULATION OF GENES IN YEAST

It will be extremely interesting to learn whether or not mammalian glucose-

VI. NUTRIENT SENSORS IN MAMMALS

Although few nutrient receptors have been characterized in vertebrates,

or mediated by locally secreted hormones. Both pathways are possible, be-

Blood-Borne Sensory Afferent

VII. TEMPORAL AND QUANTITATIVE ASPECTS

Regulatory pathways that require activation of transcription are excellent

production for a longer period by increasing synthesis of phosphoenolpyruvate

The slow time course of enzyme induction can be understood through a

This equation predicts that, at equilibrium, the concentration of a product

AMINO ACIDS Protein Pool

before a change in mRNA concentration can produce an equal change in the

kb gene, which generates a mature transcript of about 2.6 kb. Transcription

VIII. CONCLUSIONS AND PROSPECTS

The energy-yielding macronutrients are the substrates for numerous met-

relationship between structural elements in genes and functional elements in

LACTOSE INTOLERANCE AND REGULATION OF

Robert K. Montgomery, Hans A. Buller, Edmond H. H. M. Rings,