Bacterial Cell To Cell Communication

This page intentionally left blank
Bacteri al Cel l -to-Cel l Communi cati on

Many bacterial diseases are caused by organisms growing together as com-
munities or biofilms. These microorganisms have the capacity to coordin-
ately regulate specific sets of genes by sensing and communicating among
themselves by means of a variety of signals. This book examines the mechan-
isms of quorum sensing and cell-to-cell communication in bacteria and the
roles that these processes play in regulating virulence, bacterial interactions
with host tissues, and microbial development. Recent studies suggest that
microbial cell-to-cell communication plays an important role in the patho-
genesis of a variety of disease processes. Furthermore, some bacterial signal
molecules may possess immunomodulatory activity. Thus, understanding
the mechanisms and outcomes of bacterial cell-to-cell communication has
important implications for appreciating hostpathogen interactions and
ultimately may provide new targets for antimicrobial therapies that block
or interfere with these communication networks.
DONALD R. DEMUTH is Professor in the Department of Periodontics,
Endodontics and Dental Hygiene at the University of Louisville School of
Dentistry.
RI CHARD J . LAMONT is Professor of Oral Biology at the University of Florida.
Published titles
1 Bacterial Adhesion to Host Tissues. Edited by Michael Wilson 0521801079
2 Bacterial Evasion of Host Immune Responses. Edited by Brian Henderson
and Petra Oyston 0521801737
3 Dormancy in Microbial Diseases. Edited by Anthony Coates 0521809401
4 Susceptibility to Infectious Diseases. Edited by Richard Bellamy 0521815258
5 Bacterial Invasion of Host Cells. Edited by Richard Lamont 0521809541
6 Mammalian Host Defense Peptides. Edited by Deirdre Devine and Robert
Hancock 0521822203
7 Bacterial Protein Toxins. Edited by Alistair Lax 052182091X
8 The Dynamic Bacterial Genome. Edited by Peter Mullany 0521821576
9 Salmonella Infections. Edited by Pietro Mastroeni and Duncan Maskell
0521835046
10 The Influence of Cooperative Bacteria on Animal Host Biology. Edited by
Margaret McFall Ngai, Brian Henderson and Edward Ruby 0521834651
Forthcoming titles in the series
Phagocytosis of Bacteria and Bacterial Pathogenicity. Edited by Joel Ernst and
Olle Stendahl 0521845696
Over the past decade, the rapid development of an array of techniques in the fields
of cellular and molecular biology has transformed whole areas of research across
the biological sciences. Microbiology has perhaps been influenced most of all. Our
understanding of microbial diversity and evolutionary biology, and of how
pathogenic bacteria and viruses interact with their animal and plant hosts at the
molecular level, for example, have been revolutionized. Perhaps the most exciting
recent advance in microbiology has been the development of the interface discip-
line of cellular microbiology, a fusion of classical microbiology, microbial molecu-
lar biology, and eukaryotic cellular microbiology. Cellular microbiology is
revealing how pathogenic bacteria interact with host cells in what is turning out
to be a complex evolutionary battle of competing gene products. Molecular and
cellular biology are no longer discrete subject areas but vital tools and an inte-
grated part of current microbiological research. As part of this revolution in
molecular biology, the genomes of a growing number of pathogenic and model
bacteria have been fully sequenced, with immense implications for our future
understanding of microorganisms at the molecular level.
Advances in Molecular and Cellular Microbiology is a series edited by
researchers active in these exciting and rapidly expanding fields. Each volume
will focus on a particular aspect of cellular or molecular microbiology and will
provide an overview of the area, as well as examine current research. This series
will enable graduate students and researchers to keep up with the rapidly
diversifying literature in current microbiological research.
Series Editors
Professor Brian Henderson
University College, London
Professor Michael Wilson
University College, London
Professor Sir Anthony Coates
St Georges Hospital Medical School, London
Professor Michael Curtis
St Bartholemews and Royal London Hospital, London
AMCM
A
D
V
A
N
C
E
S
I
N
M
O
L
E
C
U
L
A
R
A
N
D
C
E
L
L
U
L
A
R
M
I
C
R
O
B
I
O
L
O
G
Y
Bacterial Cell-to-Cell
Communication
Role in Virulence and Pathogenesis
EDITED BY
Donald R. Demuth
University of Louisville School of Dentistry
and
Richard J. Lamont
University of Florida
caxniioci uxiviisir\ iiiss
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, So Paulo
Cambridge University Press
The Edinburgh Building, Cambridge cn: :iu, UK
First published in print format
isnx-:, ,;-c-,::-o,-;
isnx-:, ,;-c-,::-:,,c-
Cambridge University Press 2006
2005
Information on this title: www.cambridg e.org /9780521846387
This publication is in copyright. Subject to statutory exception and to the provision of
relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
isnx-:c c-,::-:,,c-
isnx-:c c-,::-o,-:
Cambridge University Press has no responsibility for the persistence or accuracy of uiis
for external or third-party internet websites referred to in this publication, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
hardback
eBook (EBL)
eBook (EBL)
hardback
Contents
List of Contributors page ix
Preface xiii
1 Quorum sensing and regulation of Pseudomonas aeruginosa
infections
Victoria E. Wagner and Barbara H. Iglewski 1
2 The Pseudomonas aeruginosa quinolone signal
Everett C. Pesci 23
3 Quorum-sensing-mediated regulation of plantbacteria
interactions and Agrobacterium tumefaciens virulence
Catharine E. White and Stephen C. Winans 39
4 Jamming bacterial communications: new strategies
to combat bacterial infections and the development
of biofilms
Michael Givskov and Morten Hentzer 65
5 Quorum-sensing-mediated regulation of biofilm growth
and virulence of Vibrio cholerae
Jun Zhu and John J. Mekalanos 101
6 LuxS in cellular metabolism and cell-to-cell signaling
Kangmin Duan and Michael G. Surette 117
7 LuxS-dependent regulation of Escherichia coli virulence
Marcie B. Clarke and Vanessa Sperandio 151
8 Quorum sensing and cell-to-cell communication in the
dental biofilm
Donald R. Demuth and Richard J. Lamont 175
9 Quorum-sensing-dependent regulation of staphylococcal
virulence and biofilm development
Jeremy M. Yarwood 199
10 Cell-density-dependent regulation of streptococcal
competence
M. Dilani Senadheera, Celine Levesque and Dennis G. Cvitkovitch 233
11 Signaling by a cell-surface-associated signal during
fruiting-body morphogenesis in Myxococcus xanthus
Lotte Sgaard-Andersen 269
Index 301
C
O
N
T
E
N
T
S
viii
Contributors
Marcie B. Clarke
Department of Microbiology
University of Texas Southwestern Medical Center
5323 Harry Hines Blvd.
Dallas, TX 74390, USA
Dennis G. Cvitkovitch
Oral Microbiology
Faculty of Dentistry
University of Toronto
124 Edward St.
Toronto, ON, Canada M5G 1G6
Donald R. Demuth
Center for Oral Health and Systemic Disease
School of Dentistry
University of Louisville
Louisville, KY 40292, USA
Kangmin Duan
Molecular Microbiology Laboratory
Northwest University
Xian, China
Michael Givskov
Centre for Biomedical Microbiology
BioCentrum Technical University of Denmark
Matematiktorvet, Bldg. 301
DK-2800 Kgs. Lyngby, Denmark
Morten Hentzer
Carlsberg Biosector
Carlsberg A/S
Gl. Carlsberg Vej 10
2500 Valby, Denmark
Barbara H. Iglewski
Department of Microbiology and Immunology
Box 672
University of Rochester School of Medicine and Dentistry
Rochester, NY 14642, USA
Richard J. Lamont
Department of Oral Biology
University of Florida
Gainesville, FL 32610, USA
Celine Levesque
Oral Microbiology
124 Edward St.
John J. Mekalanos
Department of Microbiology and Molecular Genetics
Harvard Medical School
200 Longwood Ave
Boston, MA 02115, USA
Everett C. Pesci
The Brody School of Medicine at East Carolina University
600 Moye Blvd
Greenville, NC 27834, USA
M. Dilani Senadheera
Oral Microbiology
124 Edward St.
C
O
N
T
R
I
B
U
T
O
R
S
x
Lotte Sgaard-Andersen
Max-Planck Institute for Terrestrial Microbiology
Karl-von-Frisch Str.
35043 Marburg, Germany
Vanessa Sperandio
University of Texas Southwestern Medical Center
5323 Harry Hines Blvd
Dallas, TX 74390, USA
Michael G. Surette
Department of Biochemistry and Molecular Biology
Faculty of Medicine
University of Calgary
Calgary AB, Canada T2N 4N1
Victoria E. Wagner
Box 672
University of Rochester School of Medicine and Dentistry
Rochester, NY 14642, USA
Catharine E. White
360A Wing Hall
Cornell University
Ithaca, NY 14850, USA
Stephen C. Winans
360A Wing Hall
Cornell University
Ithaca, NY 14850, USA
Jeremy M. Yarwood
3M Corporate Research Materials Laboratory
3M Center, Bldg. 201-E-01
St. Paul, MN 55144, USA
xi
C
O
N
T
R
I
B
U
T
O
R
S
Jun Zhu
University of Pennsylvania School of Medicine
201B Johnson Pavilion
3610 Hamilton Walk
Philadelphia, PA 19104, USA
C
O
N
T
R
I
B
U
T
O
R
S
xii
Preface
It is now well established that a number of bacteria communicate through
diffusible signals that may induce and/or regulate a coordinated response
by the individual organisms that make up a given population or biofilm. For
many of these organisms, it has been suggested that intercellular signaling
functions to report population density or to coordinate a response from all
cells in a microbial community. Therefore, cell-to-cell communication has
been referred to as auto-induction or quorum sensing. The response
of bacteria to quorum sensing signals is quite varied and includes, for
example, the induction of bioluminescence, the regulation of virulence
gene expression, the formation of biofilms, or the induction of horizontal
transfer of genetic material. It is also becoming increasingly apparent that
some bacteria may communicate via contact-dependent signaling mechan-
isms, and that the response to direct cell-to-cell contact influences complex
behaviors that may contribute to multicellular development or the adapta-
tion to growth in complex biofilms. In the past five to ten years, increased
interest and research in the mechanisms of bacterial cell-to-cell commu-
nication has revealed surprising complexity both in the signaling processes
themselves and in the breadth of the response of recipient cells to the signal
molecules. For example, a variety of chemical species, e.g. acyl-homoserine
lactones, oligopeptides, furan derivatives (i.e. AI-2), quinolones, butyrolac-
tones, and unsaturated fatty acids are known or have been suggested to
function as diffusible signals. Furthermore, some organisms, most notably
Pseudomonas aeruginosa and species of Vibrio, have been shown to produce
and respond to multiple diffusible signal molecules. In many cases, the
cellular proteins that are required to synthesize and respond to these
diffusible signal molecules have been identified; high-resolution crystal
structures have been determined for several of these polypeptides. In
addition, the response of several organisms to quorum sensing signals
has been investigated by using microarray technologies; at least in the case
of P. aeruginosa, the expression of hundreds of genes is influenced by
intercellular signaling. Recent research is also beginning to determine
how different signaling pathways are integrated in a given bacterial cell,
and how these cell-to-cell signaling mechanisms are linked to the processes
that control stationary-phase growth in bacteria.
In contrast to the autoinduction or quorum sensing mechanisms
described above, contact-dependent signaling in organisms such as
Myxococcus xanthus or Porphyromonas gingivalis may be initiated by non-
diffusible signals such as cell-surface polypeptides. The C-signal of M.
xanthus is probably the best-characterized cell-surface signal and it plays
an essential role in the starvation-induced development of spore-filled
fruiting bodies by this organism. Organisms that colonize the human
oral cavity may also respond to non-diffusible cell surface signals as well
as to diffusible signals. These mechanisms may facilitate the colonization
process or adaptation of cell to growth in a multispecies biofilm. Thus, the
capacity to respond to non-diffusible cell surface signals may be beneficial
for cells that exist in a complex multicellular community, or for cells that
must adapt to life in an open-flow environment that might limit the
accumulation of a diffusible signal.
In general, the greatly increased interest in bacterial cell-to-cell commu-
nication that has occurred in recent years has resulted in a much clearer
picture of the signaling mechanisms utilized by bacteria and how organ-
isms integrate various input signals to generate a specific coordinated
response as a community. This in turn, has led to efforts to exploit these
communication pathways as targets for the development of new antimicro-
bial therapies. Indeed, chemicals that jam or inhibit bacterial signaling
have the potential to be highly effective anti-biofilm agents that may over-
come the inherent resistance of biofilm-mediated diseases to treatment
with antibiotics.
Our foremost objective in compiling this book is to summarize the rapid
advances that have been made, and are continuing to be made, in the field
of intercellular communication among bacteria. Owing to the broad scope
of new information that is available in this field, we have chosen to target
this book on the biomedical community and to focus primarily on the
mechanisms of cell-to-cell communication in bacteria of medical import-
ance. A recurring theme throughout many of the chapters is the role that
cell-to-cell communication plays in regulating virulence and the develop-
ment of biofilms by these organisms. The contributors to this volume are
leading researchers in the field; each chapter reviews the most recent
P
R
E
F
A
C
E
xiv
findings, controversies, and important new information in the authors
field of expertise. Our goal is to provide the reader with a comprehensive
overview of the various mechanisms and outcomes of bacterial cell-to-cell
communication and an appreciation of how these signaling processes
may be relevant to the development of microbial communities and to
hostpathogen interactions during infection.
We thank the authors for their contributions and for their efforts and
cooperation during the assembly of this book. We also thank Katrina
Halliday of Cambridge University Press for her assistance in nurturing
this project to fruition.
xv
P
R
E
F
A
C
E
CHAPTER 1
Quorumsensing and regulation of Pseudomonas
aeruginosa infections
Victoria E. Wagner and Barbara H. Iglewski
University of Rochester School of Medicine and Dentistry,
Rochester, NY, USA
INTRODUCTION
Pseudomonas aeruginosa is a ubiquitous Gram-negative microorganism
that thrives in many environments, from soil and water to animals and
people. It is an opportunistic pathogen that can cause respiratory infec-
tions, urinary tract infections, gastrointestinal infections, keratitis, otitis
media, and bacteremia. P. aeruginosa is the fourth most common nosoco-
mial pathogen, accounting for approximately 10% of hospital-acquired
infections (www.cdc.gov). Immunocompromised patients, such as those
undergoing cancer treatment or those infected with AIDS, burn patients, or
cystic fibrosis (CF) patients, are susceptible to P. aeruginosa infections.
These infections are difficult to treat by using conventional antibiotic
therapies, and hence result in significant morbidity and mortality in such
patients. The recalcitrant nature of P. aeruginosa infections is thought to be
due to the organisms intrinsic antibiotic resistance mechanisms and its
ability to form communities of bacteria encased in an exopolysaccharide
matrix; such communities are known as biofilms.
P. aeruginosa possesses an impressive arsenal of virulence factors to ini-
tiate infection and persist in the host. These include secreted factors, such
as elastase, proteases, phospholipase C, hydrogen cyanide, exotoxin A, and
exoenzyme S, as well as cell-associated factors, such as lipopolysaccharide
(LPS), flagella, and pili. The expression of these factors is tightly regulated.
Many factors are expressed in a cell-density-dependent manner known as
quorum sensing (QS). Quorum sensing, or cell-to-cell communication,
is a means by which bacteria can monitor cell density and coordinate
population behavior. The behavior was first identified in Vibrio f ischeri
as a mechanism to induce bioluminescence (20). In P. aeruginosa, this
1
Bacterial Cell-to-Cell Communication: Role in Virulence and Pathogenesis, ed. D. R. Demuth and
R. J. Lamont. Published by Cambridge University Press. #Cambridge University Press 2005.
mechanism consists of a regulatory protein and a cognate molecule, termed
an autoinducer. At low cell densities, a small amount of autoinducer is
present in the extracellular milieu. As cell densities increase, the autoindu-
cer concentration rises in the extracellular environment and a threshold
intracellular concentration is exceeded. At this critical concentration, the
autoinducer binds to the regulatory protein and this complex acts to induce
or repress expression of target genes. The cell-density-dependent regula-
tion of virulence factor production has been suggested as a protective
means to prevent host response to invading bacteria before sufficient
bacterial numbers have accumulated (15). P. aeruginosa possesses two well-
studied QS systems, the las and rhl systems, which have been shown to be
important in its pathogenesis. The following sections describe QS in
P. aeruginosa and its contribution to P. aeruginosa virulence, and discuss
how QS may represent an attractive target to develop new antimicrobial
therapies.
P. AERUGINOSA QUORUM SENSING SYSTEMS
In 1991, Gambello and Iglewski reported that a protein encoded by a gene
termed lasR could complement elastase production in an elastase-deficient
P. aeruginosa strain (24). Homology searches showed that LasR belongs to the
LuxR family of QS transcriptional regulators (2). Subsequently, two complete
QS systems, las and rhl, have been identified and well studied in P. aeruginosa.
Each system consists of a transcriptional regulatory protein (LasR in the las
system and RhlR in the rhl system) and cognate autoinducer signal molecules
(N-(3-oxododecanoyl) homoserine lactone (3-oxo-C
12
-HSL) in the las system,
and N-butyryl homoserine lactone (C
4
-HSL) in the rhl system). LasR and
RhlR are members of the LuxR family of transcriptional regulators and
share 31% and 23% identity, respectively, with the transcriptional activator
LuxR in V. fischeri (20, 38). Each protein contains an autoinducer binding site
and a DNA binding region. Both LasR and RhlR form multimers upon
binding to their respective autoinducer (35, 36). This transcriptional regulator
protein : autoinducer complex modulates target gene expression, presumably
by binding to conserved DNA elements, termed las boxes, located upstream of
the translational start site of QS-regulated genes (91). The autoinducers are
synthesized from S-adenolsylmethioinine (SAM), which contributes the
homoserine lactone ring and the available cellular acylacyl carrier protein
(ACP) pool that is used to formthe acyl side chain (31). The lasI gene encodes
the synthase that directs the synthesis of 3-oxo-C
12
-HSL; rhlI encodes the
synthase required for C
4
-HSL production. The acyl side-chain length has
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
2
been demonstrated to be important in the specificity of the autoinducer
molecule for its cognate transcriptional regulatory protein (49). The length
of the acyl side chain is believed to contribute to the differences in transport
of the autoinducer molecules from the cell. Diffusion studies have found
that, whereas C
4
-HSL, with a relatively short acyl side chain, is able to diffuse
freely across the cell membrane, 3-oxo-C
12
-HSL, which has a much longer
acyl side chain and is more hydrophobic, is actively pumped from the cell by
the multidrug efflux pump MexAB-OprM (51).
The las and rhl systems are organized in a hierarchical manner in
which the las system regulates expression of the rhl system (Figure 1.1).
The LasR-3-oxo-C
12
-HSL complex was shown to exert both transcriptional
control, through activation of rhlR transcription, and post-translational
control, thought to be mediated by the competitive binding of 3-oxo-C
12
-
HSL to RhlR when the C
4
-HSL concentration is low, of the rhl system (53).
The las system also regulates production of elastase (lasB), LasA protease
(lasA), exotoxin A (toxA), alkaline proteases (apr), and a type two secretion
pathway (xcpR, xcpP) (3, 23, 24, 48, 50, 68, 86). Expression of rsaL, which
has been demonstrated to repress lasI transcription and alter elastase
production when over expressed on a plasmid in P. aeruginosa, is also
activated by the las QS system (14, 28, 70, 89). The rhl system has been
shown to regulate production of rhamnolipids (rhlA, rhlB), pyocyanin, and
lectin-binding protein (lecA), and to be required for maximal activation of
LasA protease (lasA) and alkaline proteases (apr) (50). The rhl system has
also been reported to regulate expression of the stationary-phase factor RpoS
(rpoS) (37), although a conflicting report stated that rpoS regulates QS (93).
A more recent analysis proposed a complex model of rpoS regulation in
which rpoS can activate the las system, which activates the rhl system, which
then activates rpoS transcription (69).
A third LuxR homolog in P. aeruginosa, with 29% identity to LasR and
32% identity to RhlR, was identified and shown to act as a repressor of
quorum sensing (5). This transcriptional regulator, termed qscR for
quorum-sensing-control repressor, has been shown to repress transcrip-
tion of lasI, rhlI, hcnAB, lasB, pqsH, and two clusters of phenazine genes
(phzA1 and phzA2) (5, 40). QscR has been demonstrated to interact directly
with both LasR and RhlR to form heterodimers in the absence of autoindu-
cer molecules and can associate with 3-oxo-C
12
-HSL and C
4
-HSL (40).
Further study is needed to define the role of QscR in the QS-regulon.
A third signal molecule, 2-heptyl-3-hydroxy-4-quinolone or the
Pseudomonas quinolone signal (PQS), has also been described (54).
Interestingly, the chemical structure of PQS is not an acylated homoserine
3
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
P
.
A
E
R
U
G
I
N
O
S
A
I
N
F
E
C
T
I
O
N
S
lactone molecule but rather a quinolone molecule. This signal has been
proposed to act as a link between the las and rhl systems (45, 54). A recent
study identified several genes, including pqsABCD and pqsH, in P. aeruginosa
PAO1 that are required for PQS biosynthesis (22). LasR is required for
rhl I
rhl R
RhlI RhlR
RhlR
RhlR
Target genes:
rhlAB rhlI lasB
LasI
las I las R
LasR
LasR
3-oxo-C
12
-HSL
3-oxo-C
12
-HSL
3-oxo-C
12
-HSL
3-oxo-C
12
-HSL
LasR
Target genes:
lasB,A lasI toxA
xcpR,P apr rhlR
C
4
-HSL
C
4
-HSL
C
4
-HSL
C
4
-HSL
Figure 1.1. The las and rhl QS systems in P. aeruginosa. Both the las and rhl systems are
composed of a transcriptional regulator protein (LasR, RhlR) and an autoinducer synthase
(LasI, RhlI). The LasI synthase produces 3-oxo-C
12
-HSL, and the RhlI synthase produces
C
4
-HSL. The autoinducer then binds to its respective cognate protein. The LasR:3-oxo-
C
12
-HSL complex activates transcription of genes involved in the production of several
virulence factors, including lasB, aprA, and toxA, as well as upregulating transcription
of lasI and of rhlR. The RhlR:C
4
-HSL complex activates transcription of several genes
involved in virulence, including rhlAB. (see also color plate section)
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
4
PQS synthesis whereas RhlR is required for PQS bioactivity, although the
precise mechanism is not clear (22). Interestingly, the rhl system has been
shown to repress synthesis of PQS (44). The balance of 3-oxo-C
12
-HSL and
C
4
-HSL concentrations appears to play a key role in the production of PQS
(44). A more thorough discussion of PQS and its place in the QS system is
presented elsewhere in this volume (see Chapter 2).
EXTENDING THE P. AERUGINOSA QS REALM
Recent efforts to map the QS regulon have elucidated a large number of
genes that are regulated by the las and rhl QS systems. A recent study used
transposon mutagenesis to identify QS regulated genes (92). A promoterless-
lacZ cassette was inserted randomly into the chromosome of a lasIrhlI
P. aeruginosa PAO1 mutant. The mutant library was screened for lacZ
induction upon exposure to 3-oxo-C
12
-HSL alone, C
4
-HSL alone, or both
autoinducer molecules. Of 7,000 mutants screened, 270 mutants pro-
duced a greater than 2-fold stimulation of b-galactosidase activity in
response to exogenous autoinducer(s). Forty-seven of these mutants that
reproducibly exhibited a greater than 5-fold b-galactosidase activity were
mapped to determine the locus of disruption. A total of 39 unique
genes were identified. These included some previously known QS-regulated
genes, such as rhlB and rhlI, and other genes involved in known
QS-regulated processes, such as phenazine synthesis, cyanide synthesis,
and pyoveridine synthesis. Seven putative operons were discovered; 14 of
the 39 genes identified as QS-regulated possessed putative upstream las-box
sequences. Interestingly, these mutations could be classified into four
distinct classes based upon the timing of activation and requirement for
3-oxo-C
12
-HSL alone, C
4
-HSL alone, or both autoinducer molecules for
maximal induction. The authors hypothesized that based upon their results
approximately 3%4% of P. aeruginosa genes are controlled by quorum
sensing.
The sequencing of the P. aeruginosa PAO1 genome allowed the devel-
opment of P. aeruginosa high-density oligonucleotide microarrays (80). The
advent of these arrays led to subsequent global analyses of QS regulation in
P. aeruginosa PAO1. The ability to probe the expression of all known
annotated genes under various experimental conditions significantly
expanded the QS regulon. One microarray experiment mapped the las
and/or rhl regulon by using RNA isolated from a lasIrhlI mutant grown in
the absence or presence of 3-oxo-C
12
-HSL and C
4
-HSL (89). In this study,
616 genes were identified as QS-regulated (p 0.05 based on three
5
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
P
.
A
E
R
U
G
I
N
O
S
A
I
N
F
E
C
T
I
O
N
S
biological replicates per condition), with 394 genes being QS-activated and
222 genes being QS-repressed (89). These included 32 of the 52 previously
known QS-regulated genes. Many of the 616 genes (34%) encoded proteins
of unclassified or unknown function. The remainders were grouped into 24
of the 26 functional categories used to annotate P. aeruginosa gene products
(www.pseudomonas.com), with genes encoding membrane proteins or
putative enzymes representing the next two largest functional categories
of QS-regulated genes. Not surprisingly, more than 50 of the 616 genes
encoded known or putative virulence factors involved in attachment, colo-
nization, dissemination, and destruction of host tissues. Of note, the
microarrays revealed a link between QS and expression of type III secretion
genes. Only 7% (43 genes) possessed upstream las boxes, suggesting that
many of the genes identified are indirectly QS-regulated. Interestingly, 37
of the 616 QS-regulated genes encode known or putative transcriptional
regulators. Therefore these genes may regulate the expression of many QS-
regulated genes that do not possess upstream las boxes. A recent analysis
has further characterized the recognition sites of LasR:3-oxo-C
12
-HSL bind-
ing and found that this complex is able to recognize and activate genes that
do not possess a putative las box and is unable to activate other genes with
putative las boxes (71). For example, although PA1897 possesses a putative
las box and has previously been shown to require the las system for activa-
tion, the las-dependence activation of PA1897 is apparently due to its direct
regulation by qscR (PA1898), which is regulated by LasR. A more complex
picture of the QS regulon, consisting of multiple layers of regulation,
emerged (Figure 1.2).
Concurrently, another group also performed a transcriptome analysis
to identify QS-regulated genes by using both a lasIrhlI signal mutant
and a lasRrhlR receptor mutant (70). Transcripts were deemed to be
QS-regulated if a fold-change difference of greater than or equal to 2.5,
derived from comparison of transcript levels by using a lasIrhlI mutant
grown in the absence or presence of exogenous autoinducer and the
lasRrhlR mutant as compared with the wild-type, was observed.
Transcripts identified as QS-regulated from both sets of experiments
were compared; those transcripts that overlapped and had consistent
regulatory expression patterns above background level were used to define
a list of QS-regulated genes. By this criterion, 315 genes were identified as
QS-activated and 38 genes were identified as QS-repressed. More than 87
possible operons were discovered. Importantly, a kinetic analysis of QS
regulation suggested that the concentration of LasR was critical in the
timing of activation or repression of QS-regulated genes.
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
6
Although both microarray studies revealed that many genes
(c.6%10% of the genome) representing diverse functional groups in
P. aeruginosa were QS-regulated, there was only approximately 50% agree-
ment between the data sets as to the identity of these genes. Interestingly,
researchers discovered that by altering the medium formulation (i.e. by
using more minimal media or rich media) and oxygen availability used to
cultivate P. aeruginosa PAO1, the detection of QS-regulated transcripts
obtained by microarray analysis varied (89). For example, the level of expres-
sion of lasR and rhlR was lower in the absence of oxygen; this decrease in
expression may have contributed to the inability to detect transcripts for certain
QS-regulated genes under anaerobic conditions (89). Therefore, the discre-
pancies between both studies may reflect not only different microarray analy-
sis tools and criteria to define QS-regulated transcripts but also the different
experimental conditions used by each group.
qscR
(PA1898)
vqsR
(PA2591)
rhlR
(PA3477)
LasR
3-oxo-C
12
-HSL
34 other transcriptional regulators
580 other genes
Figure 1.2. Increasing complexity of the QS regulon. Microarray analysis has revealed
that conservatively more than 600 genes are part of the QS regulon. Of these genes,
three transcriptional regulators, rhlR, vqsR, and qscR, possess putative las boxes that
suggest these genes are directly activated by the LasR:3-oxo-C
12
-HSL complex. These
regulators may regulate expression of other QS-regulated genes that encode
transcriptional regulators and genes that encode proteins representing other
functional categories. The QS regulon is hypothesized to have multiple layers via
several transcriptional regulatory circuits.
7
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
P
.
A
E
R
U
G
I
N
O
S
A
I
N
F
E
C
T
I
O
N
S
In a third microarray analysis of the las and rhl QS regulon, the authors
reported yet another set of QS-regulated genes (28). This study explored the
QS regulon in cells grown planktonically, as the previous studies did, and
also in cells grown as a biofilm. Taking advantage of the previous array
results, the authors compared data sets to form a cohesive list of genes
defined as the general QS regulon. The authors noted that all of these
genes, with the exception of one gene (PA0144), were expressed in
P. aeruginosa biofilms. This is not surprising: it has been demonstrated
that lasI is critical for mature biofilmformation in P. aeruginosa (13). Another
study found that during P. aeruginosa biofilm development in a once-flow-
through biofilm system, transcripts for 56 out of 72 previously identified
QS-regulated virulence genes were detected by using microarray analysis
(90). Of these, 32 genes (57%) respond to 3-oxo-C
12
-HSL alone. Interestingly,
some of these genes were detected only at day 1 (immature biofilm) or only at
day 4 (mature biofilm). These data suggest that different components of the
QS system are important during specific phases of biofilm development and
maintenance. Interestingly, several genes that have been identified as
QS-regulated, including rhlA, rpoS, pslB (PA2232), and PA50575059,
have recently been shown to be important in biofilm development and
maintenance (12, 30, 32, 55).
Recently, a fourth LuxR-transcriptional regulator that has been identi-
fied as QS-regulated by using microarrays (28, 70, 89) was found to be
required for autoinducer synthesis and extracellular virulence factor pro-
duction, as well as full pathogenicity in a Caenorhabditis elegans infection
model (34). The gene, termed vqsR, was discovered by using transposon
mutagenesis screening in the P. aeruginosa CF clinical isolate TB. vqsR
possesses a putative las box upstream of its annotated translational start
site, suggesting that it is directly regulated by the LasR3-oxo-C
12
-HSL
complex (34, 89). Microarray analysis of the vqsR mutant compared with
the wild type revealed that several genes that have been previously identi-
fied as QS-regulated, including rhlA and genes involved in the denitrifica-
tion pathway, are also regulated by vqsR when grown in the presence of
H
2
O
2
or human serum. Interestingly, vqsR regulated expression of several
genes that have been previously determined to be iron-regulated (8, 34).
Iron plays a key role in P. aeruginosa pathogenesis (88). These data suggest
that vqsR may be the link between QS and iron regulation that had been
previously put forward by other studies (2, 92).
Several genes have been identified that affect activation of the QS
regulon. These include the response regulator gacA, the CRP-homolog
vfr, the transcriptional regulator mvaT, the sigma factor rpoN, the stationary
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
8
phase sigma factor rpoS, and relA, which is involved in the stringent
response (1, 17, 64, 69, 87, 93). Polyphosphate kinase (PPK) has been
shown to be required for maximal 3-oxo-C
12
-HSL and C
4
-HSL synthesis
and elastase and rhamnolipid production (63). The regulatory protein
RsmA, has also been demonstrated to influence homoserine lactone auto-
inducer production, rhamnolipid production, and swarming (29).
Undoubtedly, the QS regulon is subject to global regulation at several
levels; this regulation may reflect the importance of tight regulation of
QS in response to environmental cues as well as to the metabolic state of
the cell. Further study is needed to elucidate the influence of these factors
on QS regulation. However, it is clear that more than just cell density
(a quorum) regulates the QS regulon in P. aeruginosa.
INFLUENCE OF QS ON P. AERUGINOSA VIRULENCE
IN PLANT AND ANIMAL MODELS
To determine the role of QS in pathogenesis, several models of infec-
tion have been developed to identify virulence genes in P. aeruginosa in
addition to the use of mammals. These include Arabidopsis thaliana,
Caenorhabditis elegans, Galleria melonella, Drosophila melanogaster, and
Dictyostelium discoideum (10, 19, 26, 33, 42, 43, 46, 56, 59, 60, 61, 62, 81).
These latter systems are attractive because they allow for rapid screening of
many mutants without subjecting large numbers of mammals to infection
and subsequent death. Importantly, data from these experiments have
shown that a number of QS-regulated P. aeruginosa virulence factors appear
to be conserved across host species. In the A. thaliana model, gacA, toxA,
plcS, and dsbA, as well as several genes of unknown function, were found to
be important in plant pathogenesis. The C. elegans model identified genes
that were important in two modes of killing, slow and fast killing. These
two modes were found to be dependent on the strain of P. aeruginosa used
as well as on media conditions. From studies of C. elegans, numerous
virulence genes were identified including gacA, lasR, rhlR, vqsR, pqsR, and
np20, as well as genes with no previously described functions. A paralytic
mode of killing was determined to be due to production of hydrogen
cyanide, a toxic virulence factor that is QS-regulated (21). By using the
D. discoideum model, components of the rhl QS (rhlR, rhlI, and rhlA) in
addition to nfxC were determined to be required for virulence (9). In a
G. melonella model, genes involved in type III secretion, including exoU and
exoT, were found to be important in killing as well as gacA, lasR, and pscD.
A Drosophila melanogaster model revealed that genes required for twitching
9
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
P
.
A
E
R
U
G
I
N
O
S
A
I
N
F
E
C
T
I
O
N
S
motility (pilGHIJKL, chpABCDE), aminoacid, nucleotide, andcentral metabolism
(pyrF, pgm, cca), and a gene of unknown function (PA5441), were important in
killing (11, 57). Both rhlR and algT have been identified as important in an alfalfa
seedling (Medicago sativa) model of infection (72). Many of the genes found to
influence pathogenesis are part of the QS regulon, either as a global regulator of
QS activation (e.g. gacAS), QS transcriptional regulators (e.g. lasR, rhlR, vqsR,
pqsR), or genes involved in QS autoinducer synthesis (e.g. lasI, rhlI, pqsCDEH)
or production of extracellular secreted factors (e.g. hcnA, rhlA) (Table 1.1).
Research using animal models of both acute and chronic infection has
supported the premise that QS significantly contributes to P. aeruginosa
pathogenesis (15, 58, 66, 83) (Table 1.2). In a mouse model of acute
pulmonary infection, lasR, lasI, rhlI, and lasIrhlI mutants were signifi-
cantly attenuated in virulence (52, 83). Analysis of the lasI and rhlI mutants
revealed that a rhlI mutant caused pneumonia in 15%of mice, as opposed to
a lasI mutant, which caused pneumonia in 30% of mice (52). This suggests
that, whereas both lasI and rhlI contribute to infection, rhlI regulates
specific factor(s) that stimulate airway inflammation and resultant pneu-
monia. In addition, pili were demonstrated to contribute to pathogenesis in
the same model of acute infection (82). Several QS genes have also been
demonstrated to be important in the rat model of acute pneumonia. A lasR
mutant was avirulent in the rat model (41). When a lasIrhlI mutant was
Table 1.1. Genes in the QS regulon found to be important in pathogenesis of
P. aeruginosa in various host infection models
Host species Genes References
Arabidopsis thaliana gacA *, toxA 62
Caenorhabditis elegans hcnA, pqsCDEH, pqsR, phnA, lasR *,
rhlR, rhlI*, gacA *, gacS, phzB *,
mexA, vqsR, PA0745, PA3032
10, 22, 34, 42
Dictyostelium
discoideum
rhlR, lasRrhlR, rhlI *,
lasIrhlI *, rhlA, pscJ
9, 59
Drosophila melanogaster relA, pilGHIJKL 11, 19
Galleria mellonella phzB *, gacA*, lasR * 46, 60
Medicago sativa seedling rhlR 72
*Genes also required for virulence in a thermal injury mouse model. Note that
not all genes have been tested in the thermal injury mouse model.
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
10
tested in this model, the mutant produced milder lung pathology, induced
a stronger serum antibody response, and increased interferon gamma
concentrations when compared with the wild-type P. aeruginosa (95). In
addition, tatC, a gene involved in a novel secretory system, the twin-argi-
nine-translocation (TAT) pathway, and that was previously reported to be
QS-repressed, has been shown to be important in a rat lung model (47). An
ambitious high-throughput screening of 7,968 P. aeruginosa mutants in a
rat model of chronic infection identified 214 mutants, representing 148
annotated open reading frames (ORFs), as attenuated in virulence (57).
Among these genes were several previously identified QS-regulated genes,
including metE (PA1927), narK2 (PA3876), modA (PA1863), and a gene
encoding a product of unknown function (PA1874). In a thermal injury
mouse model, lasR, lasI, or rhlI mutants were all less lethal and exhibited a
diminished ability to disseminate from the burn site (65, 66). However, a
double lasIrhlI deficient mutant was the most attenuated in virulence (66).
Several other genes, including gacA and phzB, have also been shown to be
important in the burn model of infection (Table 1.2).
In addition, QS-regulated factors such as elastase and the autoinducer
3-oxo-C
12
-HSL have been shown to have immunomodulatory effects. In a
mouse thermal injury model, a lasIrhlI mutant failed to stimulate produc-
tion of proinflammatory cytokines including interleukin-6 (IL-6), tumor
necrosis factor alpha (TNF-a), and transforming growth factor beta (TGF-b),
suggesting that P. aeruginosa QS plays a role in triggering the produc-
tion of these cytokines in vivo (67). In vitro cultures of human bronchial
epithelial cells exposed to a lasR P. aeruginosa mutant produced lower
amounts of interleukin-8 (IL-8), a potent chemotactic factor, than did
cells exposed to wild-type P. aeruginosa (76, 77). Subsequent studies showed
that purified 3-oxo-C
12
-HSL could stimulate significant IL-8 pro-
duction in human bronchial epithelial cells, whereas other autoinducers
Table 1.2. Genes in the QS regulon that contribute to virulence in animal models
Host Genes References
mouse model of acute
pneumonia lasR, lasI, rhlI, lasIrhlI, pilA 52, 82, 83
rat chronic lung
infection model
lasR, lasIrhlI, tatC, metE, narK2,
modA, PA1874
41, 47, 57, 95
thermal injury model lasR, lasI, rhlI, lasIrhlI, gacA, phzB 42, 65, 66
11
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
P
.
A
E
R
U
G
I
N
O
S
A
I
N
F
E
C
T
I
O
N
S
such as C
6
-HSL and C
4
-HSL did not, and that the stimulation of IL-8
production by 3-oxo-C
12
-HSL was due to induction of NFkB through activa-
tion of extracellular-signal-regulated kinases (ERKs) (76, 77). Purified
3-oxo-C
12
-HSL was also found to induce cyclooxygenase-2 (Cox-2) via the
NFkB pathway in vitro in human lung fibroblasts (78). Cox-2 was shown
to subsequently upregulate production of prostaglandin E
2
(PGE
2
), a potent
inducer of mucus secretion, vasodilation, and edema (78). This link
between 3-oxo-C
12
-HSL and PGE
2
is especially intriguing because
increased concentrations of PGE
2
have been found in P. aeruginosa infec-
tions in CF lungs and burn wounds (76). 3-oxo-C
12
-HSL can also modulate
immune cell responses, such as to activate interferon-gamma (IFN-g) in
T-cells, to inhibit interleukin-12 (IL-12) and TNF-a production in lipopoly-
saccharide (LPS)-stimulated macrophages, to promote immunoglobulin
E (IgE) in interleukin-4 (IL-4)-stimulated peripheral blood mononuclear
cells, and to cause apoptosis in macrophages and neutrophils (4, 76, 77,
84, 85). The immunomodulatory effect of 3-oxo-C
12
-HSL has led to the
hypothesis that QS can control T-cell responses away from a protective
T-helper-1 (Th1) host response, thus promoting P. aeruginosa survival in vivo
(85). Interestingly, a recent report demonstrated that both 3-oxo-C
12
-HSL
and C
4
-HSL can penetrate mammalian cells and activate their respective
transcriptional regulators, LasR and RhlR (94). Using LasR and RhlR
chimeric transcription factors, this study showed that autoinducers are
required for synthesis of functional LasR and RhlR proteins in monkey
kidney COS-1 cells. Owing to the resemblance of the autoinducer binding
domainof the LuxR-family protein, TraR, to a GAF or PAS domainpresent ina
number of mammalian signaling proteins and transcription factors,
the authors speculated that eukaryotic receptor proteins exist that are able to
interact with autoinducers.
QS PLAYS AN ACTIVE ROLE IN VIVO
Recent evidence has suggested that P. aeruginosa QS is active in vivo.
Several studies have investigated the presence of autoinducer molecules
(3-oxo-C
12
-HSL, C
4
-HSL, PQS) and detection of transcripts for known QS
regulators and QS-regulated genes in CF patient samples. Transcripts for
lasR correlated well with the detection of transcripts for lasA, lasB, and toxA
in sputum samples that were obtained from infected CF patients (79). The
highest correlation was found between lasR and lasA detection, followed by
lasR and lasB, then lasR and toxA. These results suggested that lasR may be
coordinately regulating expressionof these factors invivo. Further analysis also
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
12
revealed a relation between lasI expression and lasR, lasA, lasB, and toxA (18).
Because the LasR-3-oxo-C
12
-HSL complex is known to directly activate tran-
scription of lasI in vitro, it was not surprising that the strongest correlation was
between lasRand lasI. Interestingly, a weak but statistically significant correla-
tion between lasR and algD transcription was found that suggested that lasR
might to some extent regulate algD or be activated by a common environ-
mental trigger. AlgD catalyzes the first step in alginate biosynthesis, resulting
in the mucoid phenotype often observed in clinical isolates fromP. aeruginosa-
infected CF patients (16). In a mouse thermal injury model, algDwas required
for full virulence in a mucD PA14 background (97). Recently, algR2 (algQ), a
gene that activates algDexpression in mucoid P. aeruginosa strains, was shown
to negatively regulate the expression of lasR, rhlR, and lasB at the level of
transcription in a mucoid clinical isolate, suggesting a link between alginate
production and the QS regulon (39).
CF sputa have also been reported to contain 3-oxo-C
12
-HSL, C
4
-HSL,
and PQS. Both 3-oxo-C
12
-HSL and C
4
-HSL could be detected at relatively
low concentrations in 78% and 26% of CF sputa samples, respectively (18).
Interestingly, there was no correlation between the accumulation of lasR,
lasI, lasA, and lasB and the amount of autoinducer detected. The ratio of
3-oxo-C
12
-HSL to C
4
-HSL produced in situ in biological CF sputum samples
was found to mimic those ratios found in P. aeruginosa grown as a biofilm
under laboratory conditions (73). The authors proposed that this result
suggested the organisms were growing as biofilms in CF sputa.
Moreover, the knowledge that certain autoinducer ratios may indicate the
growth state of P. aeruginosa in vivo could be used to identify therapeutic
agents that interfere with P. aeruginosa infections. PQS has been detected at
physiologically relevant amounts in the sputum, bronchoalveolar lavage
fluid, and mucopurulent fluid from distal airways of endstage lungs
removed at transplant in CF patients (7). In P. aeruginosa isolates from
infant CF patients, PQS production was elevated compared with laboratory
strains, suggesting that PQS is important early in the establishment of
chronic infections (25).
QS AS A THERAPEUTIC TARGET
The relationship between QS and pathogenesis suggests that interfer-
ence of QS in P. aeruginosa represents an attractive target for treatment of
P. aeruginosa infections (also see Chapter 4). This is especially important as
QS interference does not seek to kill or inhibit microbial growth, which
may lead to less chance of development of antibiotic resistance. Several
13
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
P
.
A
E
R
U
G
I
N
O
S
A
I
N
F
E
C
T
I
O
N
S
means of accomplishing this goal include inhibition of LasR and RhlR
function, disruption of lasR/lasI and/or rhlR/rhlI activation, and interrup-
tion of autoinducer synthesis and activity (77). A naturally occurring com-
pound, a halogenated furanone synthesized by Delisea pulchra, and several
derivatives of this molecule, have been shown to inhibit QS systems in
P. aeruginosa (27, 28, 96). One of these derivatives, C-30, was effective in
decreasing bacterial load in a mouse model of chronic P. aeruginosa lung
infection (96). Microarray analysis revealed that this compound affected
transcription of approximately 80% of genes identified as QS-regulated
(28). A high-throughput study to identify molecules that inhibit QS has
found several candidates that are able to compete with 3-oxo-C
12
-HSL
activation of LasR and have been shown to decrease elastase production
in P. aeruginosa (74, 75). Other mechanisms of QS interference include the
use of autoinducer-specific antibodies to prevent entry into the cell, the use
of lactonases that degrade the homoserine lactone ring of 3-oxo-C
12
-HSL
and C
4
-HSL, the inactivation of genes that are global activators of QS,
such as gacA or vfr, and the use of antisense oligonucleotides that inhibit
translation of lasR, lasI, rhlR, and rhlI transcripts (77). Interestingly, human
las and
rhl QS
systems
Biofilm formation and
maintenance:
lasI
rhlA
PA50575059
rpoS
pslB
Antibiotic sensitivity:
mexAB
lasI
rhlI
Virulence:
lasI lasAB rhlAB aprA
lasR phzB hcnABC
rhlI toxA
rhlR
Stress response:
PA50575059
relA
rpoS
Figure 1.3. Role of QS in regulation of P. aeruginosa. Examples of QS-regulated genes that
have been found to be implicated in regulation of biofilm formation and maintenance,
stress response, antibiotic sensitivity, and virulence are depicted.
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
14
airway epithelial cells have been shown to inactivate 3-oxo-C
12
-HSL (6).
Determination of the identity of this cell-associated factor may prove useful
in designing novel therapies.
P. aeruginosa QS systems have been demonstrated to be important in the
establishment and maintenance of P. aeruginosa infections. Recent studies
have enhanced our current understanding of the las and rhl QS systems and
revealed a more complex picture than previously imagined. QS appears to play
a pivotal role in P. aeruginosa stress response, biofilm development and main-
tenance, antibiotic resistance, and virulence (Figure 1.3). Further investigation
of QS regulation in P. aeruginosa will elucidate the nuances of these systems.
This is especially imperative as it becomes clear that these systems represent a
novel means of therapeutic intervention.
REFERENCES
1 Albus, A., E. Pesci, L. Runyen-Janecky, S. West and B. Iglewski 1997. Vfr
controls quorum sensing in Pseudomonas aeruginosa. J. Bacteriol. 179(12):
392835.
2 Arevalo-Ferro, C., M. Hentzer, G. Reil et al. 2003. Identification of quorum-
sensing regulated proteins in the opportunistic pathogen Pseudomonas aeruginosa
by proteomics. Environ. Microbiol. 5(12): 135069.
3 Chapon-Herve, V., M. Akrim, A. Latifi et al. 1997. Regulation of the xcp secretion
pathway by multiple quorum-sensing modulons in Pseudomonas aeruginosa.
Molec. Microbiol. 24(6): 116978.
4 Chhabra, S. R., C. Harty, D. S. Hooi et al. 2003. Synthetic analogues of the
bacterial signal (quorum sensing) molecule N-(3-oxododecanoyl)-L-homoserine
lactone as immune modulators. J. Med. Chem. 46(1): 97104.
5 Chugani, S. A., M. Whiteley, K. M. Lee et al. 2001. QscR, a modulator of quorum-
sensing signal synthesis and virulence in Pseudomonas aeruginosa. Proc. Natn.
Acad. Sci. USA 98(5): 27527.
6 Chun, C. K., E. A. Ozer, M. J. Welsh, J. Zabner and E. P. Greenberg 2004.
Inactivation of a Pseudomonas aeruginosa quorum-sensing signal by human air-
way epithelia. Proc. Natn. Acad. Sci. USA 101(10): 358790.
7 Collier, D. N., L. Anderson, S. L. McKnight et al. 2002. A bacterial cell
to cell signal in the lungs of cystic fibrosis patients. FEMS Microbiol. Lett.
215(1): 416.
8 Cornelis, P. and S. Aendekerk 2004. A new regulator linking quorum sensing
and iron uptake in Pseudomonas aeruginosa. Microbiology 150(4): 7526.
9 Cosson, P., L. Zulianello, O. Join-Lambert et al. 2002. Pseudomonas aeruginosa
virulence analyzed in a Dictyostelium discoideum host system. J. Bacteriol. 184(11):
302733.
15
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
P
.
A
E
R
U
G
I
N
O
S
A
I
N
F
E
C
T
I
O
N
S
10 Darby, C., C. L. Cosma, J. H. Thomas and C. Manoil 1999. Lethal paralysis of
Caenorhabditis elegans by Pseudomonas aeruginosa. Proc. Natn. Acad. Sci. USA
96(26): 152027.
11 DArgenio, D. A., L. A. Gallagher, C. A. Berg and C. Manoil 2001. Drosophila as
a model host for Pseudomonas aeruginosa infection. J. Bacteriol. 183(4):
146671.
12 Davey, M. E., N. C. Caiazza and G. A. OToole 2003. Rhamnolipid surfactant
production affects biofilm architecture in Pseudomonas aeruginosa PAO1.
J. Bacteriol. 185(3): 102736.
13 Davies, D. G., M. R. Parsek, J. P. Pearson et al. 1998. The involvement of cell-to-
cell signals in the development of a bacterial biofilm. Science 280(5361): 2958.
14 de Kievit, T., P. C. Seed, J. Nezezon, L. Passador and B. H. Iglewski 1999. RsaL, a
novel repressor of virulence gene expression in Pseudomonas aeruginosa.
J. Bacteriol. 181(7): 217584.
15 de Kievit, T. R. and B. H. Iglewski 2000. Bacterial quorum sensing in pathogenic
relationships. Infect Immun 68(9): 483949.
16 Deretic, V., J. F. Gill and A. M. Chakrabarty 1987. Gene algD coding for
GDPmannose dehydrogenase is transcriptionally activated in mucoid
Pseudomonas aeruginosa. J. Bacteriol. 169(1): 3518.
17 Diggle, S. P., K. Winzer, A. Lazdunski, P. Williams and M. Camara 2002.
Advancing the quorum in Pseudomonas aeruginosa: MvaT and the regulation of
N-acylhomoserine lactone production and virulence gene expression. J. Bacteriol.
184(10): 257686.
18 Erickson, D. L., R. Endersby, A. Kirkham et al. 2002. Pseudomonas aeruginosa
quorum-sensing systems may control virulence factor expression in the lungs of
patients with cystic fibrosis. Infect. Immun. 70(4): 178390.
19 Erickson, D. L., J. L. Lines, E. C. Pesci, V. Venturi and D. G. Storey 2004.
Pseudomonas aeruginosa relA contributes to virulence in Drosophila melanogaster.
Infect. Immun. 72(10): 563845.
20 Fuqua, W., S. Winans and E. Greenberg 1994. Quorum sensing in bacteria: the
LuxR\LuxI family of cell density-responsive transcriptional regulators.
J. Bacteriol. 176: 26975.
21 Gallagher, L. A. and C. Manoil 2001. Pseudomonas aeruginosa PAO1 kills
Caenorhabditis elegans by cyanide poisoning. J. Bacteriol. 183(21): 620714.
22 Gallagher, L. A., S. L. McKnight, M. S. Kuznetsova, E. C. Pesci and C. Manoil
2002. Functions required for extracellular quinolone signaling by Pseudomonas
aeruginosa. J. Bacteriol. 184(23): 647280.
23 Gambello, M., S. Kaye and B. Iglewski 1993. LasR of Pseudomonas aeruginosa is a
transcriptional activator of the alkaline protease gene (apr) and an enhancer of
exotoxin A expression. Infect. Immun. 61: 11804.
24 Gambello, M. J. and B. H. Iglewski 1991. Cloning and characterization of the
Pseudomonas aeruginosa lasR gene, a transcriptional activator of elastase expres-
sion. J. Bacterial. 173(9): 30009.
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
16
25 Guina, T., S. O. Purvine, E. C. Yi et al. 2003. Quantitative proteomic analysis
indicates increased synthesis of a quinolone by Pseudomonas aeruginosa isolates
from cystic fibrosis airways. Proc. Natn. Acad. Sci. USA 100(5): 27716.
26 Hendrickson, E. L., J. Plotnikova, S. Mahajan-Miklos, L. G. Rahme and F. M.
Ausubel 2001. Differential roles of the Pseudomonas aeruginosa PA14 rpoN gene
in pathogenicity in plants, nematodes, insects, and mice. J. Bacteriol. 183(24):
712634.
27 Hentzer, M., K. Riedel, T. B. Rasmussen et al. 2002. Inhibition of quorum
sensing in Pseudomonas aeruginosa biofilm bacteria by a halogenated furanone
compound. Microbiology 148(1): 87102.
28 Hentzer, M., H. Wu, J. B. Andersen et al. 2003. Attenuation of Pseudomonas
aeruginosa virulence by quorum sensing inhibitors. EMBO J. 22(15): 380315.
29 Heurlier, K., F. Williams, S. Heeb et al. 2004. Positive control of swarming,
rhamnolipid synthesis, and lipase production by the posttranscriptional RsmA/
RsmZ system in Pseudomonas aeruginosa PAO1. J. Bacteriol. 186(10): 293645.
30 Heydorn, A., B. Ersboll, J. Kato et al. 2002. Statistical analysis of Pseudomonas
aeruginosa biofilm development: impact of mutations in genes involved in twitch-
ing motility, cell-to-cell signaling, and stationary-phase sigma factor expression.
Appl. Environ. Microbiol. 68(4): 200817.
31 Hoang, T. T., S. A. Sullivan, J. K. Cusick and H. P. Schweizer 2002. Beta-ketoacyl
acyl carrier protein reductase (FabG) activity of the fatty acid biosynthetic pathway
is a determining factor of 3-oxo-homoserine lactone acyl chain lengths.
Microbiology 148(12): 384956.
32 Jackson, K. D., M. Starkey, S. Kremer, M. R. Parsek and D. J. Wozniak 2004.
Identification of psl, a locus encoding a potential exopolysaccharide that is essen-
tial for Pseudomonas aeruginosa PAO1 biofilm formation. J. Bacteriol. 186(14):
446675.
33 Jander, G., L. G. Rahme and F. M. Ausubel 2000. Positive correlation between
virulence of Pseudomonas aeruginosa mutants in mice and insects. J. Bacteriol.
182(13): 38435.
34 Juhas, M., L. Wiehlmann, B. Huber et al. 2004. Global regulation of quorum
sensing and virulence by VqsR in Pseudomonas aeruginosa. Microbiology 150(4):
83141.
35 Kiratisin, P., K. D. Tucker and L. Passador 2002. LasR, a transcriptional activator
of Pseudomonas aeruginosa virulence genes, functions as a multimer. J. Bacteriol.
184(17): 491219.
36 Lamb, J. R., H. Patel, T. Montminy, V. E. Wagner and B. H. Iglewski 2003.
Functional domains of the RhlR transcriptional regulator of Pseudomonas aerugi-
nosa. J. Bacteriol. 185(24): 712939.
37 Latifi, A., M. Foglino, K. Tanaka, P. Williams and A. Lazdunski 1996. A hier-
archical quorum-sensing cascade in Pseudomonas aeruginosa links the transcrip-
tional activators LasR and RhlR (VsmR) to expression of the stationary-phase
sigma factor RpoS. Molec. Microbiol. 21: 113746.
17
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
P
.
A
E
R
U
G
I
N
O
S
A
I
N
F
E
C
T
I
O
N
S
38 Latifi, A., M. Winson, M. Foglino et al. 1995. Multiple homologues of LuxRand LuxI
control expression of virulence determinants and secondary metabolites through
quorum sensing in Pseudomonas aeruginosa PAO1. Molec. Microbiol. 17: 33343.
39 Ledgham, F., C. Soscia, A. Chakrabarty, A. Lazdunski and M. Foglino 2003.
Global regulation in Pseudomonas aeruginosa: the regulatory protein AlgR2
(AlgQ) acts as a modulator of quorum sensing. Res. Microbiol. 154(3): 20713.
40 Ledgham, F., I. Ventre, C. Soscia et al. 2003. Interactions of the quorum sensing
regulator QscR: interaction with itself and the other regulators of Pseudomonas
aeruginosa LasR and RhlR. Molec. Microbiol. 48(1): 199210.
41 Lesprit, P., F. Faurisson, O. Join-Lambert et al. 2003. Role of the quorum-sensing
system in experimental pneumonia due to Pseudomonas aeruginosa in rats. Am. J.
Respir. Crit. Care Med. 167(11): 147882.
42 Mahajan-Miklos, S., L. G. Rahme and F. M. Ausubel 2000. Elucidating the
molecular mechanisms of bacterial virulence using non-mammalian hosts.
43 Mahajan-Miklos, S., M. W. Tan, L. G. Rahme and F. M. Ausubel 1999. Molecular
mechanisms of bacterial virulence elucidated using a Pseudomonas aeruginosa-
Caenorhabditis elegans pathogenesis model. Cell 96(1): 4756.
44 McGrath, S., D. S. Wade and E. C. Pesci 2004. Dueling quorum sensing systems
in Pseudomonas aeruginosa control the production of the Pseudomonas quinolone
signal (PQS). FEMS Microbiol. Lett. 230(1): 2734.
45 McKnight, S. L., B. H. Iglewski and E. C. Pesci 2000. The Pseudomonas quino-
lone signal regulates rhl quorum sensing in Pseudomonas aeruginosa. J. Bacteriol.
182(10): 27028.
46 Miyata, S., M. Casey, D. W. Frank, F. M. Ausubel and E. Drenkard 2003. Use of
the Galleria mellonella caterpillar as a model host to study the role of the type III
secretion system in Pseudomonas aeruginosa pathogenesis. Infect. Immun. 71(5):
240413.
47 Ochsner, U. A., A. Snyder, A. I. Vasil and M. L. Vasil 2002. Effects of the twin-
arginine translocase on secretion of virulence factors, stress response, and patho-
genesis. Proc. Natn. Acad. Sci. USA 99(12): 831217.
48 Passador, L., J. Cook, M. Gambello, L. Rust and B. Iglewski 1993. Expression of
Pseudomonas aeruginosa virulence genes requires cell-to-cell communication.
Science 260: 112730.
49 Passador, L., K. Tucker, K. Guertin, M. Journet, A. Kende and B. Iglewski 1996.
Functional analysis of the Pseudomonas aeruginosa autoinducer PAI. J. Bacteriol.
178, 59956000.
50 Pearson, J., E. Pesci and B. Iglewski 1997. Roles of Pseudomonas aeruginosa las
and rhl quorum-sensing systems in control of elastase and rhamnolipid biosyn-
thesis genes. J. Bacteriol. 179(18): 575667.
51 Pearson, J., C. Van Delden and B. Iglewski 1999. Active efflux and diffusion are
involved in transport of Pseudomonas aeruginosa cell-to-cell signals. J. Bacteriol.
181(4): 120310.
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
18
52 Pearson, J. P., M. Feldman, B. H. Iglewski and A. Prince 2000. Pseudomonas
aeruginosa cell-to-cell signaling is required for virulence in a model of acute
pulmonary infection. Infect. Immun. 68(7): 43314.
53 Pesci, E., J. Pearson, P. Seed and B. Iglewski 1997. Regulation of las and rhl
quorum sensing in Pseudomonas aeruginosa. J. Bacteriol. 179(10): 312732.
54 Pesci, E. C., J. B. Milbank, J. P. Pearson et al. 1999. Quinolone signaling in the
cell-to-cell communication system of Pseudomonas aeruginosa. Proc. Natn. Acad.
Sci. USA 96(20): 1122934.
55 Pham, T. H., J. S. Webb and B. H. Rehm 2004. The role of polyhydroxyalkanoate
biosynthesis by Pseudomonas aeruginosa in rhamnolipid and alginate
production as well as stress tolerance and biofilm formation. Microbiology
150(10): 340513.
56 Plotnikova, J. M., L. G. Rahme and F. M. Ausubel 2000. Pathogenesis of the
human opportunistic pathogen Pseudomonas aeruginosa PA14 in Arabidopsis.
Plant Physiol. 124(4): 176674.
57 Potvin, E., D. E. Lehoux, I. Kukavica-Ibrulj et al. 2003. In vivo functional
genomics of Pseudomonas aeruginosa for high-throughput screening of
new virulence factors and antibacterial targets. Environ. Microbiol. 5(12):
1294308.
58 Preston, M. J., P. C. Seed, D. S. Toder et al. 1997. Contribution of proteases and
LasR to the virulence of Pseudomonas aeruginosa during corneal infections. Infect.
Immun. 65(8): 308690.
59 Pukatzki, S., R. H. Kessin and J. J. Mekalanos 2002. The human pathogen
Pseudomonas aeruginosa utilizes conserved virulence pathways to infect the social
amoeba Dictyostelium discoideum. Proc. Natn. Acad. Sci. USA 99(5): 315964.
60 Rahme, L. G., F. M. Ausubel, H. Cao et al. 2000. Plants and animals share
functionally common bacterial virulence factors. Proc. Natn. Acad. Sci. USA
97(16): 881521.
61 Rahme, L. G., E. J. Stevens, S. F. Wolfort et al. 1995. Common virulence factors
for bacterial pathogenicity in plants and animals. Science 268(5219): 1899902.
62 Rahme, L. G., M. W. Tan, L. Le et al. 1997. Use of model plant hosts to identify
Pseudomonas aeruginosa virulence factors. Proc. Natn. Acad. Sci. USA 94(24):
1324550.
63 Rashid, M. H., K. Rumbaugh, L. Passador et al. 2000. Polyphosphate kinase is
essential for biofilm development, quorum sensing, and virulence of
Pseudomonas aeruginosa. Proc. Natn. Acad. Sci. USA 97(17): 963641.
64 Reimmann, C., M. Beyeler, A. Latifi et al. 1997. The global activator GacA of
Pseudomonas aeruginosa PAO positively controls the production of the auto-
inducer N-butyryl-homoserine lactone and the formation of the virulence factors
pyocyanin, cyanide, and lipase. Molec. Microbiol. 24(2): 30919.
65 Rumbaugh, K. P., J. A. Griswold and A. N. Hamood 1999. Contribution of the
regulatory gene lasR to the pathogenesis of Pseudomonas aeruginosa infection of
burned mice. J. Burn Care Rehabil. 20(1): 429.
19
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
P
.
A
E
R
U
G
I
N
O
S
A
I
N
F
E
C
T
I
O
N
S
66 Rumbaugh, K. P., J. A. Griswold, B. H. Iglewski and A. N. Hamood 1999.
Contribution of quorum sensing to the virulence of Pseudomonas aeruginosa in
burn wound infections. Infect. Immun. 67(11): 585462.
67 Rumbaugh, K. P., A. N. Hamood and J. A. Griswold 2004. Cytokine induction
by the P. aeruginosa quorum sensing system during thermal injury. J. Surg. Res.
116(1): 13744.
68 Rust, L., E. C. Pesci and B. H. Iglewski 1996. Analysis of the Pseudomonas
aeruginosa elastase (lasB) regulatory region. J. Bacteriol. 178(4): 113440.
69 Schuster, M., A. C. Hawkins, C. S. Harwood and E. P. Greenberg 2004. The
Pseudomonas aeruginosa RpoS regulon and its relationship to quorum sensing.
70 Schuster, M., C. P. Lostroh, T. Ogi and E. P. Greenberg 2003. Identification,
timing, and signal specificity of Pseudomonas aeruginosa quorum-controlled
genes: a transcriptome analysis. J. Bacteriol. 185(7): 206679.
71 Schuster, M., M. L. Urbanowski and E. P. Greenberg 2004. Promoter specificity
in Pseudomonas aeruginosa quorum sensing revealed by DNA binding of purified
LasR. Proc. Natn. Acad. Sci. USA 101: 158339.
72 Silo-Suh, L., S. J. Suh, P. A. Sokol and D. E. Ohman 2002. A simple alfalfa
seedling infection model for Pseudomonas aeruginosa strains associated with
cystic fibrosis shows AlgT (sigma-22) and RhlR contribute to pathogenesis.
Proc. Natn. Acad. Sci. USA 99(24): 15699704.
73 Singh, P. K., A. L. Schaefer, M. R. Parsek et al. 2000. Quorum-sensing signals
indicate that cystic fibrosis lungs are infected with bacterial biofilms. Nature
407(6805): 7624.
74 Smith, K. M., Y. Bu and H. Suga 2003. Induction and inhibition of Pseudomonas
aeruginosa quorum sensing by synthetic autoinducer analogs. Chem. Biol. 10(1):
819.
75 Smith, K. M., Y. Bu and H. Suga 2003. Library screening for synthetic agonists and
antagonists of a Pseudomonas aeruginosa autoinducer. Chem. Biol. 10(6): 56371.
76 Smith, R. S. and B. H. Iglewski 2003. P. aeruginosa quorum-sensing systems and
virulence. Curr. Opin. Microbiol. 6(1): 5660.
77 Smith, R. S. and B. H. Iglewski 2003. Pseudomonas aeruginosa quorumsensing as
a potential antimicrobial target. J. Clin. Invest. 112(10): 14605.
78 Smith, R. S., R. Kelly, B. H. Iglewski and R. P. Phipps 2002. The Pseudomonas
autoinducer N-(3-oxododecanoyl) homoserine lactone induces cyclooxygenase-2
and prostaglandin E2 production in human lung fibroblasts: implications for
inflammation. J. Immunol. 169(5): 263642.
79 Storey, D., E. Ujack, H. Rabin and I. Mitchell 1998. Pseudomonas aeruginosa lasR
transcription correlates with the transcription of lasA, lasB, and toxA in chronic
lung infections associated with cystic fibrosis. Infect. Immun. 66(6): 25218.
80 Stover, C. K., X. Q. Pham, A. L. Erwin et al. 2000. Complete genome sequence of
Pseudomonas aeruginosa PA01, an opportunistic pathogen. Nature 406(6799):
95964.
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
20
81 Tan, M. W., L. G. Rahme, J. A. Sternberg, R. G. Tompkins and F. M. Ausubel
1999. Pseudomonas aeruginosa killing of Caenorhabditis elegans used to
identify P. aeruginosa virulence factors. Proc. Natn. Acad. Sci. USA 96(5):
240813.
82 Tang, H., M. Kays and A. Prince 1995. Role of Pseudomonas aeruginosa pili in
acute pulmonary infection. Infect. Immun. 63(4): 127885.
83 Tang, H. B., E. DiMango, R. Bryan et al. 1996. Contribution of specific
Pseudomonas aeruginosa virulence factors to pathogenesis of pneumonia in a
neonatal mouse model of infection. Infect. Immun. 64(1): 3743.
84 Tateda, K., Y. Ishii, M. Horikawa et al. 2003. The Pseudomonas aeruginosa auto-
inducer N-3-oxododecanoyl homoserine lactone accelerates apoptosis in macro-
phages and neutrophils. Infect. Immun. 71(10): 578593.
85 Telford, G., D. Wheeler, P. Williams et al. 1998. The Pseudomonas aeruginosa
quorum-sensing signal molecule N-(3-oxododecanoyl)-L-homoserine lactone has
immunomodulatory activity. Infect. Immun. 66(1): 3642.
86 Toder, D., M. Gambello and B. Iglewski 1991. Pseudomonas aeruginosa LasA;
a second elastase gene under transcriptional control of lasR. Molec. Microbiol.
5: 200310.
87 van Delden, C., R. Comte and A. M. Bally 2001. Stringent response activates
quorum sensing and modulates cell density-dependent gene expression in
Pseudomonas aeruginosa. J. Bacteriol. 183(18): 537684.
88 Vasil, M. L. and U. A. Ochsner 1999. The response of Pseudomonas aeruginosa to
iron: genetics, biochemistry and virulence. Molec. Microbiol. 34(3): 399413.
89 Wagner, V. E., D. Bushnell, L. Passador, A. I. Brooks and B. H. Iglewski 2003.
Microarray analysis of Pseudomonas aeruginosa quorum-sensing regulons: effects
of growth phase and environment. J. Bacteriol. 185(7): 208095.
90 Wagner, V. E., R. J. Gillis and B. H. Iglewski 2004. Transcriptome analysis
of quorum-sensing regulation and virulence factor expression in Pseudomonas
aeruginosa. Vaccine 22(suppl. 1): 51520.
91 Whiteley, M. and E. P. Greenberg 2001. Promoter specificity elements in
Pseudomonas aeruginosa quorum-sensing-controlled genes. J. Bacteriol. 183(19):
552934.
92 Whiteley, M., K. M. Lee and E. P. Greenberg 1999. Identification of genes
controlled by quorum sensing in Pseudomonas aeruginosa. Proc. Natn. Acad. Sci.
USA 96(24): 139049.
93 Whiteley, M., M. R. Parsek and E. P. Greenberg 2000. Regulation of quorum
sensing by RpoS in Pseudomonas aeruginosa. J. Bacteriol. 182(15): 435660.
94 Williams, S. C., E. K. Patterson, N. L. Carty et al. 2004. Pseudomonas aeruginosa
autoinducer enters and functions in mammalian cells. J. Bacteriol. 186(8):
22817.
95 Wu, H., Z. Song, M. Givskov et al. 2001. Pseudomonas aeruginosa mutations in lasI
and rhlI quorum sensing systems result in milder chronic lung infection.
Microbiology 147(5): 110513.
21
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
P
.
A
E
R
U
G
I
N
O
S
A
I
N
F
E
C
T
I
O
N
S
96 Wu, H., Z. Song, M. Hentzer et al. 2004. Synthetic furanones inhibit quorum-
sensing and enhance bacterial clearance in Pseudomonas aeruginosa lung infec-
tion in mice. J. Antimicrob. Chemother. 53(6): 105461.
97 Yorgey, P., L. G. Rahme, M. W. Tan and F. M. Ausubel 2001. The roles of mucD
and alginate in the virulence of Pseudomonas aeruginosa in plants, nematodes and
mice. Molec. Microbiol. 41(5): 106376.
V
.
E
.
W
A
G
N
E
R
A
N
D
B
.
H
.
I
G
L
E
W
S
K
I
22
CHAPTER 2
The Pseudomonas aeruginosa quinolone signal
Everett C. Pesci
The Brody School of Medicine at East Carolina University,
Greenville, NC, USA
INTRODUCTION
Pseudomonas aeruginosa is a ubiquitous Gram-negative bacterium that
is a major source of acute infections for immunocompromised individuals.
This opportunistic pathogen also infects the lungs of most cystic fibrosis
(CF) patients, causing a chronic infection that produces progressive lung
damage throughout the life of the patient. P. aeruginosas ability to survive
in almost any surroundings is augmented by an intricate cell-to-cell
signaling scheme that controls a large number of cell functions. Through
our ongoing attempts to eavesdrop on P. aeruginosa, we have learned that
communities of this organism appear to be constantly chattering among
themselves as they adapt to their environment. The las and rhl quorum
sensing systems of P. aeruginosa are acyl-homoserine lactone-based signal
systems that have been well characterized and are nicely reviewed in
Chapter 1 of this book. The focus of this chapter will be a different type of
signal, which has only recently been identified. The signal is 2-heptyl-3-
hydroxy-4-quinolone and is referred to as the Pseudomonas quinolone
signal (PQS). This signal is unique in that it is the only known quinolone
molecule used as a cell-to-cell signal and P. aeruginosa is the only organism
known to produce it.
DISCOVERY OF PQS
PQS was discovered while studying the effects of the rhl quorum
sensing system on lasB induction. The lasB gene encodes LasB elastase,
a protease considered to be a major P. aeruginosa virulence factor (1, 33).
The regulation of this gene is complex, as it is directly controlled in a
23
positive manner by both the las quorum sensing system (LasR and
N-3-oxododecanoyl-L-homoserine lactone (3-oxo-C
12
-HSL)) and the rhl
quorum sensing system [RhlR and N-butyryl-L-homoserine lactone
(C
4
-HSL)] (2, 14, 29) (see Figure 2.1 for signal structures). In a P. aeruginosa
lasR mutant, lasB is not induced because the entire quorum sensing cas-
cade is interrupted (note that the las quorumsensing systeminduces the rhl
quorum sensing system) (17, 31). The addition of C
4
-HSL to a lasR mutant
strain caused a minor induction of lasB; exogenous 3-oxo-C
12
-HSL had no
effect (28). However, when an organic extract of a spent culture supernatant
fromwild-type P. aeruginosa was added to a growing culture of the lasRmutant,
the lasB gene was significantly induced (30). Because this induction of lasB
could not be complemented with exogenous 3-oxo-C
12
-HSL or C
4
-HSL, it was
obvious that a third P. aeruginosa cell-to-cell signal was involved. The
signal was purified from P. aeruginosa; chemical analysis showed that it was
2-heptyl-3-hydroxy-4-quinolone (PQS) (30) (Figure 2.1). This structure was
confirmed by acquiring synthetic PQS and showing that its chemical properties
were identical to those of natural PQS (30). The synthetic compound also
induced lasB in a manner similar to that of natural PQS (30). Thus, a new
bacterial cell-to-cell signal had been identified.
RELATIONSHIP OF PQS TO THE P. AERUGINOSA QUORUM
SENSING CIRCUITRY
The first clues as to the relationship between PQS and the las and rhl
quorum sensing systems surfaced during the initial purification of PQS.
O O
O
N
H
O O
3-oxo-C
12
-HSL
O
N
H
O
C
4
-HSL
N
H
O
OH
PQS
Figure 2.1. Structures of the three Pseudomonas aeruginosa cell-to-cell signals.
E
.
C
.
P
E
S
C
I
24
PQS was not produced at a detectable level by a lasR mutant, indicating that
the las quorum sensing system directly or indirectly controlled a factor
necessary for PQS production (30). It was also noted that PQS could not
induce lasB in a strain that did not produce RhlR (30). Although this made it
appear as though RhlR and PQS could be an effector protein and signal team,
this has not been proven. An attempt to develop an Escherichia coli bioassay in
which RhlR was expressed in the presence of a lasB
0
lacZ reporter fusion and
exogenous PQS failed (30). In addition, radiolabelled PQS did not specifically
associate with cells expressing RhlR or LasR (E. Pesci, unpublished data) as is
the case for C
4
-HSL and 3-oxo-C
12
-HSL, respectively (26, 29). Therefore, the
specific target for PQS has remained elusive.
The link between quorum sensing systems and PQS grew stronger
with the finding that PQS could induce rhlI, the gene that encodes the
C
4
-HSL synthase (25). PQS had a smaller positive effect on lasR and rhlR
expression and no effect on lasI expression (25). Together, PQS and C
4
-HSL
were able to restore the expression of rhlI in a lasR mutant to a level similar
to that seen in the wild-type strain (25). This indicated that PQS and C
4
-HSL
have an additive effect on rhlI expression. It also suggests that PQS could be
acting as a linker between the las and rhl quorum sensing systems. Overall,
the available data showed that the las quorum sensing system must induce
a gene (or genes) required for PQS synthesis, and that PQS has a positive
effect on the rhl quorum sensing system. The link implied from this is
complicated because it has been shown that rhlI is expressed at a normal
level in a mutant that does not produce PQS (13). This leads to the speculation
that rhlI is controlled in multiple ways to ensure that this gene is induced
when needed.
THE GENETICS OF PQS SYNTHESIS
Additional evidence for quorum sensing control of PQS production
began to accumulate with the identification of genes required for PQS
synthesis. Because PQS production required an active las quorum sensing
system, a logical place to search for PQS biosynthetic genes was in a pool of
mutants harboring insertions in quorum-sensing-controlled genes (34).
This search proved fruitful with the discovery of a 3-oxo-C
12
-HSL-controlled
monooxygenase homolog, encoded by gene PA2587 (pqsH), that was required
for PQS production (B. Bullman, W. Calfee, and E. Pesci, unpublished
data). The function of PqsH has not been determined, but it is most likely
adding the hydroxy group to 2-heptyl-4-quinolone in the final step of PQS
synthesis (see Figure 2.2). Some evidence for this role was gathered by two
25
T
H
E
P
.
A
E
R
U
G
I
N
O
S
A
Q
U
I
N
O
L
O
N
E
S
I
G
N
A
L
p
q
s
A
p
q
s
C
p
q
s
B
p
q
s
D
p
q
s
E
p
q
s
R
p
h
n
A
p
h
n
B
C
O
O
H
c
h
o
r
i
s
m
a
t
e
C
H
2
O
C
O
O
H
C
P
h
n
A
B
A
C
P
C
o
A
C C
H
2
C

O
O
S
C
H
3
C
H
3
C C
H
2
C
O
O
H
O
-
k
e
t
o
-
d
e
c
a
n
o
i
c

a
c
i
d
(
C
H
2
)
6
(
C
H
2
)
6
C C
H
2
O
S C
H
3
(
C
H
2
)
6
P
q
s
B
,

P
q
s
C
,

o
r

P
q
s
D

P
q
s
A
?
N H O
N H O
?
P
q
s
H
P
Q
S
C
O
2
p
q
s
H
N
H
2
C
O
O
H
a
n
t
h
r
a
n
i
l
a
t
e
N
H
2
C
C
o
A
O
S
P
q
s
A
?
O
H
C

O
F
i
g
u
r
e
2
.
2
.
P
r
o
p
o
s
e
d
s
y
n
t
h
e
s
i
s
s
c
h
e
m
e
f
o
r
P
Q
S
.
E
n
z
y
m
a
t
i
c
a
c
t
i
v
i
t
i
e
s
f
o
r
p
r
o
t
e
i
n
s
e
n
c
o
d
e
d
b
y
t
h
e
P
Q
S
s
y
n
t
h
e
t
i
c
r
e
g
i
o
n
h
a
v
e
n
o
t
b
e
e
n
p
r
o
v
e
n
,
b
u
t
a
p
o
s
s
i
b
l
e
s
y
n
t
h
e
s
i
s
s
c
h
e
m
e
c
a
n
b
e
p
r
o
p
o
s
e
d
b
a
s
e
d
o
n
h
o
m
o
l
o
g
i
e
s
t
o
k
n
o
w
n
e
n
z
y
m
e
s
.
I
n
t
h
i
s
m
o
d
e
l
,
t
h
e
p
u
t
a
t
i
v
e
c
o
e
n
z
y
m
e
A
(
C
o
A
)
l
i
g
a
s
e
,
P
q
s
A
,
l
i
n
k
s
C
o
A
t
o
e
i
t
h
e
r
a
t
e
n
-
c
a
r
b
o
n
a
c
y
l
c
h
a
i
n
o
r
a
n
t
h
r
a
n
i
l
a
t
e
.
T
h
e
q
u
e
s
t
i
o
n
m
a
r
k
s
b
y
P
q
s
A
i
n
d
i
c
a
t
e
t
h
e
p
o
s
s
i
b
i
l
i
t
y
o
f
t
h
e
e
n
z
y
m
e
s
a
c
t
i
n
g
i
n
o
n
e
r
e
a
c
t
i
o
n
o
r
t
h
e
o
t
h
e
r
.
T
h
e
a
c
y
l
c
h
a
i
n
i
s
t
h
e
n
t
r
a
n
s
f
e
r
r
e
d
t
o
a
n
a
c
y
l
c
a
r
r
i
e
r
p
r
o
t
e
i
n
(
A
C
P
)
t
h
r
o
u
g
h
t
h
e
a
c
t
i
o
n
o
f
o
n
e
o
f
t
h
e
t
h
r
e
e
p
u
t
a
t
i
v
e
a
c
y
l
c
a
r
r
i
e
r
p
r
o
t
e
i
n
s
y
n
t
h
a
s
e
s
e
n
c
o
d
e
d
b
y
p
q
s
B
,
p
q
s
C
,
a
n
d
p
q
s
D
.
T
h
e
s
e
d
i
f
f
e
r
e
n
t
p
u
t
a
t
i
v
e
A
C
P
s
y
n
t
h
a
s
e
s
c
o
u
l
d
b
e
r
e
s
p
o
n
s
i
b
l
e
f
o
r
t
h
e
g
r
e
a
t
v
a
r
i
a
b
i
l
i
t
y
s
e
e
n
i
n
2
-
a
l
k
y
l
-
4
-
q
u
i
n
o
l
o
n
e
s
.
W
h
i
l
e
t
h
i
s
i
s
h
a
p
p
e
n
i
n
g
,
c
h
o
r
i
s
m
a
t
e
i
s
c
o
n
v
e
r
t
e
d
t
o
a
n
t
h
r
a
n
i
l
a
t
e
t
h
r
o
u
g
h
t
h
e
a
c
t
i
o
n
o
f
t
h
e
p
u
t
a
t
i
v
e
a
n
t
h
r
a
n
i
l
a
t
e
s
y
n
t
h
a
s
e
P
h
n
A
B
.
A
n
t
h
r
a
n
i
l
a
t
e
i
s
t
h
e
n
c
h
a
r
g
e
d
w
i
t
h
C
o
A
,
p
o
s
s
i
b
l
y
b
y
P
q
s
A
a
s
m
e
n
t
i
o
n
e
d
a
b
o
v
e
.
A
n
t
h
r
a
n
i
l
a
t
e
-
C
o
A
w
o
u
l
d
t
h
e
n
b
e
c
o
n
d
e
n
s
e
d
w
i
t
h
t
h
e
b
-
k
e
t
o
-
d
e
c
a
n
o
i
c
a
c
i
d
a
t
t
a
c
h
e
d
t
o
t
h
e
A
C
P
a
n
d
t
h
e
d
i
r
e
c
t
P
Q
S
p
r
e
c
u
r
s
o
r
w
o
u
l
d
r
e
s
u
l
t
.
T
h
e
q
u
e
s
t
i
o
n
m
a
r
k
b
y
t
h
i
s
r
e
a
c
t
i
o
n
i
n
d
i
c
a
t
e
s
t
h
a
t
t
h
e
c
o
n
d
e
n
s
a
t
i
o
n
o
c
c
u
r
s
v
i
a
t
h
e
a
c
t
i
o
n
o
f
a
n
u
n
i
d
e
n
t
i
f
i
e
d
e
n
z
y
m
e
.
T
h
e
p
u
t
a
t
i
v
e
m
o
n
o
o
x
y
g
e
n
a
s
e
e
n
c
o
d
e
d
b
y
p
q
s
H
w
o
u
l
d
t
h
e
n
a
d
d
t
h
e
h
y
d
r
o
x
y
g
r
o
u
p
t
o
t
h
e
t
h
i
r
d
c
a
r
b
o
n
o
f
t
h
e
p
r
e
c
u
r
s
o
r
t
o
p
r
o
d
u
c
e
P
Q
S
.
separate lines of research. First, it was determined through an agar plate
cross-feeding assay that a pqsHmutant can feed a pqsC or pqsDmutant with a
compound that allows the production of PQS (13). Because these strains are
most likely affected at an early step of PQS production, the pqsH mutant
must be able to make a compound that is close to, or the direct precursor
of, PQS. Second, radiolabeled 2-heptyl-4-quinolone is converted to PQS by
a wild-type strain (9). This result suggested that this compound is the
immediate precursor of PQS and implied that a monooxygenase should be
responsible for the conversion of the precursor into PQS.
The main PQS biosynthetic gene cluster was then discovered because
of the involvement of anthranilate in PQS synthesis. The basic synthetic
pathway for quinolone compounds has been elucidated (6). This pathway
suggested that anthranilate would be a precursor for PQS synthesis and
this was shown to be the case when radiolabeled anthranilate was converted
into PQS (3). P. aeruginosa encodes at least four anthranilate synthases, one
of which is PhnAB. The phnA and phnB genes putatively encode the large
and small subunit, respectively, of anthranilate synthase (11). PhnAB was
initially believed to be directly involved in the synthesis of the blue phena-
zine compound pyocyanin (11). This role was made less likely with the
discovery that P. aeruginosa encodes two separate, duplicate phenazine
synthetic gene clusters, each with its own anthranilate synthase homolog (23).
A survey of mutants that contained insertions in and around the phnAB region
showed that this location was required for PQS synthesis (13). These mutants
were originally pooled because they did not produce pyocyanin, indicating
that PQS controls pyocyanin production (13). This also implies that
the regulation of pyocyanin by phnAB happens indirectly through PQS.
The genes of this region that were identified as being required for
PQS production were: PA0998 and PA0999, both of which are b-keto-
acylacyl carrier protein synthase homologs; PA1001 (phnA); and PA1003,
which is homologous to members of the LysR-type regulator family (13).
This genetic region is summarized in Figure 2.2. This study also showed
that the monooxygenase homolog encoded by pqsH (gene PA2587)
was required for PQS synthesis (13). At the same time, a parallel study led
to the identification of this region as being important for the autolytic
phenotype exhibited by older, plate-grown cultures of P. aeruginosa (7). It
was shown that the phenotype of a mutant that produced colonies with
concentric zones of lysis could be suppressed by the mutation of genes
within the PQS synthetic region (7). Included in the genes that affected
autolysis were: gene PA0996, which encodes a hydroxybenzoate CoA ligase
homolog; gene PA0997, which encodes a beta-keto-acylacyl-carrier protein
E
.
C
.
P
E
S
C
I
28
synthase homolog; PA0998, PA0999, PA1003, and PA2587 (the latter four
genes were identified above) (7). All of these genes were found to be neces-
sary for PQS production (7). To simplify the nomenclature of these genes,
genes PA0996 through PA1000 were named pqsABCDE (7, 13). Gene
PA1003 was previously published as mvf R (4), and was renamed pqsR
because of its location in the PQS synthetic region and its role in regulating
PQS production (discussed below) (7, 13). Gene PA2587 was subsequently
named pqsH (7, 13). In addition, gene PA4444, which was responsible for
the autolytic phenotype discussed above and encodes a homolog of a mono-
oxygenase (like pqsH), was named pqsL (7). The loss of this gene actually
causes PQS production to be increased 4-fold (7) and it has been shown that
pqsH and pqsL may be competing to convert 2-heptyl-4-quinolone to PQS or
an alternate compound (21).
Overall, this group of genes appears to encode much of what is required
for the synthesis of PQS from basic precursor molecules. For the sake
of caution, it must be pointed out that this statement is based only on
homology, and true enzymatic roles for the encoded proteins have not been
determined. Our proposed synthesis scheme is included in Figure 2.2.
THE REGULATION OF PQS SYNTHESIS
Mutant complementation studies and reverse transcriptase PCR
experiments showed that pqsABCDE and phnAB are transcribed as two
separate polycistronic operons (13, 24). As mentioned above, insertions in
pqsA, pqsB, pqsC, pqsD, or phnA all lead to the loss of PQS (7, 13).
The necessity of the phnB gene for PQS synthesis has not been determined;
a mutant with an insertion in pqsE, which encodes a hypothetical protein,
still makes a normal amount of PQS (13). Most interestingly, the pqsE
mutant was still defective in pyocyanin production (13). This mutant
was also found to be unable to respond to exogenous PQS, indicating that
PqsE is important for the cell to respond to PQS (13). Further studies are
necessary to determine the role of pqsE in the cellular response to PQS.
Preliminary studies on the expression of the pqsABCDE operon have
shown that it is regulated in a very complex manner. A pqsA
0
lacZ fusion
was not induced in either a lasI or a pqsR mutant, indicating that pqsABCDE
expression is positively controlled by the las quorum sensing system and
PqsR (24). The production of PQS was greatly reduced in these mutant
strains, thereby confirming the reporter fusion data (24). The reliance of
pqsA on a functional pqsR gene was also confirmed in a separate study (9).
Most interestingly, when the pqsA
0
lacZ reporter was placed in a rhlI
29
T
H
E
P
.
A
E
R
U
G
I
N
O
S
A
Q
U
I
N
O
L
O
N
E
S
I
G
N
A
L
mutant, it was induced to a level that was greater than 7-fold higher than
that seen in the wild type strain (24). PQS production was also increased
6-fold in this strain (24). These results indicated that pqsABCDE is regulated
in a negative manner by the rhl quorum sensing system. Whether these
regulation events are the result of direct or indirect interactions of LasR,
RhlR, and/or PqsR with the promoter region of pqsA is not known. It is
worthy of note that the pqsA promoter region has two sequences that are
similar to quorumsensing operator sequences, which might allowfor dual or
alternate binding of both LasR and RhlR (24). Although the importance of
these sequences has not been directly investigated, there are data that suggest
that some type of competition is occurring at the pqsApromoter. The addition
of exogenous C
4
-HSL was able to inhibit the induction of pqsABCDE by
exogenous 3-oxo-C
12
-HSL in a mutant that produces neither signal (24).
In addition, the production of PQS, which was induced by 3-oxo-C
12
-HSL,
was inhibited by C
4
-HSL (24). Taken together, these data imply that
a competition between C
4
-HSL and 3-oxo-C
12
-HSL occurs during the
induction of pqsABCDE. A model for the proposed regulation scheme for
PQS is presented in Figure 2.3.
The second operon of the PQS synthetic region is also tightly
controlled. The expression of phnAB has been shown to be directly
regulated by PqsR. A pqsR mutant does not express phnAB and PqsR was
shown to bind specifically to the phnAB promoter region (4). PqsR also
appeared to have no effect on the expression of lasR or rhlR (4). Overall, the
available data suggest that the main role of PqsR is to act as a master
regulator for the two operons of the PQS synthetic region to which it is
adjacent.
Finally, the regulation of PqsR is also quite interesting. It has been
found that this regulator is inactivated through cleavage. This regulatory
event was discovered when cells expressing a PqsRGST fusion protein
were exposed for a short time to sterile, spent supernatant from a stationary-
phase culture. The spent supernatant caused PqsR to be cleaved at amino
acid 147 (total length is 332 amino acids) (4). This effect was found to
be dependent on the bacterial growth phase: spent supernatant from
log-phase cultures did not cause PqsR cleavage (4). This is apparently an
autoregulation event, because spent supernatant from a pqsR mutant had
no effect on cleavage (4). These data indicate that PqsR is negatively auto-
regulated by a stationary-phase factor responsible for site-specific cleavage of
the regulator. Overall, the regulation of PQS production is both interesting
and complex; further studies are necessary to fully understand PQS
production.
E
.
C
.
P
E
S
C
I
30
THE TIMING OF PQS PRODUCTION
One of the unique features of PQS, compared with other quorum
sensing signals, is the timing of its production. Unlike 3-oxo-C
12
-HSL and
C
4
-HSL, the major production of which is started during the mid to late log
phase of growth (27, 28), PQS production occurs at a later stage of growth.
McKnight et al. (25) showed that PQS production began during the end of
log-phase growth, and greatly increased during the stationary phase of
growth. The production was at a maximal level when cultures had grown
pqsABCDE phnAB pqsR
LasR
3-oxo-C
12
-HSL
pqsH
C
4
-HSL
RhlR
PqsR
N
H
O
OH
+
+
+
+
+
lasB
pyocyanin
hcnABC
lecA
rhamnolipid
RpoS
others?
+
PQS
+
+
Figure 2.3. Proposed model for the regulation of PQS production. The las and rhl quorum
sensing systems have a positive and negative effect, respectively, on the pqsABCDE operon.
Whether these are direct or indirect effects is not clear. The las quorumsensing systemalso
positively affects the transcription of pqsH. PqsR is required for the production of
pqsABCDE; this regulator also directly interacts with the phnAB promoter to induce
transcription. It is possible that the effects of the las and rhl quorum sensing systems are
being directed through pqsR. This is currently being determined. PqsR is autoregulated
either directly or indirectly through site-specific cleavage as indicated by the split protein.
Together, pqsH, pqsABCDE, and phnAB produce enzymes that are responsible for the
synthesis of PQS. PQS has positive effects on several genes or products including the rhl
quorum sensing system. The mechanism of action for PQS is presumably through an
unidentified regulatory protein.
31
T
H
E
P
.
A
E
R
U
G
I
N
O
S
A
Q
U
I
N
O
L
O
N
E
S
I
G
N
A
L
for 3042 h and then greatly decreased after 48 h of growth (25). These
cultures were grown in a rich medium, PTSB, and PQS was indirectly
detected by using a P. aeruginosa bioassay. Most interestingly, PQS was
produced in this manner despite the use of a P. aeruginosa strain that
contained a constitutively expressed, mutated LasR protein which induced
las quorum-sensing-controlled genes in the absence of 3-oxo-C
12
-HSL (30).
This indicated that PQS production required a stationary-phase factor that
was not regulated by the las quorum sensing system (25). Not surprisingly,
this expression pattern is similar to that of phnAB, which is also maximally
expressed during stationary phase with a large decrease in expression
thereafter (4). This decrease in phnAB expression, and thus in PQS pro-
duction, is most likely due to the stationary-phase-dependent cleavage of
PqsR (4). In addition to these data, two other studies (10, 20) have
shown that PQS production starts during the late log phase of growth,
becomes maximal during the stationary phase of growth, and decreases
after reaching a peak concentration. Although PQS production was found
to be slightly earlier in the studies of Lepine et al. (20), their more sensitive
liquid chromatography/mass spectrometry detection assay and the use
of LB media still showed a stationary-phase peak concentration of PQS
that then declined. Most interestingly, this study found that a similar
P. aeruginosa secondary metabolite, 2-heptyl-4-hydroxyquinoline N-oxide,
was produced in a similar time frame but its concentration did not decrease
after reaching a maximal level (20). This suggests that a specific
PQS alteration may occur. Finally, Diggle et al. (10) used a thin layer
chromatography detection assay to confirm that PQS production in LB
media begins in the late log phase of growth and becomes maximal later
during the stationary phase of growth. Their growth-curve study was
stopped before PQS concentrations would have begun to decrease (10).
The studies of Diggle et al. (10) also showed that the addition of exogenous
PQS would induce lecA expression by P. aeruginosa. However, unlike other
acyl homoserine lactone signals, PQS did not advance the timing of a gene
that it controlled (lecA) (10). (Note: lecA encodes for the virulence deter-
minant PA-IL lectin.) Once again, it is apparent that PQS activity relies on a
stationary-phase factor that has not yet been identified.
Overall, the later timing of PQS production compared with other
cell-to-cell signals suggests a role in responding to stationary-phase growth.
P. aeruginosa has devoted numerous control mechanisms to ensure that
cell-to-cell signaling occurs at the proper time (see (10) for a summary of
quorum sensing timing), thereby leading one to conclude that, when it
comes to cell-to-cell signaling, timing is everything.
E
.
C
.
P
E
S
C
I
32
PQS AND P. AERUGINOSA VIRULENCE
Many cell-to-cell signals have been shown to be important for virulence
(see (8) for review); PQS is not an exception. Strains that contain mutations
within genes required for PQS production have been identified in many
virulence gene identification experiments. PQS mutants are greatly
reduced in the ability to produce paralytic killing in the nematode
Caenorhabditis elegans (12, 13, 22). This was determined to be caused
by a lack of hydrogen cyanide production, owing to the positive control
that PQS exerts over the hcnABC operon (12, 13). A pqsR mutant
was also unable to produce rotting in a lettuce leaf virulence assay
and was greatly reduced in virulence for the burned mouse model of
P. aeruginosa infection (32). In addition, this strain was found to be less
virulent when tested in both an acute and a chronic model of mouse lung
infection (18).
The importance of PQS exhibited in the virulence assays described
above apparently stems from its ability to control the production of several
virulence factors. In addition to hydrogen cyanide, which was mentioned
previously, PQS controls the production of LasB elastase (3, 10, 25, 30)
and pyocyanin (4, 12, 13, 32). Pyocyanin is important in P. aeruginosa
pathogenesis (18) in that, when introduced into mouse lungs, it caused a
large influx of neutrophils. However, the virulence of a pqsR mutant could
not be complemented by the addition of exogenous pyocyanin, indicating
that PQS regulates multiple important virulence factors (18). To this
end, PQS has also been shown to control the production of hemolytic
activity (32), PA-IL lectin (10), rhamnolipid (10, 16), Rhl R (10), Rpo S (10),
and C
4
-HSL (10, 25). It has been proposed that PQS acts directly or
indirectly through RhlR (30) and controls primarily rhl quorum-sensing-
regulated genes (10), but this theory has not yet been tested in a microarray
experiment.
At the bacterial community level, PQS is known to have an effect on
the growth of P. aeruginosa on stainless steel coupons (10); the overpro-
duction of PQS causes autolysis, possibly through the activation of
lysogenic phage or bacteriocin (7). These effects imply that PQS is likely
to be involved in the formation of P. aeruginosa communities such as
biofilms.
With regard to human infections, PQS has been found in sputum
samples from cystic fibrosis patients infected with P. aeruginosa (5).
The amount of PQS in sputum correlated nicely with the density of
P. aeruginosa in the sample (5). PQS was also seen in bronchial alveolar
33
T
H
E
P
.
A
E
R
U
G
I
N
O
S
A
Q
U
I
N
O
L
O
N
E
S
I
G
N
A
L
lavage fluid from infected CF patients and in mucopurulent airway fluid
from a lung resected from a CF transplant patient (5). Not surprisingly, the
majority (9 out of 10) of P. aeruginosa isolates from CF patients produced
PQS in laboratory cultures (5). One study has also shown that 4 out of
5 P. aeruginosa CF strains actually produced more PQS than the laboratory
strain PAO1 (15). Furthermore, these strains produced PQS earlier (log
phase) than the laboratory strain (stationary phase). This same study
found that growing P. aeruginosa in low-magnesium media, which are
thought to simulate the CF lung, caused PQS production to be increased
5-fold (15). Taken together, these data can allow one to theorize that PQS
production increases, perhaps permanently, once a P. aeruginosa strain
enters the lung of a CF patient.
Overall, numerous studies have shown that PQS is important for the
virulence of P. aeruginosa. How this signal functions during infections is
not known, but it is apparently providing some type advantage as
P. aeruginosa adapts to its environment in vivo.
THE POTENTIAL OF PQS AS A DRUG TARGET
One of the most exciting aspects of cell-to-cell signaling is that
the signals and their synthetic pathways are attractive targets for drug
development. Agents that target these pathways would most likely not be
bactericidal or bacteriostatic, but they hopefully will decrease virulence and
augment a known antibiotic and/or allow the immune system to better
clear the infecting organism. With this in mind, preliminary experiments
have shown that the PQS synthetic pathway is a viable target. Calfee et al. (3)
showed that an anthranilate analog, methyl anthranilate, inhibited the
synthesis of PQS (3). This presumably would occur by the analog entering
the PQS synthetic pathway before the condensation of anthranilate and the
fatty acid chain (see Figure 2.2). Methyl anthranilate also caused a major
decrease in the production of elastase activity by P. aeruginosa, suggesting
that the analog interfered with the signaling pathway and thereby inhibited
the induction of a PQS-controlled gene (lasB) (3). The effect of methyl
anthranilate was also confirmed when Diggle et al. (10) showed this
analog inhibited lecA induction and pyocyanin production, both of which
are PQS-controlled virulence factors. These studies have been expanded
upon to show that multiple anthranilate analogs can inhibit PQS signaling
(S. McKnight and E. Pesci, unpublished data). This promising line of
research will require a better understanding of the functions of different
enzymes in the PQS synthetic pathway.
E
.
C
.
P
E
S
C
I
34
OTHER PQS-LIKE MOLECULES SYNTHESIZED BY P. AERUGINOSA
The 4-quinolones (pyo compounds) produced by P. aeruginosa were first
discovered because of the antibiotic activity they exhibited (6, 19). PQS was
overlooked because it apparently does not have antibiotic activity, and this was
at least partly provenby showing that it has no effect on E. coli or Staphylococcus
aureus (30). However, PQS is one of at least 56 quinolone-type compounds
produced by P. aeruginosa (21). These compounds all appear to fall under the
control of pqsR and are most likely synthesized by the genes of the PQS
synthetic region (9). The roles of these different secondary metabolites must
be assessed on an individual basis. At this point, many of the compounds are
known to have antibiotic activity; except for PQS, none has been shown to act
directly as a cell-to-cell signal.
CONCLUDING REMARKS
Only five years have passed since the Pseudomonas quinolone signal was
identified (30). In that time, a great deal has been learned about the synthesis
and regulation of PQS. The signal has also been found to play an important
role in the virulence of P. aeruginosa. Despite what has been learned, there is
obviously a great deal that is not yet known about this fascinating P. aeruginosa
cell-to-cell signal. The mechanism by which it causes gene induction, its role
during infections, and the exploitation of its synthetic pathway for drug devel-
opment are all future directions that will provide a wealth of knowledge about
howand why P. aeruginosa uses PQS to help ensure survival. As we continue to
keep our collective ears turned toward the P. aeruginosa culture plate, it is likely
that this bacterium will once again prove to be a fascinating conversationalist.
ACKNOWLEDGEMENTS
E. C. Pesci is supported by a research grant fromthe National Institutes
of Allergy and Infectious Disease (grant R01-AI46682). A thank-you
is extended to T. R. de Kievit, E. A. Ling, D. S. Wade, J. P. Coleman, and
A. J. Pesci for help in chapter preparation and thoughtful insight.
REFERENCES
1 Blackwood, L. L., R. M. Stone, B. H. Iglewski and J. E. Pennington 1983.
Evaluation of Pseudomonas aeruginosa exotoxin A and elastase as virulence factors
in acute lung infection. Infect. Immun. 39: 198201.
35
T
H
E
P
.
A
E
R
U
G
I
N
O
S
A
Q
U
I
N
O
L
O
N
E
S
I
G
N
A
L
2 Brint, J. M. and D. E. Ohman 1995. Synthesis of multiple exoproducts in
Pseudomonas aeruginosa is under the control of RhlR-RhlI, another set of regu-
lators in strain PAO1 with homology to the autoinducer-responsive LuxR-LuxI
family. J. Bacteriol. 177: 715563.
3 Calfee, M. W., J. P. Coleman and E. C. Pesci 2001. Interference with Pseudomonas
quinolone signal synthesis inhibits virulence factor expression by Pseudomonas
aeruginosa. Proc. Natn. Acad. Sci. USA 98: 116337.
4 Cao, H., G. Krishnan, B. Goumnerov et al. 2001. A quorum sensing-associated
virulence gene of encodes a LysR-like transcription regulator with a unique self-
regulatory mechanism. Proc. Natn. Acad. Sci. USA 98: 1461318.
5 Collier, D. N., L. Anderson, S. L. McKnight et al. 2002. A bacterial cell to
cell signal in the lungs of cystic fibrosis patients. FEMS Microbiol. Lett.
215: 416.
6 Cornforth, J. W. and A. T. James 1956. Structure of a naturally occurring antago-
nist of dihydrostreptomycin. Biochem. J. 63: 12430.
7 DArgenio, D. A., M. W. Calfee, P. B. Rainey and E. C. Pesci 2002. Autolysis
and autoaggregation in Pseudomonas aeruginosa colony morphology mutants.
relationships. Infect. Immun. 68: 483949.
9 Deziel, E., F. Lepine, S. Milot, et al. 2004. Analysis of Pseudomonas aeruginosa
4-hydroxy-2-alkylquinolines (HAQs) reveals a role for 4-hydroxy-2-heptylquinoline
in cell-to-cell communication. Proc. Natn. Acad. Sci. USA 101: 133944.
10 Diggle, S. P., K. Winzer, S. R. Chhabra et al. 2003. The Pseudomonas aeruginosa
quinolone signal molecule overcomes the cell density-dependency of the
quorum sensing hierarchy, regulates rhl-dependent genes at the onset of
stationary phase and can be produced in the absence of LasR. Molec. Microbiol.
50: 2943.
11 Essar, D. W., L. Eberly, A. Hadero and I. P. Crawford 1990. Identification and
characterization of genes for a second anthranilate synthase in Pseudomonas
aeruginosa: interchangeability of the two anthranilate synthases and evolutionary
implications. J. Bacteriol. 172: 884900.
12 Gallagher, L. A. and C. Manoil 2001. Pseudomonas aeruginosa PAO1 kills
Caenorhabditis elegans by cyanide poisoning. J. Bacteriol. 183: 620714.
13 Gallagher, L. A., S. L. McKnight, M. S. Kuznetsova, E. C. Pesci and C. Manoil
2002. Functions required for extracellular quinolone signaling by Pseudomonas
aeruginosa. J. Bacteriol. 184: 647280.
14 Gambello, M. J. and B. H. Iglewski 1991. Cloning and characterization of
the Pseudomonas aeruginosa lasR gene, a transcriptional activator of elastase
expression. J. Bacteriol. 173: 30009.
15 Guina, T., S. O. Purvine, E. C. Yi et al. 2003. Quantitative proteomic analysis
indicates increased synthesis of a quinolone by Pseudomonas aeruginosa isolates
from cystic fibrosis airways. Proc. Natn. Acad. Sci. USA 100: 27716.
E
.
C
.
P
E
S
C
I
36
16 Kohler, T., C. van Delden, L. K. Curty, M. M. Hamzehpour and J. C. Pechere
2001. Overexpression of the MexEF-OprN multidrug efflux system affects
cell-to-cell signaling in Pseudomonas aeruginosa. J. Bacteriol. 183: 521322.
17 Latifi, A., M. Foglino, K. Tanaka, P. Williams and A. Lazdunski 1996. A
hierarchical quorum-sensing cascade in Pseudomonas aeruginosa links the trans-
criptional activators LasR and RhlR (VsmR) to expression of the stationary-phase
18 Lau, G. W., H. Ran, F. Kong, D. J. Hassett and D. Mavrodi 2004. Pseudomonas
aeruginosa pyocyaninis critical for lung infectioninmice. Infect. Immun. 72: 42758.
19 Leisinger, T. and R. Margraff. 1979. Secondary metabolites of the fluorescent
pseudomonads. Microbiol. Rev. 43: 42242.
20 Lepine, F., E. Deziel, S. Milot and L. G. Rahme 2003. A stable isotope dilution
assay for the quantification of the Pseudomonas quinolone signal in Pseudomonas
aeruginosa cultures. Biochim. Biophys. Acta 1622: 3641.
21 Lepine, F., S. Milot, E. Deziel, J. He and L. G. Rahme 2004. Electrospray/mass
spectrometric identification and analysis of 4-hydroxy-2-alkylquinolines (HAQs)
produced by Pseudomonas aeruginosa. J. Am. Soc. Mass Spectrom. 15: 8629.
22 Mahajan-Miklos, S., M. W. Tan, L. G. Rahme and F. M. Ausubel 1999. Molecular
mechanisms of bacterial virulence elucidated using a Pseudomonas aeruginosa-
Caenorhabditis elegans pathogenesis model. Cell 96: 4756.
23 Mavrodi, D. V., R. F. Bonsall, S. M. Delaney et al. 2001. Functional analysis of
genes for biosynthesis of pyocyanin and phenazine-1-carboxamide from
Pseudomonas aeruginosa PAO1. J. Bacteriol. 183: 645465.
24 McGrath, S., D. S. Wade and E. C. Pesci 2004. Dueling quorum sensing systems
in Pseudomonas aeruginosa control the production of the Pseudomonas quinolone
signal (PQS). FEMS Microbiol. Lett. 230: 2734.
25 McKnight, S. L., B. H. Iglewski and E. C. Pesci 2000. The Pseudomonas quinolone
signal regulates rhl quorum sensing in Pseudomonas aeruginosa. J. Bacteriol. 182:
27028.
26 Passador, L., K. D. Tucker, K. R. Guertin et al. 1996. Functional analysis of the
Pseudomonas aeruginosa autoinducer PAI. J. Bacteriol. 178: 59956000.
27 Pearson, J. P., K. M. Gray, L. Passador et al. 1994. Structure of the autoinducer
required for expression of Pseudomonas aeruginosa virulence genes. Proc. Natn.
Acad. Sci. USA 91: 197201.
28 Pearson, J. P., L. Passador, B. H. Iglewski and E. P. Greenberg 1995. A second
N-acylhomoserine lactone signal produced by Pseudomonas aeruginosa. Proc.
Natn. Acad. Sci. USA 92: 14904.
29 Pearson, J. P., E. C. Pesci and B. H. Iglewski 1997. Roles of Pseudomonas
aeruginosa las and rhl quorum-sensing systems in control of elastase and
rhamnolipid biosynthesis genes. J. Bacteriol. 179: 575667.
Sci. USA 96: 1122934.
37
T
H
E
P
.
A
E
R
U
G
I
N
O
S
A
Q
U
I
N
O
L
O
N
E
S
I
G
N
A
L
31 Pesci, E. C., J. P. Pearson, P. C. Seed and B. H. Iglewski 1997. Regulation of las
and rhl quorum sensing in Pseudomonas aeruginosa. J. Bacteriol. 179: 312732.
32 Rahme, L. G., M. W. Tan, L. Le et al. 1997. Use of model plant hosts to identify
Pseudomonas aeruginosa virulence factors. Proc. Natn. Acad. Sci. USA94: 1324550.
33 Tamura, Y., S. Suzuki and T. Sawada 1992. Role of elastase as a virulence factor
in experimental Pseudomonas aeruginosa infection in mice. Microb. Pathogen. 12:
23744.
34 Whiteley, M., K. M. Lee and E. P. Greenberg 1999. Identification of genes
controlled by quorum sensing in Pseudomonas aeruginosa. Proc. Natn. Acad. Sci.
USA 96: 139049.
E
.
C
.
P
E
S
C
I
38
CHAPTER 3
Quorum-sensing-mediated regulation of
plantbacteria interactions and Agrobacterium
tumefaciens virulence
Catharine E. White and Stephen C. Winans
Department of Microbiology,
Cornell University, Ithaca, NY, USA
INTRODUCTION
Plant-associated bacteria have a wide range of interactions with their
hosts, from non-specific associations to more dedicated symbiotic or patho-
genic interactions. Many complex interactions take place between plant
roots and associated bacteria, fungi, and protozoa in a highly diverse and
dense community within the rhizosphere. Bacterial cell-to-cell communi-
cation systems in this ecological niche appear to affect biofilm formation,
pathogenesis, and production of siderophores and antibiotics. These activ-
ities are no doubt important in root colonization as well as in symbiosis and
pathogenesis. Exciting developments and current studies in understanding
the many complex interactions in the rhizosphere include both the char-
acterization of the microbial communities involved and the responses of
the plant hosts to these communities. Cell-to-cell signaling between
members of the community is no doubt critical for these interactions to
sense population densities and diffusion barriers in the rhizosphere. Such
studies are beyond the scope of this chapter, but we refer the reader to
recent reviews of this field (43, 65, 82).
Perhaps the best-characterized group of soil bacteria that serves as the
model for understanding plantbacteria associations is the Rhizobiaceae.
This family, in the alpha subgroup of the Proteobacteria, includes members
of the genera Rhizobium, Sinorhizobium, Mesorhizobium, Azorhizobium,
and Bradyrhizobium (collectively referred to here as rhizobia), which form
symbiotic relationships with host plants, and several pathogenic species of
the genus Agrobacterium (including A. tumefaciens, A. rhizogenes, A. vitis,
and A. rubi, here referred to as agrobacteria). Rhizobia form symbiotic
relationships with specific host plants by inducing formation of nodules
39
on the plant roots. Colonizing bacteria are restricted to these nodules, and
benefit the host by fixing atmospheric nitrogen. In contrast, pathogenic
agrobacteria cause tumors or other neoplasias on a wide variety of host
plants. Although rhizobia are beneficial to their plant hosts and agrobac-
teria are plant pathogens, their initial interactions with the host plants are
very similar. The colonizing bacteria must respond to host-released signals
during the initial stages of infection and attach to the plant surface. In both
cases, these bacteria synthesize and respond to diffusible chemical signals
called acyl-homoserine lactones (AHLs). AHL-mediated cell-to-cell com-
munication in rhizobia is reviewed elsewhere (30, 65, 84). For the remain-
der of this chapter we will focus on A. tumefaciens.
THE Ti PLASMIDS OF A. TUMEFACIENS AND THE DISCOVERY OF
AHLs AS CONJUGAL PHEROMONES
A. tumefaciens are primarily soil-dwelling bacteria, but many isolates
carry large (200kb) plasmids that are called tumor-inducing (Ti) plasmids,
which are required for the formation of crown gall tumors on infected
plants. There is significant variation between strains of A. tumefaciens;
isolates are most often classified according to the opines that they can
catabolize (despite the fact that Ti plasmids can generally catabolize more
than one type of opine) (94). The best-characterized A. tumefaciens strains
contain either so-called octopine-type Ti plasmids (including strains A6,
B6, Ach5, 15966, and R10) or nopaline-type Ti plasmids (including strains
C58 and T37). The full genome sequence of strain C58 has been completed
by two different groups (32, 87). Its circular chromosome shows striking
synteny with that of Sinorhizobium meliloti, whereas its linear chromosome
and two plasmids appear unique. A composite sequence of the octopine-
type Ti plasmid has also been published (89).
Almost all of the genes that are required for tumor formation are
located on these Ti plasmids. Tumor formation requires the products of
approximately 25 vir genes (71, 85), which process and transfer oncogenic
fragments of DNA (T-DNA) from the Ti plasmid into the host cell, where
they are integrated into the host genomic DNA. Sequence analysis indicates
that the T-DNA transfer apparatus evolved from a bacterial conjugation
system (42, 48, 49, 68). In the host cell, the T-DNA-encoded genes direct
the overproduction of phytohormones that cause the cell proliferation that
is the hallmark of this disease (56). The T-DNA also carries a number of
genes whose products synthesize opines in the infected plant cells. These
compounds are released from the plant cells, taken up by the agrobacteria
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
40
via dedicated ABC-type uptake systems, and used as sources of energy,
carbon, and, in some cases, nitrogen or phosphorus (17). The genes
required for opine uptake and catabolism are also located on the Ti plasmid
and are positively regulated by the cognate opine (17). Although there is a
remarkable diversity in the types of opine used by A. tumefaciens, there is
always a perfect match between the types of opine encoded on the T-DNA
and the opine-catabolic genes of a Ti plasmid.
Ti plasmids have a second complete set of DNA transfer genes. These
tra and trb genes direct the interbacterial conjugal transfer of the entire Ti
plasmid (24) and are organized into three operons (Figure 3.1) (1). The
traAFBH and the traCDG operons are divergently transcribed with the
origin of transfer (oriT) lying within the intergenic region between these
operons. The third operon includes the traI and trbBCDEJKL genes, most
of which are expected to be involved in mating pair formation and synthesis
of the conjugal pili. Some of these genes share a common ancestry with a
number of the vir genes, although sequence similarities between the tra
and vir genes are generally rather weak. Ti plasmid conjugation is reviewed
in more detail elsewhere (24, 86).
Ti plasmid conjugation played a critical role in research on the mechan-
ism of A. tumefaciens virulence. In early experiments, it was discovered that
the formation of crown gall tumors on plants by A. tumefaciens required a
genetic element that could be horizontally transferred by cellcell contact,
suggesting that this trait was plasmid-encoded (44, 45). This plasmid was
subsequently visualized by gel electrophoresis (79). Later, a portion of this
plasmid (the T-DNA) was found in infected plant cells in the bacteria-
induced tumors (57).
In these early studies, conjugal transfer of the Ti plasmid was detect-
able only in crown gall tumors (44, 45, 79), suggesting that these tumors
produce some factor that is essential for conjugation. Following the dis-
covery that crown gall tumors produce opines, Kerr and colleagues soon
demonstrated that conjugation could occur ex planta in the presence
of these compounds (31, 46). In these experiments, conjugation of the
octopine-type Ti plasmids (such as pTiR10) was stimulated by the opine
octopine, whereas conjugation of nopaline-type Ti plasmids (such as
pTiC58) required agrocinopine A or B. Opines that stimulate conjugation
are referred to as conjugal opines.
In 1991, Kerr and colleagues discovered a second compound that stimu-
lated conjugal transfer of Ti plasmids; however, unlike opines, this compound
was produced by the bacteria themselves (91). This diffusible compound,
originally called conjugation factor (CF), stimulated conjugation in
41
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
A. tumefaciens strains carrying octopine-type plasmids but only when they were
cultured in the presence of octopine. Conjugation of Ti plasmids that other-
wise transferred with low efficiency was dramatically stimulated by CF, but
only when it was provided to the conjugal donors. Not long after the discovery
that CF stimulates conjugation, it was shown that this compound could
specifically increase the transcription of a tralacZ fusion (66). By using this
fusionas a reporter, four bioactive compounds were identified, including 3-oxo-
C8-homoserine lactone (OOHL), 3-oxo-C6-homoserine lactone (OHHL),
Octopine Mannopine
NH
NH
HO
H
N
N
H
NH
+
NH
3
NH
2
N
H
O
O O
O
O
O
O O
O
O
O O O
O
H-(CHOH)
5
-CH
2
OccR-Octopine
Complex
MocR-Mannopine
Complex
mocR mot trlR
tral trb (Mating Pore)
occR traR traM tra (Mobilization) occ
PP
PP PP
PP
P P
repC B
Autoinducer
Synthase
TraR
Dimer
(Active)
TraR-TrlR
Heterodimer
(Inactive)
TraR-TraM
Heterodimer
(Inactive)
AttM
Active
Inactive
A
Figure 3.1. The TraRTraI quorum sensing system of Agrobacterium tumefaciens. TraI
synthesizes OOHL, which diffuses across the cellular envelope. At high population density,
OOHL accumulates in the cell and binds to TraR. TraR activates expression of the two tra
promoters, the trb operon which includes traI, traM, and repABC. OccR activates
transcription of traR in the presence of octopine, and MocR activates trlR in the presence of
mannopine. TraM and TrlR are anti-activators of TraR. Finally, at the stationary phase of
growth, attM is transcribed, resulting in inactivation of OOHL.
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
42
and two chemical derivatives of those compounds that may have been
produced during purification (90). Following the identification of these com-
pounds as the CF of octopine-type plasmids, it was shown that nopaline-type
plasmids produce OOHL as the only active compound (38).
N-acyl-homoserine lactones (AHLs) have also been discovered in many
other proteobacteria, beginning with Vibrio fischeri (21). These compounds
are released and detected by populations of bacteria and are thought to
permit cell-to-cell communication within a population, leading to coordi-
nation of a wide variety of behaviors. These signaling systems are thought
to allow cells to estimate their population densities and to carry out various
types of behavior only at high population densities, a phenomenon often
referred to as autoinduction or quorum sensing.
IDENTIFICATION OF A Ti-PLASMID-ENCODED AHL SYNTHASE
AND AHL RECEPTOR
The genes required for the synthesis and detection of the AHLs in the
nopaline- and octopine-type plasmids were discovered concurrently by two
different groups. In a study using random Tn5 mutagenesis, Farrand and
colleagues isolated an insertion that eliminated the requirement of inducing
opines for conjugation (3). Sequence analysis indicated that this transposon
had inserted upstream of a gene called traR and was probably being over-
expressed fromanoutward-reading promoter onthe transposon (66). The role
of traR in tra gene expression was verified by expressing this gene from a
constitutive promoter along with a tralacZ fusion in a Ti-plasmid-less strain
of A. tumefaciens. b-Galactosidase expression was induced only in the presence
of OOHL or OHHL, although the former inducer was far more potent. This
experiment suggested that TraR was indeed a receptor for OOHL, and that it
activated expression of conjugation genes.
Bioassays using a tralacZ reporter strain indicated that A. tumefaciens
synthesizes AHLs, suggesting that it encoded a protein homologous to LuxI of
V. fischeri. To search for such a gene, Farrand and colleagues tested random
Tn3 insertions in the nopaline-type Ti plasmid for production of OOHL (39).
All of the insertions that disrupted OOHL synthesis mapped to one gene, traI,
which was indeed a luxI homolog. TraR upregulated the expression of traI in
the presence of exogenous autoinducer, strongly increasing OOHL produc-
tion. The traI gene is the first gene in the trb operon, which encodes a type IV
secretion system that is required for conjugation.
The traR and traI genes of the octopine-type Ti plasmid were identified
at about the same time. The traR gene was discovered in a random Tn5gus
43
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
mutagenesis study designed to identify genes that were inducible by octo-
pine (912). As octopine was known to stimulate Ti plasmid conjugation,
all octopine-inducible fusions were screened for defects in conjugation.
Although none of themcaused a decrease in conjugation, one of the fusions
unexpectedly showed enhanced conjugation (29). This transposon inser-
tion had disrupted a gene called traM, which is adjacent to traR. From this
fusion, traR was eventually identified and found to stimulate conjugation
when overexpressed. The reason that a traM null mutation elevated
conjugation was unraveled in later studies (see below).
The luxI homolog traI on the octopine-type Ti plasmid was identified
by screening a cosmid library for the synthesis of OOHL (29). Like traI of
nopaline-type Ti plasmids, this traI gene appeared to be the first gene in the
trb operon. Both OOHL production and conjugation were disrupted by a Tn5
insertion that mapped to the traI gene. However, when traI was expressed in
trans in the same strain, OOHL production was restored but conjugation was
not. These results indicated that traI was indeed the first gene of the trb operon.
In an effort to identify the reaction mechanism for OOHL synthesis,
TraI was purified as a His6-tagged fusion and found to produce OOHL
in vitro when added to a complex extract of bacterial proteins and small
molecules (55). TraI also synthesized OOHL in a defined buffer supple-
mented with 3-oxooctanoyl-ACP and S-adenosylmethionine (SAM). This
was the first time that the activity of any LuxI-type protein had been
reconstituted in vitro and the substrates identified. 3-Oxooctanoyl-CoA
was inactive as a fatty acid group donor. Similarly, a variety of possible
precursors for the homoserine lactone ring were tested, but only SAM was
active. In the same study, purified TraI could transfer a radiolabeled car-
boxyl carbon of SAM to the OOHL product, providing further evidence that
SAM is the precursor for the homoserine lactone ring. A similar reaction
mechanism was later described for other LuxI-type proteins (64, 78).
Recently, the crystal structures of two proteins from this family were
solved, EsaI of Pantoea stewartii and LasI of Pseudomonas aeruginosa (33,
83). As EsaI and LasI synthesize different AHLs, 3-oxo-C6- and 3-oxo-C12-
homoserine lactone, respectively, comparisons of these two structures are
useful for understanding not only the reaction mechanism but also the
determinants of acyl-ACP specificity.
REGULATION OF traR EXPRESSION
The discovery of the TraRTraI quorum sensing system of the Ti
plasmids helped explain previous observations that conjugation of these
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
44
plasmids requires opines. For octopine-type plasmids such as pTiR10, the
traR gene is at the distal end of the occ operon, which is activated by
octopine via OccR (Figure 3.1) (28). When traR is expressed from a consti-
tutive promoter, conjugation does not require octopine. Therefore, regula-
tion of traR transcription by OccR fully explains the requirement of
octopine for conjugation. In the nopaline-type plasmids, traR expression
is also directly regulated by opines but through a different mechanism. In
pTiC58, the conjugal opines agrocinopine A and B induce expression of
the acc genes and the divergently transcribed arc operon, which includes
traR (2). In early studies of this system it was found that a null mutation of
accR causes constitutive expression of these genes, and traR overexpression
also results inconstitutive conjugation but has no effect onthe acc genes. More
recent evidence from experiments with purified AccR and promoter DNA
indicates that the expression of the arc and acc operons is directly repressed
by AccR, and that this repression is relieved by agrocinopines (67).
The regulation of traR by opines appears to have evolved independently
several times. OccR is an activator of transcription when it binds to the
conjugal opine octopine, whereas AccR acts as a repressor in the absence of
agrocinopines. OccR and AccR are not related proteins: OccR is in the LysR
family of transcriptional regulators but AccR is related to the Lac repressor
(2, 28). Finally, these traR genes are located within completely dissimilar
operons.
Regulation of traR expression by opines has since been described for a
number of other plasmids in A. tumefaciens, and in general these follow
the models described above for octopine- and nopaline-type plasmids
(Figure 3.2). On pTiChry5, traR is in a two gene operon called arc (58).
The expression of this operon is induced by the agrocinopines C and D,
which are thought to relieve repression by an AccRhomolog that is encoded
nearby (58). Agrocinopines Cand Dare also known to induce the transfer of
the agropine-type Ti plasmid pTiBo542 (23). Another nopaline-type
plasmid is pATK84b from A. radiobacter isolate K84 (15, 34, 35). This
Agrobacterium strain is unable to induce the formation of crown gall
tumors, and in fact pATK84b does not carry the vir genes or T-DNA of
the Ti plasmids. However, pATK84b does allow this strain to compete for
opines as a source of nutrients, as it has functional copies of nopaline and
agrocinopine A and B catabolic genes (15, 16). Interestingly, pATK84b also
has two copies of traR (59). One copy, traR
noc
, is the last gene of the nox
operon and is induced by nopaline, whereas traR
acc
is induced by agroci-
nopines A and B. Each of these two traR genes is required for induction of
transfer by the cognate opine (59).
45
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
POST-TRANSCRIPTIONAL REGULATION OF TraR ACTIVITY
The TraRTraI regulatory system is inhibited by a surprising number
of proteins, one of which is TraM. A null mutation in traM elevates tra gene
transcription and conjugation, whereas overexpression of traM abolishes
tra gene expression in a strain expressing wild-type levels of TraR (26, 38).
Providing high levels of OOHL did not stimulate tra gene expression.
However, tra gene transcription was restored by overexpressing TraR,
suggesting that TraM and TraR may interact stoichiometrically.
TraR was demonstrated by yeast two-hybrid assays and far western
immunoblots to interact directly with TraM (40). Deletion and point muta-
tions showed that binding is mediated by the C-termini of both proteins
(40). TraMcould both prevent the interaction of TraR with its DNAbinding
site and bind to TraR in proteinDNA complexes to cause release of TraR
Figure 3.2. Regulation of traR on an octopine-type Ti plasmid and two nopaline-type Ti
plasmids. Also shown is pAtK84b, which does not have vir genes or T-DNA but does carry
opine catabolic genes. On octopine-type Ti plasmids, traR is activated by OccR in response
to octopine. On the nopaline-type Ti plasmids, traR is transcribed when repression by AccR
is relieved by agrocinopines. pAtK84b has two traR genes: one is thought to be activated by
NocR in reponse to nopaline, and the other is repressed by AccR.
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
46
from DNA (51). Analysis by surface plasmon resonance showed that
TraRTraM binding is stable and has an affinity in the nanomolar range
(75). TraRTraM complexes eluted from a gel filtration column at a mole-
cular mass of approximately 60kDa in one study (75) or 157 kDa in a second
study (80). The former is consistent with a TraR monomer binding one or
two TraM monomers; the latter is consistent with two TraR dimers binding
two TraM dimers. This discrepancy must be due to differences in experi-
mental conditions. Several TraM point mutants were isolated that still
bound TraR but did not inhibit TraR activity (46, 70), suggesting that
binding and inactivation may occur in two sequential steps. The TraM
structure was solved by two groups using X-ray crystallography (75, 8).
TraM crystallized as a dimer, with each monomer consisting largely of
two antiparallel a-helices. The monomers were intimately associated in
the dimer, and significant hydrophobic surface was buried upon
dimerization.
It is not clear how TraMTraR interactions disrupt TraRDNA inter-
actions; however, two different mechanisms have been suggested. Chen
and colleagues proposed that the TraM and TraR homodimers may dis-
sociate upon binding to each other so that a TraRTraM anti-activation
complex can form, perhaps interacting through the surface of TraM that
is otherwise buried in the homodimer interface (8). Evidence for this model
is that TraM dimers, when their subunits are covalently cross-linked to
each other, can bind to TraR but are unable to inactivate the protein (8).
Mutational analyses of TraM also suggest that residues involved in initial
binding to TraR are different from those required for TraR inactivation
(51, 75). A second study reached different conclusions, proposing that two
dimers of TraM are clamped between two dimers of TraR, with part of
each of the four TraR DNA-binding domains bound into one of two narrow
grooves on the surface of each TraM dimer (80). This complex would
physically inhibit the TraR dimers from binding to DNA. Structural ana-
lysis of TraRTraM complexes will help to resolve the intriguing question
of how TraM mediates inactivation of TraR, although the static structure
will not provide a complete picture, inasmuch as TraR binding and inacti-
vation are thought to occur in several sequential steps.
The adaptive significance of TraMis not understood. On both octopine-
and nopaline-type plasmids, expression of traM is positively regulated by
TraR and OOHL (25, 36), indicating that this quorum-sensing regulon
induces its own antagonist and suggesting that TraM may be acting as a
governor to limit expression of this regulon (Figure 3.1). TraM is a highly
conserved member of TraRTraI quorum-sensing systems, and is even
47
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
associated with these systems on the symbiosis megaplasmids of
Rhizobium spp. (36).
Another TraR inhibitor, designated TrlR, is encoded only on octopine-
type Ti plasmids. These Ti plasmids incite tumors that synthesize several
opines, including mannopine. In the course of analyzing the mot operon,
which is required for the utilization of mannopine, a gene, designated trlR,
was identified. This gene is virtually identical to traR except for a frameshift
mutation between the N-terminal pheromone binding domain and the
C-terminal DNA binding domain (60). As might be expected, TrlR is inactive
in tra gene expression, whereas correction of the frameshift mutation by
site-directed mutagenesis resulted in a trlR mutant that encoded a fully
functional protein (95). Identical genes were subsequently identified on a
number of other octopine-type Ti plasmids, indicating that the frameshift
within trlR is not a laboratory artefact but in fact is disseminated in nature
(95). TrlR is not only non-functional but also inhibits TraR activity by
forming inactive heterodimers (7). Similar truncated alleles of LuxR (made
in the laboratory) have similar properties in that they block the function of
native LuxR, probably by forming inactive heterodimers (14).
Expression of trlR is induced in response to mannopine, probably via
the MocR protein (Figure 3.1) (60, 95). Mannopine decreases conjugation
specifically due to trlR, as expressing this gene froma constitutive promoter
also blocked conjugation (7). In the same study it was found that expression
of trlR was strongly repressed by favored catabolites, including succinate,
glutamine, and tryptone. As these nutrients could restore TraR activity by
blocking trlR gene expression, it was speculated that TrlR functions as an
inhibitor of the energetically expensive process of conjugation when nutri-
ents are limiting.
TraRactivity is regulated by a third gene, attM, although in this case the
mechanism of inhibition is more indirect than described above. This gene
is a homologue of aiiA from Bacillus cereus (32, 87); both genes direct the
inactivation of AHLs by hydrolyzing their lactone ring, forming the corres-
ponding N-acyl homoserine (Figure 3.1) (18, 19, 89). The attM gene lies
within the attKLM operon, where attK and attL encode a predicted semi-
aldehyde dehydrogenase and alcohol dehydrogenase, respectively (88). If
these enzymes form a pathway, the predicted final product of OOHL
metabolism would likely be N-(3-oxooctanoyl)aspartate.
The attKLM operon is repressed by the product of the divergent attJ
gene; expression is induced at stationary phase in response to either carbon
or nitrogen limitation (83). This provides suggestive evidence that attKLM
is a catabolic operon that is induced only if its natural substrates are present
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
48
and favored carbon sources are absent. It is likely that the natural substrates
interact with AttJ to derepress expression of this operon. However, AHLs
are not necessarily the natural substrates for these enzymes, as expression
of the attKLM operon is not induced by any tested AHLs (89) (Y. Chai and
S. C. Winans, unpublished data). Even if this pathway is not dedicated to
inactivation and catabolism of OOHL, it is possible that A. tumefaciens can
recycle these signal molecules. AHL recycling as an energy and nitrogen
source has been demonstrated in Variovorax paradoxus and in Pseudomonas
aeruginosa (37, 47).
THE ROLE OF OOHL IN TraR MATURATION
TraRactivity is checked at yet another level, in that the protein is rapidly
degraded by cytoplasmic proteases in the absence of OOHL. When TraR is
strongly overexpressed in E. coli or in A. tumefaciens, it accumulates in the
insoluble fraction of cell lysates. This led one group to suggest that, in the
absence of OOHL, TraR binds to the cytoplasmic membrane and that it is
released by OOHL to the cytoplasm, where it dimerizes and binds DNA
(70). However, another group had a different interpretation, in which this
insoluble protein accumulated in inclusion bodies that formed owing to
TraR overproduction and were biologically irrelevant. When traR was
mildly overexpressed, the protein accumulated only when OOHL was pre-
sent, and was otherwise degraded by cytoplasmic proteases. This finding
was first made by western immunoblots and confirmed by pulsechase
experiments (97). These data led us to conclude that OOHL rescues TraR
from proteolysis during or directly after translation, and may act as a
scaffold for TraR folding. In addition, purified apo-TraR was rapidly
degraded by trypsin to oligopeptides, whereas TraROOHL complexes
were more resistant to the proteases and were cleaved only at the accessible
interdomain linkers of the protein (97). Again, this indicates that TraR
absolutely requires OOHL to fold into its native and stable conformation.
Contrary to the well-established lock-and-key model of proteinligand
recognition, it is becoming increasingly evident that ligand binding
induces conformational changes in many other protein receptors, from
ordering of short loops to disorder-to-order transitions of the entire poly-
peptide chain (20). In many cases these proteins do not accumulate in the
absence of a regulatory signal. The term intrinsically unstructured was
coined by Dyson and Wright (20) to signify the importance of these pro-
teins in cellular processes and as models for studying protein-folding
mechanisms. Like TraR, many proteins that are intrinsically unstructured
49
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
are involved in highly time-dependent processes. These include cell-cycle
regulators required for inducing apoptosis, regulatory RNA binding pro-
teins, proteins involved at various points in signal transduction pathways,
and many other transcriptional regulators involved in activation or repres-
sion (20). It has also been suggested that one function of a disorder-to-order
transition upon binding is to increase the binding rates for a faster response
to the stimulus by increasing the likelihood that initial long-range interac-
tions will occur (72). Incorporation of OOHL into the folding process may
also optimize the specificity of the proteinligand interaction, as has been
suggested for some other intrinsically unstructured proteins (20).
STRUCTURE AND FUNCTION STUDIES OF TraR
Knowing that TraR requires OOHL for solubility enabled purification
of amounts sufficient for X-ray crystallography. There are now two crystal
structures of TraR available: both are ternary complexes of TraR, OOHL,
and the DNA binding site (called a tra box) for TraR (81, 92). These
structures support previous biochemical data but also have some unex-
pected features (Figure 3.3). They confirm earlier findings that TraR
binds OOHL via its N-terminal domain in a 1:1 mole ratio, and that it
binds DNA as a dimer via its C-terminal domain (70, 97). The N-terminal
domain has an aba structure, with one molecule of OOHL embedded
between a layer of a-helix and the b-sheet (81, 92). OOHL is therefore
deeply buried within the N-terminal pheromone-binding domain of each
TraR subunit (Figure 3.3). The hydrophobic atoms of OOHL pack with
hydrophobic residues in the core of the domain; the polar oxo-groups and
amine form hydrogen bonds with nearby residues. When bound to TraR,
the pheromone appears to be completely protected from solvent (81, 92).
This engulfment of the ligand in the core of the N-terminal domain sup-
ports our previous model that apo-TraR is to at least some degree unstruc-
tured and that OOHL participates in the protein-folding process.
The main dimerization interface of TraR is composed of a long a-helix
in the N-terminal domain of each subunit that is parallel with the same
helix of the opposite subunit. Contributions of residues buried in this
interface to dimer formation were confirmed by mutational analysis (52).
Although the N-terminal domains are sufficient for dimer formation, the
C-terminal domains also dimerize via two parallel helices, although the
interface is not as extensive (81, 92). The structure of the C-terminal
domains, each a four-helix bundle containing a helixturnhelix DNA
binding motif, is highly conserved and places the LuxR family members
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
50
in the larger NarLFixJ family of transcriptional regulators (53). The DNAis
in B-form and the consensus tra box, an 18 bp sequence with perfect dyad
symmetry, has a 308 bend toward the sides of the dimer (81, 92). Although
there is an extensive interface between the protein and the DNA, the main
contributors to specificity and affinity of binding are two arginine resi-
dues in the recognition helix that have hydrogen bonds with two bases in
the major groove of each half site (81, 92) (C. E. White and S. C. Winans,
(a)
(b)
(c )
T4
G5
210
1
3
1
3
9
9
N
C
C
N
206
Figure 3.3. Ternary structure of TraR, OOHL, and tra box DNA. (a) Just the N-terminal
domains of a dimer. The dimerization interface is highlighted, and is along the length of
a-helix 9 of each monomer. The OOHL and water molecule bound in each N-terminal
domain are shown in space-fill and can be seen most clearly in the right-most monomer.
The linker between the N-terminal domains and the C-terminal domains is marked with a
C. (b) Just the C-terminal domains of the TraR dimer bound to DNA. Hydrogen bonds
between the side-chains of arginine 206 and 210 of the recognition helix and bases T4 and
G5 of the binding site are shown. The dimerization interface is also highlighted, along
a-helix 13 of each monomer, and the linker between the NTD and CTD of each monomer is
marked with an N. (c) A view of a full dimer down the long axis of the DNA, showing the
asymmetry of the crystallized protein due to the flexible linkers. Coordinates for these
models are from references 81 and 92. (See also color plate section)
51
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
unpublished data). One of these residues had been previously shown as
critical for DNA binding affinity (50). TraR had also been shown to bind to
target DNA sequences with high specificity (96). In fact there is little
variation in most of the native tra boxes of the Ti plasmid, and the two
bases that make direct contacts with two arginines of TraR as described
above are absolu tely cons erved ( se e F igu re 3.4 b ).
The N-terminal pheromone-binding domain and the C-terminal DNA-
binding domain of each monomer are connected by a 12-residue unstruc-
tured linker (81, 92). This flexibility may be the reason for a pronounced
asymmetry of the crystallized protein. The C-terminal domains of each dimer
have a two-fold axis of symmetry, and the N-terminal domains also have a
two-fold axis of symmetry, but these axes lie at a 908 angle to each other (81,
92). The presence of the flexible linkers connecting the C-terminal domains of
the dimer to the N-terminal domains suggests that there is a high degree of
flexibility between the two domains in vivo. This has been proposed to play a
role at divergently transcribed promoters, as discussed in the next section.
There is significant interest in understanding how AHLs convert their
receptors from inactive to active forms, and how these proteins can dis-
criminate between their cognate ligand and similar AHLs found in nature.
When TraR is expressed at native levels in the cell it binds to its cognate
ligand, OOHL, with high specificity (93). There are four hydrogen bonds
between residues in the binding pocket of TraR and the polar groups of
OOHL, in addition to numerous hydrophobic and packing interactions
surrounding the ligand (81, 92). Mutational analyses confirm that the
polar interactions between residues of TraR and the OOHL are critical for
binding and stability of TraR (6, 52). One of these polar interactions is of
considerable interest as it involves a variable group on the OOHL. The 3-oxo
group of OOHL is not common to all AHLs and therefore could be an
important determinant of specificity of binding. In the binding pocket,
this group makes a water-mediated hydrogen bond to a threonine residue.
In an attempt to alter binding specificity for a 3-unsubstituted AHL
(C8-homoserine lactone), point mutations were made of this residue to increase
hydrophobicity and displace the water molecule from its binding site (6).
These mutations resulted in a broadened specificity of binding rather than
in altered specificity. In the same study, point mutations were also con-
structed to change the preference of binding to AHLs with shorter fatty acid
tails, by increasing hydrophobic bulk in the binding pocket. Some of the
mutations did alter specificity of binding but the stability of these mutants
was also decreased. These data suggest that the residues in the binding
pocket that contact the OOHL play dual roles in folding or stability of the
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
52
N-terminal domain and ligand recognition. As these residues are buried in
the hydrophobic core of the protein, they make many contributions to pack-
ing interactions with surrounding residues in addition to OOHL. Therefore,
incorporating OOHL into the protein-folding process may be important not
only for regulation of TraR activity but also for specificity of the protein
ligand interaction.
TraR AS AN ACTIVATOR OF TRANSCRIPTION
Seven TraR-dependent promoters have been identified on the octopine-
type Ti plasmid (Figure 3.4). These include all of the genes required
for conjugal transfer, arranged in three operons: traAFBH and traCDG-yci,
both required for conjugal DNA processing, and the traItrb operon (27).
The TraR anti-activator traM is also activated by TraR (C. Fuqua, personal
communication). More recently, three TraR-dependent promoters of the
repABC operon have been described (61). This operon is required for both
vegetative replication of the plasmid and partitioning into daughter cells.
Promoter structure at this operon is quite complex, as there is at least one
additional promoter for rep that is not TraR-dependent, but is repressed by
RepA (62). This ensures plasmid maintenance and partitioning when TraR
is inactive. Activation of repABC by TraR increases Ti plasmid copy number
about seven-fold, thereby enhancing the expression of all Ti-plasmid-
encoded genes (61). This could lead to enhanced uptake and catabolism of
opines as well as enhanced conjugation and possibly enhanced T-DNA
transfer.
The transcription start sites at all of these TraR-dependent promoters
have been mapped; each has an identifiable tra box (27, 61) (C. Fuqua,
personal communication). The traA and traC promoters are divergently
transcribed from tra box I (the consensus sequence); the traI and two rep
promoters (repAP1 and repAP2) are divergently transcribed from tra box II
(Figure 3.4). The tra boxes of PtraA, PtraC, PtraI, PrepA1, and PrepA3 are
centered at approximately 45 nucleotides upstream from the transcrip-
tion start site and overlap the 35 element of each promoter. In contrast,
the tra boxes of the PtraM and PrepA2 promoters are centered at 66
nucleotides from the transcription start site (Figure 3.5). These promoters
are reminiscent of the class I and class II promoters first described for
CRP (5). At class II promoters the activator overlaps the 35 element and
can make multiple contacts with RNA polymerase. These contacts recruit
RNA polymerase to the promoter and possibly affect later steps in tran-
scription initiation. At class I promoters the activator binds farther
53
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
Class I promoter
Class II promoter
61.5
"CTD
"CTD
Sigma
Sigma
"NTD
"NTD
41.5
35 10
10
(a)
(b)
Activator
Activator
Figure 3.5. Predicted models of TraR at class I (a) and class II (b) type promoters. These
models are based on studies of other activators (5).
(a)
(b)
<<+1
G
T T T T T T T T T T T T T T TT TTT T T T T T T T G G G G G G G G G G G G G G G G G G G C C C C C C C C C C C C T T T A
G G T TT TTT
T
T T T T
T TT T T TT T T TT T T T T TT T T T
T T T T T T TT T TT TT T TT T T G G G
G
G G
G
GTG
TG
TG
TG
G G GA
A A A A
AAAA A A A A A A AA A A A A A A G G G G G
G G G G G G GG G C
C
C C C A C
A C
A C
A C
T T
T
TT
T
T
T
T
T
T G
G
G
G
G
G A A
T
T
G A
A
AAAA
AAA
A
G
G
C
C
C
C
C
C
C
C
C
C C C C C C C CC C C CCC C C C G GG G G G GG G C C
C C C C C C CC C C C A A AAA A A A A A A A A A A A A A
A A AA A A A A GC C C C A A AT A A A A A A A AAAAA A A A A C C C C C C GG G G G G GG GG
CACG G G G G GGG A A AAA AA A A A A AA A T T T T T G G G G C C C C C C C CC C C G GGA A A T T T C CC CCG G GG G G T TTT G T T G CC CC C CC C A AAT T
10 35 35 10
10 35 35 10
10
35 10
35
35
+1>>
+1>>
+1>>
10 +1>>
+1>>
T
tra
tra
tra
tra
box
box
box
box
P traC P traA
P repA1
P repA3
P repA2
P traM
P traI
tra box I
tra box II
tra box III
tra box IV
I
II
III
IV
<<+1
Figure 3.4. (a) DNA sequences and tra boxes of TraR-dependent promoters. The tra box for
each promoter is highlighted, transcription start sites are marked with a 1, and the first
gene of the operon is marked above each start site (translation starts are not shown). Each of
the three TraR-dependent rep promoters (repA1, repA2, and repA3) is indicated.
(b) Alignment of the four known tra boxes of pTiR10. The two positions of each half-site that
make sequence-specific contacts with Arg206 and Arg210 of TraR are highlighted.
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
54
upstream on the same face of the DNA as polymerase and can contact only
the a C-terminal domain, which is connected to the rest of the polymerase
by a flexible linker (5).
As TraR can activate transcription at both class I- and class II-like pro-
moters, the activator is predicted to contact at least the a C-terminal domain of
RNA polymerase. At class II promoters TraR is thought to make additional
contacts with s70. This prediction is based on the observation that LuxR,
which also activates transcription froma class II promoter, interacts with both
the a C-terminal domain and s70 of RNA polymerase (25, 41, 73). A potential
RNA polymerase contact site, or activating region, has been identified on the
surface of the TraR C-terminal domain and overlaps with a patch on the LuxR
surface that is also critical for activation (22) (C. E. White and S. C. Winans,
unpublished data). Mutations of residues in this position do not disrupt
protein stability and DNA binding but are required for activation. However,
this patch on LuxR is predicted to interact with s70 (76), whereas the same
region of TraR is predicted to contact the a C-terminal domain of RNA
polymerase (C. E. White and S. C. Winans, unpublished data). Two additional
residues have been identified on the TraRN-terminal domain that are thought
to contact the a C-terminal domain (50, 69). Owing to the high degree of
flexibility between the TraR domains, these residues could lie very close to the
patch of residues that we identified on the C-terminal domain (81, 92). Both
sets of residues could together forma contact site for the a C-terminal domain.
This potentially novel activating region, formed from two different sets of
residues fromdifferent parts of the TraR protein, may be critical in the activity
of TraR at divergent promoters. However, further studies are required to fully
understand the interactions of TraR with RNA polymerase at TraR-dependent
promoters.
PERSPECTIVES AND FUTURE STUDIES
In this chapter we have discussed the details of how the cell-to-cell
signaling system of A. tumefaciens is regulated, what genes are activated by
TraR, and how TraR functions as a quorum-sensing transcriptional regu-
lator. However, many of the biochemical details of this system remain to be
solved. One major question involves the inactivation of TraR by its anti-
activator TraM. Although the mutagenesis studies of TraM suggest that
there are two steps in its activity, binding to TraR and then deactivation, the
mechanism of inhibition remains elusive (51, 75). These studies along with
the structural studies of TraM led to the model that, following complex
formation, the TraM dimer dissociates to inactivate TraR (8). Further
55
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
structural studies of TraMTraR complexes are critical to understanding
this mechanism.
Although the mechanisms by which TrlR acts are fairly well resolved,
we really have very little idea of why TrlR is adaptive. Why does one opine,
octopine, stimulate Ti plasmid conjugation and replication whereas
another, mannopine, inhibits both activities? The third gene we described
that inhibits TraR activity, attM, offers even more puzzles. Did this gene
evolve to degrade AHLs specifically, or some other substrate? If AttM
catalyzes the first step in OOHL degradation and AttK and AttL are also
involved, what is the enzymatic pathway? Can OOHL be mineralized and
used by A. tumefaciens as a carbon and nitrogen source, or is the function of
AttM only to regulate the cellular concentration of OOHL?
We also do not fully understand how OOHL protects TraR against
proteolysis. Even if OOHL does not act as a scaffold in the folding process,
major rearrangements must occur in the N-terminal domain of TraR upon
binding to the ligand to sequester the molecule in the core of the domain
and help the protein to evade proteolysis. Future biophysical studies of the
interaction of apo-TraR with OOHL will be very exciting and reveal not only
how OOHL contributes to TraR stability but also the specificity of the
interaction. The LuxR family members have evolved to be highly selective
in AHL binding; structural studies of other receptors and comparisons of
these to TraR are critical to understanding binding specificity and how
these proteins recognize their targets. A number of other LuxR homologs,
including LuxR and RhlR, have been purified; unlike TraR, they do not bind
to their cognate AHLs irreversibly (54, 77).
Although a possible RNA polymerase contact site has been identified
on the C-terminal domain of TraR, it is very likely that there is at least one
more patch critical to activation at class II-like promoters that has not yet
been identified. The details of how TraR interacts with RNA polymerase at
TraR-dependent promoters, particularly the divergent traAtraC and
traIrepABC promoters, remain to be determined. Although we often use
models of transcription activation that are based on CRP, it is already
apparent that there are major differences in how these two proteins interact
with RNA polymerase. The predicted contact site with the a C-terminal
domain of RNA polymerase on TraR is between the dimerization interface
and the edge of the protein on the C-terminal domain, whereas on CRP this
same activating region is on the edge of the dimer, contacting the a-subunit
as it lies adjacent to CRP in the minor groove of the DNA (4). Moreover, the
DNA bend in the activator binding site is also likely to influence how it
interacts with polymerase. CRP induces an 808 bend in the DNA, resulting
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
56
in kinks in each half-site, whereas the angle is only 308 in TraRDNA
complexes (63, 81, 92). Finally, although CRP can activate transcription
at divergent promoters, the complex promoter architecture at tra box II is
unique (5, 27, 61), with two class II-type and one class I-type promoter all
being activated from one TraR-binding site! There are also a number of
differences in the mechanism of activation between LuxR and TraR. The
LuxR C-terminal domain is sufficient for activation, whereas two residues
have been identified on the TraR N-terminal domain that are predicted to
contact RNA polymerase (13, 50, 74). TraR requires a supercoiled template
for transcription activation, but LuxR does not (77, 96).
In this chapter we have focused on the biochemistry of this system,
but there are many intriguing questions of how the quorum-sensing
system of A. tumefaciens relates to its role as a plant pathogen. As the
quorum-sensing system is induced only in the presence of opines, it is
active only on transformed plants and only after infection. This in itself
seems to be an anomaly, as most signaling systems in plant pathogens are
thought to be important in measuring a population of sufficient size before
inducing virulence so that pathogenesis might be successful. Perhaps
increase in Ti plasmid copy number by conjugation and replication in a
plant-associated population benefits A. tumefaciens by increasing gene
dose of the virulence and opine import and catabolic genes. As the traI
and traR genes are both on the Ti plasmid, then all conjugal donors in the
population both release and detect the OOHL signal. The obvious result of
this is that donors are counting other donors instead of possible recipients
that do not carry the Ti plasmid, so perhaps donors conjugate with each
other to increase overall Ti plasmid copy number in the population. In
support of this, sequence examination suggests that the Ti plasmid may
not have a potent entry-exclusion system. Conjugation between donor
cells would also serve to increase copy number of all Ti plasmid genes in
bacteria colonizing a tumor, including all opine catabolic operons, and
conjugation operons. This would also increase the gene dosage of T-DNA
and vir genes, which could stimulate further rounds of tumorigenesis on
infected plants.
ACKNOWLEDGEMENTS
The authors thank the members of their laboratory for helpful
discussions and review of the paper. Research in the authors laboratory
is supported by grants from NIH (GM42893) and NSF (MCB-9904917).
57
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
REFERENCES
1 Alt-Morbe, J., J. L. Stryker, C. Fuqua et al. 1996. The conjugal transfer system of
Agrobacterium tumefaciens octopine-type Ti plasmids is closely related to the
transfer system of an IncP plasmid and distantly related to Ti plasmid vir
genes. J. Bacteriol. 178: 424857.
2 Beck von Bodman, S., G. T. Hayman and S. K. Farrand 1992. Opine catabolism
and conjugal transfer of the nopaline Ti plasmid pTiC58 are coordinately regu-
lated by a single repressor. Proc. Natn. Acad. Sci. USA 89: 6437.
3 Beck von Bodman, S., J. E. McCutchan and S. K. Farrand 1989. Characterization
of conjugal transfer functions of Agrobacterium tumefaciens Ti plasmid pTiC58.
4 Benoff, B., H. Yang, C. L. Lawson et al. 2002. Structural basis of transcription
activation: the CAP-alpha CTD-DNA complex. Science 297: 15626.
5 Busby, S. and R. H. Ebright 1999. Transcription activation by catabolite activator
protein (CAP). J. Molec. Biol. 293: 199213.
6 Chai, Y. and S. C. Winans 2004. Site-directed mutagenesis of a LuxR-type
quorum-sensing transcription factor: alteration of autoinducer specificity.
Molec. Microbiol. 51: 76576.
7 Chai, Y., J. Zhu and S. C. Winans 2001. TrlR, a defective TraR-like protein of
Agrobacterium tumefaciens, blocks TraR function in vitro by forming inactive
TrlR:TraR dimers. Molec. Microbiol. 40: 41421.
8 Chen, G., J. W. Malenkos, M. R. Cha, C. Fuqua and L. Chen 2004. Quorum-
sensing antiactivator TraM forms a dimer that dissociates to inhibit TraR. Molec.
Microbiol. 52: 164151.
9 Cho, K., C. Fuqua, B. S. Martin and S. C. Winans 1996. Identification of
Agrobacterium tumefaciens genes that direct the complete catabolism of octopine.
J. Bacteriol. 178: 187280.
10 Cho, K., C. Fuqua and S. C. Winans 1997. Transcriptional regulation and loca-
tions of Agrobacterium tumefaciens genes required for complete catabolism of
octopine. J. Bacteriol. 179: 18.
11 Cho, K. and S. C. Winans 1993. Altered-function mutations in the Agrobacterium
tumefaciens OccR protein and in an OccR-regulated promoter. J. Bacteriol. 175:
771519.
12 Cho, K. and S. C. Winans 1996. The putA gene of Agrobacterium tumefaciens is
transcriptionally activated in response to proline by an Lrp-like protein and is not
autoregulated. Molec. Microbiol. 22: 102533.
13 Choi, S. H. and E. P. Greenberg 1991. The C-terminal region of the Vibrio fischeri
LuxR protein contains an inducer-independent lux gene activating domain. Proc.
14 Choi, S. H. and E. P. Greenberg 1992. Genetic evidence for multimerization of
LuxR, the transcriptional activator of Vibrio fischeri luminescence. Molec. Mar.
Biol. Biotechnol. 1: 40813.
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
58
15 Clare, B. G., A. Kerr and D. A. Jones 1990. Characteristics of the nopaline
catabolic plasmid in Agrobacterium strains K84 and K1026 used for biological
control of crown gall disease. Plasmid 23: 12637.
16 Dessaux, Y., A. Petit, S. K. Farrand and P. J. Murphy 1998. Opines and opine-
like molecules involved in plant/Rhizobiaceae interactions. In H. P. Spaink,
A. Kondorosi and P. J. Hooykaas (eds), The Rhizobiaceae, pp. 17397. Dordrecht,
The Netherlands: Kluwer Academic Publishers.
17 Dessaux, Y., A. Petit and J. Tempe 1992. Opines in Agrobacterium biology. In
D. P. S. Verma (ed.), Molecular Signals in Plant-Microbe Communications, pp. 10936.
Ann Arbor, MI: CRCPress.
18 Dong, Y. H., A. R. Gusti, Q. Zhang, J. L. Xu and L. H. Zhang 2002. Identification
of quorum-quenching N-acyl homoserine lactonases from Bacillus species. Appl.
Environ. Microbiol. 68: 17549.
19 Dong, Y. H., L. H. Wang, J. L. Xu et al. 2001. Quenching quorum-sensing-
dependent bacterial infection by an N-acyl homoserine lactonase. Nature 411:
81317.
20 Dyson, H. J. and P. E. Wright 2002. Coupling of folding and binding for
unstructured proteins. Curr. Opin. Struct. Biol. 12: 5460.
21 Eberhard, A., A. L. Burlingame, C. Eberhard et al. 1981. Structural identification
of autoinducer of Photobacterium fischeri luciferase. Biochemistry 20: 24449.
22 Egland, K. A. and E. P. Greenberg 2001. Quorum sensing in Vibrio fischeri:
analysis of the LuxR DNA binding region by alanine-scanning mutagenesis.
23 Ellis, J. G., A. Kerr, A. Petit and J. Tempe 1982. Conjugal transfer of nopaline
and agropine Ti-plasmids: the role of agrocinopines. Molec. Gen. Genet. 186:
26973.
24 Farrand, S. K. 1998. Conjugal plasmids and their transfer. In H. P. Spaink,
A. Kondorosi and P. J. J. Hooykaas (eds), The Rhizobiaceae: Molecular Biology of
Model Plant-associated Bacteria, pp. 199233. Dordrecht, The Netherlands: Kluwer
Academic Publishers.
25 Finney, A. H., R. J. Blick, K. Murakami, A. Ishihama and A. M. Stevens 2002.
Role of the C-terminal domain of the alpha subunit of RNA polymerase in LuxR-
dependent transcriptional activation of the lux operon during quorum sensing.
26 Fuqua, C., M. Burbea and S. C. Winans 1995. Activity of the Agrobacterium Ti
plasmid conjugal transfer regulator TraR is inhibited by the product of the traM
gene. J. Bacteriol. 177: 136773.
27 Fuqua, C. and S. C. Winans 1996. Conserved cis-acting promoter elements are
required for density-dependent transcription of Agrobacterium tumefaciens con-
jugal transfer genes. J. Bacteriol. 178: 43540.
28 Fuqua, C. and S. C. Winans 1996. Localization of OccR-activated and TraR-
activated promoters that express two ABC-type permeases and the traR gene of
Ti plasmid pTiR10. Molec. Microbiol. 20: 1199210.
59
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
29 Fuqua, W. C. and S. C. Winans 1994. A LuxR-LuxI type regulatory system
activates Agrobacterium Ti plasmid conjugal transfer in the presence of a plant
tumor metabolite. J. Bacteriol. 176: 2796806.
30 Gage, D. J. 2004. Infection and invasion of roots by symbiotic, nitrogen-fixing
rhizobia during nodulation of temperate legumes. Microbiol. Molec. Biol. Rev. 68:
280300.
31 Genetello, C., N. Van Larebeke, M. Holsters et al. 1977. Ti plasmids of
Agrobacterium as conjugative plasmids. Nature 265: 5613.
32 Goodner, B., G. Hinkle, S. Gattung et al. 2001. Genome sequence of the plant
pathogen and biotechnology agent Agrobacterium tumefaciens C58. Science 294:
23238.
33 Gould, T. A., H. P. Schweizer and M. E. Churchill 2004. Structure of the
Pseudomonas aeruginosa acyl-homoserinelactone synthase LasI. Molec. Microbiol.
53: 113546.
34 Hayman, G. T. and S. K. Farrand 1990. Agrobacterium plasmids encode struc-
turally and functionally different loci for catabolism of agrocinopine-type opines.
Molec. Gen. Genet. 223: 46573.
35 Hayman, G. T. and S. K. Farrand 1988. Characterization and mapping of the
agrocinopine-agrocin 84 locus on the nopaline Ti plasmid pTiC58. J. Bacteriol.
170: 175967.
36 He, X., W. Chang, D. L. Pierce, L. O. Seib, J. Wagner and C. Fuqua 2003. Quorum
sensing in Rhizobium sp. strain NGR234 regulates conjugal transfer (tra) gene
expression and influences growth rate. J. Bacteriol. 185: 80922.
37 Huang, J. J., J. I. Han, L. H. Zhang and J. R. Leadbetter 2003. Utilization of acyl-
homoserine lactone quorum signals for growth by a soil pseudomonad and
Pseudomonas aeruginosa PAO1. Appl. Environ. Microbiol. 69: 59419.
38 Hwang, I., D. M. Cook and S. K. Farrand 1995. A new regulatory element
modulates homoserine lactone-mediated autoinduction of Ti plasmid conjugal
transfer. J. Bacteriol. 177: 44958.
39 Hwang, I., P. L. Li, L. Zhang et al. 1994. TraI, a LuxI homologue, is responsible
for production of conjugation factor, the Ti plasmid N-acylhomoserine lactone
autoinducer. Proc. Natn. Acad. Sci. USA 91: 463943.
40 Hwang, I., A. J. Smyth, Z. Q. Luo and S. K. Farrand 1999. Modulating quorum
sensing by antiactivation: TraM interacts with TraR to inhibit activation of Ti
plasmid conjugal transfer genes. Molec. Microbiol. 34: 28294.
41 Johnson, D. C., A. Ishihama and A. M. Stevens 2003. Involvement of region 4 of
the sigma 70 subunit of RNA polymerase in transcriptional activation of the lux
operon during quorum sensing. FEMS Microbiol. Lett. 228: 193201.
42 Kado, C. I. 1994. Promiscuous DNA transfer system of Agrobacterium tume-
faciens: role of the virB operon in sex pilus assembly and synthesis. Molec.
43 Kent, A. D. and E. W. Triplett 2002. Microbial communities and their interac-
tions in soil and rhizosphere ecosystems. A. Rev. Microbiol. 56: 21136.
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
60
44 Kerr, A. 1971. Acquisition of virulence by non-pathogenic isolates of
Agrobacterium radiobacter. Physiol. Plant Pathol. 1: 2416.
45 Kerr, A. 1969. Transfer of virulence between isolates of Agrobacterium. Nature
223: 11756.
46 Kerr, A., P. Manigault and J. Tempe 1977. Transfer of virulence in vivo and in vitro
in Agrobacterium. Nature 265: 5601.
47 Leadbetter, J. R. and E. P. Greenberg 2000. Metabolism of acyl-homoserine
lactone quorum-sensing signals by Variovorax paradoxus. J. Bacteriol. 182:
69216.
48 Lessl, M., D. Balzer, W. Pansegrau and E. Lanka 1992. Sequence similarities
between the RP4 Tra2 and the Ti VirB region strongly support the conjugation
model for T-DNA transfer. J. Biol. Chem. 267: 2047180.
49 Lessl, M. and E. Lanka 1994. Common mechanisms in bacterial conjugation and
Ti-mediated T-DNA transfer to plant cells. Cell 77: 3214.
50 Luo, Z. Q. and S. K. Farrand 1999. Signal-dependent DNA binding and func-
tional domains of the quorum-sensing activator TraR as identified by repressor
activity. Proc. Natn. Acad. Sci. USA 96: 900914.
51 Luo, Z. Q., Y. Qin and S. K. Farrand 2000. The antiactivator TraM interferes
with the autoinducer-dependent binding of TraR to DNA by interacting with
the C-terminal region of the quorum-sensing activator. J. Biol. Chem. 275:
771322.
52 Luo, Z. Q., A. J. Smyth, P. Gao, Y. Qin and S. K. Farrand 2003. Mutational
analysis of TraR. Correlating function with molecular structure of a quorum-
sensing transcriptional activator. J. Biol. Chem. 278: 1317382.
53 Maris, A. E., M. R. Sawaya, M. Kaczor-Grzeskowiak et al. 2002. Dimerization
allows DNA target site recognition by the NarL response regulator. Nat. Struct.
Biol. 9: 7718.
54 Medina, G., K. Juarez, B. Valderrama and G. Soberon-Chavez 2003. Mechanism
of Pseudomonas aeruginosa RhlRtranscriptional regulation of the rhlABpromoter.
J. Bacteriol. 185: 597683.
55 More, M. I., L. D. Finger, J. L. Stryker et al. 1996. Enzymatic synthesis of a
quorum-sensing autoinducer through use of defined substrates. Science 272:
16558.
56 Morris, R. O. 1990. Genes specifying auxin and cytokinin biosynthesis in prokar-
yotes. In P. J. Davies (ed.), Plant Hormones and Their Role in Plant Growth and
Development, pp. 63655. Dordrecht, The Netherlands: Kluwer Academic
Publishers.
57 Nester, E. W., D. J. Merlo, M. H. Drummond et al. 1977. The incorporation and
expression of Agrobacterium plasmid genes in crown gall tumors. Basic Life Sci.
9: 18196.
58 Oger, P. and S. K. Farrand 2001. Co-evolution of the agrocinopine opines and the
agrocinopine-mediated control of TraR, the quorum-sensing activator of the Ti
plasmid conjugation system. Molec. Microbiol. 41: 117385.
61
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
59 Oger, P. and S. K. Farrand 2002. Two opines control conjugal transfer of an
Agrobacteriumplasmid by regulating expression of separate copies of the quorum-
sensing activator gene traR. J. Bacteriol. 184: 112131.
60 Oger, P., K. S. Kim, R. L. Sackett, K. R. Piper and S. K. Farrand 1998. Octopine-
type Ti plasmids code for a mannopine-inducible dominant-negative allele of
traR, the quorum-sensing activator that regulates Ti plasmid conjugal transfer.
61 Pappas, K. M. and S. C. Winans 2003. A LuxR-type regulator from Agrobacterium
tumefaciens elevates Ti plasmid copy number by activating transcription of plas-
mid replication genes. Molec. Microbiol. 48: 105973.
62 Pappas, K. M. and S. C. Winans 2003. The RepA and RepB autorepressors and
TraR play opposing roles in the regulation of a Ti plasmid repABC operon. Molec.
63 Parkinson, G., C. Wilson, A. Gunasekera et al. 1996. Structure of the CAP-DNA
complex at 2.5 angstroms resolution: a complete picture of the protein-DNA
interface. J. Molec. Biol. 260: 395408.
64 Parsek, M. R., D. L. Val, B. L. Hanzelka, J. E. Cronan, Jr. and E. P. Greenberg
1999. Acyl homoserine-lactone quorum-sensing signal generation. Proc. Natn.
Acad. Sci. USA 96: 43605.
65 Pierson, L. S. I., D. W. Wood and S. B. von Bodman 1999. Quorum sensing in
plant-associated bacteria. In G. M. Dunny and S. C. Winans (eds), Cell-Cell
Signaling in Bacteria, pp. 10115. Washington, DC: ASM Press.
66 Piper, K. R., S. Beck von Bodman and S. K. Farrand 1993. Conjugation factor of
Agrobacterium tumefaciens regulates Ti plasmid transfer by autoinduction. Nature
362: 44850.
67 Piper, K. R., S. Beck Von Bodman, I. Hwang and S. K. Farrand 1999. Hierarchical
gene regulatory systems arising from fortuitous gene associations: controlling
quorumsensing by the opine reguloninAgrobacterium. Molec. Microbiol. 32: 107789.
68 Pohlman, R. F., H. D. Genetti and S. C. Winans 1994. Common ancestry
between IncN conjugal transfer genes and macromolecular export systems of
plant and animal pathogens. Molec. Microbiol. 14: 65568.
69 Qin, Y., Z. Q. Luo and S. K. Farrand 2004. Domains formed within the
N-terminal region of the quorum-sensing activator TraR are required for tran-
scriptional activation and direct interaction with RpoA from agrobacterium.
J. Biol. Chem. 279: 4084451.
70 Qin, Y., Z. Q. Luo, A. J. Smyth et al. 2000. Quorum-sensing signal binding
results in dimerization of TraR and its release from membranes into the cyto-
plasm. EMBO J. 19: 521221.
71 Sheng, J. and V. Citovsky 1996. Agrobacterium-plant cell DNA transport: have
virulence proteins, will travel. Plant Cell 8: 1699710.
72 Shoemaker, B. A., J. J. Portman and P. G. Wolynes 2000. Speeding molecular
recognition by using the folding funnel: the fly-casting mechanism. Proc. Natn.
Acad. Sci. USA 97: 886873.
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
62
73 Stevens, A. M., N. Fujita, A. Ishihama and E. P. Greenberg 1999. Involvement of
the RNA polymerase alpha-subunit C-terminal domain in LuxR-dependent activ-
ation of the Vibrio fischeri luminescence genes. J. Bacteriol. 181: 47047.
74 Stevens, A. M. and E. P. Greenberg 1997. Quorum sensing in Vibrio fischeri:
essential elements for activation of the luminescence genes. J. Bacteriol. 179:
55762.
75 Swiderska, A., A. K. Berndtson, M. R. Cha et al. 2001. Inhibition of the
Agrobacterium tumefaciens TraR quorum-sensing regulator. Interactions with
the TraM anti-activator. J. Biol. Chem. 276: 4944958.
76 Trott, A. E. and A. M. Stevens 2001. Amino acid residues in LuxR critical for its
mechanismof transcriptional activation during quorumsensing in Vibrio fischeri.
77 Urbanowski, M. L., C. P. Lostroh and E. P. Greenberg 2004. Reversible acyl-
homoserine lactone binding to purified Vibrio fischeri LuxR protein. J. Bacteriol.
186: 6317.
78 Val, D. L. and J. E. Cronan, Jr 1998. In vivo evidence that S-adenosylmethionine
and fatty acid synthesis intermediates are the substrates for the LuxI family of
autoinducer synthases. J. Bacteriol. 180: 264451.
79 Van Larebeke, N., G. Engler, M. Holsters et al. 1974. Large plasmid in
Agrobacterium tumefaciens essential for crown gall-inducing ability. Nature 252:
16970.
80 Vannini, A., C. Volpari and S. Di Marco 2004. Crystal structure of the quorum-
sensing protein TraM and its interaction with the transcriptional regulator TraR.
J. Biol. Chem. 279: 242916.
81 Vannini, A., C. Volpari, C. Gargioli et al. 2002. The crystal structure of the
quorum sensing protein TraR bound to its autoinducer and target DNA. EMBO
J. 21: 4393401.
82 Walker, T. S., H. P. Bais, E. Grotewold and J. M. Vivanco 2003. Root exudation
and rhizosphere biology. Plant Physiol. 132: 4451.
83 Watson, W. T., T. D. Minogue, D. L. Val, S. Beck von Bodman and M. E. Churchill
2002. Structural basis and specificity of acyl-homoserine lactone signal produc-
tion in bacterial quorum sensing. Molec. Cell 9: 68594.
84 Whitehead, N. A., A. M. Barnard, H. Slater, N. J. Simpson and G. P. Salmond
2001. Quorum-sensing in Gram-negative bacteria. FEMS Microbiol. Rev. 25:
365404.
85 Winans, S. C. 1992. Two-way chemical signaling in Agrobacterium-plant inter-
actions. Microbiol. Rev. 56: 1231.
86 Winans, S. C., J. Zhu and M. I. More 1999. Cell density-dependent gene expres-
sion by Agrobacterium tumefaciens during colonization of crown gall tumors. In
G. M. Dunny and S. C. Winans (eds), Cell-Cell Signaling in Bacteria, pp. 11728.
Washington, DC: ASM Press.
87 Wood, D. W., J. C. Setubal, R. Kaul et al. 2001. The genome of the natural genetic
engineer Agrobacterium tumefaciens C58. Science 294: 231723.
63
P
L
A
N
T
B
A
C
T
E
R
I
A
I
N
T
E
R
A
C
T
I
O
N
S
88 Zhang, H. B., C. Wang and L. H. Zhang 2004. The quormone degradation
system of Agrobacterium tumefaciens is regulated by starvation signal and stress
alarmone (p)ppGpp. Molec. Microbiol. 52: 1389401.
89 Zhang, H. B., L. H. Wang and L. H. Zhang 2002. Genetic control of quorum-
sensing signal turnover in Agrobacterium tumefaciens. Proc. Natn. Acad. Sci.
USA 99: 463843.
90 Zhang, L., P. J. Murphy, A. Kerr and M. E. Tate 1993. Agrobacteriumconjugation
and gene regulation by N-acyl-L-homoserine lactones. Nature 362: 4468.
91 Zhang, L. H. and A. Kerr 1991. A diffusible compound can enhance conju-
gal transfer of the Ti plasmid in Agrobacterium tumefaciens. J. Bacteriol. 173:
186772.
92 Zhang, R. G., T. Pappas, J. L. Brace et al. 2002. Structure of a bacterial quorum-
sensing transcription factor complexed with pheromone and DNA. Nature
417: 9714.
93 Zhu, J., J. W. Beaber, M. I. More et al. 1998. Analogs of the autoinducer
3-oxooctanoyl-homoserine lactone strongly inhibit activity of the TraR protein of
Agrobacterium tumefaciens. J. Bacteriol. 180: 5398405.
94 Zhu, J., P. M. Oger, B. Schrammeijer et al. 2000. The bases of crown gall
tumorigenesis. J. Bacteriol. 182: 388595.
95 Zhu, J. and S. C. Winans 1998. Activity of the quorum-sensing regulator TraR
of Agrobacterium tumefaciens is inhibited by a truncated, dominant defective
TraR-like protein. Molec. Microbiol. 27: 28997.
96 Zhu, J. and S. C. Winans. 1999. Autoinducer binding by the quorum-sensing
regulator TraR increases affinity for target promoters in vitro and decreases TraR
turnover rates in whole cells. Proc. Natn. Acad. Sci. USA 96: 48327.
97 Zhu, J. and S. C. Winans 2001. The quorum-sensing transcriptional regulator
TraRrequires its cognate signaling ligand for protein folding, protease resistance,
and dimerization. Proc. Natn. Acad. Sci. USA 98: 150712.
C
.
E
.
W
H
I
T
E
A
N
D
S
.
C
.
W
I
N
A
N
S
64
CHAPTER 4
Jamming bacterial communications: new
strategies to combat bacterial infections
and the development of biofilms
Michael Givskov
Center for Biomedical Microbiology, BioCentrum Technical University of Denmark
Morten Hentzer
Carlsberg Biosector, Carlsberg A/S, Valby, Denmark
INTRODUCTION
The growth and activity of microorganisms affect our lives in both
positive and negative ways. We have, since early times, tried to combat
unwanted microbes and utilize those expressing useful traits.
Microorganisms can cause diseases and chronic infections in humans,
animals, and plants. In medicine, agriculture and fish farming, treatment
scenarios are based on antimicrobial compounds such as antibiotics, with
toxic and growth-inhibitory properties. Control of growth by eradication
of bacteria is one of the most important scientific achievements.
Unfortunately, bacteria have become gradually more and more resistant
to antibiotics, and infections caused by resistant bacteria are on a dramatic
increase. It has recently become apparent that the bacterial lifestyle also
contributes significantly to this problem. The traditional way of culturing
bacteria as planktonic, liquid cultures imprinted the view that bacteria live
as unicellular organisms. Although it must be emphasized that such test-
tube studies have led to fundamental insights into basic life processes and
have unraveled complex intracellular regulatory networks, it is now clear
that in nature microbial activity is mainly associated with surfaces and we
as scientists must therefore turn our attention to this sessile mode of
growth (33). It appears that the ability to formsurface-associated, structured
and cooperative consortia (referred to as biofilms) is one of the most
remarkable and ubiquitous characteristics of bacteria (12). In this sessile
life form, bacteria can cause various problems in industrial settings, rang-
ing from corrosion and biofouling to food contamination. In clinical
microbiology, the biofilm mode of bacterial growth has attracted particular
attention. Many persistent and chronic infections (including pulmonary
65
infections of cystic fibrosis (CF) patients, periodontitis, otitis media, biliary
tract infection, and endocarditis) as well as the colonization of medical
implants (catheters, artificial heart valves, etc.) are now believed to be
intrinsically linked to the formation of bacterial biofilms (15).
Biofilm infections are becoming more and more common owing to the
more frequent insertion of artificial medical implants. Taking the change
in age distribution (as the proportion of elderly people increases) in the
developed world into account, it will be of the utmost importance to address
this major medical and public health problem in the twenty-first century.
A recent public announcement from the US National Institutes of Health
stated that more than 80% of all microbial infections involve biofilms (17).
The capability of forming a biofilm within the human body is therefore
considered to represent a pathogenic trait per se. Biofilm infections (often
caused by opportunistic pathogenic bacteria) are particularly problematic
as they give rise to chronic infections, inflammation, and tissue damage.
For the clinician, the major problem is that bacteria living in biofilms can
often withstand the host defense systems and are markedly tolerant to
antibiotics, often exceeding the highest deliverable doses of antibiotics
and thus making an efficient treatment impossible. Treatment of Biofilm
infections therefore calls for new strategies.
WHY QS INHIBITORS?
The most obvious alternative to antibiotic-mediated killing or growth
inhibition would be to attenuate the bacteria with respect to their patho-
genicity. The ability to organize structurally and to distribute activities between
the different bacteria demands a high degree of coordinated cellcell inter-
action reminiscent of that seen in multicellular organisms. Such interactions
involve cell-to-cell communication systems to adjust the various functions
of specialized members of the population. In fact, many bacteria employ
intercellular communication systems that rely on small signal molecules to
monitor their ownpopulationdensities ina process knownas quorumsensing
(QS) as described in detail in previous chapters. Signal molecules that are
released by the cells modulate the activity of other cells in the vicinity, thus
regulating collective activities. Cell-to-cell signaling is referred to as QS
because the system enables a given bacterial species to sense when a critical
(i.e. quorate) population density has been reached in the host and in response
activate the expression of target genes required for succession (32).
A diverse range of bacterial metabolites is now recognized as intercel-
lular QS signals. The list includes peptides, butyrolactones, palmitic acid
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
66
methyl esters, quinones, and cyclic dipeptides. Although the principle
behind signal-mediated gene expression in both Gram-positive and
Gram-negative bacteria is shared, the molecular mechanisms and signal
molecules differ. Peptides serve as the signal molecules in many Gram-
positive bacteria; the signals are sensed via two component phosphorelay
systems (54). Among Gram-negative bacteria, N-acyl-L-homoserine lactone
(AHL)-dependent QS systems are particularly widespread. These systems
are used to coordinate expression of traits that are fundamental to the
interaction of bacteria with each other, with their environment, and particu-
larly with higher organisms, covering a variety of functions ranging from
pathogenic to symbiotic interactions. These include expression of biolumi-
nescence in the specialized light organs of squids and fish, surface
colonization, and biofilm development, as well as production of virulence
factors and hydrolytic enzymes during infection of eukaryotic hosts.
Back in the mid 1990s, the strictly cell-density-dependent formation of
a swarm colony observed in Serratia liquefaciens led to the hypothesis that a
QS system triggered colony expansion once a critical size, the quorum, had
been attained. The behavior of the swrI knockout mutant on swarm media
in the presence or absence of exogenously added BHL (N-butanoyl-L-
homoserine lactone) clearly supported this hypothesis (27). The swr system
controls swarming but not swimming motility and was therefore the first
published example of a surface-associated behavior that is controlled by
means of QS. Production of several virulence factors was in fact controlled
by the swr QS system by means of the QS-controlled lipB transporter (94).
lipB is part of an operon encoding a type I secretion system, which is
responsible for the secretion of extracellular lipase, metalloprotease, and
S-layer protein. Accordingly, QS systems seem to link virulence with sur-
face activity, which has opened a new perspective for controlling undesired
microbial activity. Throughout the years, we and our collaborators have
documented that blockade of AHL-mediated communication represents an
effective approach to interfere with surface colonization and to attenuate
the virulence of bacterial pathogens (34, 43, 89). We have denoted com-
pounds capable of this as anti-pathogenic drugs. QS systems regulate (in
the classical sense) non-essential phenotypes, so if the new drugs could be
based on QS inhibitor (QSI) compounds the bacteria should not, at least in
theory, meet the same hard selective pressure as is imposed by antibiotics.
The major advantage of this creative strategy for antipathogenic therapy is
that it is likely to circumvent the problem of resistance, which is intimately
connected to the use of conventional antibacterial agents such as antibio-
tics. Although resistant mutants may arise, they are not expected to have
67
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
a selective growth advantage per se and thus should not outcompete the
parental strain. Furthermore, QSI compounds are not expected to elimi-
nate communities of helpful and beneficial bacteria present in the host (for
example, the gut flora). The antipathogenic drug approach is generic in
nature and relevant to a broad spectrum of scientists such as microbio-
logists and medical doctors working in the field of infectious diseases and
artificial implants and those engineers who are involved in maintenance of
industrial facilities and water pipelines.
PSEUDOMONAS AERUGINOSA AND QS
Pseudomonas aeruginosa is an increasingly prevalent opportunistic
human pathogen and is the most common Gram-negative bacterium
found in nosocomial and life-threatening infections of immunocompro-
mised patients (114). Patients with cystic fibrosis (CF) are especially dis-
posed to P. aeruginosa infections and for these persons the bacterium is
responsible for high rates of morbidity and mortality (46, 60). P. aeruginosa
makes several virulence factors that contribute to its pathogenesis, as
described in Chapter 1 of this volume. QS plays a key role in orchestrating
the expression of many of these virulence factors, such as exoproteases,
siderophores, exotoxins, and several secondary metabolites, and partici-
pates in the development of biofilms (18, 43, 81, 121). P. aeruginosa pos-
sesses two QS systems: the LasRLasI and the RhlRRhlI, with the cognate
signal molecules OdDHL (N-[3-oxo-dodecanoyl]-L-homoserine lactone) and
BHL, respectively. The two systems do not operate independently: the las
system positively regulates expression of the rhl system (57, 86).
Intertwined in this QS hierarchy is the quinolone signal (PQS) system,
which provides a link between the las and rhl QS systems (85) (70).
In addition to mediate communication between bacteria, immuno-
assays in vitro on human leukocytes have shown that OdDHL possesses
immunomodulatory properties, e.g. inhibition of lymphocyte prolifera-
tion and downregulation of tumor necrosis factor alpha and IL-12 produc-
tion (110). OdDHL has been demonstrated to activate T-cells in vivo to
produce the inflammatory cytokine g-interferon (104), thereby potentially
promoting a Th-2 dominated response leading to increased tissue damage
and inflammation. OdDHL also possesses proinflammatory, immuno-
modulatory, and vasorelaxant properties (7). In vivo evidence has left little
doubt that QS controlled gene expression plays an important role in the
development of P. aeruginosa infections (29, 82, 125, 126).
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
68
THE INVOLVEMENT OF QS IN P. AERUGINOSA BIOFILM
DEVELOPMENT
Abiofilm is a structured community of bacterial cells enclosed in a self-
produced polymeric matrix and adherent to an inert or living surface.
Bacterial biofilms are considered ubiquitous in nature (12). In most cases
biofilms form at the interface between a solid surface and an aqueous
phase. According to the prevailing model, biofilm development is believed
to proceed through a temporal series of stages (79). This hypothesis has
gained momentum from the isolation of mutants that appear to be arrested
at certain stages of this development (37, 62, 75, 87, 113). In the initial
phase, bacteria attach to a surface, grow, and then proliferate to form
microcolonies. These microcolonies develop into hydrated structures in
which bacterial cells are enmeshed in a matrix of self-produced slime
(58). For reviews see (109, 120). This slime is commonly referred to as
exopolymeric substances (EPS) and may, dependent on the type of bacteria
and their overall metabolic status, consist of all major classes of macro-
molecules (proteins, polysaccharides, DNA, and RNA) in addition to pepti-
doglycan, lipids, and phospholipids (118, 124). Mature biofilms typically
consist of towers and mushrooms of cells enmeshed in copious
amounts of EPS, separated by channels and interstitial voids to allow
convective flow to transport nutrients to interior parts of the biofilm and
remove waste products. This structural heterogeneity is commonly referred
to as the biofilm architecture.
The involvement of QS in P. aeruginosa biofilm architecture is based on
circumstantial evidence. Davies et al. (18) demonstrated that a lasI mutant
formed flat and undifferentiated biofilms in contrast to the wild type, which
formed the characteristic biofilm architecture. The authors also found that
the flat biofilms formed by the quorum mutant were eradicated with SDS
treatment, suggesting that development of the characteristic biofilm toler-
ance to antimicrobial treatment would in fact require proper biofilm differ-
entiation, which in turn would rely on QS-dependent gene expression.
Studies by others revealed no differences between the biofilms of the
wild type and those formed by signal-negative mutants (44, 108). Purevdorj
et al. (88) reported minor structural differences between wild-type and mutant
biofilms, but these differences were only apparent when particular hydro-
dynamic conditions were used for growing the biofilms. This indicates that
the experimental settings influence the look of the biofilm.
One of the key issues for biofilm structure seems to be the growth
conditions. OToole et al. (74) showed the catabolite repression control (Crc)
69
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
protein to be involved; Klausen et al. (53) demonstrated that the availability of
different carbon sources affected P. aeruginosa biofilm structures. In general,
glucose-grown biofilms are very heterogeneous, exhibiting discrete towers and
mushroomstructures separated by water channels and voids. On the contrary,
biofilms grown under identical conditions but withcitrate as the carbon source
appear flat and uniform. In addition, when the mediumis supplemented with
nitrate as alternative electron acceptor to oxygen, compact and less differen-
tiated biofilms are obtained (authors unpublished results).
The involvement of QS in biofilm formation has also been demon-
strated for Burkholderia cepacia (48, 49), Aeromonas hydrophila (61),
Pseudomonas putida (105) and S. liquefaciens (56). AHL-negative mutants
of B. cepacia and A. hydrophila showed defects in the late stages of biofilm
development and thus were unable to form typical structured biofilms.
Recent work on S. liquefaciens MG1 has demonstrated that expression of
QS-controlled genes is crucial at a specific stage for the development of its
characteristic filamentous biofilm (56).
The advent of cDNA microarray technology has provided great insight
into differential gene expression in biofilm bacteria. Despite the striking
differences between the lifestyles of planktonic and biofilm bacteria, only
1%3%of the 5,570 predicted genes in P. aeruginosa PAO1 show differential
expression in the two modes of growth (39, 41, 43, 119). The gene expres-
sion profile of P. aeruginosa biofilms exhibited greater similarity to that of
planktonic cultures in stationary phase than to that of planktonic cultures
in mid exponential growth or early stationary phase, hence supporting the
view that biofilms are dominated by bacteria with stationary-phase physio-
logy. The genes upregulated during biofilm growth are involved in many
different cellular processes, such as transcription and translation, energy
metabolism, intermediate metabolism, transport, cofactor biosynthesis,
amino acid biosynthesis, and cell-wall synthesis. Notably, a large number
of genes related to denitrification and anaerobic respiration were upregu-
lated in biofilms, indicating that anaerobiosis is important in microbial
biofilm physiology. A study of the transient gene expression during biofilm
development showed that many genes involved in anaerobic respiration
became further activated as the biofilm matured. At late stages of biofilm
development, transcription of a gene cluster related to the filamentous
phage Pf1 became strongly activated (more than 200-fold) (39, 41, 119).
The Pf1-like genes have been shown to be involved in bacterial cell death
and biofilm dispersal (117).
The involvement of QS in biofilm formation and development has
been established by several research groups (18, 20, 95) but the identity
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
70
of a class of specific QS-controlled genes that influence biofilm formation
has remained elusive. It is noteworthy that most attempts to identify QS-
controlled genes in P. aeruginosa have been performed on planktonic
cultures and not on biofilm-growing cells, hence neglecting the possibility
that the constitution of the QS regulon can be influenced by environmental
factors. In our laboratory, we performed a parallel identification of QS-
controlled genes in planktonic and biofilm-growing P. aeruginosa. We
found that, indeed, the QS regulon is influenced by environmental condi-
tions; a number of biofilm-specific QS-controlled genes could clearly be
identified. These genes are of course prime suspects for mediating the role
of QS in biofilm development (39, 41). Interestingly, many of the biofilm-
specific quorum-sensing-controlled genes were related to responses to
iron-limitation. We hypothesize that in the biofilm mode of growth either
the availability of iron is reduced, owing to binding to the EPS matrix, or,
alternatively, biofilm cells require elevated amounts of iron relative to
planktonic cells to support growth. The increased demand for iron during
biofilm growth could be due to anaerobiosis as many of the genes compris-
ing the nitrate respiratory pathway, which we found to be differentially
upregulated in biofilms, are iron-containing proteins.
NATURAL BLOCKERS: A EUKARYOTIC DEFENSE STRATEGY
AGAINST BIOFILMS AND BIOFOULING
The ability of bacteria to form biofilms is a major challenge for living
organisms such as humans, animals, and marine eukaryotes (55, 59).
Marine plants are, in the absence of more advanced immune systems,
prone to disease (11, 30). Bacteria can be highly detrimental to marine
algae and other eukaryotes (59). The Australian red macroalga Delisea
pulchra (23) produces a range of halogenated furanone compounds (for
structures and numbering see Figure 4.1), which display antifouling and
antimicrobial properties (21, 22, 92). This particular alga originally
attracted the attention of marine biologists because it was devoid of exten-
sive surface colonization, i.e. biofouling, unlike other plants in the same
environment. Biofouling is primarily caused by marine invertebrates and
plants, but bacteria are believed to be the first colonizers of submerged
surfaces, providing an initial conditioning biofilm to which other marine
organisms may attach (93). Therefore, the abundance and composition of
the bacterial community on the surface will significantly affect the sub-
sequent development of a biofouling community (4, 38). To cope with this,
eukaryotes have developed chemical defense mechanisms (19, 22, 116)
71
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
which in several cases are secondary metabolites that inhibit bacterial
colonization-relevant phenotypes (52, 67, 100).
The effect of furanones on bacterial colonization phenotypes is due to
interference with specific cell processes rather than to general toxicity or
surface modification (52, 68). The D. pulchra furanone compounds consist
of a furan ring structure with a substituted acyl chain at the C-3 position and
a bromine substitution at the C-4 position. The substitution at the C-5
position may vary in terms of side-chain structure. The natural furanones
are halogenated at various positions by bromide, iodide, or chloride (23).
For furanone structures, see Figure 4.1. D. pulchra produces a minimum of
BHL
O
O
O
O
O
O
O
O
O
O
O
S S
S
N
Br
Br Br
Br
Br
Br
Br Br
Br
S
NO
2
O
O
O
O
OH
OH
HO
O
O
O
O
O
O
O
O
O O
O O
O
O
O
O
O
H H
N
H
N
H
N
H
N
H
N
H
H H
O
O
O
OHHL OdDHL
furanone C-2 C-4 C-8 C-30 C-56
patulin penicillic acid GC-7
C(6)
C(3)
C(1)
O(1)
C(4)
C(3)
C(5)
C(1) C(3)
2-heptylthioacetyl-homoserine lactone N -pentylsulfonyl-homoserine lactone
4-nitropyridine-N -oxide
Figure 4.1. AHL signals and QSI compounds. The upper line shows the Pseudomonas
aeruginosa signal molecules BHL and OdDHL and the Vibrio fischeri signal OHHL with
carbon atom numbering. The rest are QSI compounds. The second line shows the two
synthetic sulfur compounds developed from AHL by rational drug design. The third line
shows the fungal metabolites patulin and penicillic acid, a metabolite (GC-7) isolated from
garlic, and the synthetic compound 4-NPO. The bottom line shows three natural furanone
compounds, C-2 (with carbon atomnumbering), C-4 and C-8, isolated fromDelisea pulchra,
and two synthetic derivatives, C-30 and C-56.
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
72
30 different species of halogenated furanone compound, which are stored in
specialized vesicles. The compounds are released at the surface of the thallus at
concentrations ranging from 1 to 100 ngcm
2
(26, 93). Field experiments
have demonstrated that the surface concentration of furanones is inversely
correlated to the degree of colonization by marine bacteria (52).
Several of the furanone compounds that exhibit structural similarity to the
short-chain AHL molecules inhibited swarming motility of S. liquefaciens MG1
(34). This, taken together with inhibition of LuxR-controlled transcriptional
fusions, strongly suggested that furanone compounds competed with the
cognate signals for the SwrRand the LuxRreceptors. Based on the observation
that biofilm formation on submerged surfaces precedes the attraction of
higher fouling organisms, Givskov et al. (34) hypothesized that the D. pulchra
furanones constitute a specific means of eukaryotic interference with bacterial
communication, surface colonization and virulence factor expression, i.e.
multicellular behavior. Extensive experimental evidence in support of this
hypothesis has accumulated during the years and include the observations
that furanones inhibit QS-regulated expression of Vibrio fischeri biolumines-
cence (63), virulence factor production and pathogenesis in P. aeruginosa (42,
43), luminescence and virulence in vivo of the black tiger prawn pathogen
Vibrio harveyi (64), and finally virulence of Erwinia carotovora (66).
Furanone-mediated displacement of
3
H-labeled OHHL molecules
from LuxR-overproducing Escherichia coli cells supported the assumption
of a direct interaction between furanones and LuxR homologs (63). At the
same time it was puzzling that there was no substantial affinity of a labeled
furanone for E. coli cells overproducing LuxR (65). This apparent paradox
was resolved with the finding that halogenated furanones accelerate the
degradation of the LuxR protein (65). The authors discovered that the half-
life of the protein is reduced up to 100-fold in the presence of halogenated
furanones. This suggests that halogenated furanones modulate LuxR activ-
ity by destabilization, rather than by protecting the transcriptional activator
from interaction with the cognate signal. The furanone-dependent reduc-
tion in the cellular concentration of the LuxR protein correlated with a
reduction in expression of a plasmid encoded P
luxI
-gfp(ASV) fusion, sug-
gesting that the reduction in LuxR concentration is the mechanism by
which furanones control QS (65).
SYNTHETIC ANALOGS
AHL-dependent QS systems may be jammed inseveral ways, inparticular
by inhibition of AHL signal synthesis, increased AHL signal degradation, and
73
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
blockade of AHL signal reception. Although it is conceivable that AHL
biosynthesis could be effectively obstructed by blockade of AHL synthases,
no specific inhibitors for this class of enzymes have yet been derived (80).
Dong et al. (24) isolated an enzyme, named AiiA, from Bacillus sp. that
inactivates AHLs by hydrolyzing the lactone bond. It was shown that trans-
genic plants expressing AiiAexhibit enhanced resistance to Erwinia carotovora
infections. The authors suggested that, because this bacterium employs
AHL-dependent QS to control expression of plant pathogenic traits, the atten-
uated virulence is likely to be a direct consequence of signal degradation.
Whether this strategy is applicable to the treatment of human infections
remains to be seen.
Molecules capable of antagonizing binding of the cognate AHL signal
to the LuxR-type receptor would block signal transduction and therefore
jam the communication system. Competitive inhibitors are likely to be
structurally related to the native AHL signal in order to bind to and occupy
the AHL binding site but fail to activate the LuxR-type receptor.
Non-competitive inhibitors may show little or no structural similarity to
AHL signals, as these molecules are thought to bind differently to the
receptor protein. However, the recent finding by Chun et al. (9) that the
human airway might protect itself by producing a lactonase which targets
and inactivates certain derivatives of AHLs (in particular OdDHL) might
limit the value of putative drugs based on agonist design.
Synthetic AHL analogs described so far mainly fall into two categories:
(i) compounds differing from the cognate inducers by the length and
composition of the acyl side chain; and (ii) compounds in which the lactone
ring has been substituted by other heterocycles or carbocycles. With respect
to (i), Hanzelka and Greenberg (36) performed a series of detailed studies
of analogs to OHHL in experimental scenarios including E. coli harboring a
ptacluxR overexpression system or the lux operon devoid of the luxI gene.
From these studies, it was obvious that the length of the acyl side chain
plays a critical role in binding to the LuxR receptor and in agonistic as well
as antagonistic activity. Passador et al. (129) performed a similar series of
studies with homologs of the OdDHL signal on recombinant Escherichia
coli with plasmid-borne ptaclasR (for competitive binding assays) or lasR,
PlasBlacZ fusions (for agonistic activity). For interaction with the LasR
protein, the authors found that the lower limit for the acyl chain length
consisted of eight carbons. The substitution at the C(3
0
)-position (see
Figure 4.1), which is often referred to as 3-oxo-, is important for the
agonistic activity of AHLs, but so far no clear prediction on the antagonistic
effect of a modification of this position can be made (96, 69, 128).
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
74
The C(1
0
), on the other hand, is sensitive to substitution with sulfur such as
the n-pentylsulfonyl-homoserine lactone shown in Figure 4.1. Sulfonamide
derivatives of AHLs show QSI activity against the LuxR protein (73) but not
the LasR or RhlR proteins (authors unpublished observations). However,
substitution of the C(3
0
) atom with sulfur, as in the 2-heptylthioacetyl-
homoserine lactone (Figure 4.1) shows strong activity against both LuxR-
and LasR-controlled QS systems (84).
With respect to chirality, natural AHL signals are L-isomers whereas
D-isomers are generally devoid of biological activity (8, 50, 69). Importantly,
D-isomers do not function as antagonists, indicating that they do not bind to
the LuxR-type receptor (50).
With respect to (ii), the effects of changes in the composition and size
of the homoserine lactone ring with either a 12-carbon (carrying a 3-oxo-
substitution at the C(3
0
), or a four-carbon acyl side chain were recently
investigated by Smith et al. (102, 103). From the screening and testing of
combinatory libraries it was concluded that ring size, the keto group (at
C(1
0
)) and the presence of saturated carbons in the ring strongly affected the
inhibitory activity of the molecule on the LasR system of P. aeruginosa. Only
slight variations in these key positions, such as a change from a saturated
ring to an aromatic benzene ring, transformed the molecule from an
agonist into an antagonist (102). Based on these data, it was suggested
that the presence of an aromatic ring interferes with the ability to activate
LasR. Accordingly, we recently found that the compound 4-nitropyridine-
N-oxide inhibited the P. aeruginosa QS systems (91). Substitution of the
O(1) with sulfur creates a thiolactone structure, which also antagonizes
LuxR activation as demonstrated by Schaefer et al. (96); we have recently
isolated the garlic compound C-7, carrying two sulfur atoms in the ring,
with similar activity (84).
The natural D. pulchra furanones were also used as scaffolds for
synthetic libraries. Manefield et al. (65) found that the longer the aliphatic
side chain protruding from the furan ring, the less active are the furanone
compounds. On the other hand, removal of the side chain of the natural
furanones such as C-30 and C-56 increased their inhibitory activity
(Figure 4.1). Concurrent with this ranking was the finding that C-30
promotes a more rapid LuxR turnover than C-8 (65). Furthermore, C-56
and C-30 were the only furanone compounds capable of significant repres-
sion of QS-regulated gene expression in P. aeruginosa. This was initially
based on measurement of the virulence factors and later verified by
DNA-microarray-based transcriptomics (42, 43). The inhibitory effect was
competitive with that of OdDHL. The bromine atoms also influence
75
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
inhibitory activity (Figure 4.1). C-8 and C-56 both lack the C(4) Br-atom in
comparison with C-2 and C-30, the result of which was found to be a
decrease in the QSI activity. However, replacing the single Br atom in
C-56 with a methyl group completely abolishes the QSI function.
SCREENS FOR QS INHIBITORS
Throughout the past decade, a number of genetic constructs have been
employed to test compounds for their effect on QS systems. The majority of
these studies have been performed on the V. fischeri, Erwinia carotovora,
P. aeruginosa, Chromobacterium violaceum and Agrobacterium tumefaciens
quorum sensors. The experimental tools and scenarios include binding of
3
H-labeled AHL signals and their homologs to recombinant Escherichia coli
cells overexpressing the respective receptor proteins as well as activation or
competitive inhibition of QS target genes and their reporter fusions. A classic
reporter system is QS-controlled violacein production in C. violaceum (69).
Violacein is the purple compound with antibiotic properties produced by
certain derepressed strains of C. violaceum. The I-mutant CVO26 has played
an important role as AHL monitor for a number of laboratories and has also
been used to describe D. pulchra and furanone action as illustrated on the front
cover of Microbiology by Manefield et al. (63). Escherichia coli containing the
plasmid pSB401 has been used to measure the induction of luminescence by
various AHL analogs as well as reduction in luminescence in the presence of
both AHL signal and a QSI compound. In pSB401, the luxRand the P
luxI
(luxI
promoter from V. fischeri) have been coupled to the entire lux structural
operon (luxCDABE) from Photorhabdus luminescens (122). However, there are
limitations to these two systems: the CVO26 only responds to a few AHL
signals such as BHL and HHL (N-hexanoyl-L-homoserine lactone) and may
therefore also be narrow range with respect to various natural QS inhibitors.
Expression of bioluminescence is a complex phenotype sensitive to almost
anything that affects the physiology of the host bacterium. The A. tumefaciens
reporter composed by a traG::lacZ fusion and traRis relatively promiscuous in
its choice of AHL molecules and therefore widely used for thin-layer
chromatography overlays to identify AHL molecules. With the exception of
BHL, this monitor detects AHLs with 3-oxo-, 3-hydroxy-, and 3-unsubstituted
side chains of all lengths tested (99).
We have constructed a number of green fluorescent quorum sensors,
which consist of luxRP
luxI
fused to a modified version of the gfpmut3*
gene encoding an unstable green fluorescent protein (GFP) variant of the
jellyfish Aequorea victoria as a reporter for non-invasive, real-time detection
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
76
of gene expression without the addition of chemical substrates (see
Figure 4.2) (1). This system opened up unprecedented possibilities for
studies in situ of QS and QS inhibition in ecologically and clinically relevant
scenarios at the macroscopic as well as at the microscopic, single-cell level.
Unstable versions of the GFP had previously been constructed to enable on
line measurements in a non-cumulative fashion (2). Furthermore, it
allowed for monitoring of fluctuations in gene expression and even
repeated measurements of the same cells to continually assess bacterial
activity. Colonies of E. coli harboring a luxRP
luxI
-gfp(ASV)-cassette on a
high-copy-number pUC-derived plasmid appeared completely dark under
Green fluorescent cells Dark cells
(a)
(b)
No
signal
or
signal
block
+
signal
R* R
R
e
s
p
o
n
s
e

t
i
m
e
s
:
1
0

1
5

m
i
n
4

5

h
R
R
R
R
lasR
P
lasB
Gfp(ASV)
luxR
P
luxl
Gfp(ASV)
R
R*
R
R*
+
+
luxR
P
luxl
Gfp(ASV)
lasR
P
lasB
Gfp(ASV)
Active receptor Blocked receptor AHL signal QSI
Figure 4.2. (a) Molecular details of two green fluorescent QS monitors. The LuxR-based
monitor responds to a variety of AHL signals except for BHL, whereas the LasR-based
monitor primarily responds to OdDHL. (b) Microscopic inspection of Escherichia coli cells
carrying the LuxR-based monitor (right panels show fluorescent microscopy). The cells
respond to the addition of 10 nM OHHL by producing detectable amounts of the unstable
GFP protein. If the OHHL is removed by washing, or if a QSI compound is added which
blocks GFP synthesis, growing cells lose the GFP signal in 45 hours. (See also color plate
section)
77
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
the epifluorescence microscope. In striking contrast, the same strain pro-
duced bright green fluorescent colonies when cultivated under identical
conditions in the presence of 1 nM OHHL. The lesson learned from
numerous attempts to construct functional gfp gene fusions is that a
relatively high level of gene expression followed by efficient translation is
required for this to function at the single cell level: 15 min after addition of
10 nM OHHL to a planktonic culture, single cells appeared bright
green fluorescent under the epi-fluorescence microscope. This monitor
has a relatively broad-spectrum AHL detection limit in the range of
1 nM for OHHL, 10 nM for HHL, 10 nM for OHL, 10 nM for OdDHL
and 1000 nM for BHL, with a rapid (15 min) response time at the single-
cell level (1).
As with TraR, LuxR overexpression seems to increase agonistic activ-
ities of the various AHL analogs (128). The present GFP had a 45 min half-
life and consequently the monitor systemexhibited a less impressive 4 h lag
before the signal had been reduced to below the detection limit (126).
QS-regulated gene expression in Pseudomonas aeruginosa can be studied
in the same way by means of a QS monitor, which consists of a
fusion between the lasB elastase gene and gfp (PlasBgfp(ASV)). By means
of a mini-transposon this construct was placed on the chromosome of
P. aeruginosa to ensure stable segregation and a constant gene dosage of the
reporter system (42). Both monitor strains have been further modified to
carry a constitutively expressed dsred gene encoding the red fluorescent
protein RFP (see Figure 4.5). Red fluorescence therefore correlates to bacterial
biomass accumulation, green to bacterial communication. These systems
have been used to quantify the efficacy of a number of natural and
synthetic furanone compounds (42) as well as a number of novel QSI
compounds (90, 91).
NOVEL QSI SCREENS
The ability of the above-mentioned reporter fusions to be operated in
simple microtiter tray assays offers the researcher a reasonable high-
throughput facility. As well as being able to detect inducing signals, they
can also be used to detect the presence of QSI activity. However, the
systems are problematic in that they often require more than one test to
verify the result of the screen. The test compound may exert its effect on the
reporter itself, or alternatively interfere with precursors, substrates, or
environmental conditions for proper function of the reporter, or it might
simply affect growth without being bactericidal. Another obstacle is the
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
78
adjustment of the concentration of the test QSI compound in such a way
that it does not lead to any of the above mentioned side effects. This
requires that a single test compound must be probed at several different
concentrations. A functional QSI is a compound that blocks QS (of a
selected QS-controlled monitor system) in the low micromolar range.
Furthermore, the inhibitor should be chemically stable and non-toxic.
In our view, the last point is essential since new anti-pathogenic drugs
against, for example, P. aeruginosa should circumvent the problem of
development of resistance that is found with conventional antimicrobial
compounds.
To meet the above requirements, we recently designed screening sys-
tems based on a novel concept, i.e. the bacteria require the presence of a
QSI for growth (91). The basic principle of these so-called QSI selectors is
that a toxic gene product is placed under QS control (see Figure 4.3). For
example, QSI selector 1 (QSIS1) is based on the luxR gene and PluxI fused
to the phlA gene from S. liquefaciens MG1. Expression of phospholipase
leads to lysis of the E. coli host cell (35). Therefore, in the presence of
externally added OHHL, the phlA gene is expressed from PluxI and the
cells are killed. If, however, a QSI compound is present along with the AHL,
the cells survive. QSI selector 2 (QSIS2) is based on the las system from
P. aeruginosa: the lasB promoter has been fused to the sacB gene. When
established on the chromosome in a rhlIlasI P. aeruginosa background,
expression causes cell death in the presence of exogenous BHL OdDHL
and sucrose. Presence of a QSI rescues the selector bacteria. The screen is
performed as a diffusion assay. Test samples are added to wells made in the
agar. The components of the sample diffuse out into the agar, creating a
diffusion gradient with the highest concentration closest to the well. If QS
is blocked, the cells survive and form a ring of growth around the sample
app li cat ion point (Figure 4.3b ). In our expe rience, th e s creen works well on
everything from pure compounds and extracts with organic solvents to
freshly collected samples.
NOVEL NATURAL QSIs
Many plants and fungi have coevolved and established carefully regu-
lated symbiotic associations with bacteria. Exudates from pea (Pisum
sativum) contain several separable activities able to stimulate, whereas
others inhibited, expression of AHL-regulated traits (111). Finding healing
powers in plants is an ancient idea. Curiously, since the advent of anti-
biotics in the 1950s, the use of plant derivatives as antimicrobial agents has
79
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
been virtually non-existent. The developed world has relied on bacterial and
fungal sources for these activities. However, recent studies have shown that
a number of foods contain furanone compounds with similarity to the
D. pulchra compounds described above (101). Inspired by this, we have
surveyed various herbs, plants and fungal extracts for AHL-inhibitory
activity. It is worth while noting that both plants and fungi are devoid of
Cell death
(a)
(b)
Active receptor
R
Rescue
R
R
R
R
lasR
P
lasB
sacB
lasR
P
lasB
sacB
R
R
R
R
+
+
luxR
P
luxl
phlA
luxR
P
luxl
phlA
Blocked receptor
Sample application
Rescue
ring
AHL signal QSI
Figure 4.3. (a) Molecular details and principles of the two QSI selectors QSIS1 and QSIS2.
QSIS1 is based on the luxR gene and the target promoter PluxI fused to the phlA gene from
Streptococcus liquefaciens MG1. In Escherichia coli, expression of the PhlA phospholipase
leads to rapid lysis of the host cell in the presence of exogenous OHHL. QSIS2 is based on
the las system from Pseudomonas aeruginosa; the lasB promoter has been fused to the sacB
gene. When established on the chromosome in a rhlI lasI P. aeruginosa background,
expression causes cell death in the presence of exogenous BHLOdDHL and sucrose.
(b) Agar cast with OHHL and E. coli cells carrying the QSIS1. A50ml volume of C-30 sample
was added to the sample application point and incubated overnight. C-30 diffuses out into
the agar, creating a diffusion gradient with the highest concentration closest to the well. The
clear zone surrounding the application hole indicates that the compound is toxic at high
concentrations; this is again surrounded by a rescue ring formed by bacterial growth where
the concentration of C-30 is sufficiently high to overcome and block OHHL activation of the
phospholipase gene.
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
80
active immune systems, as seen in mammals, but rather rely on chemical
defense systems to deal with bacteria in the environment. Plants and fungi
may therefore utilize chemical compounds to inhibit (or in other cases, to
stimulate) bacterial AHL-mediated communication.
Plants have an almost limitless ability to synthesize aromatic sub-
stances, of which at least 12,000 have been isolated, a number estimated
to be less than 10% of the total (97). Bean sprouts, camomile, carrot, chili
habanero, the bee wax propolis, yellow pepper, and garlic were recently
found by us to hold QSI activity (91). Although garlic (Allium sativum) has
been used for its medicinal properties for thousands of years, investigations
into its mode of action are relatively recent. Garlic has a wide spectrum of
actions: not only is it antibacterial, antiviral, antifungal, and antiprotozoal,
but it also has beneficial effects on the cardiovascular and immune systems.
Garlic extract scored positive with the QSIS1 and QSIS2. This strongly
suggests that the extract contains a QSI compound that inhibits QS in
Pseudomonas aeruginosa. Our result was recently confirmed by Affymetrix
GeneChip
1
-based transcriptomics (91). Some of the garlic QSI com-
pounds were recently identified by us (84) but they did not show activity
against P. aeruginosa (see Figure 4.1). The structure of the active compound
against P. aeruginosa is now known by us.
In a recent screening of 55 different Penicillium species conducted by
us, one third were found to produce QSI activity (90). Two compounds,
patulin and penicillic acid (see Figure 4.1) from Penicillium species, were
isolated and demonstrated to be the biologically active QSI compounds.
Their structural similarity to C-30 and C-56 is obvious. Accordingly they
were found to block the QS systems of Pseudomonas aeruginosa as con-
firmed by DNA-microarray-based transcriptomics (90).
TRANSCRIPTOMICS: THE ULTIMATE TOOL
FOR ANALYZING DRUG SPECIFICITY
DNA microarray technology is a cutting-edge molecular approach to
study global gene expression and can be employed to decipher complex
regulatory networks in both bacteria and eukaryotes. The technique per-
mits us to take snapshots of the involved organisms global gene expres-
sion profile during different stages of growth and treatments. Until
recently, comparison of gene expression levels was limited to tracking
one or a few genes at a time. The use of reporter fusions is generally
accepted as a suitable way to monitor gene expression. It should be kept
in mind that such gene fusions are based on hybrid genes and hence
81
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
produce hybrid mRNAs. Such molecules cannot be expected to have stabil-
ity similar to that of the native messenger, and therefore information
gained from such fusions should not a priori be considered more reliable
than to the microarray hybridization data that are gained from uninter-
rupted messengers. Recently, a number of studies using the Affymetrix
P. aeruginosa microarray have been published (3, 43, 98, 115, 127). In several
of these reports, the microarray data have been verified by an independent
method such as real-time PCR. It is noteworthy that in all studies the data
obtained from microarray experiments and the independent methods were
found to correlate well. These reports provide a solid proof of concept for
the future application of GeneChips
1
.
For example, the drug specificity of the furanone C-30 was evaluated by
using the Affymetrix P. aeruginosa GeneChip
1
(43). The study was per-
formed as a parallel identification of QS-controlled genes and furanone-
target genes and included sample points from OD
450
0.5 to approxi-
mately OD
450
3. In total, 163 genes were found to be controlled by QS.
This corresponds to 2.9% of all the predicted genes in P. aeruginosa PAO1.
Among these genes, most previously reported QS genes were present.
Growing P. aeruginosa in the presence of the furanone compound caused
a change in expression of 93 genes, among which 85 genes (1.5%) were
repressed and 8 genes (0.1%) were activated in response to C-30. A com-
parative analysis of the QS regulon and C-30 target genes showed that 80%
of the furanone-repressed genes are in fact controlled by QS. Furanone-
repressed genes are not restricted to genes regulated by either the las or the
rhl encoded systems but are found throughout the continuum of QS-
induced genes. In essence, the analysis showed a clear overlap between
QS-controlled genes and furanone-repressed genes (Figure 4.4.) The study
demonstrated that by using furanone compounds it is possible to synthesize
QS-inhibitors that specifically target the QS circuit and cause a dramatic
reduction in expression of the major P. aeruginosa virulence factors, includ-
ing elastase, hydrogen cyanide, rhamnolipid, chitinase, and phenazine.
Performed under similar growth conditions, Affymetrix GeneChip
1
-
based analysis demonstrated that garlic extract affected the expression of
167 (3%) genes in total (91). At the single sample point (OD
450
2) 34% of
the QS regulon was downregulated by the garlic treatment. Affected genes
were found among LasR-, RhlR-, and LasRRhlR-controlled genes, many
of which encode the major P. aeruginosa virulence factors. It is encouraging
that relatively crude extracts, provided they do not affect the growth rate of
the culture, can be applied for GeneChip
1
-based transcriptomics. Along
either drug isolation or drug development pipelines, where natural as well
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
82
as synthetic compounds are employed as scaffolds for drug design, the
GeneChip
1
delivers the ultimate answer to the target specificity of the
compounds in question. This information is highly valuable at an early
stage of pipelines. In the former case, it allows a decision to be made as to
Figure 4.4. Gene expression atlas of QS-controlled genes and QSI target genes in
Pseudomonas aeruginosa PAO1. Upregulated genes are shown in green; downregulated
genes are red. QS-controlled gene expression is shown in the outer half of the atlas;
furanone C-30 target genes are indicated in the inner half. QS-controlled genes were
identified by growing a P. aeruginosa PAO1 lasIrhlI mutant with or without exogenous
AHL signals (2 mM OdDHL and 5 mM BHL) and retrieving samples for microarray analysis at
several culture densities. C-30 target genes were identified by growing P. aeruginosa PAO1
with or without 10 mM furanone C-30. P. aeruginosa PAO1 ORFs are shown in red (sense
strand) and blue (antisense strand) in the innermost circle. The outermost scale gives the
gene localization (in base pairs) in the genome. The color scale bar gives the correlation
between color coding and fold-change in gene expression. (see also color plate section)
83
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
whether to embark on the time-consuming process of purification of
natural compounds.
BIOFILM MODELS: WHAT QSI DRUGS DO TO A BIOFILM
During biofilm formation, cells aggregate and embed themselves in an
EPS (exopolymeric substances) matrix, which gives rise to a rigid 3-D
structure. Nickel et al. (72) demonstrated that the biofilm mode of growth
substantially increased P. aeruginosas tolerance to tobramycin treatment;
Costerton et al. (12) later expanded that to cover a number of other organ-
isms and antibiotics. Since then numerous studies have generalized this
finding; for reviews, see (15, 25, 47). Although several mechanisms (among
them diffusion barriers and special physiological conditions) have been
postulated to explain reduced susceptibility to antimicrobials in bacterial
biofilms, it is becoming evident that biofilm tolerance to antimicrobial
treatments is multifactorial.
To study the effect of QSI compounds on biofilm development, the
typical experimental scenario used by us is P. aeruginosa biofilms that have
been allowed to establish in flow chambers in the presence or absence of,
for example, furanone compounds. Scanning confocal laser microscopy
(SCLM)-based inspection of the GFP-based QS reporter systems revealed
that the compounds C-56 and C-30 penetrated the microcolonies, and
repressed QS in the majority of the biofilm cells. Importantly, the concen-
tration used was similar to what was found to inhibit QS in planktonic cells
without affecting growth or general gene expression (42, 43). By comparing
our sets of accumulated data we have found that, similar to the biofilms
formed by QS mutants (5), there is (under our experimental settings) no
obvious difference in the early stages of biofilm formation in the presence
or absence of furanones or any other QSI compound. The effect of QS
deficiency (either by mutation or by QSI treatment) becomes more obvious
at later stages. For example, at day 7 furanone-treated biofilms were less
persistent and more similar to the QS mutant biofilms (39, 42, 43) with the
impression of flat, less differentiated biofilms at day 10, which eventually
sloughed off (42).
P. aeruginosa biofilms grown in the presence of furanones were also
subsequently exposed to 0.1% SDS or tobramycin. Tobramycin is an anti-
biotic frequently used in the treatment of CF patients. Three days pre-
treatment with furanones sensitized the biofilms to dissolution such that
both the detergent and tobramycin were able to kill 90%95% of
P. aeruginosa cells (43). Similar data have been obtained with regulatory
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
84
QS mutants (5). In comparison, the effects of SDS and tobramycin on
untreated (or wild-type) biofilms were superficial. We have obtained similar
results if the biofilm was treated with both garlic and tobramycin or patulin
and tobramycin (90, 91). These results strongly suggest that biofilm
sloughing and tolerance to antibiotics and disinfectants can be controlled
with QSI drugs. This synergistic effect of QSI drugs and antibiotics is
somewhat reminiscent of antibiofilm treatment based on electrical current,
which also enhances the efficacy of antibiotics (51).
The biofilmmode of growth also offers protection against the activity of
polymorphonuclear leucocytes (PMNs). A PMNP. aeruginosa biofilm
model was recently established by us (5). The typical experiment proceeds
thus. Biofilms three days old are exposed to freshly isolated PMNs by
inoculation into the biofilm flow chambers. The sets of biofilms with
PMNs are then inspected by means of SCLM; gradually, differences in
the biomass of the types of biofilm can be observed. By using this model,
we recently found that QS mutants as well as QSI-treated wild-type biofilms
are susceptible to PMN phagocytosis (5, 90).
Activated PMNs liberate oxygen radicals in the form of H
2
O
2
, which
has a bactericidal effect. P. aeruginosa biofilm cells tolerate high concentra-
tions (50100 mM), mainly owing to two protective mechanisms: the pre-
sence of katA-encoded catalase activity (28) and failure of H
2
O
2
to fully
penetrate the biofilm owing to a reactiondiffusion interaction as described
by Stewart et al. (107). These authors concluded that some protective, yet
unknown mechanism other than incomplete penetration is operative in
P. aeruginosa biofilms treated with H
2
O
2
. We have assessed tolerance of
biofilms to H
2
O
2
by introducing 100mM to the influent medium to bio-
films 3 days old (5). LIVE/DEAD BacLight staining demonstrated that the
wild-type biofilm cells were left unharmed by the treatment in contrast to
the QS mutant, which was highly sensitive. Furthermore, we also gener-
ated data that suggested this tolerance could result from the amount of
oxygen radicals produced by the PMNs, which in turn depends on the
magnitude of activation caused by the bacteria. We found that only QS
mutants, in contrast to wild-type biofilms, caused detectable activation of
the PMNs (5). Furthermore, wild-type biofilms treated with C-30 caused
detectable activation of the PMNs similar to that of the QS mutant. Our data
support a model where QS signals directly affect the magnitude of the
oxidative PMN burst. In other words, PMN activation can be blocked by
adding OdDHL and BHL to lasRrhlR double mutants. On the other hand,
PMN activation can be promoted by treatment of wild-type biofilm with
QSI compounds.
85
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
In concurrence with these results, Wu et al. (125) found that pulmonary
infections caused by a lasIrhlI mutant induced a faster and stronger
immune response against the bacterial infection in the early phase, as
judged from the severity of lung pathology, higher lung IFN-a production,
stronger oxidative burst of blood PMNs, and faster antibody response
compared with the wild-type counterpart. Similar data have recently been
obtained in vivo with the lasRrhlR double mutant (5).
PULMONARY DOSERESPONSE MODELS
For drug design and development, it is obvious that a direct study of
QSI functionality under complex conditions in vivo has tremendous poten-
tial. The QS monitors based on GFP reporters offered that possibility.
A pseudo-chronic lung infection can be established in mice with E. coli
carrying a LuxR-controlled PluxIgfp(ASV) system (126) or P. aeruginosa
equipped with a lasB-gfp(ASV) system (43). Healthy mice normally clear
introduced bacteria efficiently by means of the mucociliary escalator. The
intratracheal introduction of bacteria in the form of alginate beads (see
Figur e 4.5a ) reduces the abili ty to rapidly clea r th e b acteria b y m eans of th e
mucociliary escalator. Instead, PMNs will be recruited by the immune
system and su rrou nd the infected ar eas in the l ung ( see Figur e 4. 5 b ). As a
result, inflammation and partial tissue destruction occur. For a period of up
to a two wee ks , th is is re minis cent o f th e situ ation i n the infecte d h uma n
CF lung (45, 71, 83).
Usually the mice are infected one or two days before the doseresponse
experiment. On the day of the experiment combinations of OHHL and QSI
drugs are injected in the tail vein and animals are sacrificed at different
time points after the treatments. Sections of frozen lung tissue are prepared
for SCLM. Both types of QS monitor express RFP for easy detection and
localization in the lung tissue and additionally express GFP in response to
AHL-dep ende nt QS (F igu re 4.5c , d ) If, h owever, QS is blocked, no or only
little GFP signals can be recorded. As shown by Hentzer et al. (43) intro-
duction of OHHL into the mouse blood circulation caused activation of the
E. coli LuxR QS monitor in a concentration-dependent manner. This
demonstrates that OHHL is transported by the blood, penetrates the lung
tissue, and induces LuxR-controlled PluxIgfp(ASV) expression in the
infecting bacteria. This model system has been used by us to evaluate the
efficacy of the furanones as well as a number of other potential QSI drugs in
vivo. For example, furanone C-30 (given as 2 mg/g body mass (BM) or 10 mM
if the entire mouse volume is considered to be 20 ml), co-administered
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
86
with OHHL, caused repression of OHHL-dependent activation of the QS
sensor (43). The furanone inhibition could be partly relieved by increasing
dosages of OHHL. This experimental setup shows whether a QSI test drug
is transported by the blood circulation to the lungs, penetrates the lung
tissues, enters the bacteria and in turn represses the OHHL-controlled
QS monitor, and it allows for calibration of the effective concentration of
the drug.
(a) (c) (b)
RFP GFP RFP GFP
(d )
R
R
+
R
dsred dsred lasR
P
lasB
P
lasB
R
R
Active receptor Blocked receptor OdDHL QSI
Time 0 6 h after injection
gfp(ASV) P
tac
gfp(ASV) P
tac
lasR
(e)
Figure 4.5. The pulmonary doseresponse model. (a) Mice are challenged intratracheally
with alginate beads containing P. aeruginosa. (b) Photomicrographs of mouse lung tissue
infected with P. aeruginosa. The arrow points at an alginate bead surrounded by numerous
PMNs. Adapted fromreference (126). (c) QSI drugs can be injected intravenously in the tail
vein. (d, e) Mouse lung tissue infected with P. aeruginosa carrying the LasR-based monitor
PlasBgfp for detection of cell-to-cell signaling (green fluorescence) and a tag for simple
identification in tissue samples (red fluorescence) examined by SCLM. (d) Mice were
administered C-30 via intravenous injection at time zero. (e) Infected animals were
sacrificed in groups of three at the time point indicated and the lung tissue samples were
examined by SCLM according to Hentzer et al. (43). Parts (d) and (e) are adapted from (43).
(See also color plate section)
87
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
As shown by Hentzer et al. (43), the ability of a test QSI drug to
suppress P. aeruginosa QS in vivo can be tested by infecting mouse lungs
with the P. aeruginosa QS monitor. The infection was allowed to establish
for 2 days before furanone C-30 (2 mg/g-BM) was introduced through the
tail ve in (Figure 4.5c ). In this model, no QS s ignals ar e intr odu ced be caus e
of the en dogeno us P. aeruginosa synt hesis (Figure 4.5d ). Over a time span of
46h after administration of C-30, the GFP signal from the P. aeruginosa
equipped with a lasBgfp(ASV) monitor was significantly reduced
(Figure 4. 5 e ). After 8 1 0 h, the GFP signal rea ppear ed, i ndica tin g th at th e
furanone was cleared from the mouse blood circulation and, hence,
de novo GFP synthesis recommenced. This experimental scenario reveals
important information about the mode of action of the test compound:
first, the effective concentration required to block wild-type P. aeruginosa
QS can be determined; and second, the stability of the compounds in the
mouse organism can be assessed. For example, this type of experiment
told us that a single furanone C-30 injection lasts approximately 6h (43).
This indicates that three doses per day are required for continuous
blocking of QS-controlled gene expression of the infecting bacteria.
There are many more questions that need to be answered by this
model before the treatment model is employed, such as whether the
test drug can or should be administered intravenously, intraperitoneally
or subcutaneously.
TREATMENT MODEL
The previous doseresponse models can be used to establish the con-
ditions required for QS inhibition in vivo. To assess the potential of a test
QSI compound in the treatment of P. aeruginosa infections, we have devel-
oped a mouse pulmonary infection model. Groups of mice are infected with
P. aeruginosa as described above. After this, the mice are typically split in
two groups, which receive injections of either the test QSI compound or
placebo for the following 35 days. The dosages and daily intervals are as
determined by the doseresponse models above. Groups of animals are
sacrificed on a daily basis; the lungs are removed and homogenized and the
contents spread on agar plates for determination of the number of bacteria.
Hentzer et al. (43) demonstrated convincing treatment with C-30: the
furanone-treated groups of animals displayed a bacterial content on average
three orders of magnitude lower than that of the placebo group, and the
efficiency of bacterial clearing correlated positively with the concentration
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
88
of the drug. We have tested several other potential QSI compounds in this
model, together with garlic extract and a patulin; the former proved even
more efficient than C-30 (6).
ANTIBIOTICS AND QSI DRUGS: ALONE OR TOGETHER?
More than half of the infectious diseases that affect compromised
patients involve bacterial species that are commensal or are common
in the environment, such as Staphylococcus epidermidis and P. aeruginosa
(13, 14). Biofilm-based infections share clinical characteristics. Biofilms
develop commonly on inert surfaces of medical devices or on dead tissue,
but they can also form on living tissues, as in endocarditis. Numerous
investigations of the surfaces of medical devices (arteriovenous shunts,
urinary and other types of catheters, orthopedic devices, and mechanical
heart valves) that have been surgically removed owing to device-related
infections show the presence of exopolymer-embedded bacteria. Tissues
taken from non-device-related chronic infections also show the presence of
biofilm bacteria surrounded by an exopolymeric matrix. Such diseases
include dental caries, periodontitis, otitis media, biliary tract infections,
and bacterial prostatitis (for reviews see (15, 25, 47)).
The best investigated example of a disease in which biofilms play a
prominent role is the occurrence of chronic endobronchiolitis caused by
mucoid (alginate over-producing) variants of P. aeruginosa in patients
suffering from CF. CF is caused by mutations in the CFTR gene encoding
the 1480 amino acid CF transmembrane conductance regulator (CFTR)
(16). CFTR functions as a chloride ion channel at the apical membrane of
epithelial cells of the sweat ducts, pancreatic ducts, the crypts of the small
intestine, the reproductive organs, and the respiratory epithelial cells,
which impairs ciliary function (16). Although the endobronchial chronic
infection of the biofilm-growing P. aeruginosa is impossible to eradicate,
repetitive, aggressive antibiotic chemotherapy at the early stages (as prac-
ticed at the Danish CF-Centre) has proven to be efficient in postponing the
onset of the chronic infection, which is accompanied by deterioration of
pulmonary function (31, 112).
The flip side of the coin is the development of antibiotic-resistant
bacteria (10). The frequency with which resistant mutants develop in the
CF lung may be higher than normal. Oliver et al. (78) reported that 36% of a
group of CF patients were colonized by hypermutable (mutator) strains in
contrast to non-CF patients acutely infected with P. aeruginosa. Mutator
strains can appear due to inactivating mutations in DNA repair genes,
89
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
e.g. mutS, mutT (76, 77, 78). Such mutations are in turn caused by
reactive oxygen metabolites released from the activated PMNs in the lungs
(123). The high mutation frequency, 3 10
6
in CF patients versus 3 10
8
in non-CF patients, were associated with a significantly higher frequency
of resistance to several groups of antibiotics (78). Such increased
antibiotic resistance adds to the protection already offered by the biofilm
mode of growth (106).
This highlights the importance of development of non-pathogenic
therapies that limit the formation of persistent biofilms in the lungs. In
the biofilm mode of growth the colonizing bacteria are protected from the
host defense system and the action of antibiotics. It is therefore highly
interesting that QSI compounds that were found to efficiently eradicate
pulmonary infections also possess PMN-activating properties. The admin-
istration of QSI compounds will then be expected to lead to development of
less persistent biofilms but also inhibit expression of bacterial exoenzymes
that actively degrade components of the immune system. Taken together
with the synergistic effect of QS blockage and PMN activity, this might
promote clearance, which in turn will reverse the severity of infection
and improve the lung function. In addition, the synergistic effect of anti-
biotics and QSI drugs might turn out to be a useful combination in
future chemotherapies. In relation to the insertion of medical implants,
short-term prophylactic therapy or alternatively special surface coatings
liberating QSI compounds might prove efficient in reducing the risk of
developing detrimental biofilms on the devices. The recent finding by us
that QSI drugs are effective in treatment of vibriosis in a fish model system
illustrates that such compounds might be useful alternatives to antibiotics
in fish farming (89).
ACKNOWLEDGEMENT
Our work on QS inhibition has been supported by grants from the
Danish Technical Research Council and the Villum Kann Rasmussen
Foundation to MG.
REFERENCES
1 Andersen, J. B., A. Heydorn, M. Hentzer et al. 2001. gfp based N-acyl-homoserine-
lactone monitors for detection of bacterial communication. Appl. Environ.
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
90
2 Andersen, J. B., C. Sternberg, L. K. Poulsen et al. 1998. New unstable variants of
green fluorescent protein for studies of transient gene expression in bacteria.
Appl. Environ. Microbiol. 64: 224046.
3 Bagge, N., M. Schuster, M. Hentzer et al. 2004. Pseudomonas aeruginosa
biofilms exposed to imipenem exhibit changes in global gene expression and
beta-lactamase and alginate production. Antimicrob. Agents Chemother. 48:
117587.
4 Belas, M. R. 2003. The swarming phenomenon of Proteus mirabilis. ASM News
58: 1522.
5 Bjarnsholt, T., P. . Jensen, M. Burmlle et al. 2005. Pseudomonas aeruginosa
tolerance to tobramycin, hydrogen peroxide and polymorphonuclear leucocytes is
quorum sensing depended. Microbiology 151: 37383.
6 Bjarnsholt, T., P. . Jensen, T. B. Rasmussen et al. 2005. Garlic jams Pseudomonas
aeruginosa communication and cures pulmonary infections. Infect. Immun.
(in press.)
7 Camara, M., P. Williams and A. Hardman 2002. Controlling infection by tuning
in and turning down the volume of bacterial small-talk. Lancet Infect. Dis. 2:
66776.
8 Chhabra, S. R., P. Stead, N. J. Bainton et al. 1993. Autoregulation of carbapenem
biosynthesis in Erwinia carotovora by analogues of N-(3-oxohexanoyl)-L-
homoserine lactone. J Antibiot.(Tokyo) 46: 44154.
9 Chun, C. K., E. A. Ozer, M. J. Welsh, J. Zabner and E. P. Greenberg 2004.
Inactivation of a Pseudomonas aeruginosa quorum-sensing signal by human
airway epithelia. Proc. Natn. Acad. Sci. USA 101: 358790.
10 Ciofu, O., B. Giwercman, S. S. Pedersen and N. Hoiby 1994. Development of
antibiotic resistance in Pseudomonas aeruginosa during two decades of antipseu-
domonal treatment at the Danish CF Center. Acta Pathol. Microbiol. Immunol.
Scand. 102: 67480.
11 Correa, J. A. 1996. Diseases in seaweeds: an introduction. Hydrobiologia 326: 878.
12 Costerton, J. W., K. J. Cheng, G. G. Geesey et al. 1987. Bacterial biofilms in nature
and disease. A. Rev. Microbiol. 41: 43564.
13 Costerton, J. W., Z. Lewandowski, D. E. Caldwell, D. R. Korber and H. M. Lappin-
Scott 1995. Microbial biofilms. A. Rev. Microbiol. 49: 71145.
14 Costerton, J. W., Z. Lewandowski, D. DeBeer et al. 1994. Biofilms, the custom-
ized microniche. J. Bacteriol. 176: 213742.
15 Costerton, J. W., P. S. Stewart and E. P. Greenberg 1999. Bacterial biofilms:
a common cause of persistent infections. Science 284: 131822.
16 Davidson, D. J. and D. J. Porteous 1998. Genetics and pulmonary medicine.
1. The genetics of cystic fibrosis lung disease. Thorax 53: 38997.
17 Davies, D. 2003. Understanding biofilm resistance to antibacterial agents. Nat.
Rev. Drug Discov. 2: 11422.
18 Davies, D. G., M. R. Parsek, J. P. Pearson et al. 1998. The involvement of cell-to-
cell signals in the development of a bacterial biofilm. Science 280: 2958.
91
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
19 Davis, A. R., Targett, N. M, O. J. McConnell and C. M. Young 1989. Epibiosis of
marine algae and benthic invertebrates: natural products chemistry and other
mechanisms inhibiting settlement and overgrowth. Bioorg. Mar. Chem. 3: 86114.
20 De Kievit, T. R., R. Gillis, S. Marx, C. Brown and B. H. Iglewski 2001. Quorum-
sensing genes in Pseudomonas aeruginosa biofilms: their role and expression
patterns. Appl. Environ. Microbiol. 67: 186573.
21 de Nys, R., P. Steinberg, C. N. Rogers, T. S. Charlton and M. W. Duncan 1996.
Quantitative variation of secondary metabolites in the sea hare Aplysia parvula
and its host plant, Delisea pulchra. Mar. Ecol. Prog. Ser. 130: 135146.
22 de Nys, R., P. D. Steinberg, P. Willemsen et al. 1995. Broad spectrum effects of
secondary metabolites from the red alga Delisea pulchra in antifouling assays.
Biofouling 8: 25971.
23 de Nys, R., A. D. Wright, G. M. Konig and O. Sticher 1993. New halogenated
furanones from the marine alga Delisea pulchra. Tetrahedron 49: 1121320.
24 Dong, Y. H., L. H. Wang, J. L. Xu et al. 2001. Quenching quorum-sensing-
dependent bacterial infection by an N-acyl homoserine lactonase. Nature 411:
81317.
25 Drenkard, E. 2003. Antimicrobial resistance of Pseudomonas aeruginosa biofilms.
Microbes Infect. 5: 121319.
26 Dworjanyn, S., R. de Nys, and P. D. Steinberg. 1999. Localization of secondary
metabolites in the red alga Delisea pulchra. Mar. Biol. 133: 72736.
27 Eberl, L., M. K. Winson, C. Sternberg et al. 1996. Involvement of N-acyl-L-
homoserine lactone autoinducers in controlling the multicellular behaviour of
Serratia liquefaciens. Molec. Microbiol. 20: 12736.
28 Elkins, J. G., D. J. Hassett, P. S. Stewart, H. P. Schweizer and T. R. McDermott
1999. Protective role of catalase in Pseudomonas aeruginosa biofilm resistance to
hydrogen peroxide. Appl. Environ. Microbiol. 65: 4594600.
29 Favre-Bonte, S., J. C. Pache, J. Robert et al. 2002. Detection of Pseudomonas
aeruginosa cell-to-cell signals in lung tissue of cystic fibrosis patients. Microb.
Pathogen. 32: 1437.
30 Fenical, W. 1997. New pharmaceuticals from marine organisms. Trends
Biotechnol. 15: 33941.
31 Frederiksen, B., C. Koch and N. Hoiby 1997. Antibiotic treatment of initial coloni-
zation with Pseudomonas aeruginosa postpones chronic infection and prevents
deterioration of pulmonary function in cystic fibrosis. Pediatr. Pulmonol. 23: 3305.
32 Fuqua, C., S. C. Winans and E. P. Greenberg 1994. Quorum sensing in bacteria:
the LuxR-LuxI family of cell density-responsive transcriptional regulators.
33 Geesey, G. G., W. T. Richardson, H. G. Yeomans, R. T. Irvin and J. W Costerton
1997. Microscopic examination of natural sessile bacterial populations from an
alpine stream. Can. J. Microbiol. 23: 17336.
34 Givskov, M., R. de Nys, M. Manefield et al. 1996. Eukaryotic interference with
homoserine lactone-mediated prokaryotic signaling. J. Bacteriol. 178: 661822.
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
92
35 Givskov, M. and S. Molin 1993. Secretion of Serratia liquefaciens phospholipase
from Escherichia coli. Molec. Microbiol. 8: 22942.
36 Hanzelka, B. L. and E. P. Greenberg 1995. Evidence that the N-terminal region of
the Vibrio fischeri LuxR protein constitutes an autoinducer-binding domain.
37 Heilmann, C., C. Gerke, F. Perdreau-Remington and F. Gotz 1996.
Characterization of Tn917 insertion mutants of Staphylococcus epidermidis
affected in biofilm formation. Infect. Immun. 64: 27782.
38 Henschel, J. R. and P. A. Cook 1990. The development of a marine fouling
community in relation to the primary film of microorganisms. Biofouling
2: 111.
39 Hentzer, M., L. Eberl and M. Givskov 2004. Quorum sensing in biofilms: Gossip
in the slime world? In M. Ghannoum and G. OToole (eds.), Microbial Biofilms,
pp. Washington, DC: ASM Press.
41 Hentzer, M., L. Eberl and M. Givskov 2005. Transcriptome analysis of Pseudomonas
aeruginosa biofilms: anaerobic respiration and iron limitation. Biofilms. (In press.)
42 Hentzer, M., K. Riedel, T. B. Rasmussen et al. 2002. Inhibition of quorum
sensing in Pseudomonas aeruginosa biofilm bacteria by a halogenated furanone
compound. Microbiology 148: 87102.
43 Hentzer, M., H. Wu, J. B. Andersen et al. 2003. Attenuation of Pseudomonas
aeruginosa virulence by quorum sensing inhibitors. EMBO J. 22: 113.
44 Heydorn, A., B. Ersboll, J. Kato et al. 2002. Statistical analysis of Pseudomonas
aeruginosa biofilm development: impact of mutations in genes involved in twitch-
ing motility, cell-to-cell signaling, and stationary-phase sigma factor expression.
Appl. Environ. Microbiol. 68: 200817.
45 Hoiby, N. 1993. Antibiotic therapy for chronic infection of Pseudomonas in the
lung. A. Rev. Med. 44: 110.
46 Hiby, N. and B. Frederiksen 2000. Microbiology of cystic fibrosis. In M. E. Hodson
and D. M. Geddes (eds.), Cystic Fibrosis, pp. 83107. London: Edward Arnold.
47 Hoiby, N., J. H. Krogh, C. Moser et al. 2001. Pseudomonas aeruginosa and the in
vitro and in vivo biofilm mode of growth. Microbes Infect. 3: 2335.
48 Huber, B., K. Riedel, M. Hentzer et al. 2001. The cep quorum-sensing system of
Burkholderia cepacia H111 controls biofilm formation and swarming motility.
Microbiology 147: 251728.
49 Huber, B., K. Riedel, M. Kothe et al. 2002. Genetic analysis of functions involved
in the late stages of biofilm development in Burkholderia cepacia H111. Molec.
50 Ikeda, T., K. Kajiyama, T. Kita et al. 2001. The synthesis of optically pure
enantiomers of N-acyl-homoserine lactone autoinducers and their analogues.
Chem. Lett. 30: 31415.
51 Jass, J., J. W. Costerton and H. M. Lappin-Scott 1995. The effect of electrical
currents and tobramycin on Pseudomonas aeruginosa biofilms. J. Ind. Microbiol. 15:
23442.
93
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
52 Kjelleberg, S., P. D. Steinberg, M. Givskov et al. 1997. Do marine natural
products interfere with prokaryotic AHL regulatory systems? Aquat. Microb.
Ecol. 13: 8593.
53 Klausen, M., A. Aaes-Jorgensen, S. Molin and T. Tolker-Nielsen 2003.
Involvement of bacterial migration in the development of complex multicellular
structures in Pseudomonas aeruginosa biofilms. Molec. Microbiol. 50: 618.
54 Kleerebezem, M., L. E. Quadri, O. P. Kuipers and W. M. de Vos 1997. Quorum
sensing by peptide pheromones and two-component signal-transduction systems
in Gram-positive bacteria. Molec. Microbiol. 24: 895904.
55 Kushmaro, A., Y. Loya, E. Fine and E. Rosenberg 1996. Bacterial infection and
coral bleaching. Nature 380: 396.
56 Labatte, M., S. Y. Queck, S. A. Rice, M. Givskov and S. Kjelleberg 2004. Quorum
sensing controlled biofilm development in Serratia liquefaciens MG1. J. Bacteriol.
186: 6928.
57 Latifi, A., M. Foglino, K. Tanaka, P. Williams and A. Lazdunski 1996. A hier-
archical quorum-sensing cascade in Pseudomonas aeruginosa links the transcrip-
tional activators LasR and RhIR (VsmR) to expression of the stationary-phase
58 Lawrence, J. R., D. R. Korber, B. D. Hoyle, J. W. Costerton and D. E. Caldwell
1991. Optical sectioning of microbial biofilms. J. Bacteriol. 173: 655867.
59 Littler, M. M. and D. S. Littler 1995. Impact of CLOD pathogen on pacific coral
reefs. Science 267: 135660.
60 Lyczak, J. B., C. L. Cannon and G. B. Pier 2002. Lung infections associated with
cystic fibrosis. Clin. Microbiol. Rev. 15: 194222.
61 Lynch, M. J., S. Swift, D. F. Kirke et al. 2002. The regulation of of biofilm
development by quorum sensing in Aeromonas hydrophila. Environ. Microbiol.
4: 1828.
62 Mack, D., M. Nedelmann, A. Krokotsch et al. 1994. Characterization of trans-
poson mutants of biofilm-producing Staphylococcus epidermidis impaired
in the accumulative phase of biofilm production: genetic identification of a
hexosamine-containing polysaccharide intercellular adhesin. Infect. Immun.
62: 324453.
63 Manefield, M., R. de Nys, N. Kumar et al. 1999. Evidence that halogenated
furanones from Delisea pulchra inhibit acylated homoserine lactone (AHL)-
mediated gene expression by displacing the AHL signal from its receptor protein.
64 Manefield, M., L. Harris, S. A. Rice, R. de Nys and S. Kjelleberg 2000. Inhibition
of luminescence and virulence in the black tiger prawn (Penaeus monodon)
pathogen Vibrio harveyi by intercellular signal antagonists. Appl. Environ.
65 Manefield, M., T. B. Rasmussen, M. Henzter et al. 2002. Halogenated furanones
inhibit quorum sensing through accelerated LuxR turnover. Microbiology 148:
111927.
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
94
66 Manefield, M., M. Welch, M. Givskov, G. P. Salmond and S. Kjelleberg 2001.
Halogenated furanones from the red alga, Delisea pulchra, inhibit carbapenem
antibiotic synthesis and exoenzyme virulence factor production in the phyto-
pathogen Erwinia carotovora. FEMS Microbiol. Lett. 205: 1318.
67 Maximilien, R., R. de Nys, C. Holmstrom 1998. Chemical mediation of
bacterial surface colonisation by secondary metabolites of the red alga Delisea
pulchra. Aquat. Microb. Ecol. 15: 23346.
68 Maximilien, R. R., R. de Nys, C. Holmstrom et al. 1998. Chemical mediation of
bacterial surface colonisation by secondary metabolites from the red alga Delisea
pulchra. Aquat. Microb. Ecol. 15: 23346.
69 McClean, K. H., M. K. Winson, L. Fish et al. 1997. Quorum sensing and
Chromobacterium violaceum: exploitation of violacein production and inhibition
for the detection of N-acylhomoserine lactones. Microbiology 143: 370311.
70 McKnight, S. L., B. H. Iglewski and E. C. Pesci 2000. The Pseudomonas quinolone
signal regulates rhl quorum sensing in Pseudomonas aeruginosa. J. Bacteriol. 182:
27028.
71 Moser, C., H. K. Johansen, Z. J. Song et al. 1997. Chronic Pseudomonas aeruginosa
lung infection is more severe in Th2 responding BALB/c mice compared to Th1
responding C3H/HeNmice. Acta Pathol. Microbiol. Immunol. Scand. 105: 83842.
72 Nickel, J. C., I. Ruseska, J. B. Wright and J. W. Costerton 1985. Tobramycin
resistance of Pseudomonas aeruginosa cells growing as a biofilm on urinary
catheter material. Antimicrob. Agents Chemother. 27: 61924.
73 Nielsen, J. and Givskov, M. 2003. Compounds and methods for controlling
bacterial virulence. PCT WO 03/106445. [Patent]
74 OToole, G. A., K. A. Gibbs, P. W. Hager, P. V. Phibbs, Jr. and R. Kolter 2000. The
global carbon metabolism regulator Crc is a component of a signal transduction
pathway required for biofilm development by Pseudomonas aeruginosa. J.
Bacteriol. 182: 42531.
75 OToole, G. A. and R. Kolter. 1998. Initiation of biofilm formation in
Pseudomonas fluorescens WCS365 proceeds via multiple, convergent signaling
pathways: a genetic analysis. Molec. Microbiol. 28: 44961.
76 Ochsner, U. A., M. L. Vasil, E. Alsabbagh, K. Parvatiyar and D. J. Hassett 2000.
Role of the Pseudomonas aeruginosa oxyR-recG operon in oxidative stress defense
and DNA repair: OxyR-dependent regulation of katB-ankB, ahpB, and ahpC-
ahpF. J. Bacteriol. 182: 453344.
77 Oliver, A., F. Baquero and J. Blazquez 2002. The mismatch repair system (mutS,
mutL and uvrD genes) in Pseudomonas aeruginosa: molecular characterization of
naturally occurring mutants. Molec. Microbiol. 43: 164150.
78 Oliver, A., R. Canton, P. Campo, F. Baquero and J. Blazquez 2000. High
frequency of hypermutable Pseudomonas aeruginosa in cystic fibrosis lung infec-
tion. Science 288: 12514.
79 Palmer, R. J. Jr. and D. C. White 1997. Developmental biology of biofilms:
implications for treatment and control. Trends Microbiol. 5: 43540.
95
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
80 Parsek, M. R., D. L. Val, B. L. Hanzelka, J. E. J. Cronan and E. P. Greenberg 1999.
Acyl homoserine-lactone quorum-sensing signal generation. Proc. Natn. Acad.
Sci. USA 96: 43605.
81 Passador, L., J. M. Cook, M. J. Gambello, L. Rust and B. H. Iglewski 1993.
Expression of Pseudomonas aeruginosa virulence genes requires cell-to-cell com-
munication. Science 260: 112730.
82 Pearson, J. P., M. Feldman, B. H. Iglewski and A. Prince 2000. Pseudomonas
aeruginosa cell-to-cell signaling is required for virulence in a model of acute
pulmonary infection. Infect. Immun. 68: 43314.
83 Pedersen, S. S., G. H. Shand, B. L. Hansen and G. N. Hansen 1990. Induction of
experimental chronic Pseudomonas aeruginosa lung infection with P. aeruginosa
entrapped in alginate microspheres. Acta Pathol. Microbiol. Immunol. Scand.
98: 20311.
84 Persson, P., H. H. Hansen, T. B. Rasmussen et al. 2005. Rational design
and synthesis of new quorum sensing inhibitors derived from acylated
homoserine lactones and natural products from garlic. Org. Biomolec. Chem.
3: 25362.
Sci. USA 96: 1122934.
86 Pesci, E. C., J. P. Pearson, P. C. Seed and B. H. Iglewski 1997. Regulation of las
and rhl quorum sensing in Pseudomonas aeruginosa. J. Bacteriol. 179: 312732.
87 Pratt, L. A. and R. Kolter 1998. Genetic analysis of Escherichia coli biofilm
formation: roles of flagella, motility, chemotaxis and type I pili. Molec.
88 Purevdorj, B., J. W. Costerton and P. Stoodley 2002. Influence of hydrodynamics
and cell signaling on the structure and behavior of Pseudomonas aeruginosa
biofilms. Appl. Environ. Microbiol. 68: 445764.
89 Rasch, M., C. Buch, B. Austin et al. 2004. An inhibitor of bacterial quorum
sensing reduces mortalities caused by Vibriosis in rainbow trout
(Oncorhynchus mykiss, Walbaum). Syst. Appl. Microbiol. 27: 3509.
90 Rasmussen, T., M. E. Skinders, T. Bjarnsholt et al. 2005. Idendity and effects of
quorum sensing inhibitors produced by Penicillum species. Microbiology.
(in press.)
91 Rasmussen, T. B., T. Bjarnsholt, M. E. Skinders et al. 2005. Screening for
quorum sensing inhibitors using a novel genetic system the QSI selector.
J. Bacteriol. 187: 1799814.
92 Reichelt, J. L. and M. A. Borowitzka. 1984. Antimicrobial activity from marine
algae: results of a large-scale screening programme. Hydrobiology 116/117:
15868.
93 Rice, S. A., M. Givskov, P. Steinberg and S. Kjelleberg 1999. Bacterial signals
and antagonists: the interaction between bacteria and higher organisms. J. Molec.
Microbiol. Biotechnol. 1: 2331.
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
96
94 Riedel, K., T. Ohnesorg, K. A. Krogfelt et al. 2001. N-acyl-L-homoserine lactone-
mediated regulation of the lip secretion system in Serratia liquefaciens MG1.
95 Sauer, K., A. K. Camper, G. D. Ehrlich, J. W. Costerton and D. G. Davies 2002.
Pseudomonas aeruginosa displays multiple phenotypes during development as a
biofilm. J. Bacteriol. 184: 114054.
96 Schaefer, A. L., B. L. Hanzelka, A. Eberhard and E. P. Greenberg 1996. Quorum
sensing in Vibrio fischeri: probing autoinducer-LuxR interactions with autoindu-
cer analogs. J. Bacteriol. 178: 2897901.
97 Schultes, R. E. 1978. The kingdomof plants, p. 208. In W. A. R. Thompson (ed.),
Medicines from the Earth, p. 208. New York, NY: McGraw-Hill.
98 Schuster, M., C. P. Lostroh, T. Ogi and E. P. Greenberg 2003. Identification,
timing, and signal specificity of Pseudomonas aeruginosa quorum-controlled
genes: a transcriptome analysis. J. Bacteriol. 185: 206679.
99 Shaw, P. D., G. Ping, S. L. Daly et al. 1997. Detecting and characterizing N-acyl-
homoserine lactone signal molecules by thin-layer chromatography. Proc. Natn.
Acad. Sci. USA 94: 603641.
100 Slattery, M., J. B. McClintoch and J. N. Heine 1995. Chemical defences in
Antarctic soft corals: evidence for antifouling compounds. J. Exp. Mar. Biol.
Ecol. 190: 6177.
101 Slaughter, J. C. 1999. The naturally occurring furanones: formation and func-
tion from pheromone to food. Biol. Rev. 74: 25976.
102 Smith, K. M., Y. Bu and H. Suga 2003. Library screening for synthetic agonists
and antagonists of a Pseudomonas aeruginosa autoinducer. Chem. Biol.
10: 56371.
103 Smith, K. M., Y. Bu and H. Suga 2003. Induction and inhibition of Pseudomonas
aeruginosa quorum sensing by synthetic autoinducer analogs. Chem. Biol.
10: 819.
104 Smith, R. S., E. R. Fedyk, T. A. Springer et al. 2001. IL-8 production in human
lung fibroblasts and epithelial cells activated by the Pseudomonas autoinducer N-
3-oxododecanoyl homoserine lactone is transcriptionally regulated by NF-kappa
B and activator protein-2. J. Immunol. 167: 36674.
105 Steidle, A., K. Sigl, R. Schuhegger 2001. Visualization of N-acylhomoserine
lactone-mediated cell-cell communication between bacteria colonizing the
tomato rhizosphere. Appl. Environ. Microbiol. 67: 576170.
106 Stewart, P. S. and J. W. Costerton 2001. Antibiotic resistance of bacteria in
biofilms. Lancet 358: 1358.
107 Stewart, P. S., F. Roe, J. Rayner et al. 2000. Effect of catalase on hydrogen
peroxide penetration into Pseudomonas aeruginosa biofilms. Appl. Environ.
108 Stoodley, P., F. Jrgensen, P. Williams and H. M. Lappin-Scott 1999. The role of
hydrodynamics and AHL signaling molecules as determinants of the structure
of Pseudomonas aeruginosa biofilms. In R. Bayston, M. Brading, P. Gilbert,
97
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
J. Walker and J. W. T. Wimpenny (eds), Biofilms: the Good, the Bad, and the Ugly,
pp. 32330. Cardiff: J. W. T. Bioline.
109 Sutherland, I. W. 2001. Biofilm exopolysaccharides: a strong and sticky frame-
work. Microbiology 147: 39.
110 Telford, G., D. Wheeler, P. Williams et al. 1998. The Pseudomonas aeruginosa
quorum-sensing signal molecule N-(3-oxododecanoyl)-L-homoserine lactone
has immunomodulatory activity. Infect. Immun. 66: 3642.
111 Teplitski, M., J. B. Robinson and W. D. Bauer 2000. Plants secrete substances
that mimic bacterial N-acyl homoserine lactone signal activities and affect
population density-dependent behaviors in associated bacteria. Molec. Plant
Microbe Interact. 13: 63748.
112 Valerius, N. H., C. Koch and N. Hoiby 1991. Prevention of chronic Pseudomonas
aeruginosa colonisation in cystic fibrosis by early treatment. Lancet 338: 7256.
113 Vallet, I., J. W. Olson, S. Lory, A. Lazdunski and A. Filloux 2001. The chaperone/
usher pathways of Pseudomonas aeruginosa: identification of fimbrial gene clus-
ters (cup) and their involvement in biofilmformation. Proc. Natn. Acad. Sci. USA
98: 691116.
114 Van Delden, C. and B. H. Iglewski 1998. Cell-to-cell signaling and Pseudomonas
aeruginosa infections. Emerg. Infect. Dis. 4: 55160.
115 Wagner, V. E., D. Bushnell, L. Passador, A. I. Brooks and B. H. Iglewski 2003.
Microarray analysis of Pseudomonas aeruginosa quorum-sensing regulons:
effects of growth phase and environment. J. Bacteriol. 185: 208095.
116 Wahl, M. 2003. Marine epibiosis. Fouling and antifouling: some basic aspects.
Mar. Ecol. Prog. Ser. 58: 17589.
117 Webb, J. S., L. S. Thompson, S. James et al. 2003. Cell death in Pseudomonas
aeruginosa biofilm development. J. Bacteriol. 185: 458592.
118 Whitchurch, C. B., T. Tolker-Nielsen, P. C. Ragas and J. S. Mattick 2002.
Extracellular DNA required for bacterial biofilm formation. Science 295: 1487.
119 Whiteley, M., M. G. Bangera, R. E. Bumgarner et al. 2001. Gene expression in
Pseudomonas aeruginosa biofilms. Nature 413: 8604.
120 Wimpenny, J. W. T. and R. Colasanti 1997. A more unifying hypothesis for
biofilm structures a reply. FEMS Microbiol. Ecol. 24: 1856.
121 Winson, M. K., M. Camara, A. Latifi et al. 1995. Multiple N-acyl-L-homoserine
lactone signal molecules regulate production of virulence determinants and
secondary metabolites in Pseudomonas aeruginosa. Proc. Natn. Acad. Sci. USA
92: 942731.
122 Winson, M. K., S. Swift, L. Fish et al. 1998. Construction and analysis of
luxCDABE-based plasmid sensors for investigating N-acyl homoserine lactone-
mediated quorum sensing. FEMS Microbiol. Lett. 163: 18592.
123 Worlitzsch, D., G. Herberth, M. Ulrich and G. Doring 1998. Catalase, myeloper-
oxidase and hydrogen peroxide in cystic fibrosis. Eur. Respir. J. 11: 37783.
124 Wozniak, D. J., T. J. O. Wyckoff, M. Starkey et al. 2003. Alginate is not a
significant component of the extracellular polysaccharide matrix of PA14 and
M
.
G
I
V
S
K
O
V
A
N
D
M
.
H
E
N
T
Z
E
R
98
PAO1 Pseudomonas aeruginosa biofilms. Proc. Natn. Acad. Sci. USA 100:
790712.
125 Wu, H., Z. Song, M. Givskov et al. 2001. Pseudomonas aeruginosa mutations in
lasI and rhlI quorum sensing systems result in milder chronic lung infection.
126 Wu, H., Z. Song, M. Hentzer et al. 2000. Detection of N-acylhomoserine
lactones in lung tissues of mice infected with Pseudomonas aeruginosa.
127 Yoon, S. S., R. F. Hennigan, G. M. Hilliard et al. 2002. Pseudomonas aeruginosa
anaerobic respiration in biofilms: relationships to cystic fibrosis pathogenesis.
Dev. Cell 3: 593603.
128 Zhu, J., J. W. Beaber, M. I. More et al. 1998. Analogs of the autoinducer 3-oxo-
octanoyl-homoserine lactone strongly inhibit activity of the TraR protein of
Agrobacterium tumefaciens. J. Bacteriol. 180: 5398405.
129 Passador, L., K. D. Tucker, K. R. Guertin et al. 1996. Functional analysis of the
Pseudomonas aeruginosa autoinducer PAI. J. Bacterial. 178: 59956000.
99
J
A
M
M
I
N
G
B
A
C
T
E
R
I
A
L
C
O
M
M
U
N
I
C
A
T
I
O
N
S
CHAPTER 5
Quorum-sensing-mediated regulation of biofilm
growth and virulence of Vibrio cholerae
Jun Zhu
University of Pennsylvania School of Medicine
John J. Mekalanos
Harvard Medical School, Boston, MA, USA
INTRODUCTION
Many species of bacterium exchange chemical signals to help them
monitor their population densities, a phenomenon referred to as quorum
sensing. Quorum sensing was first described over two decades ago in two
luminescent marine bacterial species, Vibrio fischeri and V. harveyi (40),
which have served as model species for studies of cell-density-dependent
gene expression. In both species, the enzymes responsible for light produc-
tion are encoded by the luciferase structural operon luxCDABE (13, 39) and
light emission occurs only at high cell density in response to the accumula-
tion of secreted autoinducer signaling molecules (40). In the 1980s,
Eberhard et al. (11) purified the first homoserine lactone autoinducer
from V. fischeri and showed that it was indeed a specific inducer of lumi-
nescence. In 1983, the basic features of the autoinduction system were
revealed at the molecular level when the lux genes of V. fischeri were
successfully cloned and expressed in Escherichia coli (12).
Although quorum sensing regulation has been analyzed in great detail
in V. harveyi and V. fischeri, the study of quorum sensing phenotypes in the
clinically important Vibrio species V. cholerae was virtually non-existent
until quite recently. This was partly because, unlike V. fischeri and
V. harveyi, V. cholerae does not possess luciferase genes and it was therefore
unclear whether it possessed any genes that were regulated by quorum
sensing. However, when the V. cholerae genome sequence was completed
(22) it was revealed that V. cholerae contains several quorum-sensing genes
similar to those of V. harveyi. In this chapter, we summarize recent studies
on V. cholerae quorum sensing from various laboratories and discuss the
significance of quorum-sensing-mediated regulation of virulence gene
expression and biofilm formation.
101
VIBRIO CHOLERAE AS A HUMAN PATHOGEN AND ITS
INFECTIOUS CYCLE
Cholera is an acute dehydrating diarrheal disease caused principally by
a potent enterotoxin (cholera toxin) produced by toxigenic V. cholerae dur-
ing infection of the human host (15). Epidemic cholera caused by toxigenic
V. cholerae belonging to the O1 or O139 serogroup is a major public health
problem in developing countries, causing an estimated 120,000 deaths
worldwide and many non-fatal cases each year, the vast majority of which
occur in children (41). Cholera is also endemic in many parts of developing
countries, where outbreaks occur widely in a seasonal pattern and are
particularly associated with poverty and poor sanitation.
The pathogenesis of cholera is a complex process and involves a num-
ber of genes encoding virulence factors that aid the pathogen in reaching
and colonizing the epithelium of the small intestine and in producing
cholera toxin (CT) that disrupts ion transport in intestinal epithelial cells (42).
In V. cholerae, the major virulence genes exist in clusters ( 27, 46). These
include the CTX genetic element, which is the genome of a lysogenic
bacteriophage (CTX) that carries the CT genes, and the TCP pathogenicity
island, which carries the genes encoding a pilus colonization factor known as
the toxin coregulated pilus (TCP).
Virulence gene expression in V. cholerae is controlled by multiple
regulatory systems (46). Several critical virulence genes are coordinately
regulated by a cascading system of regulatory factors to respond in a
concerted fashion to specific environmental conditions (Figure 5.1). The
ToxRS and TcpPH membrane complexes are believed to sense environ-
mental cues and to modulate the expression of the AraC-like transcriptional
activator ToxT in response to these cues. ToxT in turn controls the tran-
scription of genes involved in the synthesis of virulence factors such as CT
and TCP. Recent work has identified AphA and AphB as additional activa-
tors of tcpPH expression (30, 46).
Analysis of these regulatory cascades has identified pH, osmolarity,
and temperature as environmental signals that affect virulence gene
expression in vitro. Several other regulatory genes and processes (e.g.
motility and iron acquisition) also influence expression of the ToxR regulon
(32). These different regulatory systems presumably allow V. cholerae bac-
teria to sense their environment and accordingly vary gene expression to
optimize survival under different conditions. Because V. cholerae is both a
human pathogen and part of the normal aquatic flora in estuarine and
brackish waters, the functional repertoire of the V. cholerae genome must be
J
.
Z
H
U
A
N
D
J
.
J
.
M
E
K
A
L
A
N
O
S
102
unusually broad to accommodate two very different lifestyles: as a pathogen
in the human intestine and as a long-term resident of aquatic habitats
during interepidemic periods (45). However, little is known about the
extracellular signals that affect virulence in vivo or survival in the aquatic
environment.
The ToxRS, TcpPH, and ToxT regulators were previously thought to
regulate only virulence factor expression. However, a recent study (3)
attempted to define the entire ToxR regulon by comparing the transcrip-
tional profiles of toxRS, tcpPH, and toxT mutants with those of their
isogenic parent strain under conditions in vitro that are optimized for
production of virulence factors. This study found that TcpPH and ToxRS
control genes involved in metabolism and nutrient uptake, as well as many
genes of unknown function, suggesting that these regulators are involved
in adaptive responses to the host environment as well as in virulence gene
regulation. In contrast, ToxT appears to regulate only the CT and TCP
genes, suggesting that ToxT is involved only in virulence processes.
Environmental signals
(pH, temperature, bile, amino acids, osmolarity)
TcpH
TcpP
AphA
ToxT
ToxS
ToxR
tcpl tcpPH tcpA-F toxT acf ctxA ctxB
TCP
(toxin coregulated pilus)
CT (cholera toxin)
Figure 5.1. Regulatory pathways influencing the expression of virulence genes in Vibrio
cholerae. Signal transduction cascades coordinately regulate virulence in V. cholerae. The
ToxRToxS and TcpPTcpHsignaling circuits detect and respond to environmental stimuli
and activate the expression of toxT. ToxT, in turn, activates the transcription of a variety of
genes required for virulence, including cholera toxin genes and genes in the TCP
pathogenicity island involved in TCP pilus biosynthesis.
103
B
I
O
F
I
L
M
G
R
O
W
T
H
A
N
D
V
I
R
U
L
E
N
C
E
O
F
V
.
C
H
O
L
E
R
A
E
V. cholerae infections begin with the ingestion of contaminated food or
water. The bacteria that survive passage through the acid barrier of the
stomach then colonize epithelial surfaces in the small intestine. They
multiply extensively at this site and produces toxic proteins that alter the
normal function of cells lining the intestine, causing these cells to secrete
water and solutes into the lumen. The resulting diarrhea is believed to
facilitate the spread of V. cholerae back into the environment to reinitiate the
cycle. V. cholerae also maintains a permanent environmental reservoir in
aqueous environments in association with various aquatic organisms.
More recently, Merrell and colleagues (36) reported that V. cholerae present
in stools shed by cholera patients were hyperinfectious for infant mice
compared with a laboratory strain grown in vitro to stationary phase. This
work provides evidence that passage of V. cholerae through the human host
may alter critical virulence phenotypic properties that promote the trans-
mission of the organism to new human hosts.
QUORUM SENSING REGULATION IN V. CHOLERAE
Bacteria utilize quorum-sensing regulation systems to control a var-
iety of physiological functions (17, 37). Many Gram-negative bacteria use
acyl-homoserine lactones, which are synthesized by LuxI-family proteins,
as interbacterial signals and LuxR-type proteins as transcriptional reg-
ulators. In these systems, the autoinducer is usually synthesized by a
protein related to the LuxI protein of V. fischeri and binds to an intracel-
lular receptor protein that resembles the V. fischeri LuxR protein.
Receptorautoinducer complexes are thought to bind to target promoters
to activate their transcription. These autoinducers diffuse rapidly across the
cell envelope and accumulate intracellularly only at high bacterial popula-
tion densities. In Gram-positive species, modified oligopeptides are
secreted and serve as signals when they accumulate at high cell density.
The detectors for the oligopeptide signals are two-component response
regulator proteins.
V. harveyi possesses a complicated, hybrid quorum-sensing regulation
system (1) that has characteristics of both Gram-negative and Gram-positive
QS systems. V. harveyi uses an acyl-HSL autoinducer (AI-1) similar to that
of other Gram-negative quorum sensing bacteria, but the signal
detection and relay apparatus consists of two-component proteins similar
to the quorum sensing systems of Gram-positive bacteria (2). AI-1 is an acyl-
HSL (3-hydroxyl-C4-HSL), but its synthesis is dependent on LuxLM,
which does not share any homology with the LuxI family of autoinducer
J
.
Z
H
U
A
N
D
J
.
J
.
M
E
K
A
L
A
N
O
S
104
synthases. V. harveyi also produces and responds to a second quorum-sensing
autoinducer (AI-2) (48), a furanosyl borate diester that is not closely related to
the homoserine lactones found in other bacteria (6, 43). Synthesis of AI-2 is
dependent on the LuxS protein (55).
Bassler and colleagues used the V. harveyi luciferase operon as a
heterologous reporter to examine quorum sensing in V. cholerae and
found that, like V. harveyi, V. cholerae has at least two quorum sensing
systems (38). System 2 (LuxSPQ) is similar to the V. harveyi system 2,
whereas system 1 (CqsAS) is novel (Figure 5.2). At low cell density, signal
concentration is low, and CqsS and LuxQ act as autophosphorylating
kinases. These phosphates are then transferred to the cytoplasmic
phosphotransferase LuxU, which in turn transfers the phosphates to
LuxO, a s
54
-dependent activator. Upon phosphorylation, LuxO is activated
and functions to repress hapR expression (see below). HapR belongs to
the TetR-family proteins and is a homolog of V. harveyi LuxR (although
unrelated to LuxR from V. fischeri). At high cell density, the CAI-1 and AI-2
signal molecules are produced at a high level and interact with their cognate
sensors. This interaction converts CqsS and LuxQ from kinases to
Figure 5.2. Current model for quorum-sensing regulation in Vibrio cholerae. At low cell
density, LuxO protein is phosphorylated and actively represses hapR expression through
Hfq and 4 sRNAs. Virulence genes are expressed. At high cell density, accumulation of
autoinducers leads to LuxO dephosphorylation; HapR is produced and represses virulence
gene expression and biofilm formation, and activates H/A protease production. A putative
signal system 3 may also be involved in this process. P: phosphate group.
105
B
I
O
F
I
L
M
G
R
O
W
T
H
A
N
D
V
I
R
U
L
E
N
C
E
O
F
V
.
C
H
O
L
E
R
A
E
phosphatases, resulting in the loss of phosphate from LuxU and LuxO.
Consequently, LuxO is inactivated and HapR is expressed. A gene called
cqsA is responsible for the production of CAI-1 molecules. As predicted by
domain conservation, cqsA likely encodes an aminotransferase, but the
structure and biosynthesis of the CAI-1 autoinducer are not similar to
those of previously identified autoinducer molecules.
The quorum sensing apparatus in V. cholerae is unusually complex.
Study of system1 and system2 double mutants revealed that a third sensory
circuit exists in V. cholerae (38). This system 3, which is hypothesized to
respond to an intracellular signal, does not seem to transmit its signal
through LuxU but rather directly through LuxO, the central regulator for
all three systems.
In a recent study, Lenz et al. (35) found that repression of hapR by
phosphorylated LuxO is indirect. At low cell density, phosphorylated LuxO
activates the expression of the small RNAbinding protein Hfq (47) and four
redundant small regulatory RNAs (sRNAs). The HfqsRNA repressor com-
plexes then act to destabilize V. cholerae hapRmRNAs. All four sRNAs must
be inactivated to eliminate Hfq-mediated quorum-sensing repression,
whereas overexpression of only one sRNA is sufficient for repression.
Control via sRNAs may permit an ultrasensitive response to the
concentration of active LuxO, as base pairing of a sRNA with its target
message is known to promote degradation of both the sRNA and the
mRNA. The use of sRNAs to accomplish an ultrasensitive response may
be particularly apt for processes such as quorumsensing in which an all-or-
nothing response is required.
QUORUM SENSING NEGATIVELY REGULATES VIRULENCE
GENE EXPRESSION
Unlike in V. harveyi, there is no luciferase operon in V. cholerae. In an
effort to determine the targets of quorum sensing in V. cholerae, mutations
in quorum-sensing regulators luxO and hapR were tested for their roles in
V. cholerae pathogenesis (60). The hapR mutant colonizes as well as the
wildtype, but the luxO mutant is profoundly defective in colonization in an
infant mouse model. V. cholerae colonization of the infant mouse
(46 d old) intestine is an animal model commonly used for the human
diarrheal disease cholera. This model has been extremely useful in the
identification and characterization of proven and putative virulence factors
involved in human cholera (28). In addition, the expression of the ToxR
regulon was significantly depressed in the luxOmutant, as determined by a
J
.
Z
H
U
A
N
D
J
.
J
.
M
E
K
A
L
A
N
O
S
106
whole-genome microarray analysis. Overexpression of TcpP or ToxT, but
not of ToxR, restores CT production in a luxOmutant, indicating that LuxO
controls the ToxR regulon by repressing TcpP. As predicted, because LuxO
represses hapR, HapR acts downstream of LuxO to repress tcpP expression,
leading to repression of the ToxR virulence regulon. A follow-up study (31)
suggests that, at high cell density, HapR decreases tcpPH transcription
indirectly by binding directly to the aphA promoter and repressing aphA
transcription. Owing to decreased intracellular levels of AphA, hapR
strains exhibit reduced virulence gene expression at high cell density.

Quorum-sensing-mediated virulence repression is achieved through
the autoinducers CAI-1 and AI-2 (38). Addition of cell-free culture super-
natant from the wild type, but not from an autoinducer-deficient mutant,
represses tcpP expression. Autoinducers do not affect tcpP expression when
added to a recipient lacking HapR, showing that CAI-1 and AI-2 contribute
to repression of virulence factor expression specifically through the
quorum sensing circuit described above. Each autoinducer is capable of
partly inhibiting virulence gene expression, but together the autoinducers
have a more than additive repressive effect, suggesting that the V. cholerae
autoinducers function synergistically.
QUORUM SENSING INHIBITS BIOFILM FORMATION
A biofilm is a surface-associated microbial community embedded in a
self-produced, extracellular polymeric matrix (9). These compact microbial
3-D structures are found in numerous environmental sites, such as aquatic
reservoirs and tooth surfaces. Biofilmformation is recognized as a bacterial
developmental process that requires a series of discrete and well-regulated
steps. Biofilms are also clinically significant, as biofilm-associated bacteria
are less susceptible to host immune responses and antimicrobial agents
than are free-swimming bacteria (10). In addition, biofilms are often asso-
ciated with chronic infections, most notably in P. aeruginosa infections of
cystic fibrosis patients and catheter-associated biofilms of Staphylococcus
epidermidis (18).
Several studies have suggested that biofilm-mediated attachment to
abiotic surfaces may be important for V. cholerae survival in the environ-
ment (51, 58). Biofilm formation in V. cholerae is a multistep developmental
process that is controlled by several regulatory pathways (52, 53). In the first
step of biofilm formation, motility is reduced as the bacterium approaches
the surface. The bacterium may then form a transient association with
the surface. The next step is the formation of a microcolony, and
107
B
I
O
F
I
L
M
G
R
O
W
T
H
A
N
D
V
I
R
U
L
E
N
C
E
O
F
V
.
C
H
O
L
E
R
A
E
exopolysaccharide is subsequently produced to form a three-dimensional
biofilm. Finally, when environmental conditions change, some of the
bacteria may detach and swim away to find a surface in a more favorable
environment.
The two-component response regulator VpsR is required for expres-
sion of the Vibrio polysaccharide synthesis (vps) genes, which are essential
for V. cholerae biofilm formation (56). VpsT, another regulator, also is
required for vps gene expression (5). VpsT and VpsR positively regulate
their own expression and also form a complex regulatory network by
positively regulating each others expression. However, the environmental
signals that govern activation by VpsR and VpsT have not been identified.
Interestingly, a report showed that the CytR protein negatively regulates
biofilm formation by repressing vps gene expression (21). CytR also regu-
lates nucleoside catabolism genes in E. coli in response to cytidine concen-
trations. Cyclic diguanylate (c-di-GMP) synthetase and phosphodiesterase
domain-containing proteins have also been found to regulate biofilm for-
mation by controlling c-di-GMP concentration (49). Vibrio polysaccharide
synthesis and biofilm formation are affected by mutation of flagellar struc-
tural genes and a sodium-driven motor, suggesting that V. cholerae may
monitor flagellar torque to sense when a surface is encountered (33, 54).
Biofilm formation may therefore be controlled by both extracellular and
intracellular cues.
Quorum sensing plays a role in biofilm formation in various micro-
organisms (26, 44), including V. cholerae. Quorum-sensing-deficient
V. cholerae hapR mutants produce biofilms thicker than those formed by
wild-type bacteria (20, 50, 59, 60). Microarray analysis and lacZ reporter
fusions of biofilm-associated bacteria demonstrated that this effect results
from enhanced expression of the Vibrio polysaccharide synthesis operons
in hapR mutants (20, 59). Deletion of the AI-2 synthase gene luxS does not
affect biofilm formation, whereas deletion of the CAI-1 synthase gene cqsA
results in the formation of thicker biofilms. These results suggest that AI-2
signals are largely dispensible, whereas CAI-1 signaling is important for
regulating biofilm formation. cqsA mutants grown in the presence of CAI-1
restore hapR expression and formwild-type biofilms; vps expression in cqsA
mutant biofilms is higher than in wild-type biofilms at high cell densities (59).
These data suggest that quorum sensing in V. cholerae also negatively reg-
ulates transcription of the vps genes and keeps biofilm synthesis in check.
Whole-genome expression profiling revealed that HapR may indirectly
repress vps expression by negatively regulating expression of the two positive
vps regulators VpsR and VpsT (57).
J
.
Z
H
U
A
N
D
J
.
J
.
M
E
K
A
L
A
N
O
S
108
There are two points during V. cholerae infection of the host that may be
particularly affected by biofilm formation. First, the bacteria must survive
the acidic environment of the stomach before reaching their site of
replication in the upper intestine; inclusion of the bacterial cells in a
biofilm may provide protection from this harsh environment. In fact,
biofilm-associated V. cholerae are strikingly more resistant to low-pH
shock than are planktonic cells (59). Second, V. cholerae may need to release
from the biofilm in order to colonize the intestinal surface individually.
Quorum-deficient mutants form thicker biofilms and may detach from
biofilms less efficiently than do wild-type bacteria. Because virulence
genes are repressed at high cell density, individual cells released from
the biofilm may be more capable of colonizing the intestinal epithelium
than are cells remaining associated with the biofilm. In fact, the bacteria in
hapR mutant biofilms show a 10-fold lower colonization defect than wild-
type V. cholerae, whereas planktonic hapR mutants can colonize as well as
free-living wild-type cells, suggesting that quorum sensing may affect
intestinal colonization by a mechanism that involves a HapR-dependent
phenotype expressed in biofilm, such as detachment (59). Altogether, the
comparable resistance of mutant and wild-type biofilms to acid shock,
and the HapR-dependent ability of V. cholerae to detach from biofilms,
might contribute to the efficiency of infection of mammalian hosts via
the oralgastric route.
Taken together, we have proposed a working model for the role of
quorum sensing in the V. cholerae infectious cycle (Figure 5.3) (59). The
biofilm structure may be critical during entry into the host in order to
protect against acid shock in the stomach. After the bacteria reach the
intestine, dispersal of individual cells from the biofilm enables these cells
to shut off quorum-sensing-mediated repression of the CT and TCP genes,
thus permitting maximal colonization of intestinal sites. The dispersal of
cells frombiofilms in the intestine may be caused by unknown host signals,
or by the natural dynamic equilibrium of bacterial cells entering and
leaving biofilm structures, or by the combination of the above two.
Failure to disassemble the biofilm reduces the overall level of colonization.
Later in the infection, the number of V. cholerae in the intestine increases
and quorum sensing again represses CT and TCP production and activates
protease production. It has been shown that HapR activates hemagglutinin
protease (H/A protease) at high cell density (25, 50, 60). This protease was
proposed to serve as a detachase during V. cholerae intestinal colonization
(16). Detachment from the epithelium could permit individual cells to
establish new infection foci in the intestine or to exit the host.
109
B
I
O
F
I
L
M
G
R
O
W
T
H
A
N
D
V
I
R
U
L
E
N
C
E
O
F
V
.
C
H
O
L
E
R
A
E
Quorum sensing and biofilm formation may also be important in the
environmental phase of the V. cholerae life cycle. During the environmental
phase, the bacterium resides in diverse aquatic environments, often in associa-
tion with marine plankton (24). Plankton blooms often precede cholera out-
breaks, suggesting a role for these blooms in the epidemiology of disease (7). In
a recent report (8), a simple, sari-clothfiltrationprocedure that removes particles
larger than 20 mmfromwater was found to reduce cholera cases in Bangladesh
by over 50%. This report suggests that particulate material plays an important
role in V. cholerae transmission; the particulate material removed by sari-cloth
filtration may in fact be V. cholerae biofilm material. This biofilm material may
be associated with zooplankton (7, 23, 24), chironomid egg masses (4, 19) or
other organic material. If so, the removal of acid-resistant V. cholerae biofilms
might explain the utility of sari-cloth filtration. Finally, Vibrio biofilms may also
be shed by cholera patients, and this in turn could help explain the interesting
finding that V. cholerae from human stool appears more infectious than
V. cholerae from stationary-phase cultures grown in vitro (36).
brush border
low density:
virulence gene
expressed
growth growth Quorum Sensing:
virulence-off
protease-on
colonization of
intestinal epithelium
detachment coincident with
diarrhea
Quorum Sensing:
vps-off
resistance to the
gastric acid barrier
oral ingestion
Biofilms
shed by victims or formed in
the aquatic reservoir
Figure 5.3. Models for the function of quorum sensing regulation and biofilm formation in
Vibrio cholerae infectious cycles. V. cholerae resides in various aquatic environments in the
form of biofilms. Upon ingestion, the biofilm structure protects V. cholerae cells from acid
shock in the gastric environment. After passing through the stomach, quorum sensing
represses vps production and individual cells dispersed from the biofilm show increased
virulence gene expression, which enhances intestinal colonization. As bacteria multiply to
high density on the intestinal epithelium, quorum sensing promotes detachment by
repressing virulence gene expression and increasing H/A protease expression. Detached
vibrios are shed and reinduce biofilm formation at low cell density.
J
.
Z
H
U
A
N
D
J
.
J
.
M
E
K
A
L
A
N
O
S
110
The formation of V. cholerae biofilms might also improve the environ-
mental fitness of V. cholerae in other ways. In addition to protecting bacteria
from physicalchemical insults such as changes in pH or the presence of
antibiotics, biofilms likely make bacteria more resistant to biological
threats as well. Recent data support the conclusion that environmental
bacteriophages play a role in the epidemiology of cholera in Bangladesh
(14). Thus, the sensitivity of V. cholerae to environmental vibriophages or
predatory eukaryotic protozoa may be modulated by biofilm formation. If
so, biofilm formation may play complex roles in the seasonality of cholera,
its endemicity, and its transmission.
PERSPECTIVES AND FUTURE STUDIES
Quorum sensing negatively regulates V. cholerae virulence gene
expression in vitro (38, 60). There remain, however, numerous unanswered
questions regarding howquorumsignals and other environmental cues are
integrated to regulate virulence in vivo. For example, in the induction
medium in vitro, HapR must be produced through quorum sensing path-
ways in early log phase to exert its inhibitory effects on virulence (60).
Otherwise, once the ToxR-regulon is activated, the autoregulation of toxT
overrides the HapReffect. Lee et al. employed a recombination-based in vivo
expression technology (RIVET) to show that the temporal expression pat-
terns of critical V. cholerae virulence genes in vivo are clearly different from
expression patterns in vitro (34). Thus, it will be critical to examine the
temporal profiles of quorum sensing and virulence gene regulation in the
context of true infection.
Interestingly, several toxigenic V. cholerae strains (e.g. El Tor strain
N16961 and classical strain O395) possess a natural frameshift mutation in
the hapR gene (60). These strains express low levels of H/A protease. In
addition, a mutationinluxOinthese strains does not result ina CTproduction
defect (29). A high mutation rate at the hapR locus was also observed in
V. cholerae biofilms, thus promoting the formation of more robust biofilms
(20). We speculate that elimination of HapR, either by mutation or by LuxO
repression, may improve the adaptation of V. cholerae to different environ-
ments. However, in the infant mouse model, planktonic hapR mutant cells
colonize normally. This is consistent with our observation that HapR is not
involved in long-termcolonization (59). These findings suggest that the role of
HapR needs to be tested in some other animal models. Also, as we have
demonstrated, hapR mutants form thicker biofilms, and this may aid their
survival in various environments.
111
B
I
O
F
I
L
M
G
R
O
W
T
H
A
N
D
V
I
R
U
L
E
N
C
E
O
F
V
.
C
H
O
L
E
R
A
E
Another important and unresolved question is the chemical nature of
the Vibrio autoinducers. It has been established that V. cholerae produces at
least two distinct autoinducers, one of which is AI-2, a borate diester
produced by the LuxS protein that is conserved in many Gram-positive
and Gram-negative bacteria, including V. cholerae. Although the other
V. cholerae autoinducer has yet to be identified, it is more critically involved
than AI-2 in regulating virulence and biofilm formation (59). Therefore, it
is important to characterize the chemical properties of this autoinducer and
the exact roles it plays in regulating virulence. The second autoinducer may
also represent a potential therapeutic molecule, since it can decrease viru-
lence gene expression by causing elevated HapR expression and thus
decreased tcpPH expression (38, 60).
Finally, quorum sensing apparently controls multiple cellular func-
tions in V. cholerae. In addition to regulating virulence-factor expression
and biofilm formation, LuxO and HapR also control genes involved in
chemotaxis and motility, protein secretion, and protease production, as
well as a number of genes encoding conserved hypothetical proteins (60).
Moreover, comparison of the transcriptional profiles of a V. cholerae cqsA
mutant grown under different conditions in the presence and absence of
CAI-1 (J. Zhu and J. J. Mekalanos, unpublished) revealed that oxygen
content apparently affects gene regulation of CAI-1 signal pathways. For
example, ABC transporters of lipoproteins are induced by CAI-1 during
aerobic growth, but repressed during anaerobic growth. More interestingly,
transcriptional profiling of cqsA mutants attached to the intestinal epithe-
lial cells shows that CAI-1 signals repress the virulence regulatory genes
tcpPH and genes involved in biofilm formation, but activate genes involved
in twitching motility and fimbrial assembly (authors unpublished data).
In conclusion, further study of quorum sensing in V. cholerae should
enable us to understand better the natural role of quorum sensing
during V. cholerae infection. We believe that characterizing the relation-
ships among biofilm development, quorum sensing, infectivity, and
pathogenesis in V. cholerae will advance our goal of understanding the
hostpathogen interaction and discovering potential novel treatments for
cholera disease.
ACKNOWLEDGEMENTS
We thank Drs Su Chiang and Deb Hung for critical reading of the
manuscript. Research in the Mekalanos laboratory is supported by NIH
grants AI18045 and AI26289. JZ is supported by a NRSA fellowship.
J
.
Z
H
U
A
N
D
J
.
J
.
M
E
K
A
L
A
N
O
S
112
REFERENCES
1 Bassler, B. L. 1999. Howbacteria talk to each other: regulation of gene expression
by quorum sensing. Curr. Opin. Microbiol. 2: 5827.
2 Bassler, B. L., M. Wright and M. R. Silverman 1994. Multiple signaling systems
controlling expression of luminescence in Vibrio harveyi: sequence and function
of genes encoding a second sensory pathway. Molec. Microbiol. 13: 27386.
3 Bina, J., J. Zhu, M. Dziejman et al. 2003. ToxR regulon of Vibrio cholerae and its
expression in vibrios shed by cholera patients. Proc. Natn. Acad. Sci. USA 100:
28016.
4 Broza, M. and M. Halpern 2001. Pathogen reservoirs. Chironomid egg masses
and Vibrio cholerae. Nature 412: 40.
5 Casper-Lindley, C. and F. H. Yildiz 2004. VpsT is a transcriptional regulator
required for expression of vps biosynthesis genes and the development of rugose
colonial morphology in Vibrio cholerae O1 El Tor. J. Bacteriol. 186: 15748.
6 Chen, X., S. Schauder, N. Potier 2002. Structural identification of a bacterial
quorum sensing signal containing boron. Nature 415: 5459.
7 Colwell, R. R. 1996. Global climate and infectious disease: the cholera paradigm.
Science 274: 202531.
8 Colwell, R. R., A. Huq, M. S. Islam et al. 2003. Reduction of cholera in
Bangladeshi villages by simple filtration. Proc. Natn. Acad. Sci. USA 100: 10515.
9 Costerton, J. W., Z. Lewandowski, D. E. Caldwell, D. R. Korber and H. M. Lappin-
Scott 1995. Microbial biofilms. A. Rev. Microbiol. 49: 71145.
10 Costerton, J. W., P. S. Stewart and E. P. Greenberg 1999. Bacterial biofilms: a
common cause of persistent infections. Science 284: 131822.
11 Eberhard, A., A. L. Burlingame, C. Eberhard et al. 1981. Structural identification
of autoinducer of Photobacterium fischeri luciferase. Biochemistry 20: 24449.
12 Engebrecht, J., K. Nealson and M. Silverman 1983. Bacterial bioluminescence:
isolation and genetic analysis of functions from Vibrio fischeri. Cell 32: 77381.
13 Engebrecht, J. and M. Silverman 1984. Identification of genes and gene products
necessary for bacterial bioluminescence. Proc. Natn. Acad. Sci. USA 81: 41548.
14 Faruque, S., I. B. Nasser, M. J. Islamet al. 2005. Seasonal epidemics of cholera are
inversely correlated with the prevalence of environmental cholera phages. Proc.
15 Faruque, S. M., M. J. Albert and J. J. Mekalanos 1998. Epidemiology, genetics,
and ecology of toxigenic Vibrio cholerae. Microbiol. Molec. Biol. Rev. 62: 130114.
16 Finkelstein, R. A., M. Boesman-Finkelstein, Y. Chang and C. C. Hase 1992.
Vibrio cholerae hemagglutinin/protease, colonial variation, virulence, and detach-
ment. Infect. Immun. 60: 4728.
17 Fuqua, C. and E. P. Greenberg 2002. Listening in on bacteria: acyl-homoserine
lactone signaling. Nat. Rev. Molec. Cell Biol. 3: 68595.
18 Hall-Stoodley, L., J. W. Costerton and P. Stoodley 2004. Bacterial biofilms: from
the natural environment to infectious diseases. Nat. Rev. Microbiol. 2: 95108.
113
B
I
O
F
I
L
M
G
R
O
W
T
H
A
N
D
V
I
R
U
L
E
N
C
E
O
F
V
.
C
H
O
L
E
R
A
E
19 Halpern, M., Y. B. Broza, S. Mittler, E. Arakawa, and M. Broza 2004. Chironomid
egg masses as a natural reservoir of Vibrio cholerae non-O1 and non-O139 in
freshwater habitats. Microb. Ecol. 47: 3419.
20 Hammer, B. K. and B. L. Bassler 2003. Quorum sensing controls biofilm forma-
tion inVibrio cholerae. Molec. Microbiol. 50: 1014.
21 Haugo, A. J. and P. I. Watnick 2002. Vibrio cholerae CytR is a repressor of biofilm
development. Molec. Microbiol. 45: 47183.
22 Heidelberg, J. F., J. A. Eisen, W. C. Nelson et al. 2000. DNA sequence of both
chromosomes of the cholera pathogen Vibrio cholerae. Nature 406: 47783.
23 Huq, A., S. A. Huq, D. J. Grimes et al. 1986. Colonization of the gut of the blue
crab (Callinectes sapidus) by Vibrio cholerae. Appl. Environ. Microbiol. 52: 5868.
24 Huq, A., E. B. Small, P. A. West et al. 1983. Ecological relationships between Vibrio
cholerae and planktonic crustacean copepods. Appl. Environ. Microbiol. 45: 27583.
25 Jobling, M. G. and R. K. Holmes 1997. Characterization of hapR, a positive
regulator of the Vibrio cholerae HA/protease gene hap, and its identification as a
functional homologue of the Vibrio harveyi luxR gene. Molec. Microbiol.
26: 102334.
26 Kjelleberg, S. and S. Molin 2002. Is there a role for quorum sensing signals in
bacterial biofilms? Curr. Opin. Microbiol. 5: 2548.
27 Klose, K. E. 2001. Regulation of virulence in Vibrio cholerae. Int. J. Med. Microbiol.
291: 818.
28 Klose, K. E. 2000. The suckling mouse model of cholera. Trends Microbiol.
8: 18991.
29 Klose, K. E., V. Novik and J. J. Mekalanos 1998. Identification of multiple sigma54-
dependent transcriptional activators in Vibrio cholerae. J. Bacteriol. 180: 52569.
30 Kovacikova, G., W. Lin and K. Skorupski 2004. Vibrio cholerae AphA uses a novel
mechanism for virulence gene activation that involves interaction with the LysR-
type regulator AphB at the tcpPH promoter. Molec. Microbiol. 53: 12942.
31 Kovacikova, G. and K. Skorupski 2002. Regulation of virulence gene expression
in Vibrio cholerae by quorum sensing: HapR functions at the aphA promoter.
32 Krukonis, E. S. and V. J. DiRita. 2003. From motility to virulence: sensing and
responding to environmental signals in Vibrio cholerae. Curr. Opin. Microbiol.
6: 18690.
33 Lauriano, C. M., C. Ghosh, N. E. Correa and K. E. Klose 2004. The sodium-driven
flagellar motor controls exopolysaccharide expression in Vibrio cholerae.
J. Bacteriol. 186: 486474.
34 Lee, S. H., D. L. Hava, M. K. Waldor and A. Camilli 1999. Regulation and
temporal expression patterns of Vibrio cholerae virulence genes during infection.
Cell 99: 62534.
35 Lenz, D. H., K. C. Mok, B. N. Lilley et al. 2004. The small RNA chaperone Hfq
and multiple small RNAs control quorum sensing in Vibrio harveyi and Vibrio
cholerae. Cell 118: 6982.
J
.
Z
H
U
A
N
D
J
.
J
.
M
E
K
A
L
A
N
O
S
114
36 Merrell, D. S., S. M. Butler, F. Qadri et al. 2002. Host-induced epidemic spread of
the cholera bacterium. Nature 417: 6425.
37 Miller, M. B. and B. L. Bassler 2001. Quorum sensing in bacteria. A. Rev.
38 Miller, M. B., K. Skorupski, D. H. Lenz, R. K. Taylor and B. L. Bassler 2002. Parallel
quorum sensing systems converge to regulate virulence in Vibrio cholerae.
Cell 110: 30314.
39 Miyamoto, C. M., M. Boylan, A. F. Graham and E. A. Meighen 1988.
Organization of the lux structural genes of Vibrio harveyi. Expression under the
T7 bacteriophage promoter, mRNAanalysis, and nucleotide sequence of the luxD
gene. J. Biol. Chem. 263: 133939.
40 Nealson, K. H. and J. W. Hastings 1979. Bacterial bioluminescence: its control
and ecological significance. Microbiol. Rev. 43: 496518.
41 World Health Organization 1995. Report meeting: The Potential Role of New
Cholera Vaccine in the Prevention and Control of Cholera During Acute
Emergencies. World Health Organization.
42 Reidl, J. and K. E. Klose 2002. Vibrio cholerae and cholera: out of the water and
into the host. FEMS Microbiol. Rev. 26: 12539.
43 Schauder, S., K. Shokat, M. G. Surette and B. L. Bassler 2001. The LuxS family of
bacterial autoinducers: biosynthesis of a novel quorum-sensing signal molecule.
44 Schembri, M. A., M. Givskov and P. Klemm 2002. An attractive surface: gram-
negative bacterial biofilms. Sci. Signal Transduct. Knowl. Environ. 2002: RE6.
45 Schoolnik, G. K. and F. H. Yildiz 2000. The complete genome sequence of Vibrio
cholerae: a tale of two chromosomes and of two lifestyles. Genome Biol 1: Reviews,
1016 13.
46 Skorupski, K. and R. K. Taylor 1997. Control of the ToxR virulence regulon in
Vibrio cholerae by environmental stimuli. Molec. Microbiol. 25: 10039.
47 Storz, G., J. A. Opdyke and A. Zhang 2004. Controlling mRNA stability and
translation with small, noncoding RNAs. Curr. Opin. Microbiol. 7: 1404.
48 Surette, M. G., M. B. Miller and B. L. Bassler 1999. Quorum sensing in
Escherichia coli, Salmonella typhimurium, and Vibrio harveyi: a new family of
genes responsible for autoinducer production. Proc. Natn. Acad. Sci. USA
96: 163944.
49 Tischler, A. D. and A. Camilli 2004. Cyclic diguanylate (c-di-GMP) regulates
Vibrio cholerae biofilm formation. Molec. Microbiol. 53: 85769.
50 Vance, R. E., J. Zhu and J. J. Mekalanos 2003. Aconstitutively active variant of the
quorum-sensing regulator LuxO affects protease production and biofilm forma-
tion in Vibrio cholerae. Infect. Immun. 71: 25716.
51 Wai, S. N., Y. Mizunoe, A. Takade, S. I. Kawabata and S. I. Yoshida 1998. Vibrio
cholerae O1 strain TSI-4 produces the exopolysaccharide materials that determine
colony morphology, stress resistance, and biofilm formation. Appl. Environ.
115
B
I
O
F
I
L
M
G
R
O
W
T
H
A
N
D
V
I
R
U
L
E
N
C
E
O
F
V
.
C
H
O
L
E
R
A
E
52 Watnick, P. and R. Kolter 2000. Biofilm, city of microbes. J. Bacteriol.
182: 26759.
53 Watnick, P. I. and R. Kolter 1999. Steps in the development of a Vibrio cholerae El
Tor biofilm. Molec. Microbiol. 34: 58695.
54 Watnick, P. I., C. M. Lauriano, K. E. Klose, L. Croal and R. Kolter 2001. The
absence of a flagellum leads to altered colony morphology, biofilm development
and virulence in Vibrio cholerae O139. Molec. Microbiol. 39: 22335.
55 Xavier, K. B. and B. L. Bassler 2003. LuxS quorum sensing: more than just a
numbers game. Curr. Opin. Microbiol. 6: 1917.
56 Yildiz, F. H., N. A. Dolganov and G. K. Schoolnik 2001. VpsR, a member of the
response regulators of the two-component regulatory systems, is required for
expression of vps biosynthesis genes and EPS(ETr)-associated phenotypes in
Vibrio cholerae O1 El Tor. J. Bacteriol. 183: 171626.
57 Yildiz, F. H., X. S. Liu, A. Heydorn and G. K. Schoolnik 2004. Molecular analysis
of rugosity in a Vibrio cholerae O1 El Tor phase variant. Molec. Microbiol.
53: 497515.
58 Yildiz, F. H. and G. K. Schoolnik 1999. Vibrio cholerae O1 El Tor: identification of
a gene cluster required for the rugose colony type, exopolysaccharide production,
chlorine resistance, and biofilm formation. Proc. Natn. Acad. Sci. USA
96: 402833.
59 Zhu, J. and J. J. Mekalanos 2003. Quorum sensing-dependent biofilms enhance
colonization in Vibrio cholerae. Dev. Cell 5: 64756.
60 Zhu, J., M. B. Miller, R. E. Vance et al. 2002. Quorum-sensing regulators control
virulence gene expression in Vibrio cholerae. Proc. Natn. Acad. Sci. USA
99: 312934.
J
.
Z
H
U
A
N
D
J
.
J
.
M
E
K
A
L
A
N
O
S
116
CHAPTER 6
LuxS in cellular metabolism
and cell-to-cell signaling
Kangmin Duan
Molecular Microbiology Laboratory, Northwest University, Xian, China
Michael G. Surette
Faculty of Medicine, University of Calgary, Canada
INTRODUCTION
Many bacteria regulate gene expression in a cell-density-dependent
manner; this behaviour has been collectively referred to as quorum
sensing or cell-to-cell communication. In its simplest form this process
results from the production and accumulation of signaling molecules
in the surrounding environment. At some threshold concentration,
the signaling molecules (also referred to as autoinducers) bind to
receptors on or in the bacterial cell that regulate gene expression. In recent
years the significance of cell-to-cell signaling in bacteria has become widely
appreciated and has been extensively reviewed (2, 21, 25, 64, 84, 106).
The concept that bacterial cells can communicate through small-
molecule chemical signals first arose in the pioneering studies in the
Hastings laboratory in the early 1970s. Vibrio fischeri produces light;
the light production is induced as a culture reaches high cell density,
i.e. in a cell-density-dependent manner. In the seminal paper by Nealson
et al. (69) it was demonstrated that cell-free cultures of V. fischeri contained
a substance (termed an autoinducer) which, when added back to cell
cultures, induced light production in the cells at a much earlier stage in
the culture. Since that time the field of cell-to-cell signaling (quorum
sensing) has grown dramatically; in recent years there has been an explosion
of research in this area. In general there remain two well-established
paradigms of cell-to-cell signaling: the N-acyl homoserine lactones used by
some Gram-negative bacteria (see Chapters 1 and 3) and the oligopeptide
signals used by some Gram-positive bacteria (see Chapters 9 and 10).
These systems are described in detail elsewhere in this book and briefly
outlined below.
117
HOMOSERINE LACTONE QUORUM SENSING SYSTEMS
The classical example of quorumsensing is the autoinduction of lumines-
cence in the marine bacterium V. fischeri (32, 69). In this system an N-acyl
homoserine lactone (AHL) interacts directly witha regulatory protein, LuxR, to
regulate expression of luminescence genes. The LuxI protein carries out
synthesis of the signaling molecule. Over 40 members of the luxI/luxR gene
family have been identified in Gram-negative bacteria and are involved in a
wide range of responses (30, 53, 64, 74, 79, 102). A second family of AHL
biosynthetic enzymes is known; these probably employ the same biosynthesis
mechanismas LuxI but share no sequence homology with it. The best-studied
examples of this family are the LuxLM proteins from V. harveyi (5). Since
bacteria produce species-specific autoinducers and each species usually
responds to its own autoinducers, the AHL-mediated quorum sensing is
believed to be an intraspecies signaling system.
OLIGOPEPTIDE QUORUM SENSING PATHWAYS
Many Gram-positive bacteria communicate in a cell-density-dependent
manner by using oligopeptides (see recent reviews (52, 94)). These systems
involve synthesis of a precursor polypeptide that is cleaved during or after
export; in some cases the mature signaling peptide may be further modi-
fied. Receptors for the peptides are typically found in the inner membrane
and belong to the histidine kinase family of two-component signal trans-
duction systems. In Bacillus subtilis (58) and Streptococcus pneumoniae (35),
oligopeptide signaling is involved in regulation of competence. In
Enterococcus faecalis and related species, oligopeptide signaling regulates
aggregation and conjugational transfer between strains (68). In
Enterococcus both stimulatory and inhibitory peptides are made and the
active signaling occurs between related but not identical cells. In
Staphylococcus aureus, a modified peptide cyclized into a thiolactone ring
is involved in cell-to-cell signaling; many virulence factors are under control
of this global regulatory system (44). Different groups of S. aureus produce
different signaling peptides. Each group is distinguished by the observation
that the peptides produced by isolates within the group are recognized as
signaling molecules but the peptides from other groups are competitive
inhibitors (43). Similar competition has been demonstrated between
S. aureus and S. epidermidis strains (71, 72).
Although the AHL and oligopeptide cell-to-cell signaling systems
remain the most recognized and studied signaling systems, there are
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
118
other very well established signaling pathways such as the butyryl lactone
signals of Streptomyces (26, 38) and the two signaling systems (A-signal and
C-signal) of Myxococcus xanthus (41, 47, 48). There seems also to be an
increasing list of cell-to-cell signaling molecules of distinct chemical
properties.
The focus of this chapter will be on another pathway that involves the
production of a signal, called autoinducer-2 (AI-2), by the enzyme LuxS.
The identification of the luxS gene and a readily available bioassay for AI-2
has resulted in the identification of AI-2 production by a wide variety of
bacteria. The elucidation of the biochemical pathway involved in AI-2
production and the lack of a luxS phenotype in many bacteria has also led
to the question as to whether AI-2 is indeed involved in cell-to-cell
signaling. The chapter will review the biology of luxS and AI-2 production
and examine the recent evidence for and against a role in cell-to-cell
signaling.
IDENTIFICATION OF LuxS
The identification of the luxS/AI-2 pathway originated in the marine
bacterium V. harveyi. In this bacterium it was recognized that two
independent signals were produced and detected by the organism.
Although the species is Gram-negative, V. harveyi has a cell-to-cell signal-
ing system with elements of both Gram-negative and Gram-positive para-
digms, as described above. Two signaling molecules, autoinducer-1 (AI-1)
and autoinducer-2 (AI-2), are involved. The first signal is b-hydroxy-butyryl
homoserine lactone (10), typical of the Gram-negative bacteria. However, the
synthesis of this signal is not by a LuxI protein homologue but by a distinct
class of enzymes, designated LuxLM(5). The detectionof the signals was found
to be through two-component histidine kinase receptors; the mechanism is
similar to that of oligopeptide sensing in the Gram-positive bacteria. The
homoserine lactone AI-1 is sensed at the cell surface by the histidine kinase
receptor LuxN (5). The detection of AI-2 is through a second histidine kinase
(LuxQ); in this instance the ligand is a complex of a periplasmic binding
protein, LuxP, and AI-2 (6). The signal transduction pathways for the two
histidine kinases converge on a common phosphorelay cascade, outlined in
Figure 6.1 (67).
The identification of the signal pathway led to the generation of
V. harveyi reporter strains that could detect either AI-1 or AI-2 (luxN and
luxQ mutants, respectively) (6). These reporter strains or variants thereof
remain the primary method of detection of AI-2 today. Using these reporter
119
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
strains, Bassler and collaborators demonstrated that, among a wide variety
of bacteria samples, none produced a detectable AI-1 signal but many
produced measurable AI-2 activity (4). This led to the hypothesis that in
V. harveyi System 1 is a species-specific signaling system and System 2 is a
more general signaling system. This has been to some extent borne out and
is reflected in the nature of species-specific acyl homoserine lactone
signals, although some crosstalk between organisms can be observed
together with the apparent universal detection of AI-2 activity by V. harveyi.
This will be elaborated on below.
Despite many years of screening, a gene responsible for AI-2 pro-
duction in V. harveyi had not been identified. That changed when it was
discovered that both Salmonella typhimurium and many Escherichia coli
sRNAs
LuxR
Hfq
luxCDABE
LuxO
LuxU
LuxQ
LuxN
LuxS
LuxLM
AI-2
AI-1
LuxP
Figure 6.1. The schematic phosphorelay of the V. harveyi quorum sensing system. The
cascade of phosphate transfer is influenced by cell density through signal-molecule
concentration. AI-1 and AI-2 are produced by LuxLM and LuxS, respectively. At low cell
density LuxQ and LuxN act as kinase and transfer P to LuxU, and in turn to the response
regulator LuxO, which regulates gene transcription through small regulatory RNA
molecules (sRNA) and Hfq. The sRNAHfq complex is a negative regulator of luxR, the
activator of the luciferase gene expression. At high cell density LuxQ and LuxN bind with
cognate autoinducers, increasing their phosphatase activity. Phosphate flow is reverted
from LuxO to LuxU to LuxQ and LuxN. Dephosphorylation of LuxO relieves the inhibition
of luxRexpression, leading to light production. The histidine kinase and response regulator
domains are shown as black rectangles and grey circles, respectively. The response
regulator LuxO contains a DNA binding domain (hatched). AI-1 (HSL) and AI-2 are
depicted as filled circles and squares, respectively. Adapted from Lenz et al. (54a).
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
120
strains (but not the common laboratory strain DH5a) produce AI-2 activity
in a growth-condition and phase-dependent manner (95, 97). Significantly
no AI-2 activity was detected in cell-free culture supernatants of these
bacteria in the original strain survey by Bassler et al. (4). By using
the V. harveyi AI-2 specific reporter strains, a transposon mutant of
S. typhimurium was identified that did not produce AI-2 (95). The insertion
was in a gene of unknown function (ygaG); it was shown that the corre-
sponding gene in DH5a contained a 1 bp deletion inactivating the gene.
AI-2 production was restored in the Salmonella transposon mutant and
in E. coli DH5a was restored by adding back the Salmonella gene in trans.
This complementation assay was used to screen for the V. harveyi gene
responsible for AI-2 production, which was, not surprisingly, in a ygaG
homolog. Given the role of this enzyme in regulating light production in
V. harveyi, it was designated luxS (96, 97). The luxS gene was found to
be widespread in many bacterial species, including both Gram-negative
and Gram-positive species (97) (Figure 6.2); to our knowledge, AI-2
activity can be detected from culture-free supernatants of all species with a
functional luxS.
METABOLIC PATHWAY OF AI-2 PRODUCTION
Despite the identification of the gene responsible for AI-2 production, the
chemical identity of AI-2 and the nature of its production remained elusive. In
most bacteria, luxS is in a single-gene operon, giving no clues to its function.
The first hint at the pathway came from a comparative genomics analysis.
In the spirochete Borrelia burgdorferi it was observed that the luxS gene was
in an operon with metK and pfs (80). MetK is responsible for the production
of the universal metabolite S-adenosylmethionine (SAM); Pfs is a bifunctional
enzyme with methylthioadenosine/S-adenosylhomocysteine nucleotidase
activity (19). SAM is utilized by a number of methyltransferases in the cell;
the by-product of methyltransferase reaction is S-adenosylhomocysteine
(SAH). SAH is a potent inhibitor of methyltransferases and is actively
degraded in all cells. In eukaryotes and some bacteria, the removal of SAH
is mediated by a SAH hydrolase generating homocysteine and adenosine.
In other bacteria SAH is converted to S-ribosylhomocysteine (SRH) and
adenine by Pfs (33, 103). Based on the comparative genomics, Schauder et al.
reasoned that LuxS catalyzed the cleavage of SRH to 4,5-dihydroxy-2,3-
pentanedione (DPD) and homocysteine. This enzymatic activity was known
to be present in E. coli (24, 33, 62), but the gene responsible had not been
identified. Similar reasoning by Winzer et al. (103) came to the same
121
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
LUXS_BACAN
LUXS_BACAN#2
LUXS_BACCR
LUXS_BACCR#2
LUXS_BACSU
LUXS_BACSU#2
LUXS_BACHD
LUXS_BACHD#2
LUXS_HELPJ
LUXS_HELPY
LUXS_LISIN
LUXS_LISMO
LUXS_STAAM
LUXS_STAEP
LUXS_OCEIH
LUXS_ENTFA
LUXS_LACLA
LUXS_LACPL
LUXS_STRA3
LUXS_STRGN
LUXS_STRPN
LUXS_STRMU
LUXS_STRPY
LUXS_CLOPE
LUXS_CLOAB
LUXS_CLOAB#2
LUXS_DEIRA
LUXS_CAMJE
LUXS_CAMJE#2
LUXS_ECO57
LUXS_ECOLI
LUXS_SHIFL
LUXS_ECOL6
LUXS_SALTI
LUXS_SALTY
LUXS_YERPE
LUXS_PHOLL
LUXS_PROMI
LUXS_SHEON
LUXS_VIBCH
LUXS_VIBHA
LUXS_VIBPA
LUXS_VIBVU
LUXS_VIBVY
LUXS_HAEIN
LUXS_PASMU
LUXS_NEIMA
LUXS_NEIMB
LUXS_HAEDU
LUXS_HELHP
LUXS_WOLSU
Figure 6.2. The phylogenetic tree of the LuxS protein, indicating the relatedness of the LuxS
protein from 46 bacteria. The corresponding organisms are: BACAN, Bacillus anthracis;
BACCR, B. cereus strain ATCC 14579 / DSM 31; BACHD, B. halodurans; BACSU,
B. subtilis; BORBU, Borrelia burgdorferi; CAMJE, Campylobacter jejuni; CLOAB,
Clostridium acetobutylicum; CLOPE, C. perfringens; DEIRA, Deinococcus radiodurans;
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
122
conclusion. The reconstitution of the production in vitro of AI-2 from SAH
by using Pfs and LuxS or from SRH by using only LuxS confirmed the
predictions (80, 103). Homocysteine was shown not to be the AI-2 signal,
clearly implicating DPD or a derivative. DPD is chemically unstable and
spontaneously cyclizes to a mixture of furanone rings that are more stable
than the linear DPD. Knowing the biochemical pathway still left the chemical
identity unknown.
There has been significant progress on the characterization of LuxS
and its enzymatic mechanism in the past few years (73, 75). There are at
least four LuxS orthologs for which X-ray crystal structures have been
reported (37, 55, 78). The enzyme is a homodimer with two active sites.
The crystal structures contain a zinc ion near the active site, suggesting that
the molecule is a zinc metalloenzyme. Recent enzymatic studies, however,
suggest that Fe may be preferred over zinc, with the metal playing an active
role in catalysis (108). A more detailed understanding of the reaction
mechanism (108, 109) has led to the synthesis of LuxS inhibitors (1),
which will prove useful in further characterization of LuxS.
THE STRUCTURES OF AI-2
The chemical identity of AI-2 had remained elusive for many years.
The chemical properties associated with the activity, a small polar mole-
cule, contributed to the difficulty in purifying the activity from complex
Figure 6.2. (cont.)
ECO57, Escherichia coli O157:H7; ECOL6, E. coli O6; ECOLI, E. coli; ENTFA, Enterococcus
faecalis (Streptococcus faecalis); HAEDU, Haemophilus ducreyi; HAEIN, H. influenzae;
HELHP, Helicobacter hepaticus; HELPJ, H. pylori J99 (Campylobacter pylori J99); HELPY,
H. pylori (C. pylori); LACLA, Lactococcus lactis (subsp. lactis) (Streptococcus lactis); LACPL,
L. plantarum; LISIN, Listeria innocua; LISMO, L. monocytogenes; NEIMA, Neisseria
meningitidis (serogroup A); NEIMB, N. meningitidis (serogroup B); OCEIH, Oceanobacillus
iheyensis; PASMU, Pasteurella multocida; PHOLL, Photorhabdus luminescens (subsp.
laumondii); PROMI, Proteus mirabilis; SALTI, Salmonella typhi; SALTY, S. typhimurium;
SHEON, Shewanella oneidensis; SHIFL, Shigella flexneri; STAAM, Staphylococcus aureus (strain
Mu50/ATCC 700699, N315, and MW2); STAEP, S. epidermidis; STRA3, Streptococcus
agalactiae (serotype III and serotype V); STRGN, S. gordonii; STRMU, S. mutans; STRPN,
S. pneumoniae (strain ATCC BAA-255/R6); STRPY, S. pyogenes (serotype M3 and serotype
M18); VIBCH, Vibrio cholerae; VIBHA, V. harveyi; VIBPA, V. parahaemolyticus; VIBVU,
V. vulnificus; VIBVY, V. vulnificus (strain YJ016); WOLSU, Wolinella succinogenes; YERPE,
Yersinia pestis.
123
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
culture media. With hindsight, this was further complicated by the hetero-
geneity in molecular species. Surprisingly, the structure(s) of the active
signal molecule were resolved not by small molecule purification and
identification but by crystallization and structure determination of the
receptor proteins in two seminal papers by the groups of Fred Hughson
and Bonnie Bassler at Princeton University (11, 66). Each of these studies
yielded quite unexpected results.
Using LuxP, which is the ligand-binding receptor of AI-2, Chen et al.
(11) trapped AI-2 out of the mixture of furanone molecules and successfully
determined the structure of AI-2 by crystallizing LuxP in a complex
with the autoinducer. As shown in Figure 6.3, the AI-2 in V. harveyi is a
borate containing (2S,4S)-2-methyl-2,3,3,4- tetrahydroxytetrahydrofuran
(S-THMF-borate). The identification of borate in the LuxP structure
was surprising. The use of borate in biological systems is very uncommon;
however, given that the concentration of borate in seawater is in millimolar
ranges, it would be an abundant anion for V. harveyi to utilize.
By using the same strategy, the crystal structure of S. typhimurium AI-2
binding protein LsrB was determined together with the bound AI-2.
Unlike that in V. harveyi the active form of AI-2 is the unborated form
of (2R,4S)-2-methyl-2,3,3,4-tetrahydroxytetrahydrofuran (R-THMF) (66)
(Figure 6.3). Importantly, the R-THMF is in equilibrium with DPD,
which is in turn in equilibrium with S-THMF-borate, and the equilibrium
may change toward one of the active AI-2 molecules depending on the
environmental conditions. The chemical equilibrium between DPD and its
derivatives also indicated that these AI-2 molecules, whether borated or
not, are interconvertible. This at least in part explains why AI-2 produced
in many other bacteria, including S. typhimurium, can be detected by
the V. harveyi reporter system even though it only responds to the borated
S-THMF form of AI-2.
It is important to note two clearly distinctive characteristics of LuxP and
LsrB. The former appears to be involved solely in sensing the extracellular
concentration of signal whereas the latter is involved in transport. This
is likely reflected in the nature of the ligandreceptor interactions, with
many more coordinating interactions in the case of LuxP. This tighter
interaction coupled with the complexing of S-THMF with borate in the
V. harveyi sensory pathway may be incompatible with the transport
function in the Salmonella system. In Salmonella the response to AI-2 in
the lsr system is mediated by the cytoplasmic protein LsrR (98); the nature
of AI-2 in its interaction with LsrR is not known but may be a phosphory-
lated derivative (98).
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
124
It has become clear that at least two derivatives of the LuxS product
DPD are sensed in different bacteria. The identification of chemically
distinct forms of AI-2 in S. typhimurium raises the question about
the chemical entity of the AI-2 in other bacteria. It is plausible that the
S-THMF-borate AI-2 signal molecule is common in marine bacteria whose
growth environment contains borate. However, whether the AI-2 in other
O
OH
OH
O
O
O
CH
3
HO
OH
O
O
OH
HO
CH
3
O
H
2
O
CH
3
HO
OH
OH
HO
O
OH
HO
CH
3
OH
HO
H
2
O
DPD
S -DHMF
S -THMF
R-DHMF
R-THMF
Figure 6.3. Two chemically distinctive AI-2 structures. Proposed solution structures of the
AI-2 signaling molecules recognized by Vibrio harveyi (S-THMF) and Salmonella
typhimurium (R-THMF). In V. harveyi, the S-THMF complexes with borate in the binding
pocket of the receptor LuxP. R-THMF interacts with the receptor LsrB in S. typhimurium.
DPD, 4,5-dihydroxy-2,3-pentanedione; S-DHMF and R-DHMF, stereoisomers of
2,4-dihydroxy-2-methyldihydrofuran-3-one; S-THMF and R-THMF, stereoisomers of
2-methyl-2,3,3,4-tetrahydroxyfuran. Adapted from Miller et al. (66).
125
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
bacteria is the same as that used by S. typhimurium remains to be deter-
mined. Nevertheless, it is evident that what has generically been termed
AI-2 is actually a mixture of molecules, more than one of which can be
used for interspecies signaling. In this chapter we use luxS/AI-2-mediated
quorum sensing to include the AI-2 in V. harveyi, that in S. typhimurium,
and other active DPD derivatives yet to be identified.
THE CONTROVERSY OF AI-2 AND CELL-TO-CELL SIGNALING
The identification of the biochemical pathway leading to AI-2 produc-
tion has highlighted controversy regarding the role of LuxS in bacterial
physiology. Based on the metabolic pathway (Figure 6.4) its role in the
removal of SAH from the cell places it clearly in primary metabolism,
whereas more traditional cell-to-cell signaling pathways such as AHLs are
more typical of secondary metabolites. The biochemical pathway has two
implications regarding possible physiological roles for LuxS. The first is in
the removal of SAH. The strong inhibitory effect of SAH on methyl trans-
ferase makes this an important step; pfs mutants show significantly
reduced growth rates in many bacteria, and second-site revertants that
restore growth rates are readily isolated. In contrast in many bacteria luxS
has no observable growth phenotypes. The second possible role is in the
production of homocysteine for methionine recycling. This is further
hinted at by the organization of operons with which luxS is associated in
some bacteria (Figure 6.5).
As the number of sequenced genomes increases, information from
comparative genomic studies can add to our hypothetic roles for genes. Just
as the organization with metK and pfs implicated luxS in the SAH degrada-
tion pathway, its association with metB and cysK in some bacteria further
supports a role for LuxS in primary metabolism (see Figure 6.4). However,
the lack of a luxS growth phenotype in Salmonella even in methionine-free
media (A. L. Beeston and M. G. Surette, unpublished data) makes the
significance of this pathway unclear. The importance of luxS in methionine
recycling through production of homocysteine from S-ribosylhomocysteine
remains to be determined in most if not all bacteria. It is likely that under
some growth conditions this pathway will be important and that pleiotropic
effects on metabolism in a luxS mutant may account for many of the
observed phenotypes in some bacteria. This is likely to be particularly true
in those examples where complementation by exogenous AI-2 does not
restore the wild-type phenotypes. It is not clear how addition of exogenous
AI-2 can complement the luxS mutant with respect to methionine recycling
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
126
since the DPD would not be involved in the salvage pathway for homo-
cysteine to methionine. One possible explanation may be that homocys-
teine is released into the media with AI-2 in conditioned media that is used
for complementation, but to our knowledge this has not been tested.
However, this seems unlikely, as it would be significantly diluted. It is
not surprising that many bacteria can utilize AI-2 as a sole carbon source
NH
2
CH
C
C
H
2
OH
O
H
2
C
S
H
3
C
N
N
N
N
NH
2
O
H OH
H
H
H H
O
NH
2
CH
C
CH
2
HO
O
H
2
C
+S H
3
C
O
OH
OH
O
N
N
N
N
NH
2
O
H OH
H
H
H
H
O
NH
2
CH
C
CH
2
HO
O
H
2
C
S
OH
O
H
H
H
H
H
O
NH
2
CH
C
CH
2
HO
O
H
2
C
S
O
OH
NH
2
HS
ATP PPi + P
adenine
adenosine
AI-2
Pfs
LuxS
SAH Hydrolase
MetK
Methionine
Synthase
(MetE / MetH)
(Methionine
Biosynthesis
MetA,MetB,MetC)
Met
SAM
SRH
DPD
HC
SAH
methyltransferases
Figure 6.4. The metabolic pathways involved in AI-2 production and methionine
metabolism. The pathway is based on Escherichia coli and Salmonella. Not all genes are
present in all bacteria. Methionine (Met) is converted to S-adenosylmethionine (SAM) by
the enzyme MetK. Numerous methyltransferases in the cell use SAM, generating
S-adenosylhomocysteine (SAH) as a by-product. Removal of SAHis important, because it is
a potent inhibitor of methyltransferases. The pathway of SAH degradation in eukaryotes
and some bacteria by SAH hydrolase to homocysteine and adenosine is shown in gray. In
other bacteria, SAH is removed by the action of Pfs to generate S-ribosylhomocysteine
(SRH) and adenine. SRH is converted to homocysteine (HC) and 4,5-dihydroxy-2,3-
pentanedione (DPD) by LuxS. DPD spontaneously cyclizes into AI-2 (see Figure 6.3).
Recycling of homocysteine into methionine is carried out by methionine synthase (MetE or
MetH). In the biosynthesis of methionine, homocysteine is generated from oxaloacetate by
MetA, MetB, and MetC. Adapted from Greene (33) and the Ecocyc database and Metacyc
database (49, 50).
127
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
(as has been observed for AHLs as well) as many bacteria are efficient at
utilizing a large variety of compounds for growth. However, most experi-
ments where exogenous AI-2 restores phenotypes are carried out in media in
which the exogenous AI-2 would contribute a very minor amount to the
available nutrients and it seems unlikely that the observed effects are
mediated through altered metabolism. When examining luxS/AI-2-
mediated quorum sensing in a bacterium, one of the key results to look for
is complementation of the luxS mutant by exogenous AI-2 or culture
supernatant.
AI-2 AS A CLASSICAL QUORUM SENSING SIGNAL
The role of AI-2 as a bona fide cell-to-cell signaling molecule is well
established among the Vibrio spp.; this is not surprising, because it was first
characterized for its role in cell-density-dependent light production in V.
harveyi as described above. The V. harveyi bioassay remains the primary
method of detecting AI-2 activity in culture supernatants. The role of AI-2
in signaling is now well established in several Vibrio species in addition to
V. harveyi. The general architecture of the signaling network seems to be
conserved, with two or three signaling systems apparently acting in each
species.
The sequencing of the V. cholerae genome (36) revealed the presence
of an intact AI-2 signaling pathway homologous to that in V. harveyi and
that the regulators in this pathway play an important role in regulating
virulence gene expression (110). The realization that the V. harveyi and
V. cholerae systems were very similar suggested that the V. harveyi
luxCDABE operon could be used as a heterologous reporter of quorum
sensing in V. cholerae; this was indeed the case (65). By using this strategy it
metB cysK luxS
luxS
metK pfs
luxS
A
B
C
Figure 6.5. Gene location and operon context of luxS. (a) In most bacterial strains, luxS
appears to be in a single-gene operon. (b) In at least two bacterial strains luxS is found in an
operon with pfs and metK. (c) In at least three bacterial strains luxS is found in an operon
with metB and cysK.
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
128
was revealed that, in addition to a luxS/AI-2 pathway and convergent AHL
pathway, named CqsA/CqsS as in V. harveyi, a third cell-to-cell signaling
pathway was also present (65). The genes involved in this third pathway
have yet to be identified. In addition to the regulation of virulence genes,
cell-to-cell signaling has been shown to control biofilm formation in
V. cholerae (34).
V. fischeri, the strain in which the LuxI/R homoserine lactone system
and cell-to-cell signaling was initially described, has two AHL systems and
luxS. These signals regulate light production and symbiotic colonization of
the bobtail squid. The predominant signal is produced by LuxI; however,
both AinS (the second AHL pathway) and LuxS contribute to the timing and
magnitude of the response. Colonization of the host and light production is
dependent on the LuxIR pathway and a modest reduction is observed in an
ainS mutant. A luxS mutant had minimal effect on colonization of the host;
however, an ainSluxS double mutant had a synergistic decrease in colo-
nization and light production (57).
In V. vulnificus the expression of extracellular virulence determinants is
under control of the two systems SmcR (an AHL system) and LuxS.
Mutations in either system reduce the production of extracellular cytotoxic
factors (51, 81). Spent media could complement the luxS defects. The luxS
mutation also resulted in attenuation of lethality in a mouse infection
model (51).
AI-2 AS A QUORUM SENSING SIGNAL IN BACTERIA OTHER
THAN VIBRIO SPP.
In E. coli, AI-2 signaling was initially reported together with that in
V. harveyi and S. typhimurium (97). A role for LuxS/AI-2 in pathogenic
E. coli has been well demonstrated. Both enterohemorrhagic E. coli (EHEC)
and enteropathogenic E. coli (EPEC) produce a lesion on epithelial cells called
the attaching and effacing (AE) lesion; the genes responsible for the lesion
are located in the locus of enterocyte effacement (LEE). Kaper and colleagues
have reported the involvement of AI-2 signals in the expression of type III
secretion genes and virulence genes in EPEC and EHEC (31, 82, 86, 87; see
also Chapter 7). Conditioned media were shown to increase the transcription
of the LEE1 and LEE2 operons from EPEC and of the LEE1 operon from
EHEC 23-fold. In EPEC, AI-2-mediated quorum sensing regulates the
expression of a number of virulence-associated genes (82, 90) through
luxS-dependent regulation of per, an early transcriptional regulator of the
LEE virulence genes.
129
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
The spectrum of luxS-mediated gene regulation in E. coli was further
characterized by using genome-scale examinations (18, 87). From DNA
microarrays, Sperandio et al. (87) reported that approximately 10% of the
array genes of E. coli O157:H7 were regulated by luxS at least 5-fold.
Widespread effects on the regulation of flagella and motility genes by
AI-2 signaling were observed. A total of 32 genes including the master
regulator FlhDC, genes encoding flagella biosynthetic genes, motility pro-
teins, and chemotaxis proteins were all downregulated significantly in the
luxS mutant compared with the wild-type strain (87). These results suggest
that AI-2 signaling may have diverse roles in cell division, growth, SOS
response, shiga toxin production, and motility and chemotaxis (87).
However, a further report from Kaper and colleagues has demon-
strated that the signal-dependent responses on LEE and flagellar gene
expression are not, as previously reported, due directly to AI-2 (89).
Another uncharacterized signal, termed AI-3, regulates the expression of
the type III secretion genes and the two-component system qseBCA, which
in turn regulates the flagella regulon (90) through flhDC, the master
regulator for the flagella and motility genes. Interestingly, AI-3 signal
production is also dependent on luxS (89; see also Chapter 7). It remains
to be seen what chemical entity AI-3 is and how it is synthesized via LuxS.
Another microarray study by DeLisa et al. (18) using E. coli strain W3110
indicated that 242 genes, or approximately 5.6% of the E. coli genome, are
regulated by AI-2. This report also shows AI-2 regulation of genes whose
products are involved in diverse cellular processes including virulence,
biofilm formation, motility, surface and membrane-associated functions,
and small molecule metabolism, as well as a number of genes of unknown
function. Although these findings also show AI-2 regulation of genes
involved in cell division, only one gene ( ftsE) was reported by both
Sperandio et al. (87) and DeLisa et al. (18); the remaining genes on both
lists are mutually exclusive. Changes in transcription ranged from 2.5-fold
to 33-fold; the majority of activated or repressed genes showed less than a
5-fold difference.
In Serratia marcescens, luxS mutants are defective in antibiotic and
virulence factor production (81); importantly, the defects can be comple-
mented by exogenous AI-2 by using conditioned culture media. The
mutant shows reduced virulence in a Caenorhabditis elegans infection
model (81). Similarly, Coulthurst et al. (14) reported the detection of AI-2
in S. marcescens ATCC 274 and Serratia ATCC 39006 and demonstrated a
role for AI-2 in the regulation of carbapenem antibiotic production in
Serratia ATCC 39006 and of prodigiosin and secreted haemolysin
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
130
production in S. marcescens ATCC 274. The phenotypic defect of luxS
mutants in the strains was complemented by the addition of the culture
supernatant of a wild-type strain. The complementation is effective even
if the supernatant is substantially diluted. It is clear that luxS-mediated
signaling operates in these bacteria as a classic quorum sensing system.
A low amount of signal molecule regulates gene expression, and the
regulation defects in signal synthesis mutants can be complemented by
addition of exogenous signal molecule.
In Shigella flexneri the luxS mutation decreases in the expression of
virB, a transcriptional regulator of invasion gene loci; however, invasive-
ness and dissemination of Shigella to adjacent cells remained unaffected by
the lack of AI-2 (15). The addition of AI-2 upregulated the expression of virB
in this bacterium.
AI-2-mediated quorum sensing has also been reported in many Gram-
positive bacteria. In Clostridium perfringens, extracellular toxin production
is reduced in a luxS mutant; this defect can be complemented by addition
of culture media from a wild-type strain or from DH5a expressing
C. perfringens luxS. The levels of all three toxins are affected; however,
only one of these (pfoA) appears to be affected at the level of transcription
(70). Interestingly, luxS appears to be in an operon with metB and cysK in
C. perfringens. These enzymes are involved in the synthesis of methionine
and cysteine, respectively. Importantly, the expression of the pfoA is not
affected by mutations in either metB or cysK, indicating that luxS and
presumably AI-2 are responsible for increased expression of toxin (70).
The luxS gene in Actinobacillus actinomycetemcomitans appears to play a
role in aerobic growth in iron limiting conditions (see Chapter 8). Reduced
expression of iron scavenging, transport and storage proteins was observed
in a luxS mutant; a number of phenotypes in a luxS mutant are restored by
exogenous AI-2 (27). In a search for homolog of the V. harveyi AI-2 sensory
pathway, ArcB was identified as the closest LuxQ homolog (27). The arcB
mutant was phenotypically similar to the luxS mutant.
Within the genus Streptococcus, several species are important human
pathogens, and many have been shown to have a functional luxS/AI-2-
mediated quorumsensing system. In Streptococcus gordonii DL1, a common
human oral bacterium, inactivation of luxS lowered the expression of a
number of genes (59) although growth of the bacterium was unaffected
under several tested conditions (8). These genes include gtfG, encoding
glucosyltransferase; fruA, encoding extracellular exo-b-D-fructosidase;
and lacD, encoding tagatose 1,6-diphosphate aldolase (59). The luxS
mutant also forms altered microcolony architecture within S. gordonii
131
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
biofilms (8) and is unable to form normal mixed-species biofilms with
S. gingivalis (59) (see below). S. mutans is another important component
in the human oral biofilm community and a primary pathogen of human
dental caries. Similar to S. gordonii, inactivation of luxS gene in this
bacterium does not result in any apparent growth defect in liquid culture
(60). No obvious difference in growth rate, nutrient requirements, acid
production, or colony morphology could be observed between the wild type
and the mutant. Although earlier observations by Wen and Burne indicated
that in S. mutans strain UA159 the luxS mutation had no effect on its ability
to form biofilms (100), marked alterations in the biofilm structure and its
resistance to adverse agents were seen in the luxS mutant of S. mutans 25175
strain as a result of loss of AI-2 production (60). This apparent discrepancy
was found to be due to the methods and growth medium of biofilm growth
and evaluation. In a later study by Wen and Burne (101), inactivation of
luxS resulted in a downregulation of fructanase, a demonstrated virulence
determinant, by more than 50%, and the expression of other genes including
recA, smnA encoding AP endonuclease, and nth-encoding endonuclease were
downregulated in the luxS mutant, especially in early-exponential-phase cells.
The brpA encoding a biofilmregulatory protein was also downregulated in the
mutant. On a different supporting matrix, i.e. hydroxyapatite disks, the ability
of the bacteriumto formbiofilms was shown to be greatly decreased, especially
when biofilm was grown in medium with sucrose; the structure of such
biofilms was also different from that of the wild type. Importantly, the luxS
mutant phenotypes were able to be complemented by supernatants fromwild-
type cultures (101).
BACTERIA WITH AMBIGUOUS ROLES FOR AI-2
In each of the species and experiments described above, luxS/AI-2
appears to play an important role in the organism. LuxS defects can be
complemented by exogenous AI-2 (or at least by conditioned media from
wild-type but not luxS mutant strains), strongly suggesting a role for AI-2 in
cell-to-cell communication. However, there are many systems in which no
such role is apparent or the existing experimental evidence is ambiguous.
The gastric pathogen Helicobacter pylori possesses a luxS ortholog and
produces AI-2 (28, 45). An earlier search for AI-2-regulated genes in
this bacterium failed to detect any, either by two-dimensional protein
profiling comparing the wild type and luxS mutant, or by a test of the
known virulence genes (45). However, it has been demonstrated that the
growth-phase-dependent expression of flaA, the major flagellin gene, is
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
132
also luxS-dependent (56). Inactivation of luxS abolished the induction of
flaA expression in higher cell density, but conditioned media from a wild-
type culture resulted in only a slight (though statistically significant)
increase in flaA expression (56). Biofilm formation in a luxS mutant is
about two times greater than in a wild-type strain (13), but the effect
of conditioned medium was not tested in this study. These results are
consistent with AI-2 as a signal as well as with luxS playing an important
role in metabolism through homocysteine pathways.
Interestingly, H. pylori is one of the two bacteria where luxS is located
downstream of metB or cysK genes and likely forms an operon with these
genes, as discussed previously. The H. pylori strain J99 has the gene order
of cysKmetBluxS and the strain 26695 has metBcysKluxS orientation
(GenBank). The operon structure of luxS, with genes involved in amino
acid synthesis pathways, seems also to be in agreement with the dual role of
luxS in this bacterium.
Other ambiguous examples of AI-2 signaling are found in Photorhabdus
luminescens and Campylobacter jejuni. In the first case, the production of
the antibiotic carbapenem has been reported as being repressed by luxS/AI-2,
but no complementation by exogenous AI-2 was done (20). In C. jejuni,
the expression of flaA was reduced by about half in a luxS mutant, but
no difference at protein level was observed (42). The motility and autoagglu-
tination also experienced small reductions, but, considering the lack of
exogenous complementation experiments and the magnitude of the altera-
tion, the signaling role of luxS/AI-2 is yet to be established in this bacterium.
Borrelia burgdorferi is the causative agent of Lyme disease. The spiro-
chete contains a luxS gene that could complement an E. coli mutant (91).
The luxS system appears to control the expression of a number of bacterial
proteins of which some are involved in the infection process; the addition of
exogenous AI-2 altered the expression of these genes (63, 91, 92). However, it
has recently been reported that luxS is not required for tick colonization,
transmission to a mammalian host, or the induction of disease (9, 40),
strongly suggesting that AI-2 signaling plays no role in disease processes
or bacterial virulence in this species. When infections of a susceptible host
were initiated by intraperitoneal injection rather than by transmission by a
tick, similar results were obtained (i.e. no effect on virulence of the luxS
mutant) (40). The luxS mutants can colonize and cause disease efficiently,
but whether they are at a competitive disadvantage compared with wild-type
strains was not addressed. Although competition experiments might
have been useful to clarify the infectiveness of luxS mutants, it should be
noted that even if the wild-type strain were more competitive than the
133
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
mutant this would not necessarily indicate that the mechanism was through
cell-to-cell signaling.
Similar to H. pylori, the luxS gene in B. burgdorferi forms an operon
with other genes in the genome (40), but in this case with metK, encoding
the SAM synthase and pfs. As the genes in this pfsmetKluxS operon
encoding the enzymes catalyze continuous reactions in one metabolic
pathway, it is likely that they all have metabolic functions in this bacterium.
Consistent with this possibility is the fact that no AI-2 activity could be
detected in the culture supernatant or the concentrated cell lysate by using
the V. harveyi reporter system (9, 40) even though the luxS gene has been
shown to be active and functional (91). It is important to note, however, that
neither metE or metH homolog are predicted in the B. burgdorferi genome.
Considering this organisms lack of biosynthetic pathways (29) and of
apparent methionine salvage enzymes, it seems unlikely that the role of
LuxS would be in methionine recycling.
In Neisseria menigitidis, an important Gram-negative bacterium caus-
ing septicemia and meningitis, it has been shown that luxS/AI-2 plays a
role in virulence (104). A luxS mutant of Serogroup B of N. meningitidis had
attenuated virulence in an infant rat model of bacteremia (104); however, it
was later demonstrated that AI-2 appears to have no effect on concerted
regulation of gene expression, by using microarray analysis (22).
Many of these studies are complicated by the lack of extracellular
complementation by AI-2; this clearly is at odds with the conventional
view of cell-to-cell signaling. These observations argue against a role for
AI-2 in intraspecies signaling, at least in these bacteria. There always
remains the caveat that the appropriate conditions have not been examined
but the accumulating evidence would suggest that for many bacteria AI-2
may not play a role as a typical quorum-sensing signal.
AI-2 AS AN INTERSPECIES QUORUM SENSING SIGNAL
An important paper indicating that AI-2 activity might be more general
than the better characterized species/strain-specific AHL signaling mole-
cules came from Bassler et al. in 1997 (4). Using the V. harveyi reporter
strains specific for V. harveyi AI-1 (butyryl homoserine lactone) and AI-2,
they observed that, whereas system 1 of V. harveyi responded only to culture
supernatants from V. harveyi, system 2 responded to culture supernatants
from a number of different bacterial species, primarily Vibrio species but
also Yersinia enterocolytica. Notable with hindsight was the absence of signal
from a number of bacterial species, including E. coli and S. typhimurium.
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
134
This was because the conditions for AI-2 accumulation in culture media
were not used in the original study; because AI-2 is actively removed by the
bacteria as the culture approaches stationary phase, no or very little detect-
able activity remains in overnight culture media. The observation that the
AI-2 signal was more widely produced than AI-1 led to the hypothesis that
perhaps AI-2 was a general signaling mechanism indicating bacterial cell
density among different species whereas AI-1 was indicative of the species
that produces it itself (3, 4, 25, 79).
Although AHL-mediated quorum sensing is generally species-specific,
crosstalk among closely related species through AHL pathways has been
demonstrated for a number of bacteria (61, 76, 83); the use of reporter
strains for identification of signaling molecules (AHL and AI-2) implies the
ability of signals to cross species boundaries. In the case of AI-2 signaling
the V. harveyi reporter strains remain the primary assay used to detect
signal production. Most strains containing a luxS gene have been
shown to produce a signal under some growth condition that activates the
AI-2-specific V. harveyi reporters. As described above, however, the nature
of the AI-2 signal is not straightforward; in V. harveyi the active molecule
is a furanyl borate diester (11) whereas in S. typhimurium the active signal is
R-THMF (66). The chemical equilibriumof products will be affected by the
chemical environment; different forms of the molecule may predominate
under particular assay conditions. Nonetheless, the use of different deriva-
tives of the LuxS product as AI-2 molecules in different species does not
disagree with the suggestion that AI-2 may be a generic bacterial signaling
molecule shared by a wide variety of species.
The implication of AI-2 as an interspecies signal is that AI-2 may
contribute to the establishment and behavior of mixed microbial com-
munities. The hypothesized role of AI-2 in interspecies signaling has
been borne out in a number of studies, most significantly in the mixed
microbial communities forming dental plaque and also as part of the
interspecies interactions between Pseudomonas aeruginosa and oropharyn-
geal flora in cystic fibrosis. There is strong evidence in support of this from
studies of these two polymicrobial communities.
Cell-to-cell signaling and AI-2 in dental plaque
Dental plaque is a complex biofilm community that comprises more
than 500 bacterial species. The bacteria normally exist in harmony with the
host when the dynamic yet stable microbial community is maintained.
However, overgrowth by some microorganisms, especially a group of
135
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
Gram-negative anaerobic species, e.g. Porphyromonas gingivalis, causes the
initiation and progression of periodontitis, an oral disease that is among the
most prevalent human diseases (85). The biofilm community has a well-
differentiated, organized structure formed in a spatiotemporal order by
cooperative actions of multispecies bacteria. The biofilm formation on the
surface of teeth is initiated by the early colonizer bacteria, among which
60% are Streptococcus spp., and develops through coaggregation and coad-
herence by late-colonizing bacteria via interactions with the early coloni-
zers. It is suggested that the coaggregation and coadherence are essential to
communication between species and that the communication between
bacterial species are likewise integral to this process and help to establish
patterns of spatiotemporal development. An AHL quorum-sensing signal,
however, has not been detected and is therefore probably not produced, at
least not in a significant quantity, in this community. In contrast, some of
the key components in this community have been shown to produce AI-2,
including S. mutans, S. gordonii, Actinobacillus actinomycetemcomitans,
Porphyromonas gingivalis, Fusobacterium nucleatum, and Prevotella intermedia.
It is clear that LuxS-mediated interspecies quorum sensing plays important
roles in this complex community.
In S. mutans, which is among the early colonizers of dental plaque
biofilm, a luxS mutant forms a biofilm remarkably different from that
formed by the wild type. Instead of being smooth and confluent, the biofilm
formed by the luxS mutant is rather rough and granular. Furthermore, the
biofilm formed by the luxS mutant has increased resistance to detergent
and antibiotics (60). Inactivation of the luxS gene in Porphyromonas gingivalis
and A. actinomycetemcomitans resulted in changes in the expression of genes
involved in hemin acquisition and LtxA leukotoxin production, respectively. In
addition, the virulence of P. gingivalis was attenuated in luxS mutants.
Importantly, conditioned medium from E. coli DH5a expressing the luxS
gene of A. actinomycetemcomitans was able to complement the defects in gene
expression in the P. gingivalis luxS mutant, indicating that LuxS-dependent
signaling potentially mediates interspecies communication in mixed-species
biofilms. Experiments with S. gordonii and P. gingivalis, mixed-species biofilms
showed that LuxS-dependent intercellular communication between P. gingivalis
and S. gordonii is essential for biofilm formation (59). S. gordonii is one other
member of the early colonizers. A substratum of S. gordonii attached to solid
surfaces supports the adhesion of P. gingivalis, which is initiated via the inter-
actions between S. gordonii cell-surface proteins and the fimbriae of P. gingivalis.
The adhered P. gingivalis then progresses to form biofilm microcolonies dis-
tinctive from the simple adhered microaggregates of P. gingivalis itself.
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
136
Inactivation of the luxS gene in S. gordonii altered the expression of a number of
genes, among which is a group of genes involved in carbohydrate metabolismin
this organism. More importantly, this luxS mutant was unable to form normal
mixed-species biofilms with LuxS-deficient P. gingivalis. However, if one of the
partner strains possessed a functional luxS gene the mixed-species biofilms
microcolonies were formed normally, indicating that the bacteria could respond
to a heterologous AI-2 signal. These observations strongly support a role for
AI-2-mediated interspecies quorumsensing in mixed-species biofilmformation
in dental plaque.
The CF community
The other example of AI-2 playing a role in microbial communities is
the mixed-species signaling betweenPseudomonas and the oropharyngeal flora
(OF) in cystic fibrosis (CF) lungs. In patients with CF, altered ion transport
gives rise to a more viscous pulmonary mucous layer, thereby impairing
ciliary clearance and permitting the existence of a biofilm-forming microbial
community in the lungs (99). Among them, the leading pathogenic agent is
P. aeruginosa, whose infection ultimately causes pulmonary failure, resulting
in premature mortality (93). In addition to P. aeruginosa, these patients
respiratory tracts commonly experience infections by other pathogens such
as S. aureus, Haemophilus influenzae and Burkholderia cepacia, as well as a
variety of other avirulent microorganisms (12, 23) including OF bacteria such
as viridans-group streptococci and coagulase-negative staphylococci, which
only colonize the upper respiratory tracts in healthy adults. Using an in vitro
system it has been shown that in this microbial community multifactorial
interactions occur among the microorganisms, which significantly influence
the pathogenesis of P. aeruginosa (23). One of the factors was determined to
be AI-2 produced by non-pseudomonad strains. Although P. aeruginosa does
not possess a luxS gene in its genome and does not produce AI-2, a large
amount of AI-2 was readily detectable in all the sputum samples of the CF
patients tested and in the culture supernatants of most of the OF isolates
using the V. harveyi assay. Transcriptional profiling of a set of defined P.
aeruginosa virulence factor promoters indicated that OF and exogenous AI-2
could upregulate overlapping subsets of these genes. Six out of the nine genes
that were affected by the presence of OF strains were also regulated by AI-2
added to the media. It seems that P. aeruginosa is modulating its behavior by
monitoring the environmental conditions and by eavesdropping on the
other bacteria via AI-2 and probably other signals. P. aeruginosa may have
evolved to use AI-2 as a signal indicative of a polymicrobial environment in
an animal host. The phenomenon that AI-2 from one microorganism affects
137
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
the gene expression of the other indicates that the luxS-mediated quorum
sensing system does indeed play a role in the dynamics of this microbial
community.
It has been observed that P. aeruginosa could remove AI-2 in the
medium (103). This is not surprising: degradation of AI-2 has been
observed in a number of bacteria that respond to the signal, including
P. aeruginosa (39, 77). It is not known, however, whether this represents a
strategy that P. aeruginosa employs to counter other bacteria or whether it is
just one aspect of how P. aeruginosa responds to exogenous cell-to-cell
signals. Because P. aeruginosa could be modulated by AI-2, it is possible
that it has a mechanism to control the signal. Indeed, in the classical
example of V. harveyi the molecule decreases rapidly after it reaches its
peak amount. It is normal for a regulatory system to have a component to
degrade and turn over its signal (107). In a microbial community it is
conceivable that a molecule being used by one species as a signal molecule
could be attacked or degraded by another species. Some of the bacteria may
have evolved to sense the signals produced by others; some may have
evolved to produce the signals just to affect the behavior of others.
Because of AI-2s important role in gene regulation and its relatively
small amount, this is probably more an aspect of cell-to-cell interaction
than of syntrophic interactions.
AI-2 AS A REPORTER OF METABOLIC STATUS
Because the LuxS protein is not only responsible for the synthesis of
AI-2 precursor but may also function to recycle S-adenosylhomocysteine, a
question has been raised about its role as a signal molecule (103). However,
for the same reasons, it has been suggested as being a signal in some
bacteria which measures the metabolic status of the cell instead of
just cell density (7, 105). Although the role of AI-2 as a cell-to-cell communi-
cation signal is clear in many bacteria, as discussed above, there are
increasing data that suggest its role as a signal reflecting the metabolic
status of the cell instead of simply cell number.
In the classical definition of quorum sensing, autoinducer production
increases with the increase in cell numbers, and the signals accumulate in
the culture environment. AI-2 accumulation in the extracellular environ-
ment is often very growth-phase- and media-dependent. In the original
V. harveyi studies, AI-2 accumulates throughout growth, reaching maximum
accumulation near the end of growth, presumably because there is little or no
turnover of the signal. In the original study demonstrating AI-2 activity in a
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
138
wide variety of Vibrio spp. (4), no activity was detected in E. coli or Salmonella.
It was subsequently shown that both these bacteria produce AI-2 but in a
growth-phase-dependent manner (96).
AI-2 production peaks before stationary phase and decreases after the
peak in a number of bacteria. We have observed significant effects of
growth conditions on signal production and degradation in the luxS system
for numerous bacteria (95, 96). Expression data also show that neither luxS
nor pfs expression is regulated by AI-2, suggesting that AI-2 production is
not regulated at the level of luxS expression. This result means the AI-2
production differs from classical autoinducers in the sense that it is not
autoinduced. It should be noted that autoinduction is a common but not
universal feature of cell-to-cell signaling systems. The amount of AI-2
produced is instead regulated at the level of LuxS substrate availability,
which reflects the metabolic status of the cells but not the cell number.
These results indicate that AI-2-dependent signaling is a reflection of the
metabolic state of the cell and not of cell density (7). The concentration of
AI-2 accumulating in the extracellular environment is also a function of its
degradation or active removal by bacteria; in Salmonella, production and
removal via the Lsr system seem to be oppositely regulated (A. L. Beeston
and M. G. Surette, unpublished results).
As discussed above, quorum sensing is prevalent and important in
pathogenic E. coli. In non-pathogenic, laboratory-domesticated E. coli, it has
been demonstrated that the AI-2 signaling pathway communicates the
stress or burden of overexpressing heterologous genes in the bacterium
(16). The activity of AI-2 decreases significantly following induction of
several plasmid-encoded genes in the bacterium at both low- and high-
cell-density culture conditions. The AI-2 signaling level was linearly related
to the accumulation level of each protein product. In this case the metabolic
status that AI-2 signals is reflected as the stress or burden of gene
overexpression.
The other example of a metabolic status reporting role for AI-2 comes
from the findings of the linkage between luxS and relA in S. mutans (54). In
a study investigating the role of the relA gene, which codes for a guanosine
tetraphosphate and guanosine pentaphosphate [(p)ppGpp] synthetase/
hydrolase, in biofilm formation and acid tolerance, Lemos et al. found
that a relA mutant showed significant reductions in biofilm formation
after the induction of a stringent response and altered acid resistance in
biofilms. Interestingly, the expression of the luxS gene was increased as
much as 5-fold in the relA mutants, suggesting a link between AI-2 quorum
sensing and the stringent response. Because of the role of the stringent
139
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
response in sensing cellular amino acid shortage, etc., this connection
strongly suggests a role for luxS-mediated QS in monitoring cell metabolic
status in this bacterium.
In S. pneumoniae, the luxS mutant shows a significantly decreased ability
to persist in a murine model of nasopharyngeal carriage (46). The mutation
also affects at least five operons that are involved in fatty-acid biosynthesis
and virulence-factor production. However, luxS-mediated regulation does not
follow the typical quorum-sensing paradigm: it does not occur at high cell
density (46). LuxS activity in this bacterium instead modulates the fitness of
the organism in a discrete host niche and LuxS functions in a mechanism
independent of cell density. However, as noted above, it is important to
distinguish between cell-to-cell signaling and cell density.
Although quorum sensing and cell-to-cell signaling in bacteria are
normally interchangeable terms and cell density dependence is one of the
hallmarks of this process, it has become clear that cell-to-cell signaling
is not simply cell-density-dependent. A more general definition of quorum
sensing as a gene regulation coordinated or mediated by secreted small
signal molecules which function at low concentrations (i.e. at concen-
trations lower than that which would mediate physiological changes
through the metabolism of the substrate) seems appropriate. The responding
bacteria do so when the autoinducer molecule reaches a threshold concen-
tration. In many experimental systems this occurs late in the growth of liquid
cultures as the cells reach high cell density. However, this concentration of
signal can also be reached at lower cell densities. It is now clear that in many
cell-to-cell signaling systems (including AHLs and oligopeptides) both the
production and the degradation of the signaling molecule can be tightly
regulated and can be influenced by growth conditions. Moreover, in most
laboratory studies it is routine practice to vigorously shake liquid cultures,
because this results in a homogeneous environment for all the cells. This
does remove any local concentration gradients and in effect dilutes signaling
molecules. The dynamics of response as well as the growth phase can be
significantly altered by simply not shaking the cultures! The concept of cell
density should be used cautiously in the context of cell-to-cell signaling. In
homogenous growth conditions such as in shaking liquid culture, all cells
respond to essentially the same conditions. When the environment is not
homogenized, different cells within the population are exposed to very
different environments.
Cell-to-cell signaling is often interpreted under the paradigm established
for V. fischeri, where cell density in a culture correlates with signal accumulation
and response. The culture tube offers a reasonable approximation of V. fischeri
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
140
life as a symbiont in the light organ of the squid with respect to quorum
sensing. However, this should not be extrapolated to all circumstances. The
importance of growth conditions for production and response is often under-
estimated. We have observed significant effects of growth conditions on
signal production and degradation in the luxS system in numerous bacteria
(95, 96) and in the production of acyl homoserine lactone signals in
P. aeruginosa (K. Duan and M. G. Surette, unpublished results). Localized
signaling and coordinated behaviour in small groups of cells in micro-
colonies, aggregates, microniches within colonies, biofilms, and consortia
of multiple species are likely to be more natural situations for cell-to-cell
signaling for most bacteria. Coordinated behaviour can also arise from
bacterial manipulation of the environment through the utilization of nutri-
ents and the release of toxic metabolites, enzymes (e.g. proteases), and
extracellular matrices (e.g. capsular material). This can give the appearance
of coordinated behaviour. Bacterial chemotaxis in soft agar provides an
excellent example of this. There is coordination of cells within the popula-
tion at the macroscopic level (well-defined chemotactic rings) arising from
the manipulation of the environment and a common response (i.e. chemo-
taxis up a gradient of nutrients). However, each cell within the population
behaves independently. Similar microenvironments will arise under any
condition where the cells are not in homogeneous environments (i.e. almost
any condition other than well-mixed liquid cultures). This is one mechanism
that can give rise to spatial and temporal patterns of gene expression in
microniches within bacterial colonies, swarming bacteria, and biofilms.
In the natural world, where bacteria exist primarily in mixed populations
and often in well-defined consortia of species, exchange of metabolites
and scavenging of other cells waste products may be a central feature of
their natural history. Cell-to-cell signaling is distinct from that behavior
described above in that it is mediated by signaling molecules that act at low
concentrations and not directly as metabolites. This is an important distinc-
tion and the two are not always readily distinguished. This is particularly an
issue when using conditioned media (even more so when using complex
media) and an obvious caveat to such experiments that can complicate
interpretations.
CONCLUSIONS
For many organisms, in particular the Vibrio spp., the role of AI-2 in
typical cell-to-cell signaling seems indisputable. For others, there seem to
be equally compelling data to suggest that luxS functions as part of primary
141
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
metabolic pathways and that AI-2 production is simply a by-product
of metabolism. We should be careful not to be constrained by current
paradigms trying to fit AI-2 into one domain or another: biology is full
of exceptions to general rules. LuxS/AI-2 is perhaps a little more blatant
than others in its disregard for established paradigms. Although we can
consider AI-2 as a typical intraspecies signaling molecule for many
bacteria, there is evidence (or more often the lack of evidence in support
thereof ) that argues against that role in many other bacteria. Observations
of AI-2 playing a role in interspecies communication in polymicrobial
communities are increasing. The current data suggest that AI-2 will
be found to be one of many contributing interactions that play a role
in community structure and dynamics. Understanding polymicrobial
communities will be a rich source of research in cellcell communication;
the evidence to date suggests that interspecies signaling via AI-2 will be a
player in many of these systems.
REFERENCES
1 Alfaro, J. F., T. Zhang, D. P. Wynn, E. L. Karschner and Z. S. Zhou 2004.
Synthesis of LuxS inhibitors targeting bacterial cell-cell communication. Org.
Lett. 6: 30436.
2 Bassler, B. L. 1999. Howbacteria talk to each other: regulation of gene expression
by quorum sensing. Curr. Opin. Microbiol. 2: 5827.
3 Bassler, B. L. 2002. Small talk. Cell-to-cell communicationinbacteria. Cell 109: 4214.
4 Bassler, B. L., E. P. Greenberg and A. M. Stevens 1997. Cross-species induction of
luminescence in the quorum-sensing bacterium Vibrio harveyi. J. Bacteriol. 179:
40435.
5 Bassler, B. L., M. Wright, R. E. Showalter and M. R. Silverman 1993. Intercellular
signaling in Vibrio harveyi: sequence and function of genes regulating expression
of luminescence. Molec. Microbiol. 9: 77386.
6 Bassler, B. L., M. Wright and M. R. Silverman 1994. Multiple signaling systems
controlling expression of luminescence in Vibrio harveyi: sequence and function
of genes encoding a second sensory pathway. Molec. Microbiol. 13: 27386.
7 Beeston, A. L. and M. G. Surette 2002. pfs-Dependent regulation of autoinducer
2 production in Salmonella enterica serovar Typhimurium. J. Bacteriol. 184: 34506.
8 Blehert, D. S., R. J. Palmer, Jr., J. B. Xavier, J. S. Almeida andP. E. Kolenbrander 2003.
Autoinducer 2 production by Streptococcus gordonii DL1 and the biofilm phenotype
of a luxS mutant are influenced by nutritional conditions. J. Bacteriol. 185: 485160.
9 Blevins, J. S., A. T. Revel, M. J. Caimano et al. 2004. The luxS gene is not required
for Borrelia burgdorferi tick colonization, transmission to a mammalian host, or
induction of disease. Infect. Immun. 72: 48647.
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
142
10 Cao, J. G. and E. A. Meighen 1989. Purification and structural identification of an
autoinducer for the luminescence system of Vibrio harveyi. J. Biol. Chem. 264:
216706.
11 Chen, X., S. Schauder, N. Potier et al. 2002. Structural identification of a bacterial
quorum-sensing signal containing boron. Nature 415: 5459.
12 Coenye, T., J. Goris, T. Spilker, P. Vandamme and J. J. LiPuma 2002.
Characterization of unusual bacteria isolated from respiratory secretions of cystic
fibrosis patients and description of Inquilinus limosus gen. nov., sp. nov. J. Clin.
13 Cole, S. P., J. Harwood, R. Lee, R. She and D. G. Guiney 2004. Characterization
of monospecies biofilm formation by Heliobacter pylori. J. Bacteriol. 186: 312432.
14 Coulthurst, S. J., C. L. Kurz and G. P. Salmond 2004. luxS mutants of Serratia
defective in autoinducer-2-dependent quorum sensing show strain-dependent
impacts on virulence and production of carbapenem and prodigiosin.
15 Day, W. A., Jr. and A. T. Maurelli 2001. Shigella flexneri LuxS quorum-sensing
system modulates virB expression but is not essential for virulence. Infect.
Immun. 69: 1523.
16 DeLisa, M. P., J. J. Valdes and W. E. Bentley 2001. Quorum signaling via AI-2
communicates the Metabolic Burden associated with heterologous protein pro-
duction in Escherichia coli. Biotechnol. Bioeng. 75: 43950.
18 DeLisa, M. P., C. F. Wu, L. Wang, J. J. Valdes and W. E. Bentley 2001. DNA
microarray-based identification of genes controlled by autoinducer 2-stimulated
quorum sensing in Escherichia coli . J. Bacteriol. 183: 523947.
19 Della Ragione, F., M. Porcelli, M. Carteni-Farina, V. Zappia and A. E. Pegg 1985.
Escherichia coli S-adenosylhomocysteine/5-methylthioadenosine nucleosidase.
Purification, substrate specificity and mechanism of action. Biochem. J. 232:
33541.
20 Derzelle, S., E. Duchaud, F. Kunst, A. Danchin and P. Bertin 2002. Identification,
characterization, and regulation of a cluster of genes involved in carbapenem
biosynthesis in Photorhabdus luminescens. Appl. Environ. Microbiol. 68: 37809.
21 Donabedian, H. 2003. Quorum sensing and its relevance to infectious diseases.
J. Infect. 46: 20714.
22 Dove, J. E., K. Yasukawa, C. R. Tinsley and X. Nassif 2003. Production of the
signaling molecule, autoinducer-2, by Neisseria meningitidis: lack of evidence for a
concerted transcriptional response. Microbiology 149: 185969.
23 Duan, K., C. Dammel, J. Stein, H. Rabin and M. G. Surette 2003. Modulation of
Pseudomonas aeruginosa gene expression by host microflora through interspecies
communication. Molec. Microbiol. 50: 147791.
24 Duerre, J. A. and C. H. Miller 1966. Cleavage of S-ribosyl-L-homocysteine by
extracts from Escherichia coli. J. Bacteriol. 91: 121017.
25 Federle, M. J. and B. L. Bassler 2003. Interspecies communication in bacteria.
J. Clin. Invest. 112: 12919.
143
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
26 Folcher, M., H. Gaillard, L. T. Nguyen et al. 2001. Pleiotropic functions of a
Streptomyces pristinaespiralis autoregulator receptor in development, antibiotic
biosynthesis, and expression of a superoxide dismutase. J. Biol. Chem. 276:
44297306.
27 Fong, K. P., L. Gao and D. R. Demuth 2003. luxS and arcB control aerobic growth
of Actinobacillus actinomycetemcomitans under iron limitation. Infect. Immun.
71: 298308.
28 Forsyth, M. H. and T. L. Cover 2000. Intercellular communication in Helicobacter
pylori: luxS is essential for the production of an extracellular signaling molecule.
Infect. Immun. 68: 31939.
29 Fraser, C. M., S. Casjens, W. M. Huang et al. 1997. Genomic sequence of a Lyme
disease spirochaete, Borrelia burgdorferi. Nature 390: 5806.
30 Fuqua, C., M. R. Parsek and E. P. Greenberg 2001. Regulation of gene expression
by cell-to-cell communication: acyl-homoserine lactone quorum sensing. A. Rev.
Genet. 35: 43968.
31 Giron, J. A., A. G. Torres, E. Freer and J. B. Kaper 2002. The flagella of entero-
pathogenic Escherichia coli mediate adherence to epithelial cells. Molec. Microbiol.
44: 36179.
32 Gray, K. M. and E. P. Greenberg 1992. Physical and functional maps of the
luminescence gene cluster in an autoinducer-deficient Vibrio fischeri strain iso-
lated from a squid light organ. J. Bacteriol. 174: 438490.
33 Greene, R. 1996. Biosynthesis of methionine. In F. C. Neidhardt (ed.), Escherichia
coli and Salmonella: Cellular and Molecular Biology, vol. 1, pp. 54260.
Washington, DC: ASM Press.
34 Hammer, B. K. and B. L. Bassler 2003. Quorum sensing controls biofilm forma-
tion in Vibrio cholerae. Molec. Microbiol. 50: 1014.
35 Havarstein, L. S., G. Coomaraswamy and D. A. Morrison 1995. An unmodified
heptadecapeptide pheromone induces competence for genetic transformation in
Streptococcus pneumoniae. Proc. Natn. Acad. Sci. USA 92: 111404.
36 Heidelberg, J. F., J. A. Eisen, W. C. Nelson et al. 2000. DNA sequence of both
chromosomes of the cholera pathogen Vibrio cholerae. Nature 406: 47783.
37 Hilgers, M. T. and M. L. Ludwig 2001. Crystal structure of the quorum-sensing
protein LuxS reveals a catalytic metal site. Proc. Natn. Acad. Sci. USA 98:
1116974.
38 Horinouchi, S. and T. Beppu 1994. A-factor as a microbial hormone that controls
cellular differentiation and secondary metabolism in Streptomyces griseus. Molec.
39 Huang, J. J., J. I. Han, L. H. Zhang and J. R. Leadbetter 2003. Utilization of acyl-
homoserine lactone quorum signals for growth by a soil pseudomonad and
Pseudomonas aeruginosa PAO1. Appl. Environ. Microbiol. 69: 59419.
40 Hubner, A., A. T. Revel, D. M. Nolen, K. E. Hagman and M. V. Norgard 2003.
Expression of a luxS gene is not required for Borrelia burgdorferi infection of mice
via needle inoculation. Infect. Immun. 71: 28926.
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
144
41 Jelsbak, L. and L. Sogaard-Andersen 2003. Cell behavior and cell-cell commu-
nication during fruiting body morphogenesis in Myxococcus xanthus. J. Microbiol.
Methods 55: 82939.
42 Jeon, B., K. Itoh, N. Misawa and S. Ryu 2003. Effects of quorum sensing on flaA
transcription and autoagglutination in Campylobacter jejuni. Microbiol. Immunol.
47: 8339.
43 Ji, G., R. Beavis and R. P. Novick 1997. Bacterial interference caused by auto-
inducing peptide variants. Science 276: 202730.
44 Ji, G., R. C. Beavis and R. P. Novick 1995. Cell density control of staphylococcal
virulence mediated by an octapeptide pheromone. Proc. Natn. Acad. Sci. USA 92:
120559.
45 Joyce, E. A., B. L. Bassler and A. Wright 2000. Evidence for a signaling system in
Helicobacter pylori: detection of a luxS-encoded autoinducer. J. Bacteriol. 182:
363843.
46 Joyce, E. A., A. Kawale, S. Censini et al. 2004. LuxS is required for persistent
pneumococcal carriage and expression of virulence and biosynthesis genes.
47 Kaiser, D. 2003. Coupling cell movement to multicellular development in myxo-
bacteria. Nat. Rev. Microbiol. 1: 4554.
48 Kaplan, H. B. 2003. Multicellular development and gliding motility in
Myxococcus xanthus. Curr. Opin. Microbiol. 6: 5727.
49 Karp, P. D., M. Arnaud, J. Collado-Vides et al. 2004. The E. coli EcoCyc database:
no longer just a metabolic pathway database. ASM News 70: 2530.
50 Karp, P. D., M. Riley, S. M. Paley and A. Pellegrini-Toole 2002. The MetaCyc
Database. Nucleic Acids Res. 30: 5961.
51 Kim, S. Y., S. E. Lee, Y. R. Kimet al. 2003. Regulation of Vibrio vulnificus virulence
by the LuxS quorum-sensing system. Molec. Microbiol. 48: 164764.
52 Kleerebezem, M., L. E. Quadri, O. P. Kuipers and W. M. de Vos 1997. Quorum
sensing by peptide pheromones and two-component signal-transduction systems
in Gram-positive bacteria. Molec. Microbiol. 24: 895904.
53 Lazdunski, A. M., I. Ventre and J. N. Sturgis 2004. Regulatory circuits and
communication in Gram-negative bacteria. Nat. Rev. Microbiol. 2: 58192.
54 Lemos, J. A., T. A. Brown, Jr. and R. A. Burne 2004. Effects of RelA on key
virulence properties of planktonic and biofilm populations of Streptococcus
mutans. Infect. Immun. 72: 143140.
54a Lenz, D. H., K. C. Mok, B. N. Lilley, R. V. Kulkarni, N. S. Wingreenand B. L. Bassler
(2004). The small RNA chaperone Hfq and multiple small RNAs control quorum
sensing in Vibrio harveyi and Vibrio cholerae. Cell 118: 6982.
55 Lewis, H. A., E. B. Furlong, B. Laubert et al. 2001. A structural genomics
approach to the study of quorum sensing: crystal structures of three LuxS
orthologs. Structure (Camb.) 9: 52737.
56 Loh, J. T., M. H. Forsyth and T. L. Cover 2004. Growth phase regulation of flaA
expression in Helicobacter pylori is luxS dependent. Infect. Immun. 72: 550610.
145
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
57 Lupp, C. and E. G. Ruby 2004. Vibrio fischeri LuxS and AinS: comparative study of
two signal synthases. J. Bacteriol. 186: 387381.
58 Magnuson, R., J. Solomon and A. D. Grossman 1994. Biochemical and
genetic characterization of a competence pheromone from B. subtilis. Cell 77:
20716.
59 McNab, R., S. K. Ford, A. El-Sabaeny et al. 2003. LuxS-based signaling in
Streptococcus gordonii: autoinducer 2 controls carbohydrate metabolism and bio-
film formation with Porphyromonas gingivalis. J. Bacteriol. 185: 27484.
60 Merritt, J., F. Qi, S. D. Goodman, M. H. Anderson and W. Shi 2003. Mutation of
luxS affects biofilm formation in Streptococcus mutans. Infect. Immun. 71: 19729.
61 Michael, B., J. N. Smith, S. Swift, F. Heffron and B. M. Ahmer 2001. SdiA of
Salmonella enterica is a LuxR homolog that detects mixed microbial communities.
J. Bacteriol. 183: 573342.
62 Miller, C. H. and J. A. Duerre 1968. S-ribosylhomocysteine cleavage enzyme
from Escherichia coli. J. Biol. Chem. 243: 927.
63 Miller, J. C. and B. Stevenson 2004. Increased expression of Borrelia burgdorferi
factor H-binding surface proteins during transmission from ticks to mice. Int. J.
Med. Microbiol. 293 (suppl. 37): 1205.
65 Miller, M. B., K. Skorupski, D. H. Lenz, R. K. Taylor and B. L. Bassler 2002.
Parallel quorum sensing systems converge to regulate virulence in Vibrio
cholerae. Cell 110: 30314.
66 Miller, S. T., K. B. Xavier, S. R. Campagna et al. 2004. Salmonella typhimurium
recognizes a chemically distinct form of the bacterial quorum-sensing signal
AI-2. Molec. Cell 15: 67787.
67 Mok, K. C., N. S. Wingreen and B. L. Bassler 2003. Vibrio harveyi quorum
sensing: a coincidence detector for two autoinducers controls gene expression.
EMBO J. 22: 87081.
68 Mori, M., Y. Sakagami, Y. Ishii et al. 1988. Structure of cCF10, a peptide sex
pheromone which induces conjugative transfer of the Streptococcus faecalis tetra-
cycline resistance plasmid, pCF10. J. Biol. Chem. 263: 145748.
69 Nealson, K. H., T. Platt and J. W. Hastings 1970. Cellular control of the synthesis
and activity of the bacterial luminescent system. J. Bacteriol. 104: 31322.
70 Ohtani, K., H. Hayashi and T. Shimizu 2002. The luxS gene is involved in cell-
cell signaling for toxin production in Clostridium perfringens. Molec. Microbiol. 44:
1719.
71 Otto, M. 2001. Staphylococcus aureus and Staphylococcus epidermidis peptide
pheromones produced by the accessory gene regulator agr system. Peptides
22: 16038.
72 Otto, M., R. Sussmuth, C. Vuong, G. Jung and F. Gotz 1999. Inhibition of
virulence factor expression in Staphylococcus aureus by the Staphylococcus epider-
midis agr pheromone and derivatives. FEBS Lett. 450: 25762.
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
146
73 Pappas, K. M., C. L. Weingart and S. C. Winans 2004. Chemical communication
in proteobacteria: biochemical and structural studies of signal synthases and
receptors required for intercellular signaling. Molec. Microbiol. 53: 75569.
74 Parsek, M. R. and E. P. Greenberg 2000. Acyl-homoserine lactone quorum
sensing in gram-negative bacteria: a signaling mechanism involved in associa-
tions with higher organisms. Proc. Natn. Acad. Sci. USA 97: 878993.
75 Pei D. and J. Zhu 2004. Mechanismof action of S-ribosylhomocysteinase (LuxS).
Curr. Opin. Chem. Biol. 8: 4927.
76 Riedel, K., M. Hentzer, O. Geisenberger et al. 2001. N-acylhomoserine-lactone-
mediated communication between Pseudomonas aeruginosa and Burkholderia
cepacia in mixed biofilms. Microbiology 147: 324962.
77 Roche, D. M., J. T. Byers, D. S. Smith et al. 2004. Communications blackout? Do
N-acylhomoserine-lactone-degrading enzymes have any role in quorum sensing?
78 Ruzheinikov, S. N., S. K. Das, S. E. Sedelnikova et al. 2001. The 1.2 Astructure of a
novel quorum-sensing protein, Bacillus subtilis LuxS. J. Molec. Biol. 313: 11122.
79 Schauder, S. and B. L. Bassler 2001. The languages of bacteria. Genes Dev. 15:
146880.
81 Shin, N. R., D. Y. Lee, S. J. Shin, K. S. Kim and H. S. Yoo 2004. Regulation
of proinflammatory mediator production in RAW264.7 macrophage by Vibrio
vulnificus luxS and smcR. FEMS Immunol. Med. Microbiol. 41: 16976.
82 Sircili, M. P., M. Walters, L. R. Trabulsi and V. Sperandio. 2004. Modulation
of enteropathogenic Escherichia coli virulence by quorum sensing. Infect. Immun.
72: 232937.
83 Smith, J. N. and B. M. Ahmer 2003. Detection of other microbial species by
Salmonella: expression of the SdiA regulon. J. Bacteriol. 185: 135766.
84 Smith, R. S. and B. H. Iglewski 2003. P. aeruginosa quorum-sensing systems and
virulence. Curr. Opin. Microbiol. 6: 5660.
85 Socransky, S. S. and A. D. Haffajee 1992. The bacterial etiology of destructive
periodontal disease: current concepts. J. Periodontol. 63: 32231.
86 Sperandio, V., J. L. Mellies, W. Nguyen, S. Shin and J. B. Kaper 1999. Quorum
sensing controls expression of the type III secretion gene transcription and
protein secretion in enterohemorrhagic and enteropathogenic Escherichia coli.
Proc. Natn. Acad. Sci. USA 96: 15196201.
87 Sperandio, V., A. G. Torres, J. A. Giron and J. B. Kaper 2001. Quorum sensing is
a global regulatory mechanism in enterohemorrhagic Escherichia coli O157:H7.
J. Bacteriol. 183: 518797.
89 Sperandio, V., A. G. Torres, B. Jarvis, J. P. Nataro and J. B. Kaper 2003. Bacteria-
host communication: the language of hormones. Proc. Natn. Acad. Sci. USA
100: 89516.
147
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
90 Sperandio, V., A. G. Torres and J. B. Kaper 2002. Quorum sensing Escherichia
coli regulators B and C (QseBC): a novel two-component regulatory system
involved in the regulation of flagella and motility by quorum sensing in E.coli.
91 Stevenson, B. and K. Babb 2002. LuxS-mediated quorum sensing in Borrelia
burgdorferi, the lyme disease spirochete. Infect. Immun. 70: 4099105.
92 Stevenson, B., K. von Lackum, R. L. Wattier et al. 2003. Quorum sensing by the
Lyme disease spirochete. Microbes Infect. 5: 9917.
93 Stover, C. K., X. Q. Pham, A. L. Erwin et al. 2000. Complete genome sequence
of Pseudomonas aeruginosa PA01, an opportunistic pathogen. Nature 406:
95964.
94 Sturme, M. H., M. Kleerebezem, J. Nakayama et al. 2002. Cell to cell communi-
cation by autoinducing peptides in gram-positive bacteria. Antonie Van
Leeuwenhoek 81: 23343.
95 Surette, M. G. and B. L. Bassler 1998. Quorum sensing in Escherichia coli and
Salmonella typhimurium. Proc. Natn. Acad. Sci. USA 95: 704650.
96 Surette, M. G. and B. L. Bassler 1999. Regulation of autoinducer production in
Salmonella typhimurium. Molec. Microbiol. 31: 58595.
97 Surette, M. G., M. B. Miller and B. L. Bassler 1999. Quorum sensing in
Escherichia coli, Salmonella typhimurium and Vibrio harveyi: a new family of
genes responsible for autoinducer production. Proc. Natn. Acad. Sci. USA 96:
163944.
98 Taga, M. E., S. T. Miller and B. L. Bassler 2003. Lsr-mediated transport and
processing of AI-2 in Salmonella typhimurium. Molec. Microbiol. 50: 141127.
99 Tummler, B. and C. Kiewitz 1999. Cystic fibrosis: an inherited susceptibility to
bacterial respiratory infections. Molec. Med. Today 5: 3518.
100 Wen, Z. T. and R. A. Burne 2002. Functional genomics approach to identifying
genes required for biofilm development by Streptococcus mutans. Appl. Environ.
Microbiol. 68: 1196203.
101 Wen, Z. T. and R. A. Burne 2004. LuxS-mediated signaling in Streptococcus
mutans is involved in regulation of acid and oxidative stress tolerance and biofilm
formation. J. Bacteriol. 186: 268291.
102 Whitehead, N. A., A. M. Barnard, H. Slater, N. J. Simpson and G. P. Salmond 2001.
Quorum-sensing in Gram-negative bacteria. FEMS Microbiol. Rev. 25: 365404.
103 Winzer, K., K. R. Hardie, N. Burgess et al. 2002. LuxS: its role in central
metabolism and the in vitro synthesis of 4-hydroxy-5-methyl-3(2H)-furanone.
104 Winzer, K., Y. H. Sun, A. Green et al. 2002. Role of Neisseria meningitidis luxS in
cell-to-cell signaling and bacteremic infection. Infect. Immun. 70: 22458.
106 Yarwood, J. M. and P. M. Schlievert 2003. Quorum sensing in Staphylococcus
infections. J. Clin. Invest. 112: 16205.
K
.
D
U
A
N
A
N
D
M
.
G
.
S
U
R
E
T
T
E
148
107 Zhang, H. B., L. H. Wang and L. H. Zhang 2002. Genetic control of quorum-
sensing signal turnover in Agrobacterium tumefaciens. Proc. Natn. Acad. Sci. USA
99: 463843.
108 Zhu, J., E. Dizin, X. Hu et al. 2003. S-Ribosylhomocysteinase (LuxS) is a mono-
nuclear iron protein. Biochemistry 42: 471726.
109 Zhu, J., X. Hu, E. Dizin and D. Pei 2003. Catalytic mechanism of
S-ribosylhomocysteinase (LuxS): direct observation of ketone intermediates by
13C NMR spectroscopy. J. Am. Chem. Soc. 125: 1337981.
110 Zhu, J., M. B. Miller, R. E. Vance et al. 2002. Quorum-sensing regulators control
virulence gene expression in Vibrio cholerae. Proc. Natn. Acad. Sci. USA 99:
312934.
149
L
u
x
S
I
N
C
E
L
L
U
L
A
R
M
E
T
A
B
O
L
I
S
M
A
N
D
S
I
G
N
A
L
I
N
G
CHAPTER 7
LuxS-dependent regulation of Escherichia coli
virulence
Marcie B. Clarke and Vanessa Sperandio
University of Texas Southwestern Medical Center,
Dallas, TX, USA
INTRODUCTION
Escherichia coli is the most abundant facultative anaerobe found in the
human intestinal microbial flora. This organism resides in the mucus layer
of the mammalian colon, and typically colonizes the gastrointestinal tract
of humans a few hours after birth. However, there are several clones of
E. coli that have acquired virulence traits that allow them to cause a broad
spectrum of disease. These virulence traits are usually encoded within
mobile genetic elements, such as plasmids and pathogenicity islands,
that have evolved to be stable within these clones. Three general clinical
syndromes result from the infection with these pathotypes: diarrheal dis-
ease, urinary tract infections, and meningitis/sepsis. Among the intestinal
pathogens there are six well-described categories: enterohemorrhagic
E. coli (EHEC), enteropathogenic E. coli (EPEC), enterotoxigenic E. coli
(ETEC), enteroaggregative E. coli (EAEC), enteroinvasive E. coli (EIEC),
and diffusely adherent E. coli (DAEC) (59). This chapter will focus primarily
on EHEC and EPEC, given that quorum sensing has been mostly described
within these pathotypes.
ENTEROHEMORRHAGIC E. COLI (EHEC)
Enterohemorrhagic E. coli (EHEC) O157:H7 is responsible for major
outbreaks of bloody diarrhea and hemolytic uremic syndrome (HUS)
throughout the world. EHEC causes an estimated 73,000 illnesses, 2,000
hospitalizations, and 60 deaths in the United States alone each year. EHEC
has a very low infectious dose (as few as 50 cfu); this is one of the major
contributing factors to EHEC outbreaks. Treatment and intervention
151
strategies for EHEC infections are still very controversial, with conven-
tional antibiotics usually having little clinical effect and possibly even
being harmful (by increasing the chances of patients developing hemolytic
uremic syndrome (HUS))(41, 42).
EHEC colonizes the large intestine, where it causes attaching and
effacing (AE) lesions. The AE lesion is characterized by the destruction of
the microvilli and the rearrangement of the cytoskeleton to forma pedestal-
like structure, which cups the bacteria individually. The genes involved in
the formation of the AE lesion are encoded within a chromosomal patho-
genicity island named the Locus of Enterocyte Effacement (LEE) (37). The
LEE region contains five major operons: LEE1, LEE2, LEE3, tir (LEE5), and
LEE4 (4), which encode a type III secretion system (TTSS), an adhesin
(intimin), and this adhesins receptor (Tir), which is translocated to the
epithelial cell through the bacterial TTSS (16, 55). The LEE genes are
directly activated by the LEE-encoded regulator (Ler), which is the first
gene in the LEE1 operon (5, 15, 55, 70, 84). Transcription of the LEE genes
is further positively and negatively modulated by GrlA and GlrR, respec-
tively, which are encoded in a small operon downstreamof LEE1 (13). EHEC
also produces a potent Shiga toxin (Stx) that is responsible for the major
symptoms of hemorrhagic colitis and HUS. There are two types of Stx, Stx1
and Stx2, which are most frequently associated with human disease. Both
of the genes encoding Stx1 and Stx2 are located within the late genes of a
l-like bacteriophage, and are transcribed when the phage enters its lytic
cycle (60). Disturbances in the bacterial membrane, DNA replication, or
protein synthesis (which are the targets of conventional antibiotics) may
trigger a SOS response in the bacterial cells that signals the bacteriophage
to enter the lytic cycle (41, 42). The phage replicates, Shiga toxin is pro-
duced, and the phage lyses the bacteria, thereby releasing Shiga toxin into
the host.
LuxS AND CELL-TO-CELL SIGNALING IN BACTERIA
The phenomenon of cell-to-cell signaling in bacteria has historically
been referred to as quorum sensing. The most widespread quorumsensing
system is the luxS system, first described as being involved in biolumines-
cence in Vibrio harveyi (90). Among the diverse bacterial species that
contain the luxS quorum sensing system is E. coli, including EHEC sero-
type O157:H7 (85, 89, 90). LuxS is an enzyme involved in the metabolismof
S-adenosylmethionine (SAM); it converts ribose-homocysteine into homo-
cysteine and 4,5-dihydrody-2,3-pentanedione (DPD). DPD is a very
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
152
unstable compound that reacts with water and cyclizes into several fura-
nones (73, 87, 98), one of which is thought to be the precursor of auto-
inducer-2 (AI-2) (73). The AI-2 structure has been solved by co-crystallizing
this ligand with its receptor LuxP (a periplasmic protein that resembles
the ribose binding protein RbsB) in Vibrio harveyi, and reported to be a
furanosyl-borate-diester (6). However, LuxP homologs, as well as homologs
from this signaling cascade, have only been found in Vibrio spp. Several
other bacterial species harbor the luxS gene and have AI-2 activity as
measured by a Vibrio harveyi bioluminescence assay (72, 100). However,
the only genes shown to be regulated by AI-2 in other species encode for
an ABC transporter in Salmonella typhimurium named Lsr (LuxS-regulated),
responsible for the AI-2 uptake (93). This ABC transporter is also present in
E. coli and shares homology with sugar transporters. Once inside the cell,
AI-2 is modified by phosphorylation and proposed to interact with LsrR,
which is a SorC-like transcription factor involved in repressing expression of
the lsr operon (92, 93) (Fig. 7.1). Several groups have been unable to detect
the furanosyl-borate-diester, proposed to be AI-2, in purified fractions con-
taining AI-2 activity from Salmonella spp. and E. coli (as measured by using
the V. harveyi bioluminescence assay) (73, 87, 98). These fractions only
yielded the identification of several furanosyl compounds that did not
contain boron. These results can be explained now that AI-2 has been
co-crystallized with its receptor (the periplasmic protein LsrB) in Salmonella.
In these studies the LsrB ligand was not a furanosyl-borate-diester, but a
furanone (2R, 4S-2-methyl-2,3,3,4-tetrahydrofuran (R-THMF)), consistent
with what has been observed in AI-2 fractions of Salmonella and E. coli
(58, 87, 98). This is fundamentally different from AI-2 detection in Vibrio
harveyi and raises the question whether all bacteria may actually use AI-2 as a
signaling compound, or whether it is released as a waste product or used as a
metabolite by some bacteria, rather than as a signal.
Diverse roles in signaling have been attributed to AI-2 in other organ-
isms by comparing luxS mutants with wild-type strains, and complement-
ing these mutants either genetically or with spent supernatants. Among
these roles are the LEE-encoded type III secretion system and flagellar
expression in EHEC (85, 86), expression of VirB in Shigella flexneri (10),
secretion of SpeB cysteine protease in Streptococcus pyogenes (49), type III
secretion in V. harveyi and V. parahaemolyticus (31), etc. The most compre-
hensive studies concerning the role for a LuxS-dependent autoinducer in
virulence have been performed in enterohemorrhagic E. coli (EHEC)
(83 , 85 8 8). However, by using purified and AI-2 synthesized in vitro, it has
been demonstrated that the signaling molecule activating type III secretion
153
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
and the flagellar regulon in EHEC is not the AI-2 autoinducer (87). The
autoinducer responsible for this signaling is dependent on the presence of
the luxS gene for its synthesis, but is different from AI-2. AI-2 is a very polar
furanone that does not bind to C-18 columns. The signaling compound
activating the EHEC virulence genes, which was renamed autoinducer-3
(AI-3), binds to C-18 columns and can only be eluted with methanol (87).
Electrospray Mass Spectrometry analysis of the AI-3 fraction showed a major
peak with a molecular mass of 213.1 Da and minor peaks at 109.1, 164.9,
lsrR lsrA lsrC lsrD lsrB lsrF lsrG lsrE
Salmonella
EHEC
ydeW Z2192 ydeY ydeZ Z2189 yneB Z2187
rbsD rbsA rbsC rbsB rbsK rbsR
rbs E. coli/Salmonella
RbsB
+
ribose
RbsB
(a)
(b)
R
b
s
C
R
b
s
C
RbsA
ATP
ADP
RbsK
ATP
ADP
P
LsrB
+
LsrB
L
s
rC
L
s
rC
LsRA
ATP
ADP
(c)
AI-2
LsRK
ATP
ADP
P
lsrK
ydeV
Figure 7.1. The Lsr ABC transporter system (LuxS regulated genes) from Salmonella
typhimurium and enterohemorrhagic Escherichia coli (EHEC) serotype O157:H7. All of the lsr
genes from Salmonella are present in EHEC, with the exception of lsrE. The lsrACDBFGE
genes are transcribed in an operon; in the opposite direction to this operon is lsrR (encoding
the SorC-like transcription factor that represses expression of the LsrABC transporter) and
lsrK, which phosphorylates AI-2 upon its entry in to the bacterial cell (92, 93). LsrB shares
homology with the ribose-binding periplasmic protein RbsB and is the receptor for AI-2 in
Salmonella and E. coli (58). Upon binding to LsrB, AI-2 is transported through the LsR ABC
transporter, which closely resembles the ribose ABC transporter. Once inside the cell AI-2 is
phosphorylated by LsrK. Phospho-AI-2 is thought to interact with LsrR to relieve the
repression of the lsr operon.
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
154
176.1, 196.1, 211.1, 214.1 and 222.9 Da (87). All of these are different from
those of AI-2 (6), suggesting that AI-3 is a novel compound.
These results suggest that some of the phenotypes attributed to AI-2
signaling need to be revised in light of the fact that LuxS is not devoted to
AI-2 production; it is infact an enzyme involved in the biochemical pathway for
metabolism of SAM. Consequently, altered gene expression due to a luxS
mutation will involve both genes affected by quorum sensing per se and genes
differentially expressed because of the interruption of this metabolic pathway.
Furthermore, one also has to take into consideration that a knockout of luxS
seems to affect the synthesis of at least two autoinducers, AI-2 and AI-3 (87).
The activity of both signals can be differentiated by utilizing biological tests
specific to each signal. For example, AI-3 shows no activity for the AI-2 bioassay
(87), whichis predicated onthe production of bioluminescence inVibrio harveyi
(89) and is the gold standard for AI-2 production. On the other hand, AI-3
activates the transcription of the EHEC LEE virulence genes, whereas AI-2 has
no effect in this assay (87). The only two phenotypes shown to be AI-2-
dependent, using either purified or in vitro synthesized AI-2, are biolumines-
cence in V. harveyi (73) and expression of the lsr operon in S. typhimurium (93).
The luxS gene is present in an array of bacterial species, including several
members of the human commensal microflora (90). It has recently been
shown, by using anaerobically cultured stools fromhealthy human volunteers,
that the microbial intestinal flora produce both AI-2 (detected with the
V. harveyi bioluminescence assay) and AI-3 (detected with the LEE1 transcrip-
tion AI-3-dependent bioassay) (87). To obtain further information regarding
which intestinal commensals and pathogens are able to produce AI-2 and
AI-3, freshly isolated strains from patients were tested (M. P. Sircili and
V. Sperandio, unpublished results). Using the bioassays described above,
AI-2 and AI-3 activity was observed in spent supernatants from enteropatho-
genic E. coli strains fromserogroups O26:H11 and O111ac:H9, Shigella sp., and
Salmonella sp. Activity from both autoinducers was also detected in normal
flora bacteria such as a commensal E. coli, Klebsiella pneumoniae, and
Enterobacter clocae (M. P. Sircili and V. Sperandio, unpublished results).
These results suggest that interspecies signaling may be involved in the
pathogenesis of disease caused by these other bacteria, and also in signaling
by the intestinal microflora.
CELL-TO-CELL SIGNALING IN EHEC
Sperandio et al. (85) reported that transcription of all of the LEE
operons is activated by the presence of autoinducers in supernatants from
155
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
wild-type EHEC, commensal E. coli, and MG1655 (K-12) strains, but not
from an isogenic EHEC luxS mutant or from K-12 strain DH5a (which has
a frameshift mutation in the luxS gene) (85). Analysis of luxS mutants in
EPEC and EHEC strains demonstrated an effect on type III secretion in
both of them. Type III secretion was diminished in the EPEC luxS mutant
(85) and could not be detected in the EHEC luxS mutant (87). In both cases,
this phenotype could be restored by genetic complementation with the luxS
gene cloned on a plasmid, or by providing AI-3 exogenously (85, 87).
In addition to activation of type III secretion, Sperandio et al. (86)
reported that about 10% of the common genome between EHEC and
E. coli K-12 is differentially expressed between a wild-type EHEC and its
isogenic luxS mutant (EHEC has 1.3 Mb of DNA absent in K-12, and K-12
has 0.53 Mb of DNA that is absent in EHEC (62)) (86). DeLisa and
colleagues (12) also reported, by using gene arrays, that about 5.6% of the
K-12 genome was differentially regulated between a wild-type K-12 strain
and its isogenic luxS mutant. The difference in the numbers of genes
regulated in both reports may be due to differences in methodology (growth
temperature: Sperandio et al. (86) used 37 8C whereas DeLisa and collea-
gues (12) used 308C; nutrient availability: Sperandio et al. (86) grew their
strains in DMEM, which they have previously shown to give better expres-
sion of luxS-controlled genes (85), whereas DeLisa and colleagues (12) used
LB broth) and in strains utilized (Sperandio et al. (86) used an EHEC strain,
whereas DeLisa and colleagues (12) used a K-12 strain).
The observation that about 10% of the array genes are differentially
regulated between an EHEC wild type and its isogenic luxS mutant is not
surprising if one considers the pleiotropic nature of a luxS mutation. LuxS
is a metabolic enzyme involved primarily in the conversion of ribosyl-
homocysteine into homocysteine and 4,5-dihydroxy-2,3-pentanedione,
which is the precursor of AI-2 (73). A luxS mutation will interrupt this
metabolic pathway, changing the whole metabolism of the bacterium.
A luxS mutant will accumulate S-ribosyl-homocysteine because it is
unable to catalyze its conversion to homocysteine. This could cause the
concentration of homocysteine to diminish within the cell. Inasmuch as
homocysteine is used for the de novo synthesis of methionine, the cell will
use a salvage pathway: it will use oxaloacetate to produce homocysteine to
synthesize methionine. Given that oxaloacetate is necessary, together with
L-glutamate, to synthesize aspartate, by using this salvage pathway for the
de novo synthesis of methionine, other amino-acid synthetic and catabolic
pathways will be changed within the cell (www.ecosal.org/ecosal/
index.jsp).
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
156
Among the quorum-sensing-regulated genes and phenotypes noted in
these studies were the genes encoding flagella, and motility (which may
also be involved in pathogenesis) (86). Specifically, it has been shown that
transcription of flhDC(the master regulator of the flagellar regulon) and the
mot operon (encoding motility genes) is decreased in a luxS mutant com-
pared with wild-type and complemented strains. It has also been shown
that transcription of these genes, as well as motility, could be restored by
addition of signals exogenously, further confirming that regulation of
flagellar expression and motility is being controlled by a quorum sensing
signaling mechanism (86, 87). Quorum-sensing regulation of flhDC
expression has far-reaching implications beyond flagellar expression,
given that FlhDC has been shown to also regulate bacterial cell division
(65, 66) and several metabolic processes (64). Quorum-sensing regulation
of the LEE-encoded type III secretion and the flagellar regulon in EHEC is
dependent on the AI-3 signal; the role of AI-2 signaling in EHECremains to
be established (87). Given the widespread nature of the luxS/AI-3 system in
bacteria, an interesting extrapolation is that the AI-3/luxS quorum sensing
system might have evolved to mediate microflorahost interactions, but
evolved to be exploited by EHEC to activate its virulence genes. In this
manner, the AI-3/luxS system alerts EHEC as to when it has reached the
large intestine, where large numbers of commensal E. coli, Enterococcus,
Clostridium, and Bacteroides, all of which contain the AI-3/luxS quorum
sensing system, are resident.
BACTERIALHOST CELL-TO-CELL SIGNALING
It is estimated that humans have about 10
13
eukaryotic cells and 10
14
prokaryotic cells (comprising our endogenous bacterial flora). The gastro-
intestinal (GI) tract is the site of the largest and most complex environment
in the mammalian host. The density of bacteria along the GI tract can vary
greatly, with the majority of the flora residing in the colon (10
11
10
12
bacterial cells ml
1
). Given the enormous number and diversity of bacteria
in the GI environment, it should not be surprising that the members of this
community somehow communicate among themselves and with the host
itself to coordinate various processes. The observation that the human
bacterial flora is extremely important to development, as well as in shaping
the innate immune system, further reinforces this suggestion (34).
However, some interactions among eukaryotes and microbes are detrimen-
tal and culminate in disease. Given these polar relationships, it may be
asked, at what levels do prokaryotes and eukaryotes communicate?
157
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
The bacterial signal AI-3 is not only involved in bacterialbacterial
communication, but also in bacterialhost communication, crosstalking
with the human hormones epinephrine and norepinephrine (87). An
EHEC luxS mutant is unable to signal to itself in an in vitro culture and
therefore cannot express the LEE genes, which are essential for EHEC
virulence. These results suggest that type III secretion may be abrogated
in the luxS mutant in vitro. Based on these data, it was expected that the
luxS mutant would be unable to produce attaching and effacing (AE)
lesions on cultured epithelial cells. However, the luxS mutant was still
able to produce AE lesions on epithelial cells, indistinguishable from
those produced by the wild type. Since quorum sensing in bacteria is a
cell-to-cell signaling system, it was hypothesized that a eukaryotic signaling
compound could complement the bacterial mutation. Eukaryotic cell-to-
cell signaling occurs through hormones. There are three major groups of
endocrine hormones: polypeptide hormones, steroid hormones, and hor-
mones derived from the amino acid tyrosine, which include the catechola-
mines norepinephrine and epinephrine (30). Two of the Gram-negative
bacterial autoinducers (acyl-homoserine lactones and the AI-2) are also
derived from amino-acid metabolism (72). Norepinephrine has been
demonstrated to induce bacterial growth (52) and to be taken into bacteria
(43). Using purified epinephrine and norepinephrine, Sperandio et al. (87)
showed that the luxS mutant still responds to these eukaryotic signals.
It has been shown that there is a considerable amount of epinephrine and
norepinephrine in the human GI tract (14) and that these hormones induce
chloride and potassium secretion in the colon (35). The neuronally mediated
response to epinephrine in the distal colon can be suppressed by the non-
selective b-adrenergic receptor antagonist propranolol and by the non-selective
a-adrenergic receptor antagonist phentolamine in the proximal colon (35).
Finally, it has been demonstrated that the EHEC response to epinephrine
and norepinephrine signaling is specific, given that it can be blocked by
b-adrenergic antagonists (such as propranolol). Epinephrine and norepine-
phrine can substitute for AI-3 to activate transcription of the LEE genes, type III
secretion and AE lesions on epithelial cells. Taken together, these results
suggest that AI-3 and epinephrine/norepinephrine crosstalk and that these
compounds may use the same signaling pathway. As further evidence,
regulation of the flagellar regulon is also under AI-3 and epinephrine/
norepinephrine control and one can block EHEC response to both AI-3
and epinephrine/norepinephrine by using the b-adrenergic antagonist
propranolol. Specifically, propranolol can prevent formation of the AE lesions
by the wild-type EHEC, and by the luxS mutant in epithelial cells (87).
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
158
Norepinephrine has been reported to induce bacterial growth (19, 52);
there are reports in the literature, albeit conflicting, that imply that nor-
epinephrine might function as a siderophore (20, 43). Recently, norepi-
nephrine has been implicated as inducing expression of enterobactin and
iron uptake in E. coli, suggesting that this is the mechanism involved in
growth induction (4). However, the role of norepinephrine in bacterial
pathogenesis seems to be more complex, as several reports suggested that
this signal also activates virulence gene expression in E. coli, such as
production of fimbriae and Shiga toxin (50, 51), by an unknown mechanism
of induction. Sperandio et al. (87) show that both epinephrine and norepi-
nephrine seem to crosstalk with a bacterial quorum sensing system to
regulate virulence gene expression in EHEC. This signaling is not due to
enterobactin and is TonB-independent, suggesting that it is not dependent
on the FepA outer membrane receptor for this siderophore. Finally, this
signaling is dependent on a novel autoinducer, AI-3, which is produced by
intestinal flora (87). The line dividing quorum-sensing signaling and iron
uptake is becoming increasingly blurred, especially with the discovery that
the siderophore pyoverdine from P. aeruginosa also acts as a signaling
molecule (47).
In conclusion, EHECcould respond to both a bacterial quorumsensing
signaling system and a mammalian signaling system to fine tune tran-
scription of virulence genes at different stages of infection and/or different
sites of the gastrointestinal tract (Figure 7.2). Given that eukaryotic cell-to-
cell signaling occurs through hormones, and bacterial cell-to-cell signaling
occurs through quorum sensing, it is tempting to propose that quorum
sensing might be a language by which bacterial and host cells communi-
cate. Inasmuch as the host hormones epinephrine and norepinephrine
signal to EHEC, it remains to be determined whether AI-3 exerts any
functional effects on eukaryotic cell signaling.
THE EHEC QUORUM-SENSING SIGNALING CASCADE
Quorum-sensing regulatory cascades have been extensively studied in
organisms such as Pseudomonas aeruginosa and Vibrio harveyi, and have
proven to be very complex (11, 72). Concerning the EHEC AI-3epinephrine
norepinephrine signaling cascade, a transcriptional regulator from the
LysR family, designated as QS E. coli regulator A (QseA) (83) has been
recently identified. QseA is transcriptionally activated through quorum
sensing and, in turn, binds to and directly activates transcription of the
159
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
L
e
r
I
n
t
i
m
i
n
E
s
p
A

f
i
l
a
m
e
n
t
T
i
r
E
s
p
B
D
T
y
p
e

I
I
I

s
e
c
r
e
t
i
o
n
T
i
r
,

e
f
f
e
c
t
o
r
s
I
n
t
i
m
a
t
e

a
t
t
a
c
h
m
e
n
t
I
n
t
i
m
i
n

/

T
i
r
L
e
r
E
H
E
C
R
e
s
i
d
e
n
t

f
l
o
r
a
A
I
-
3
F
l
a
g
e
l
l
a
m
o
t
i
l
i
t
y
L
E
E
L
E
E
1
L
E
E
2
L
E
E
3
I
L
E
E
5
t
i
r
L
E
E
4
L
E
E
1
L
E
E
2
L
E
E
3
I
L
E
E
5
t
i
r
L
E
E
4
S
t
x
B
l
o
o
d
s
t
r
e
a
m
K
i
d
n
e
y
s
H
U
S
E
p
i
n
e
p
h
r
i
n
e
/
N
o
r
e
p
i
n
e
p
h
r
i
n
e
F
i
g
u
r
e
7
.
2
.
M
o
d
e
l
o
f
c
e
l
l
-
t
o
-
c
e
l
l
s
i
g
n
a
l
i
n
g
r
e
g
u
l
a
t
i
o
n
o
f
E
H
E
C
v
i
r
u
l
e
n
c
e
d
u
r
i
n
g
i
n
f
e
c
t
i
o
n
.
E
H
E
C
r
e
a
c
h
e
s
t
h
e
l
a
r
g
e
i
n
t
e
s
t
i
n
e
w
h
e
r
e
t
h
e
m
i
c
r
o
b
i
a
l
f
l
o
r
a
i
s
i
n
h
i
g
h
d
e
n
s
i
t
y
a
n
d
p
r
o
d
u
c
i
n
g
A
I
-
3
.
I
t
c
o
m
b
i
n
e
s
t
h
e
A
I
-
3
s
i
g
n
a
l
d
e
r
i
v
e
d
f
r
o
m
t
h
e
f
l
o
r
a
w
i
t
h
t
h
e
e
p
i
n
e
p
h
r
i
n
e
/
n
o
r
e
p
i
n
e
p
h
r
i
n
e
s
i
g
n
a
l
s
d
e
r
i
v
e
d
f
r
o
m
t
h
e
h
o
s
t
t
o
a
c
t
i
v
a
t
e
t
h
e
A
I
-
3
q
u
o
r
u
m
-
s
e
n
s
i
n
g
s
i
g
n
a
l
i
n
g
c
a
s
c
a
d
e
.
I
t
i
s
h
y
p
o
t
h
e
s
i
z
e
d
t
h
a
t
t
h
e
f
l
a
g
e
l
l
a
r
r
e
g
u
l
o
n
i
s
a
c
t
i
v
a
t
e
d
f
i
r
s
t
,
a
l
l
o
w
i
n
g
E
H
E
C
t
o
s
w
i
m
p
r
o
f
i
c
i
e
n
t
l
y
t
h
r
o
u
g
h
t
h
e
i
n
t
e
s
t
i
n
a
l
m
u
c
u
s
l
a
y
e
r
a
n
d
a
t
t
a
i
n
c
l
o
s
e
c
o
n
t
a
c
t
w
i
t
h
t
h
e
e
n
t
e
r
i
c
e
p
i
t
h
e
l
i
a
.
T
h
e
n
E
H
E
C
a
c
t
i
v
a
t
e
s
t
h
e
L
E
E
-
e
n
c
o
d
e
d
t
y
p
e
I
I
I
s
e
c
r
e
t
i
o
n
s
y
s
t
e
m
,
f
o
r
m
i
n
g
t
h
e
a
t
t
a
c
h
i
n
g
a
n
d
e
f
f
a
c
i
n
g
l
e
s
i
o
n
s
r
e
s
p
o
n
s
i
b
l
e
f
o
r
t
h
e
b
e
g
i
n
n
i
n
g
o
f
t
h
e
d
i
a
r
r
h
e
a
l
p
r
o
c
e
s
s
.
L
a
t
e
r
i
n
t
h
e
c
o
u
r
s
e
o
f
d
i
s
e
a
s
e
,
s
o
m
e
E
H
E
C
c
e
l
l
s
w
i
l
l
a
c
t
i
v
a
t
e
t
h
e
S
t
x
p
h
a
g
e
,
u
n
d
e
r
g
o
l
y
s
i
s
a
n
d
r
e
l
e
a
s
e
S
h
i
g
a
t
o
x
i
n
.
S
h
i
g
a
t
o
x
i
n
i
s
k
n
o
w
n
t
o
b
e
t
r
a
n
s
l
o
c
a
t
e
d
t
h
r
o
u
g
h
t
h
e
i
n
t
e
s
t
i
n
a
l
b
a
r
r
i
e
r
a
n
d
t
r
a
v
e
l
t
o
t
h
e
u
r
i
n
a
r
y
t
r
a
c
t
a
n
d
o
t
h
e
r
o
r
g
a
n
s
t
h
r
o
u
g
h
t
h
e
b
l
o
o
d
s
t
r
e
a
m
.
I
f
e
n
o
u
g
h
S
h
i
g
a
t
o
x
i
n
i
s
r
e
l
e
a
s
e
d
,
t
h
e
p
a
t
i
e
n
t
m
a
y
d
e
v
e
l
o
p
h
e
m
o
l
y
t
i
c
u
r
e
m
i
c
s
y
n
d
r
o
m
e
(
H
U
S
)
.
LEE-encoded regulator (Ler) (encoded within the LEE1 operon) (F. Sharp and
V. Sperandio, unpublished results) (83). Ler is the activator for all of the other
genes within the LEE island (55). In addition, QseA also activates transcrip-
tion of the grlRA operon (R. Russell and V. Sperandio, unpublished results).
GrlA has been reported to activate transcription of LEE1 (ler), whereas GrlR
seems to repress it (13). These results suggest that QseA regulates transcrip-
tion on LEE1 at more than one level. Consequently, an EHEC qseA mutant
has a striking reduction in type III secretion but has no defect in flagellation
or motility, suggesting that QseA regulates only the LEE genes and plays no
role in the flagellar regulon (83). QseA belongs to the LysR family of tran-
scription factors; it is therefore not surprising that it autorepresses its own
transcription (79).
In addition, a two-component system renamed QseBC has been iden-
tified. QseBCis responsible for the transcriptional activation of the flagellar
regulon in response to quorum sensing (88). The QseBC system is present
in UPEC, EPEC, EHEC, E. coli K12, Salmonella typhimurium, S. typhi,
Pasteurella multocida, and Haemophilus influenzae. QseB has three amino-
acid changes between EHECand E. coli K12, whereas QseC has eight amino
acid changes. In addition, QseB shares a high level of homology with
S. typhimurium PmrA (46% identity and 62% similarity over 222 amino-
acids) and QseC shares homology with S. typhimurium PmrB (28% identity
and 45% similarity over 269 amino acids). PmrAB is involved in gene
regulation in response to extracytoplasmic ferric iron (99) and genes that
confer resistance to antimicrobial peptides such as polymyxin (28, 81).
Traditional two-component systems consist of the sensor protein,
which acts as a histidine kinase to transfer a phosphoryl group upon
sensing of an environmental signal to an aspartate residue of its cognate
response regulator, which goes on to act as a transcription factor. Both
QseBand QseCcontain conserved domains characteristic of a two-component
system. QseC, the putative sensor kinase, has two conserved trans-
membrane domains and a conserved histidine kinase domain, indicating
that its membrane localization may allow autophosphorylation upon the
recognition of its specific environmental cue. QseC also contains an
ATPase domain, which may allow it to exhibit phosphatase activity toward
QseB. In addition, QseC has a conserved EAL domain, commonly found in
signaling proteins, which consists of several acidic residues that could be
important for metal binding and may make up an active site for a phospho-
diesterase of cyclic diguanylate (c-di-GMP), a cyclic nucleotide (56, 94). The
EAL domain of the VieA response regulator in V. cholerae has recently been
implicated in the formation of biofilms by controlling c-di-GMP
161
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
concentration in V. cholerae (94). QseB contains typical response regulator
and DNA binding domains, which may allow it to receive a phosphate from
QseC and undergo a conformational change allowing it to bind efficiently to
DNA and regulate gene transcription.
The translational stop codon of qseB overlaps with the translational
start codon of qseC; reverse transcriptase PCR experiments further con-
firmed that the qseBC genes comprise an operon (8). It is well known that
many two-component systems act to positively regulate their own transcrip-
tion (3). QseBC is no exception to this rule and has also been shown to
autoactivate its own transcription (8). Transcriptional autoregulation could
serve several purposes, including the amplification of signal or providing
an additional threshold for gene activation. Signal amplification could
allow bacteria to respond extremely quickly to an environmental signal.
This scenario has been observed in other two-component systems, such as
PmrB/PmrA of S. typhimurium (28), CpxAR of E. coli (67), BvgAS of
B. pertussis (69), and PhoQ/PhoP of S. typhimurium (82). In addition,
Hoffer et al. (2001) has suggested that autoregulation of a two-component
system may be responsible for a learning system in which bacteria can
respond quickly and more effectively to a signal that has been seen in the
recent past. The PhoB/PhoR two-component system of Salmonella, in
which previous exposure to a signal appears to boost reaction during the
second exposure, appears to exhibit this type of learning (33). As a final
point, autoregulation of a two-component system could provide an addi-
tional threshold for gene activation, as is observed with CpxAR in E. coli. In
this two-component system, signal persistence is essential for the autoam-
plification and accumulation of the CpxR response regulator to a threshold
concentration before transcription can be activated (67). This additional
level of control could allow the bacterial cell to activate the energetically
expensive production of flagella through QseBC only under appropriate
conditions.
In addition to auto-activating its own transcription, QseBC is involved
in activating transcription of the flagellar regulon (88). The expression and
synthesis of the flagella and motility genes is a highly complex process,
requiring the products of more than 50 genes that are organized in 17
operons (54). These operons are organized into a hierarchy of transcrip-
tional classes: class 1, class 2, and class 3 (46). Class 1 consists solely of the
master regulator of the flagellar regulon, FlhDC, which is a required trans-
criptional activator of the class 2 genes (48). The second class of genes
includes proteins that form the hook and basal body of the flagellar appa-
ratus, s
28
, and FlgM. During early time points of flagellar activation, FlgM
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
162
acts as an anti-sigma factor to forma complex with s
28
in order to sequester
it from RNA polymerase (RNAP). The s
28
protein is an alternative sigma
factor, which has been shown to associate with RNAP and is necessary for
the transcriptional activation of the class 3 genes, which include the motility
proteins (mot operon) and the flagellin subunit (fliC) (7). Upon completion
of the hook and basal body, FlgM is exported through the immature
flagellar apparatus and depleted from the cytoplasm, freeing s
28
to activate
class 3 gene expression (22, 36).
The expression of the flagellar regulon has been shown to be regulated
through several environmental factors, including cAMP-CRP (78, 102),
temperature (1), osmolarity (77), cell-cycle control (61), and bacterial cell
density (quorum sensing) (7, 86). Currently, six transcriptional start sites
have been mapped for flhDC (101). It has recently been reported that the
expression of flagella and motility in EHEC and E. coli K12 is regulated by
quorum sensing through QseBC (88). An isogenic mutant in qseC in both
EHEC and K12 produced fewer flagella and was less motile than wild-type
and complemented strains. In addition, transcriptional fusions of flagella
class 1 (flhDC), class 2 (fliA), and class 3 (fliC and motA) were reduced
compared with the wild type. These data may suggest that QseBC acts to
regulate flagellar expression through the master regulator flhDC.
Additional studies (M. B. Clarke and V. Sperandio, unpublished results)
indicate that QseB directly binds to the flhDC and qseBC promoters in order
to regulate flagellar and its own expression.
The QseBC two-component system is activated by quorum sensing
through AI-3 (8688). Early studies indicated that an isogenic mutant in
the qseC sensor kinase was unable to respond to bacterial autoinducers or
epinephrine given exogenously (8688). Interestingly, the motility of a
luxS mutant can be restored by the addition of either autoinducers con-
tained in preconditioned supernatants or epinephrine (86, 87). In addition,
the transcription of flhDC is activated by both epinephrine and AI-3 in the
luxS mutant. Motility and flhDC transcription in a qseC mutant, however,
are unable to respond to the presence of either AI-3 or epinephrine, indicat-
ing that QseCmay possibly be sensing the presence of these cross-signaling
compounds (87).
Although QseBC regulates both its own transcription and that of
flhDC, it plays no role in the regulation of other quorum-sensing pheno-
types, such as the LEE genes (88). Because flagella and motility are not the
only phenotypes controlled by quorum sensing in EHEC, it is hypothesized
that there are several other regulators involved in this quorum-sensing
system.
163
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
Finally, three other genes in this signaling cascade have also been
identified recently: qseD (encoding another regulator of the LysR family),
and qseE and qseF (encoding a second two-component system), which are
involved in regulating expression of the LEE genes (F. Sharp et al. and
N. Reading et al., unpublished results). These data suggest that both AI-3
and epinephrine/norepinephrine are recognized by the same receptor,
which is probably in the outer membrane of the bacterium (owing to the
non-polar nature of both AI-3 and epinephrine) (87). These signaling
molecules might be imported into the periplasmic space, where they then
would interact either with one major sensor kinase (that directs the tran-
scription of other sensor kinases) or with more than one sensor kinase. The
latter hypothesis is favored, given the results that a qseCmutant, which does
not respond to either AI-3 or epinephrine, only affects the quorum-sensing
regulation of the flagellar regulon and not of the LEE genes (88), and that a
qseEF mutation only affects transcription of the LEE genes and not the
flagellar regulon (Reading, unpublished results). The interaction of AI-3
and epinephrine with more than one sensor kinase would also impart a
timing mechanism to this system, which is a desirable feature given that
it would be inefficient for EHEC to produce both the LEE type III secretion
system and flagella simultaneously. Therefore, it is hypothesized that
EHEC activates expression of the flagellar regulon first through QseBC,
and then the LEE genes at a later time through QseEF. A model of the
EHECAI-3 quorumsensing signaling cascade is depicted in Figure 7.3. The
AI-3-dependent quorum-sensing signaling cascade is present in all
Enterobacteriaceae (E. coli, Salmonella spp., Shigella spp., and Yersinia
spp.). The most striking feature is that the genes encoding the transcrip-
tional factors of this cascade are always in exactly the same context in the
chromosome of all these strains and share high levels of identity among
these different species, suggesting that this signaling cascade is function-
ally conserved in Enterobacteriaeceae.
E. coli and Salmonella also have a LuxR homolog, SdiA (96), but do not
have a luxI gene and do not produce acyl-homoserine lactones (57, 91). The
E. coli sdiA was initially isolated as a regulator of the cell-division genes
ftsQAZ (96). However, the precise role of SdiA in quorum sensing was
elusive for several years until Michael et al. (57) recently reported that SdiA
is not sensing an autoinducer produced by Salmonella itself, but rather acyl-
homoserine lactones produced by other bacterial species. Kanamaru et al.
(39) reported that overexpression of SdiA from a high-copy-number plas-
mid in EHEC caused abnormal cell division, reduced adherence to cultured
epithelial cells, and reduced expression of the intimin adhesion protein and
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
164
the EspD protein, both of which are encoded within the LEE. However, no
sdiA EHEC mutant was constructed and tested; consequently, the pheno-
types observed could be artefacts due to the abnormally high expression of
SdiA. Because no E. coli genes from either EHEC or K-12 have yet been
demonstrated to be regulated by a single chromosomal copy of sdiA, Ahmer
(2) recently concluded that there are no confirmed members of an SdiA
regulon in this species.
LEE2 LEE3 LEE4
OM
IM
flhDC
fliA
fliC
fliFGHIJK
fliMNOPQR
fliE
flgMA motABcheWA
tartapEcheRBYZ
fliDST
Flagella and
motility
P
QseBC
AI-3
P
QseEF
Epinephrine
QseA
LEE5
ATP ADP
P-AI2?
AI-2
LsrB
LEE1
Ler
OM
IM
QseD
grlRA
-
Figure 7.3. Model of the AI-3 quorum-sensing signaling cascade in EHEC. Both AI-3 and
epinephrine seem to be recognized by the same receptor, which is probably in the outer
membrane of the bacterium, owing to the non-polar nature of both signals. These signals
might be imported to the periplasmic space where they interact with two major sensor
kinases. QseC might be the sensor kinase transducing these signals towards activation of
the flagellar regulon, whereas QseE might transduce these signals to activate transcription
of the LEE genes. QseC phosphorylates the QseB response regulator, which binds to the
promoter of flhDC(encoding the flagellar master regulators FlhDC) to activate expression of
the flagellar regulon. QseB also binds to its own promoter to positively autoregulate its own
transcription. QseE is the sensor kinase and its predicted response regulator is QseF. At
what levels QseF regulates transcriptionof the LEEgenes remains to be established. QseAis
one of the transcriptional factors involved in the regulation of ler (LEE1) transcription at two
levels, by binding and activating transcription of LEE1 and by activating transcription of the
grlRA operon, where GrlA and GrlR positively and negatively regulate expression of ler,
respectively. Then, in a cascade fashion, Ler activates transcription of the other LEE genes.
QseD is a second LysR-like regulator, involved in modulating expression of the LEE and
flagellar genes. EHECalso possess an lsr operon involved in recognition and uptake of AI-2;
however, the role of AI-2 signaling in EHEC remains to be addressed.
165
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
QUORUM SENSING IN EPEC
EPEC colonizes the proximal small intestine and causes profuse and
persistent watery diarrhea lasting up to 120 days in infants (18, 68). EPEC
pathogenesis has several steps. First the bacteria adhere to the intestinal
epithelial cells, probably through the EspA filament secreted by the LEE-
encoded TTSS (32, 45). Then, Tir is translocated through the LEE-encoded
type III secretion system and inserts itself into the mammalian cell mem-
brane, where it serves as the intimin receptor, allowing the intimate attach-
ment characteristic of AE lesion formation (40). Other EPEC cells then
interact with each other, forming large microcolonies (32). The successful
formation of these microcolonies requires the bundle-forming pili (BFP)
and flagella (23, 24). The present knowledge about EPEC pathogenesis
suggests that expression of EPEC virulence genes is dependent upon the
concerted action of several regulatory factors.
One of the hallmarks of EPEC is its characteristic adherence to epithe-
lial cells, forming microcolonies, usually referred to as localized adherence
(LA) (71). EPEC produces a type IV pilus called a bundle-forming pilus
(BFP), which is responsible for interbacterial interactions, leading to micro-
colony formation (23). Recent work from Giron and collaborators (24)
suggests that EPEC flagella are also involved in adhesion and are essential
for microcolony formation.
EPEC contains a large plasmid, referred to as the EPEC adherence
factor (EAF) plasmid. The EAF plasmid encodes a regulator of virulence
genes called Per (Plasmid-Encoded Regulator) consisting of three ORFs:
perA, perB and perC. PerA is an AraC homolog (26) and activates the
expression of the bfp operon encoding the bundle-forming pilus (95). The
per loci also activate the expression of ler (LEE-encoded regulator, the first
gene in the LEE1 operon), which activates expression of the LEE2, LEE3,
LEE5, and LEE4 operons in EPEC in a regulatory cascade (29, 55, 84).
Transcription of ler is also regulated by IHF (21), Fis (25), and BipA (27).
Transcription of per is auto-activated by Per (53) and modulated by
GadX (76).
It must be noted that quorum-sensing regulation differs between
EPEC and EHEC. EPEC colonizes the proximal small intestine, which is
thought to have very few or no resident flora. Therefore, whereas quorum
sensing is primarily an interspecies signaling system during EHEC infec-
tion, it seems to be used for intraspecies signaling during EPEC infection.
In contrast to EHEC, type III secretion in EPEC is diminished, but never
abrogated, in a luxS mutant (85, 87). This differential regulation can be
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
166
explained by the additional control of the LEE genes through Per, which is
absent in EHEC (26, 55). In EPEC, GadX also represses transcription of the
LEE genes through Per in acid pH (possibly when EPEC is crossing the
stomach) and activates their transcription in alkaline pH (possibly when
EPEC reaches the small intestine) (76). EPEC has to coordinate transcrip-
tion of the LEE genes with microcolony formation. This is when quorum-
sensing regulation may play an active role. Furthermore, flagellation and
motility are also altered in both luxS and qseA mutants. Disruption of
quorum-sensing signaling affects expression of the LEE genes, BFP, and
the flagellar regulon, thereby interfering with microcolony formation and
adherence to epithelial cells (79). As a result, luxS and qseA mutants form
smaller microcolonies and adhere two and four orders of magnitude,
respectively, less than the wild-type strain to cultured epithelial cells. In
EPEC, quorum sensing is probably involved in the spatialtemporal reg-
ulation of virulence genes, allowing successful colonization of the host.
Microcolony formation is one of the first steps towards biofilm devel-
opment. Inasmuch as EPEC pathogenesis is modulated by a quorum-
sensing regulatory mechanism, and EPEC adheres to epithelial cells to
form microcolonies, it can be hypothesized that EPEC may be forming a
biofilm during infection. The observation that the luxS quorum-sensing
system in EPEC regulates expression of antigen 43, type 1 fimbriae, and
flagella (C. G. Moreira and V. Sperandio, unpublished results), structures
that have been extensively associated with biofilm formation (9, 38, 44, 63,
74, 75, 97), further supports this hypothesis. Localized adhesion could be a
step towards biofilm maturation in EPEC pathogenesis, especially given
the observation that espA, bfp, and fliC mutants are altered for biofilm
formation compared with the wild-type strain (C. G. Moreira and
V. Sperandio, unpublished results). In conclusion, given that EPEC patho-
genesis is controlled by a quorum-sensing regulatory mechanism, and
EPEC adheres in the small intestine to form microcolonies, it could be
hypothesized that EPEC forms a biofilm in the small intestine. This
would be an explanation for the persistent diarrhea associated with EPEC
infections (17).
CONCLUDING REMARKS
Treatment of EHEC infections with conventional antimicrobials is
highly controversial, because it is well documented that antimicrobials
activate the Stx phage to enter the lytic cycle, thereby producing and
releasing Shiga toxin (41, 42). There are now preliminary data (87)
167
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
indicating that b-adrenergic antagonists, such as propranolol, can inhibit
the entire signaling cascade in EHEC, rendering it unable to induce flagel-
lation, motility, and AE lesion formation in response to either AI-3 and/or
epinephrine/norepinephrine (87). These results thus suggest an exciting
possible alternative for the treatment of EHECinfections by using b-adrenergic
antagonists. In addition, once the AI-3 structure is solved, it will allow the
design of antagonists to AI-3. These studies may help generate a whole new
class of antimicrobials that can block both AI-3 and epinephrine signaling to
bacterial pathogens. Finally, these antimicrobials will be useful not only against
EHEC but possibly also against other pathogens such as enteropathogenic
E. coli (EPEC), Salmonella, Shigella, and Yersinia pestis, all of which harbor this
signaling cascade.
REFERENCES
1 Adler, J. and B. Templeton 1967. The effect of environmental conditions on the
motility of Escherichia coli. J. Gen. Microbiol. 46: 17584.
2 Ahmer, B. M. 2004. Cell-to-cell signaling in Escherichia coli and Salmonella
enterica. Molec. Microbiol. 52: 93345.
3 Bijlsma, J. J. and E. A. Groisman 2003. Making informed decisions: regulatory
interactions between two-component systems. Trends Microbiol. 11: 35966.
4 Burton, C. L., S. R. Chhabra, S. Swift et al. 2002. The growth response of
Escherichia coli to neurotransmitters and related catecholamine drugs requires
a functional enterobactin biosynthesis and uptake system. Infect. Immun.
70: 591323.
5 Bustamante, V. H., F. J. Santana, E. Calva and J. L. Puente 2001. Transcriptional
regulation of type III secretion genes in enteropathogenic Escherichia coli: Ler
antagonizes H-NS-dependent repression. Molec. Microbiol. 39: 66478.
quorum-sensing signal containing boron. Nature 415: 5459.
7 Chilcott, G. S. and K. T. Hughes 2000. Coupling of flagellar gene expression to
flagellar assembly in Salmonella enterica serovar typhimurium and Escherichia coli.
Microbiol. Molec. Biol. Rev. 64: 694708.
8 Clarke, M. B. and V. Sperandio 2005. Transcriptional autoregulation by quorum
sensing E. coli regulators B and C (QseBC) in enterohemorrhagic E. coli (EHEC).
(in press.)
9 Danese, P. N., L. A. Pratt, S. L. Dove and R. Kolter 2000. The outer membrane
protein, antigen 43, mediates cell-to-cell interactions within Escherichia coli bio-
films. Molec. Microbiol. 37: 42432.
10 Day, W. A., Jr. and A. T. Maurelli 2001. Shigella flexneri LuxS quorum-sensing
system modulates virB expression but is not essential for virulence. Infect.
Immun. 69: 1523.
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
168
relationships. Infect. Immun. 68: 483949.
12 DeLisa, M. P., C. F. Wu, L. Wang, J. J. Valdes and W. E. Bentley 2001. DNA
microarray-based identification of genes controlled by autoinducer 2-stimulated
quorum sensing in Escherichia coli. J. Bacteriol. 183: 523947.
13 Deng, W., J. L. Puente, S. Gruenheid et al. 2004. Dissecting virulence: systematic
and functional analyses of a pathogenicity island. Proc. Natn. Acad. Sci.USA 101:
3597602.
14 Eisenhofer, G., A. Aneman, P. Friberg et al. 1997. Substantial production of dop-
amine in the human gastrointestinal tract. J. Clin. Endocrinol. Metab. 82: 386471.
15 Elliott, S. J., V. Sperandio, J. A. Giron et al. 2000. The locus of enterocyte
effacement (LEE)-encoded regulator controls expression of both LEE- and non-
LEE-encoded virulence factors in enteropathogenic and enterohemorrhagic
Escherichia coli. Infect. Immun. 68: 611526.
16 Elliott, S. J., L. A. Wainwright, T. K. McDaniel et al. 1998. The complete sequence
of the locus of enterocyte effacement (LEE) fromenteropathogenic Escherichia coli
E2348/69. Molec. Microbiol. 28: 14.
17 Fagundes Neto U., L. G. Schmitz and I. Scaletsky 1995. Clinical and epidemio-
logical characteristics of acute diarrhea by classical enteropathogenic Escherichia
coli. Rev. Assoc. Med. Bras. 41: 25965.
18 Fagundes-Neto, U. 1996. Enteropathogenic Escherichia coli infection in infants:
clinical aspects and small bowel morphological alterations. Rev. Microbiol. 27:
11719.
19 Freestone, P. P., R. D. Haigh, P. H. Williams and M. Lyte 1999. Stimulation of
bacterial growth by heat-stable, norepinephrine-induced autoinducers. FEMS
Microbiol. Lett. 172: 5360.
20 Freestone, P. P., M. Lyte, C. P. Neal et al. 2000. The mammalian neuroendocrine
hormone norepinephrine supplies iron for bacterial growth in the presence of
transferrin or lactoferrin. J. Bacteriol. 182: 60918.
21 Friedberg, D., T. Umanski, Y. Fang and I. Rosenshine 1999. Hierarchy in the
expression of the locus of enterocyte effacement genes of enteropathogenic
Escherichia coli. Molec. Microbiol. 34: 94152.
22 Gillen, K. L. and K. T. Hughes 1991. Molecular characterization of flgM, a gene
encoding a negative regulator of flagellin synthesis in Salmonella typhimurium.
23 Giron, J. A., A. S. Ho and G. K. Schoolnik 1991. An inducible bundle-forming
pilus of enteropathogenic Escherichia coli. Science 254: 71013.
24 Giron, J. A., A. G. Torres, E. Freer and J. B. Kaper 2002. The flagella of entero-
pathogenic Escherichia coli mediate adherence to epithelial cells. Molec. Microbiol.
44: 36179.
25 Goldberg, M. D., M. Johnson, J. C. Hinton and P. H. Williams 2001. Role of the
nucleoid-associated protein Fis in the regulation of virulence properties of entero-
pathogenic Escherichia coli. Molec. Microbiol. 41: 54959.
169
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
26 Gomez-Duarte, O. G. and J. B. Kaper 1995. A plasmid-encoded regulatory region
activates chromosomal eaeA expression in enteropathogenic Escherichia coli.
27 Grant, A. J., M. Farris, P. Alefounder et al. 2003. Co-ordination of pathogenicity
island expression by the BipA GTPase in enteropathogenic Escherichia coli
(EPEC). Molec. Microbiol. 48: 50721.
28 Gunn, J. S. and S. I. Miller 1996. PhoP-PhoQ activates transcription of pmrAB,
encoding a two-component regulatory system involved in Salmonella typhimur-
ium antimicrobial peptide resistance. J. Bacteriol. 178: 685764.
29 Haack, K. R., C. L. Robinson, K. J. Miller, J. W. Fowlkes and J. L. Mellies 2003.
Interaction of Ler at the LEE5 (tir) operon of enteropathogenic Escherichia coli.
30 Henderson, B. W. M., R. McNab and A. J. Lax 2000. Prokaryotic and eukaryotic
signaling mechanisms. In B. W. M. Henderson, R. McNab and A. J. Lax (eds),
Cellular Microbiology: Bacteria-Host Interactions in Health and Disease, 1st edn.,
pp. 89162, West Sussex, England: John Wiley and Sons.
31 Henke, J. M. and B. L. Bassler 2004. Quorumsensing regulates type III secretion
in Vibrio harveyi and Vibrio parahaemolyticus. J. Bacteriol. 186: 3794805.
32 Hicks, S., G. Frankel, J. B. Kaper, G. Dougan and A. D. Phillips 1998. Role of
intimin and bundle-forming pili in enteropathogenic Escherichia coli adhesion to
pediatric intestinal tissue in vitro. Infect. Immun. 66: 15708.
33 Hoffer, S. M., H. V. Westerhoff, K. J. Hellingwerf, P. W. Postma and
J. Tommassen 2001. Autoamplification of a two-component regulatory system
results in learning behavior. J. Bacteriol. 183: 491417.
34 Hooper, L. V. and J. I. Gordon 2001. Commensal host-bacterial relationships in
the gut. Science 292: 111518.
35 Horger, S., G. Schultheiss and M. Diener 1998. Segment-specific effects of
epinephrine on ion transport in the colon of the rat. Am. J. Physiol. 275:
G136776.
36 Hughes, K. T., K. L. Gillen, M. J. Semon and J. E. Karlinsey 1993. Sensing
structural intermediates in bacterial flagellar assembly by export of a negative
regulator. Science 262: 127780.
37 Jarvis, K. G., J. A. Giron, A. E. Jerse et al. 1995. Enteropathogenic Escherichia coli
contains a putative type III secretion system necessary for the export of proteins
involved in attaching and effacing lesion formation. Proc. Natn. Acad. Sci.USA
92: 79968000.
38 Jones, K. and S. B. Bradshaw 1996. Biofilm formation by the enterobacteriaceae:
a comparison between Salmonella enteritidis, Escherichia coli and a nitrogen-fixing
strain of Klebsiella pneumoniae. J. Appl. Bacteriol. 80: 45864.
39 Kanamaru, K., I. Tatsuno, T. Tobe and C. Sasakawa 2000. SdiA, an Escherichia
coli homologue of quorum-sensing regulators, controls the expression of viru-
lence factors in enterohaemorrhagic Escherichia coli O157:H7. Molec. Microbiol.
38: 80516.
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
170
40 Kenny, B., R. DeVinney, M. Stein et al. 1997. Enteropathogenic E. coli (EPEC)
transfers its receptor for intimate adherence into mammalian cells. Cell 91:
51120.
41 Kimmitt, P. T., C. R. Harwood and M. R. Barer 1999. Induction of type 2 Shiga
toxin synthesis in Escherichia coli O157 by 4-quinolones. Lancet 353: 15889.
42 Kimmitt, P. T., C. R. Harwood and M. R. Barer 2000. Toxin gene expression by
Shiga toxin-producing Escherichia coli: the role of antibiotics and the bacterial SOS
response. Emerg. Infect. Dis. 6: 45865.
43 Kinney, K. S., C. E. Austin, D. S. Morton and G. Sonnenfeld 2000.
Norepinephrine as a growth stimulating factor in bacteria mechanistic studies.
Life Sci. 67: 307585.
44 Kjaergaard, K., M. A. Schembri, H. Hasman and P. Klemm 2000. Antigen 43
from Escherichia coli induces inter- and intraspecies cell aggregation and changes
in colony morphology of Pseudomonas fluorescens. J. Bacteriol. 182: 478996.
45 Knutton, S., I. Rosenshine, M. J. Pallen et al. 1998. A novel EspA-associated
surface organelle of enteropathogenic Escherichia coli involved in protein trans-
location into epithelial cells. EMBO J. 17: 216676.
46 Kutsukake, K., Y. Ohya and T. Iino 1990. Transcriptional analysis of the flagellar
regulon of Salmonella typhimurium. J. Bacteriol. 172: 7417.
47 Lamont, I. L., P. A. Beare, U. Ochsner, A. I. Vasil and M. L. Vasil 2002.
Siderophore-mediated signaling regulates virulence factor production in
Pseudomonas aeruginosa. Proc. Natn. Acad. Sci.USA 99: 70727.
48 Liu, X. and P. Matsumura 1994. The FlhD/FlhC complex, a transcriptional
activator of the Escherichia coli flagellar class II operons. J. Bacteriol. 176: 734551.
49 Lyon, W. R., J. C. Madden, J. C. Levin, J. L. Stein and M. G. Caparon 2001.
Mutation of luxS affects growth and virulence factor expression in Streptococcus
pyogenes. Molec. Microbiol. 42: 14557.
50 Lyte, M., B. P. Arulanandam and C. D. Frank. 1996. Production of Shiga-like
toxins by Escherichia coli O157:H7 can be influenced by the neuroendocrine
hormone norepinephrine. J. Lab. Clin. Med. 128: 3928.
51 Lyte, M., A. K. Erickson, B. P. Arulanandam et al. 1997. Norepinephrine-induced
expression of the K99 pilus adhesin of enterotoxigenic Escherichia coli. Biochem.
Biophys. Res. Commun. 232: 6826.
52 Lyte, M., C. D. Frank and B. T. Green 1996. Production of an autoinducer of
growth by norepinephrine cultured Escherichia coli O157:H7. FEMS Microbiol.
Lett. 139: 1559.
53 Martinez-Laguna, Y., E. Calve and J. L. Puente 1999. Autoactivation and environ-
mental regulation of bfpT expression, the gene coding for the transcriptional
activator of bfpA in enteropathogenic Escherichia coli. Molec. Microbiol. 33: 15366.
54 McNab, R. M., 1996. Flagella and motility. In F. C. Neidhardt (ed.), Escherichia
coli and Salmonella, 2nd edn, vol.1, pp. 12345. Washington, DC: ASM Press.
55 Mellies, J. L., S. J. Elliott, V. Sperandio, M. S. Donnenberg and J. B. Kaper 1999.
The Per regulon of enteropathogenic Escherichia coli: identification of a regulatory
171
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
cascade and a novel transcriptional activator, the locus of enterocyte effacement
(LEE)-encoded regulator (Ler). Molec. Microbiol. 33: 296306.
56 Merkel, T. J., C. Barros and S. Stibitz 1998. Characterization of the bvgR locus of
Bordetella pertussis. J. Bacteriol. 180: 168290.
57 Michael, B., J. N. Smith, S. Swift, F. Heffron and B. M. Ahmer 2001. SdiA of
Salmonella enterica is a LuxR homolog that detects mixed microbial communities.
J. Bacteriol. 183: 573342.
recognizes a chemically distinct form of the bacterial quorum-sensing signal
AI-2. Molec. Cell 15: 67787
59 Nataro, J. P. and J. B. Kaper 1998. Diarrheagenic Escherichia coli. Clin. Microbiol.
Rev. 11: 142201.
60 Neely, M. N. and D. I. Friedman 1998. Functional and genetic analysis of
regulatory regions of coliphage H-19B: location of Shiga-like toxin and lysis
genes suggest a role for phage functions in toxin release. Molec. Microbiol.
28: 125567.
61 Nishimura, A. and Y. Hirota 1989. Acell division regulatory mechanismcontrols
the flagellar regulon in Escherichia coli. Molec. Gen. Genet. 216: 3406.
62 Perna, N. T., G. Plunkett III, V. Butland 2001. Genome sequence of enterohae-
morrhagic Escherichia coli O157:H7. Nature 409: 52933.
63 Pratt, L. A. and R. Kolter 1998. Genetic analysis of Escherichia coli biofilmformation:
roles of flagella, motility, chemotaxis and type I pili. Molec. Microbiol. 30: 28593.
64 Pruss, B. M., J. W. Campbell, T. K. Van Dyk et al. 2003. FlhD/FlhC is a regulator
of anaerobic respiration and the Entner-Doudoroff pathway through induction of
the methyl-accepting chemotaxis protein Aer. J. Bacteriol. 185: 53443.
65 Pruss, B. M., D. Markovic and P. Matsumura 1997. The Escherichia coli flagellar
transcriptional activator flhD regulates cell division through induction of the acid
response gene cadA. J. Bacteriol. 179: 381821.
66 Pruss, B. M. and P. Matsumura 1996. A regulator of the flagellar regulon of
Escherichia coli, flhD, also affects cell division. J. Bacteriol. 178: 66874.
67 Raivio, T. L., D. L. Popkin and T. J. Silhavy 1999. The Cpx envelope stress response
is controlled by amplification and feedback inhibition. J. Bacteriol. 181: 526372.
68 Rothbaum, R., A. J. McAdams, R. Giannella and J. C. Partin 1982. A clinico-
pathologic study of enterocyte-adherent Escherichia coli: a cause of protracted
diarrhea in infants. Gastroenterology 83: 44154.
69 Roy, C. R., J. F. Miller and S. Falkow 1990. Autogenous regulation of the
Bordetella pertussis bvgABC operon. Proc. Natn. Acad. Sci.USA 87: 37637.
70 Sanchez-SanMartin, C., V. H. Bustamante, E. Calva and J. L. Puente 2001.
Transcriptional regulation of the orf19 gene and the tir-cesT-eae operon of entero-
pathogenic Escherichia coli. J. Bacteriol. 183: 282333.
71 Scaletsky, I. C., M. L. Silva and L. R. Trabulsi 1984. Distinctive patterns of
adherence of enteropathogenic Escherichia coli to HeLa cells. Infect. Immun.
45: 5346.
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
172
72 Schauder, S. and B. L. Bassler 2001. The languages of bacteria. Genes Dev.
15: 146880.
74 Schembri, M. A., K. Kjaergaard and P. Klemm 2003. Global gene expression in
Escherichia coli biofilms. Molec. Microbiol. 48: 25367.
75 Schembri, M. A. and P. Klemm 2001. Biofilm formation in a hydrodynamic
environment by novel fimH variants and ramifications for virulence. Infect.
Immun. 69: 13228.
76 Shin, S., M. P. Castanie-Cornet, J. W. Foster et al. 2001. An activator of glutamate
decarboxylase genes regulates the expression of enteropathogenic Escherichia coli
virulence genes through control of the plasmid-encoded regulator, Per. Molec.
77 Shin, S. and C. Park 1995. Modulation of flagellar expression in Escherichia coli by
acetyl phosphate and the osmoregulator OmpR. J. Bacteriol. 177: 4696702.
78 Silverman, M. and M. Simon 1974. Characterization of Escherichia coli flagellar
mutants that are insensitive to catabolite repression. J. Bacteriol. 120: 1196203.
79 Sircili, M. P., M. Walters, L. Trabulsi and V. Sperandio 2004. Modulation of
enteropathogenic E. coli (EPEC) virulence by quorum sensing. Infect. Immun. 72:
232937.
81 Soncini, F. C. and E. A. Groisman 1996. Two-component regulatory systems
can interact to process multiple environmental signals. J. Bacteriol. 178:
6796801.
82 Soncini, F. C., E. G. Vescovi and E. A. Groisman 1995. Transcriptional autoregu-
lation of the Salmonella typhimurium phoPQ operon. J. Bacteriol. 177: 436471.
83 Sperandio, V., C. C. Li and J. B. Kaper 2002. Quorum-sensing Escherichia coli
regulator A(QseA): a regulator of the LysRfamily involved in the regulation of the
LEE pathogenicity island in enterohemorrhagic Escherichia coli. Infect. Immun.
70: 308593.
84 Sperandio, V., J. L. Mellies, R. M. Delahay et al. 2000. Activation of enteropatho-
genic Escherichia coli (EPEC) LEE2 and LEE3 operons by Ler. Molec. Microbiol.
38: 78193.
85 Sperandio, V., J. L. Mellies, W. Nguyen, S. Shin and J. B. Kaper 1999. Quorum
sensing controls expression of the type III secretion gene transcription and
protein secretion in enterohemorrhagic and enteropathogenic Escherichia coli.
Proc. Natn. Acad. Sci.USA 96: 15196201.
86 Sperandio, V., A. G. Torres, J. A. Giron and J. B. Kaper 2001. Quorumsensing is a
global regulatory mechanism in enterohemorrhagic Escherichia coli O157:H7.
J. Bacteriol. 183: 518797.
87 Sperandio, V., A. G. Torres, B. Jarvis, J. P. Nataro and J. B. Kaper 2003. Bacteria-
host communication: the language of hormones. Proc. Natn. Acad. Sci. USA
100: 89516.
173
L
u
x
S
R
E
G
U
L
A
T
I
O
N
O
F
E
.
C
O
L
I
V
I
R
U
L
E
N
C
E
88 Sperandio, V., A. G. Torres and J. B. Kaper 2002. Quorum sensing Escherichia
coli regulators B and C (QseBC): a novel two-component regulatory system
involved in the regulation of flagella and motility by quorum sensing in E. coli.
89 Surette, M. G. and B. L. Bassler 1998. Quorum sensing in Escherichia coli and
Salmonella typhimurium. Proc. Natn. Acad. Sci.USA 95: 704650.
90 Surette, M. G., M. B. Miller and B. L. Bassler 1999. Quorumsensing inEscherichia
coli, Salmonella typhimurium, and Vibrio harveyi: a newfamily of genes responsible
for autoinducer production. Proc. Natn. Acad. Sci.USA 96: 163944.
91 Swift, S., M. J. Lynch, L. Fish et al. 1999. Quorum sensing-dependent regulation
and blockade of exoprotease production in Aeromonas hydrophila. Infect. Immun.
67: 51929.
93 Taga, M. E., J. L. Semmelhack and B. L. Bassler 2001. The LuxS-dependent
autoinducer AI-2 controls the expression of an ABCtransporter that functions in
AI-2 uptake in Salmonella typhimurium. Molec. Microbiol. 42: 77793.
94 Tischler, A. D. and A. Camilli 2004. Cyclic diguanylate (c-di-GMP) regulates
Vibrio cholerae biofilm formation. Molec. Microbiol. 53: 85769.
95 Tobe, T., G. K. Schoolnik, I. Sohel, V. H. Bustamante and J. L. Puente 1996.
Cloning and characterization of bfpTVW, genes required for the transcriptional
activation of bfpA in enteropathogenic Escherichia coli. Molec. Microbiol. 21:
96375.
96 Wang, X. D., P. A. de Boer and L. I. Rothfield 1991. A factor that positively
regulates cell division by activating transcription of the major cluster of essential
cell division genes of Escherichia coli. EMBO J. 10: 336372.
97 Watnick, P. I., C. M. Lauriano, K. E. Klose, L. Croal and R. Kolter 2001. The
absence of a flagellum leads to altered colony morphology, biofilm development
and virulence in Vibrio cholerae O139. Molec. Microbiol. 39: 22335.
98 Winzer, K., K. R. Hardie, N. Burgess et al. 2002. LuxS: its role in central
metabolism and the in vitro synthesis of 4-hydroxy-5-methyl-3(2H)-furanone.
99 Wosten, M. M., L. F. Kox, S. Chamnongpol, F. C. Soncini and E. A. Groisman 2000.
A signal transduction system that responds to extracellular iron. Cell 103: 11325.
101 Yanagihara, S., S. Iyoda, K. Ohnishi, T. Iino and K. Kutsukake 1999. Structure
and transcriptional control of the flagellar master operon of Salmonella typhi-
murium. Genes. Genet. Syst. 74: 10511.
102 Yokota, T. and J. S. Gots 1970. Requirement of adenosine 3, 5-cyclic phosphate
for flagella formation in Escherichia coli and Salmonella typhimurium. J. Bacteriol.
103: 51316.
M
.
B
.
C
L
A
R
K
E
A
N
D
V
.
S
P
E
R
A
N
D
I
O
174
CHAPTER 8
Quorum sensing and cell-to-cell communication
in the dental biofilm
Donald R. Demuth
School of Dentistry, University of Louisville, KY, USA
Richard J. Lamont
Department of Oral Biology, University of Florida, Gainesville, FL, USA
INTRODUCTION
The microbial community that exists in the oral cavity is perhaps the
most accessible, complex and pathogenic of the naturally occurring human
biofilms. Over 500 different species of bacteria have been identified
in the mature biofilm that forms on tooth surfaces (38). This complex
community tenaciously adheres to and develops on the acquired salivary
pellicle, a conditioning filmof salivary proteins and glycoproteins adsorbed
to oral tissue surfaces. The initial colonizers of the salivary pellicle are
predominantly Gram-positive facultative anaerobes such as the strepto-
cocci; these organisms normally exist in commensal harmony with the
host. However, as the oral biofilm matures, there is a change in the micro-
bial composition, with an increasing presence of Gram-negative organ-
isms. The two most common oral diseases in humans, dental caries and
periodontal disease, arise from populational shifts in the biofilm in
response to a variety of host and/or environmental stimuli. This results
in over-representation of pathogenic organisms in the biofilm at afflicted
sites in the oral cavity. For example, excessive consumption of dietary
sucrose favors the overgrowth of highly fermentative acidophilic organisms
such as Streptococcus mutans. The acidic local environment generated by
these organisms promotes demineralization of the hydroxyapatite matrix of
enamel, thus increasing the risk of dental caries. In contrast, periodontal
disease is caused by a biofilm that thrives in the subgingival pocket and
induces a chronic inflammatory condition that results in the destruction of
the connective tissues and bone that support the teeth (23). This subgingi-
val biofilm comprises predominantly of Gram-negative obligate anaerobes,
including high numbers of periodontal pathogens such as Porphyromonas
175
gingivalis, Actinobacillus actinomycetemcomitans, Tannerella forsythensis,
Treponema spp., and Prevotella spp.
Although the composition of the dental biofilm may vary among
individuals and from site to site in the oral cavity, its development and
maturation follows an ordered progression of events. Temporally distinct
patterns of microbial colonization are consistently observed in biofilms on
all surfaces in the oral cavity. For example, the initial colonizers of the
salivary pellicle on the coronal tooth surface are principally Gram-posi-
tive bacteria such as the viridans streptococci, actinomyces and related
Gram-positive organisms. Establishment of these organisms facilitates
the subsequent colonization of additional actinomyces and related Gram-
positive rods, along with Gram-negative species such as Veillonella and
Fusobacterium nucleatum. Further maturation of the oral biofilm is char-
acterized by the colonization of additional Gram-negative anaerobes,
including the periodontal pathogens mentioned above (22). Spatially, one
of the most dramatic features of the oral biofilm is the presence of column-
like microcolonies of early colonizers during its early stages of development
(18). In addition, distinct clusters of organisms are often recovered from
symptomatic sites of periodontitis patients, suggesting that biofilms com-
prising defined subsets of bacteria are associated with the initiation and
progression of oral disease (41, 42). These reproducible characteristics of
biofilm development and the populational shifts that can occur in the
mature biofilm have long been taken as a priori evidence for the occurrence
of intra- and/or interspecies communication among the bacterial residents
of dental plaque. However, until recently, the identification of these puta-
tive signaling pathways and the outcomes of bacterial communication have
been elusive. This chapter will focus on recent work that elucidates two
mechanisms of interspecies communication among the oral bacteria, cell-
contact-dependent communication and intra- and interspecies signaling
that is mediated by a soluble quorum sensing signal related to autoinducer
2, the cyclic borate diester produced by Vibrio harveyi.
CONTACT-DEPENDENT SIGNALING IN ORAL BACTERIA
The survival and persistence of bacteria in the human oral cavity
requires that the organisms rapidly adhere to a tissue surface and adapt
to growth in a biofilm (22). Indeed, recent studies suggest that the act of
adhering to the salivary pellicle can alter patterns of gene expression and
initiate a response in the bacterium that may facilitate biofilm growth.
Du and Kolenbrander (11) showed that the interaction of Streptococcus
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
176
gordonii with saliva induced the expression of several genes, including
sspAB, and decreased expression of various metabolic genes encoding
glucose kinase, dihydropiccolinate synthetase, and an oligopeptide-binding
lipoprotein. Interestingly, sspA and sspB encode streptococcal receptors for
a mucin-like glycoprotein, salivary gp-340 (10). The interaction of SspAand
SspB with gp-340 promotes colonization by mediating the adherence of
streptococci to the salivary pellicle. SspA and SspB also promote coaggre-
gative interactions between S. gordonii and actinomyces that may facilitate
the subsequent colonization of streptococcal substrates by Actinomyces
spp., leading to increased complexity in the developing oral biofilm (12).
More recently, Zhang et al. (54) showed that streptococci that were bound to
saliva-coated hydroxyapatite (sHA) beads exhibited increased expression of
a two-component system, termed bfrAB. Streptococcal cells that were
unable to express bfrAB were defective in colonizing sHA and did not
form biofilms on polystyrene surfaces. These results suggest that signal
transduction mediated by the bfrAB two-component system may control
important aspects of streptococcal adherence and biofilmgrowth. Although
the gene targets that are regulated by bfrAB have not yet been identified, the
studies above suggest that oral streptococci may utilize specific signal
transduction pathways to respond to saliva or other extracellular signals
in order to regulate the expression of genes required for adherence to the
salivary pellicle and/or for adapting to biofilm growth.
Maturation of the oral biofilm inevitably involves numerous intimate
interactions among the various constituent microbial species. Many studies
have shown that these interactions are not passive events but can serve
specific functions, either synergistic or antagonistic in nature. In some
cases, the interbacterial interactions are driven by physiological compat-
ibility between the interacting organisms. Thus, it becomes beneficial for
bacteria to assemble into groups that can utilize a complex substrate to
maximum efficiency. For example, oral streptococci and actinomyces pro-
duce lactate as the end product of fermentation of carbohydrates. In turn,
the Gram-negative coccus Veillonella, which adheres tightly to oral strepto-
cocci, utilizes lactate as a fermentable substrate (35). A similar nutritional
cross-feeding occurs through the close association of Porphyromonas
gingivalis with Treponema denticola. Here, P. gingivalis utilizes succinate
that is produced by T. denticola, whereas growth of T. denticola is in turn
enhanced by isobutyric acids generated by P. gingivalis. Althoughthese interac-
tions are well documented, it is unlikely that the cellular responses of the
participating organisms involve specific cell-to-cell communication that
extends beyond the physiological adaptation to nutrient availability.
177
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
S. gordonii P. gingivalis
FimA
Mfa1 SspAB
saliva-coated
surface
BfrAB
(a )
(b)
GAPDH
(c)
Figure 8.1. Contact-dependent and quorum sensing signaling in the formation of a
Porphyromonas gingivalisStreptococcus gordonii mixed-species biofilm. (a) S. gordonii cells
adhere to saliva-coated surfaces of oral tissues and the teeth. Contact with saliva induces
changes in gene expression within S. gordonii that are mediated at least in part by the
two-component system BfrAB (54). One of the genes that is induced on streptococcal
contact with saliva encodes the Ssp surface adhesin (11). The Ssp polypeptide functions as a
ligand for the attachment of P. gingivalis (9). Thus, the adherent S. gordonii cells serve as a
substratum for the subsequent colonization of the developing biofilm by P. gingivalis cells.
The initial association of P. gingivalis with S. gordonii (b) is complex and requires the
interaction of the major fimbrial subunit protein (FimA) with cell-surface glyceraldehyde-3-
phosphate dehydrogenase on S. gordonii (29). This interaction does not promote the
development of P. gingivalis microcolonies in the absence of the minor fimbriae.
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
178
In other cases, interbacterial cell-to-cell adherence among oral bacteria
has been shown to elicit a response that does appear to extend beyond a
simple adaptation to nutrient availability (see Figure 8.1). For example,
using confocal microscopy, Cook et al. (8) showed that P. gingivalis adheres
to S. gordonii cells that had previously colonized a saliva-coated coverglass
in a flow chamber. P. gingivalis subsequently developed into biofilms con-
sisting of microcolonies approximately 6080mm in depth. Such biofilms
were not observed when P. gingivalis was incubated with the saliva-
conditioned surface in the absence of S. gordonii or when P. gingivalis was
exposed to a saliva-coated surface colonized by mutans streptococci (9).
These studies indicate that the interaction of P. gingivalis with oral strepto-
cocci exhibits a high degree of selectivity and suggest that communication
occurs between these organisms that may facilitate the adaptation of
P. gingivalis to a biofilm lifestyle. Indeed, preliminary microarray analyses
of the P. gingivalis transcriptome has identified over 40 genes that are
differentially regulated in P. gingivalis upon contact with S. gordonii
(R. J. Lamont, unpublished data). These genes span several functional
groups, indicating that a broad change in the physiology of the organism
occurs as it prepares for a biofilm mode of existence. Furthermore, there is
no apparent overlap between these contact-regulated genes and the genes
controlled by the LuxS-dependent quorum sensing system during biofilm
growth (see below). Collectively, these data indicate that the regulation of
biofilm-related genes in P. gingivalis is complex and can occur at several
levels, which may correspond to distinct developmental steps in biofilm
initiation and maturation.
Contact-dependent signaling may also be involved in antagonistic
interactions that occur between organisms in the oral biofilm, as exempli-
fied by P. gingivalis adherence to Streptococcus cristatus. Studies in flow
chambers similar to those described above suggest that the interaction of
P. gingivalis and S. cristatus is of low affinity and that P. gingivalis cells are
easily washed off the streptococcal substratumby relatively lowshear forces
(flow rates). As a result, mixed P. gingivalisS. cristatus biofilms do not
develop to the extent observed with P. gingivalis and S. gordonii (51).
Figure 8.1. (cont.)
Engagement of the P. gingivalis minor fimbrial subunit (Mfa1) with the streptococcal Ssp
polypeptide necessary for the formation of P. gingivalis micro-colonies (c). Adherence of
P. gingivalis and S. gordonii (b) does not require the quorumsensing signal AI-2, whereas the
development of microcolonies (c) occurs only in the presence of AI-2 (31). Updated and
reproduced with permission from Lamont et al. 2002 (24).
179
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
A mechanistic basis for these results arises from the observation that the
initial contact between P. gingivalis and S. cristatus results in a significant
decrease in the transcription of fimA, encoding the major fimbrial adhesin
that mediates adherence of P. gingivalis to several streptococcal species (23).
Downregulation of the FimA adhesin would likely weaken the association
between P. gingivalis and S. cristatus, presumably by reducing the valency of
the interaction. Downregulation of fimA appears to be independent of the
fimRS two-component system that has been implicated in the regulation of
fimA expression (21). Instead, this process is initiated by the interaction of
P. gingivalis cells with a protein of approximately 60kDa on the surface
of S. cristatus. Xie et al. (52) found that contact of P. gingivalis with S. cristatus
regulates the expression of genes encoding two outer-membrane lipopro-
teins, pg2167 and pg2131, as well as ptpA, encoding a putative prolyl tripep-
tidyl peptidase, and pg0707, encoding a TonB-dependent outer-membrane
protein. However, genes encoding other P. gingivalis virulence factors, such
as the cysteine proteases, were not affected by interaction with S. cristatus.
Furthermore, Xie et al. (52) showed that S. cristatus-dependent regulation of
fimA expression did not occur in a null mutant of pg2167. Indeed, this
mutant exhibited constitutively high expression of FimA. This suggests
that pg2167 is involved in a regulatory pathway that may function to repress
fimA. In contrast, mutation of pg2131 did not influence the transcription of
fimA or the contact-dependent regulation of fimA. In this mutant, the
concentrations of the FimA polypeptide were significantly reduced. Thus,
PG2131 appears to be involved in the post-transcriptional processing and/or
transport of FimA (52). Interestingly, the primary sequences of PG2167 and
PG2131 exhibit significant similarity (e-64) and may be paralogs
(www.LANL). Further elucidation of their roles in the contact-dependent
regulation of fimA will likely reveal much about regulatory networks that
control the production of fimbriae in P. gingivalis in the context of a multi-
species biofilm.
QUORUM-SENSING-DEPENDENT COMMUNICATION AMONG
ORAL BACTERIA
Communication that is mediated by secreted diffusible molecules is
widespread among bacteria and regulates a number of important physio-
logical and virulence-related properties in these organisms. As highlighted
in the first several chapters of this volume, acyl-homoserine lactone (AHL)-
dependent quorum sensing systems are essential for aspects of virulence
and biofilmgrowth in a variety of organisms. Some of the first evidence that
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
180
soluble signal molecules may regulate the growth of oral bacteria arose
from the studies of Liljemark et al. (26). Measurements of the dynamics of
bacterial growth in developing dental biofilms that formed on implanted
enamel chips suggested that there was a rapid period of cell division during
the first 24 h of biofilm growth and that this period of rapid growth
accounted for 90% of the total biomass in the mature biofilm (26). In
addition, this period of rapid growth occurred in a cell-density-dependent
manner. Cell-free supernatants of medium in contact with biofilm bacteria
were shown to increase the incorporation of labeled thymidine on a per cell
per time basis when compared with control medium that was not exposed
to bacteria. This result suggested that conditioned medium from biofilm
cultures contained a soluble molecule (designated START) that influenced
the rate of bacterial cell division in the biofilm. Unfortunately, there are no
subsequent studies that characterize the START component and determine
its relationship to known quorum-sensing signal molecules (e.g. AHLs
or AI-2).
More recently, Frias et al. (15) examined 33 strains representing 16
different oral bacterial species for the production of quorum-sensing sig-
nals, by using sensor 1- or sensor 2-deficient Vibrio harveyi reporter organ-
isms (which cannot respond to AHL or AI-2 signals, respectively). None of
the oral bacterial strains that were tested induced bioluminescence of
V. harveyi B886 (sensor 1
sensor 2
), suggesting that AHL-like autoindu-

cers are not produced by these organisms. This result is consistent with the
studies of Whittaker et al. (47), who utilized three different reporter systems
but failed to detect the production of AHL-like autoinducers from Gram-
negative oral bacteria. In addition, the recent availability of complete gen-
ome sequences of several oral bacteria has facilitated analyses in silico to
identify genes that are similar to those known to be essential for the
production of, detection of, and response to AHL signals in other organ-
isms, such as homologs of LuxR, LuxI, SdiA. These studies have also failed
to demonstrate the presence of AHL quorum sensing systems in oral
bacteria. Thus, the available data suggest that quorum sensing systems
dependent on AHL signals may not be common among organisms that
populate the dental biofilm. However, it should be noted that AHL auto-
inducers that direct species-specific signaling may not be readily detected
by the available reporter strains. Furthermore, genome sequences are
currently available for only a small number of the bacterial species that
exist in the dental biofilm. Thus, it is possible that AHL-dependent quorum
sensing systems that are species-specific may exist in the dental biofilm but
cannot be identified by using the current detection methodologies.
181
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
QUORUM SENSING BY AUTOINDUCER 2 IN ORAL
BACTERIA
In contrast to the results obtained by using the sensor 2-deficient
V. harveyi reporter, Frias et al. (15) showed that conditioned medium obtained
from several oral pathogens, e.g. P. gingivalis, Prevotella intermedia, and
Fusobacterium nucleatum, induced bioluminescence of V. harveyi BB170 (sen-
sor 1
sensor 2
). This suggests that these organisms produce a signal that

stimulates the autoinducer 2 (AI-2) quorum sensing system of V. harveyi.
Surette et al. (44) had previously shown that the luxS gene of V. harveyi
encoded the polypeptide that functions to produce AI-2 and that genes related
to luxSwere present inthe genomes of Escherichia coli, Salmonella typhimurium,
and a variety of other Gram-negative and Gram-positive organisms.
Subsequent to the functional bioluminescence data of Frias et al. (15),
genes encoding polypeptides related to V. harveyi LuxS were identified
from the complete or partial genome sequences of the Gram-negative oral
pathogens P. gingivalis, A. actinomycetemcomitans, and F. nucleatum (4, 7, 13)
as well as several species of oral streptococcus (2, 33, 34, 46). Each of these
genes, when introduced into the LuxS-deficient E. coli strain DH5a, gener-
ated a recombinant organism that produced and secreted a soluble signal
capable of inducing bioluminscence in V. harveyi. Thus, although the oral
organisms examined to date appear to lack AHL-based quorum sensing
pathways, several oral pathogens clearly produce and respond to AI-2.
The widespread distribution of LuxS has led to speculation that AI-2 is a
universal, or species-non-specific, signal. AI-2 has been suggested to promote
interspecies communication in microbial communities or may report the total
bacterial biomass of a microbial community (36). Clearly, the ability to commu-
nicate across species barriers would be beneficial for oral organisms as they
thrive in a complex multi-species biofilm; indeed, AI-2 has been shown to be
required for the development of Porphyromonas gingivalis Streptococcus gordonii
biofilms (see below). However, the overall contribution of LuxS-dependent
signaling inthe development of the dental biofilmin vivo is not fully understood,
in part because the LuxS-dependent quorum-sensing regulons of oral bacteria
have not been fully defined. In addition, the AI-2 signal transduction cascade(s)
utilized by oral organisms such as Actinobacillus actinomycetemcomitans, or by
E.coli, may differ from the well-characterized signal transduction systems that
operate in Vibrio species (see below and Chapter 7). Thus, it is possible that the
physiological role of LuxS-dependent quorum sensing in oral bacteria, and in
other organisms that lack AHL-dependent quorum sensing pathways, may be
quite different fromthat of Vibrio. Indeed, the wide variety of bacterial functions
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
182
that appear to be regulated by AI-2, including diverse activities such as protein
secretion, iron uptake, and carbohydrate metabolism, suggests that different
organisms may utilize the AI-2 signaling pathway for specific physiological
requirements. The remainder of this chapter will describe the specific sets of
genes and the cellular processes that are regulated by LuxS-dependent quorum
sensing in the oral pathogens A. actinomycetemcomitans and P. gingivalis. In
addition, studies to identify the components of the signal transduction cascade
that mediates the response of oral bacteria to AI-2 will be discussed.
Is AI-2 a quorum sensing signal in oral bacteria?
In many organisms, the expression of luxS does not appear to be tran-
scriptionally regulated (49). However, Chung et al. (7) have shown that expres-
sion of luxS in P. gingivalis is environmentally controlled and varies inversely
with the osmolarity of the culture medium. Expression was greatest when cells
were cultured in medium with an osmolarity of approximately half human
physiological levels; it decreased significantly at physiological salt concentra-
tion. Thus, the production of LuxS and AI-2 by P. gingivalis may be maximal
under hypotonic conditions such as occur in saliva. In A. actinomycetemcomi-
tans, the amount of AI-2 secreted into the mediumis highest at mid-log phase
and decreases by approximately 50% when cultures reach stationary phase.
This is consistent with the kinetics of AI-2 production reported for many other
bacteria, including E. coli and Salmonella typhimurium (49). One explanation
for the decrease in AI-2 concentrations during later stages of growth is that the
signal may be degraded or internalized as cells approach and enter stationary
phase (45). Indeed, S. typhimurium has been shown to possess a LuxS-
regulated ABC transport system encoded by the lsrACDBFGE operon. This
transporter functions to actively import AI-2 into the cell (45). Once internalized,
AI-2 is modified by phosphorylation by LsrK (45). An operon corresponding
to lsrACDBFGE is also present in A. actinomycetemcomitans, suggesting that it
may also internalize and process AI-2 during the later stages of cell growth in
culture. These observations clearly contrast with the general paradigm of
quorum sensing, since the extracellular level of AI-2 is not maximal at the
highest A. actinomycetemcomitans cell density. These results (along with others
discussed in Chapter 6) have led to the suggestion that AI-2-dependent
signaling may not simply report bacterial cell density but may represent a
signal of the overall metabolic activity of a microbial community (36; see also
Chapter 6). Another interesting possibility is that different sets of responses
and behaviors are induced by extracellular and internalized AI-2 signals.
183
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
Physiologic role of LuxS-dependent signaling: regulation
of iron acquisition
Analyses of wild-type and LuxS-deficient strains of the oral pathogens
A. actinomycetemcomitans and P. gingivalis suggested that AI-2 may regulate
aspects of iron acquisition in both organisms. The growth rates of wild-type
and LuxS-deficient strains of A. actinomycetemcomitans are indistinguish-
able when cultured in iron-replete medium. In addition, wild-type
A. actinomycetemcomitans exhibits normal growth when cultured in the
presence of the ferric ion chelator EDDHA, indicating that wild-type cells
effectively compete with the chelator for iron in the culture medium. In
contrast, the luxS mutant fails to divide when cultured in the presence of
the chelator, but complementation of the mutant with a functional plasmid
borne copy of luxS restores normal growth under iron limitation. Lastly,
stunted cultures of the mutant strain cultured under iron limitation
resume normal growth when inoculated back into iron-replete medium
(14). Studies using real-time PCR revealed that the expression of several
genes involved in iron acquisition and intracellular storage are significantly
altered in the luxS mutant. Steady-state levels of afuA mRNA, encoding a
periplasmic ferric ion transport protein, and ftnAB, encoding the intracel-
lular iron storage protein ferritin, are reduced by 8-fold and more than
50-fold, respectively, in the luxS mutant. Furthermore, genes encoding
putative A. actinomycetemcomitans receptors for heme, transferrin, and
hemoglobin were also shown to be downregulated in the mutant strain,
albeit at more modest levels (2- to 3-fold). Interestingly, the expression of a
ferric citrate transport operon and a second operon encoding a putative
enterobactin receptor and an ABC-type transporter are upregulated (by 3- and
10-fold, respectively) in the luxS mutant (14). This pattern of regulation is
intriguing: it suggests that AI-2 may play a role in the adaptation of
A. actinomycetemcomitans to the biofilm environment in the host by facili-
tating a switch from the acquisition of ionic iron via chelators, e.g.
enterobactin-like siderophores, to the acquisition of iron from host proteins,
e.g. transferrin and hemoglobin. Consistent with this possibility, the
Neisseria meningitidis homolog of AfuA has been shown to represent the
periplasmic iron-binding polypeptide involved in the transport of iron
extracted from receptor-bound transferrin (16).
P. gingivalis obtains iron almost exclusively through the acquisition of
hemin; it utilizes multiple receptors and pathways to accomplish this task.
Indeed, the acquisition of hemin is essential for the expression of virulence
by P. gingivalis. A P. gingivalis luxS mutant grows poorly when cultured
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
184
under hemin-limiting conditions that support growth of the wild-type
organism. The mutant also exhibits delayed recovery in growth when
transferred back into hemin replete medium (R. J. Lamont, unpublished
data). Consistent with these observations, the LuxS-deficient strain exhibits
differential regulation (both up and down) of genes involved in various
aspects of hemin acquisition and uptake. For example, inactivation of LuxS
reduces the expression of the TonB-dependent hemin/hemoglobin recep-
tor HemR/HmuR, and of TonB itself (7). Expression of the major
P. gingivalis proteases Rgp and Kgp is also altered in the mutant strain
(4, 7); these proteases are thought to play a role in the acquisition of
hemin by degrading host hemin-sequestering proteins (43). In addition, a
reduction in hemagglutination titer has been reported in a LuxS mutant of
P. gingivalis (4). In contrast, the P. gingivalis homolog of HasF, a putative
component of the hemophore heme acquisition system, is upregulated in
the LuxS-deficient organism. The differential regulation of these hemin
acquisition pathways by AI-2 suggests that LuxS-dependent signaling may
function to fine-tune hemin uptake in P. gingivalis in response to con-
straints such as the total amount of hemin available and/or the types of
host hemin-containing molecule that are present in the local environment.
The studies described above suggest that LuxS-dependent signaling con-
trols aspects of ironacquisitionby the oral pathogens A. actinomycetemcomitans
and P. gingivalis. Indeed, increasing evidence suggests that bacterial cell-
signaling mechanisms are intimately associated with the pathways utilized
in the acquisition of iron by a variety of other organisms as well. For example,
in addition to the regulation of bioluminescence, the quorum sensing path-
ways of V. harveyi have been shown to regulate the production of a siderophore
via the response regulator LuxO (27). In addition, pyoveridine, a siderophore
produced by Pseudomonas aeruginosa, may itself function as a signaling mole-
cule that induces the expression of at least three virulence factors by altering
the activity of the extracytoplasmic sigma factor PvdS (1). Finally, the human
hormone norepinephrine has been previously suggested to possess sidero-
phore activity; Sperandio et al. have suggested that norepinephrine influences
quorum sensing in enterohemorrhagic E. coli (see Chapter 7).
LuxS-dependent regulation of virulence and biofilm
development
In many cases, the acquisition of iron is intimately involved in the expres-
sion of virulence by pathogenic organisms. However, there are conflicting data
on the role that LuxS-dependent signaling may play in regulating virulence. In
185
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
oral pathogens, this may be due in large part to the lack of a suitable animal
model systemthat accurately reflects the multispecies environment that occurs
in the subgingival pocket of the oral cavity. In A. actinomycetemcomitans, a
leukotoxin of the RTX family of Gram-negative cytolysins is known to be an
important virulence factor in early-onset forms of aggressive periodontitis.
Indeed, a specific clonal population of A. actinomycetemcomitans which is
hyperleukotoxic is found almost exclusively inearly-onset periodontitis patients
of African origin (20). Fong et al. (13) showed that the level of leukotoxin protein
is decreased by approximately 3-fold in the luxS mutant; real-time PCR
indicates a concomitant reduction of toxin mRNA in the mutant strain.
These results suggest that AI-2 induces the transcription of the leuko-
toxin operon. Consistent with these observations, early exponential-phase
A. actinomycetemcomitans cells exhibit a 2- to 3-fold increase in leukotoxin
expression and cytotoxicity after exposure to conditioned medium from
mid to late exponential-phase A. actinomycetemcomitans cultures (13). Thus,
the expression of leukotoxin may increase when A. actinomycetemcomitans
colonizes and grows to high cell density in the dental biofilm.
The role of AI-2 in modulating the virulence of Porphyromonas
gingivalis has not been clearly established. One study of P. gingivalis
virulence using a murine subcutaneous abscess model did not reveal
attenuated virulence of a P. gingivalis LuxS mutant (4). In other studies,
Chung et al. (7) showed that the capacity of a LuxS-deficient strain of
P.gingivalis to invade gingival epithelial cells did not differ appreciably
from that of wild-type cells. However, hemin levels are known to influence
the expression of a number of genes involved in the growth, survival and
pathogenicity of P. gingivalis, e.g. proteases, RecA, fimbriae, LPS, and
various outer-membrane proteins (3, 5, 23, 28, 40, 50). Thus, it is possible
that a link exists between LuxS-dependent signaling and the expression of
P. gingivalis virulence through its regulation of hemin uptake mechanisms.
Whereas the role of AI-2 in modulating P. gingivalis virulence has yet to
be confirmed, the involvement of LuxS-dependent signaling in promoting
the development of mixed-species biofilms with P. gingivalis has been
recently established. The adherence of P. gingivalis to species of viridans
streptococci but not to mutans streptococci may represent one mechanism
by which P. gingivalis identifies a suitable niche for initial colonization of
the dental biofilm. Using flow cells to mimic the open flow environment of
the oral cavity in vitro, Cook et al. (8) showed that P. gingivalis adhered to
Streptococcus gordonii cells that had previously colonized a saliva-coated
coverglass and subsequently accumulated in towering columnar micro-
colonies. However, inactivation of luxS had no effect on P. gingivalis
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
186
adherence or its formation of microcolonies. Subsequently, McNab et al.
(33) showed that S. gordonii also expressed luxS and secreted AI-2. Given the
widespread distribution of luxS and its suggested role as a universal signal,
McNab et al. (33) suggested that AI-2 produced by S. gordonii complemented
the luxS mutation in P. gingivalis. Indeed, when LuxS-deficient strains of both
organisms were analyzed for biofilmdevelopment inflowcells, P. gingivalis did
not form towering microcolonies (33). Thus, the development of P. gingivalis
microcolonies required a signal produced by luxS but biofilm development
was not strictly dependent upon the cognate P. gingivalis signal. Interestingly,
the adherence of P. gingivalis and S. gordonii was not affected by inactivation
of luxS, suggesting that the adhesion of P. gingivalis and its accretion into
towering microcolonies are independent processes and that AI-2 dependent
signaling is essential only for the latter. This is consistent with the contact-
dependent gene expression studies, described earlier in this chapter, which
showed that no significant overlap existed between the genes regulated by
contact with S. gordonii cells and the genes that exhibit differential expression
in the LuxS mutant. NcNab et al. (33) also showed that LuxS controls aspects of
carbohydrate metabolismin S. gordonii. However, it remains to be determined
whether defective biofilm development occurs as a consequence of the
disruption of carbohydrate metabolism in S. gordonii, or of hemin uptake in
P. gingivalis, or via an as yet undefined mechanism.
AI-2 signal transduction in A. actinomycetemcomitans
and P. gingivalis
The signal transduction cascades that mediate the cellular responses to
AI-2 have been identified in their entirety only for V. harveyi and V. cholerae
(25). The LuxS protein of V. harveyi is well conserved in many organisms;
indeed, comparison of the V. harveyi and A. actinomycetemcomitans LuxS
polypeptides indicates that they exhibit greater than 80% amino-acid
sequence identity (13). A. actinomycetemcomitans also possesses the Hfq
protein, an RNA chaperone that functions in V. harveyi and V. cholerae in
conjunction with four small RNAs as the repressor that acts downstream of
LuxO (25). However, a bioinformatics approach using the V. harveyi LuxP,
LuxQ, LuxU and LuxO sequences as probes failed to identify direct homo-
logs of these proteins in the A. actinomycetemcomitans genome that exhibit
the high level of sequence identity that was observed for LuxS and Hfq.
Furthermore, this finding is not unique to A. actinomycetemcomitans: simi-
lar results were obtained in searches of the E. coli, Salmonella typhimurium,
and P. gingivalis genomes. One explanation for these findings is that AI-2
187
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
functions as a quorum sensing signal only in Vibrio species and may not
mediate cell-to-cell communication in other bacteria that lack the complete
signal transduction pathway of Vibrio. Indeed, Winzer et al. (48) suggest
that for many organisms AI-2 may simply represent a toxic metabolite that
is generated as a by-product of the activated methyl cycle or, alternatively,
may be a nutrient that is excreted and subsequently consumed by these
bacteria. Evidence for and against this hypothesis is discussed in detail
in Chapter 6. However, the response of A. actinomycetemcomitans (and
other organisms as well) to AI-2 involves genes that would not be expected
to play a role in the secretion/excretion, uptake, or utilization of AI-2.
Therefore, an alternative explanation is that some organisms, such as
A. actinomycetemcomitans and E. coli, may communicate via AI-2 but do so by
responding to a different form of AI-2 than Vibrio (see below and Chapter 6)
and/or by utilizing sensor kinases and response regulators that differ from the
components of the Vibrio signal transduction pathway. Several different experi-
mental approaches have recently identified proteins of A. actinomycetemcomitans,
E. coli, and S. typhimuriumthat may represent potential components of the AI-2
signal transduction cascade in these organisms. For example, two different
two-component systems, QseBC and QseEF, have been suggested to regulate
gene expression in response to AI-2 in E. coli (see Chapter 7). In addition,
searching the A. actinomycetemcomitans genome at reduced stringency has
identified several proteins that exhibit lower, but significant primary sequence
identity (c.30%40%) to LuxP and LuxQ. These proteins and their potential
roles in AI-2 signal transduction in A. actinomycetemcomitans are summarized
in Figure 8.2 and discussed below.
The A. actinomycetemcomitans AI-2 receptor
Consistent with previous search results of the E. coli genome reported
by Surette et al. (44), the putative ribose-binding periplasmic RbsB poly-
peptide of A. actinomycetemcomitans was found to be most similar to LuxP,
the V. harveyi AI-2 receptor. Furthermore, our recent studies have shown
that A. actinomycetemcomitans RbsB, expressed and purified from a LuxS-
deficient E. coli host, binds to AI-2 in conditioned medium from both
A. actinomycetemcomitans and V. harveyi cultures (D. R. Demuth, unpub-
lished). Indeed, addition of purified RbsB to conditioned medium to a
concentration of 7 mM was sufficient to completely deplete AI-2: the treated
samples did not induce bioluminescence of the V. harveyi BB170 reporter.
In contrast, the untreated conditioned medium induced light production
200- to 2000-fold. These findings suggest that RbsB is capable of interact-
ing with both cognate and heterologous AI-2 signals and that RbsB may
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
188
function as a periplasmic AI-2 receptor in A. actinomycetemcomitans. RbsB
is part of an operon in A. actinomycetemcomitans comprising five genes,
rbsDACBK, that form a putative ABC-type ribose transport system. This
suggests that AI-2 bound by RbsB may be actively transported into the
cytoplasm via RbsC and RbsA and subsequently modified by phosphoryla-
tion via ribokinase (RbsK). Thus, the rbsDACBK operon may be function-
ally related to the lsrACDBFGE operon of S. typhimurium. Taga et al. (45)
demonstrated that lsrACDBFGE encodes an ABC-type transporter that
actively internalizes AI-2. These genes reside downstream from a second
operon in S. typhimurium, lsrRK, that encodes a regulator of lsrACDBFGE
expression (lsrR) and a kinase (lsrK) that phosphorylates AI-2 after it is
internalized (45). Interestingly, A. actinomycetemcomitans may also be
capable of internalizing and modifying AI-2 via an Lsr transport system
since it possesses divergently transcribed operons that are highly related to
LuxS
AI-2
AI-2
Lsr
AI-2
ArcB
Response
Regulator
Hfq ?
Iron Uptake, Ltx
AI-2P
LsrK
RbsB
RbsAC
AI-2
AI-2P
RbsK
Figure 8.2. The response of A. actinomycetemcomitans to autoinducer 2 (AI-2) may require
internalization of the signal. A. actinomycetemcomitans may internalize AI-2 by two different
ABC transport systems. AI-2 may be bound by the periplasmic protein RbsB (D. Demuth
et al., unpublished) and is presumably transported into the cytoplasm via the RbsC and
RbsA polypeptides. Internalized AI-2 may subsequently be phosphorylated by RbsK.
A. actinomycetemcomitans also possesses the Lsr ABC transport system, which has been
shown to internalize and phosphorylate AI-2 in S. typhimurium (44). In addition, genetic
studies suggest that the ArcB sensor kinase is involved in the cellular response to AI-2 and
may regulate genes involved in iron acquisition and storage (14; see text). The response
regulator acted upon by ArcB has not yet been identified.
189
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
the AI-2 regulated lsrACDBFGE and lsrRK gene clusters of S. typhimurium
(45). These observations suggest that A. actinomycetemcomitans may be
capable of transporting cognate and heterologous AI-2 signals into the
cell via two independent ABC transport mechanisms.
FromX-ray crystallographic studies, Chen et al. (6) determined that the
structure of AI-2 bound by LuxP is a borate diester of a hydrated furan
diulose. However, Miller et al. (37) recently showed that S. typhimurium
LsrB binds to (2R,4S)-2-methyl-2,3,3,4-tetrahydroxytetrahydrofuran, a form
of AI-2 that lacks boron and is structurally distinct from AI-2 of V. harveyi.
These studies suggest that AI-2 is present in multiple structural forms that
may exist in equilibrium (see Chapter 6), and that at least one form of AI-2
may be actively transported into the bacterial cell. The structure of the
A. actinomycetemcomitans AI-2 signal(s) that is bound by RbsB (or LsrB) is
not currently known, although the experiments described above clearly
indicate that RbsB interacts with the borate diester form produced by
V. harveyi. Furthermore, the biological consequences of the interaction of
AI-2 with RbsB or LsrB have not yet been determined. Genetic evidence
suggests that the sensor kinase ArcB may mediate some aspects of the
A. actinomycetemcomitans response to AI-2 (see below). However, ArcB does
not possess an extracellular domain that could facilitate interactions with a
periplasmic receptor and we cannot exclude the possibility that RbsB
interacts with a membrane-bound sensor kinase to facilitate the cellular
response to signal. Therefore, some important questions remain to be
answered. Is internalization of AI-2 required to induce a response in
A. actinomycetemcomitans? Does the engagement of AI-2 by RbsB or LsrB
elicit different cellular responses? Analyses of RbsB- and LsrB-deficient
A. actinomycetemcomitans mutants are currently being carried out to
address these questions.
The A. actinomycetemcomitans AI-2 sensor kinase
A low-stringency search of the A. actinomycetemcomitans genome
indicated that the ArcB sensor kinase was most similar to the V. harveyi
LuxQ sensor kinase (c.30% sequence identity over 500 residues).
However, as stated above, ArcB does not possess a large periplasmic
domain (8 amino acids) and instead is thought to respond to an intracel-
lular signal rather than interacting with a periplasmic molecule (30).
ArcB activity is also influenced by several intracellular allosteric effectors
(39). These observations are interesting given the apparent capability of
A. actinomycetemcomitans to internalize AI-2 (discussed above). To deter-
mine whether the ArcB sensor kinase plays a role in the response of
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
190
A. actinomycetemcomitans to AI-2, an isogenic arcB mutant was con-
structed and analyzed. In these experiments, it was assumed that if ArcB
contributes to the response to AI-2, the knockout mutant should exhibit a
phenotype similar to that displayed by the luxS knockout strain that cannot
produce AI-2. Interestingly, both the arcB and luxS mutant strains exhib-
ited growth rates under iron-replete conditions that were similar to those
of wild-type A. actinomycetemcomitans. However, both of the mutants grew
poorly and failed to attain high cell density under iron limitation (14).
Complementing the mutant strains with a plasmid-borne copy of the
appropriate gene (arcB or luxS) restored normal growth under iron limita-
tion to both mutants. Furthermore, expression of the LuxS-regulated
genes afuA and ftnAB were reduced 16- and 24-fold, respectively, in the
arcB mutant, consistent with the reduced expression previously reported
for these genes in a LuxS-deficient background (14). Thus, the growth and
gene expression studies indicate that inactivation of arcB and luxS have
similar effects, suggesting that ArcB may mediate aspects of the response
of A. actinomycetemcomitans to AI-2.
Several additional characteristics of ArcB are interesting in the context
of contributing to the cellular response to AI-2. First, ArcB is a tripartite
sensor that contains an integral histidine phosphotransfer domain (32).
This function is not present in the LuxQ sensor of V. harveyi but is instead
carried out by a reversible intermolecular transfer of phosphate between
LuxQ and LuxU. AI-2-dependent signal transduction through ArcB would
therefore not require an independent phosphotransfer protein related to
LuxU. This may explain our inability to detect a homolog of LuxU in the
A. actinomycetemcomitans genome, even when low stringency searches
were performed. ArcB-dependent phosphotransfer reactions to response
regulators also appear to be complex and ArcB may crosstalk with multiple
signaling pathways. For example, ArcB is capable of activating non-cognate
response regulators such as CheY (53) and EnvZ (31) in addition to its
cognate ArcA. Thus, ArcB is not stringently coupled to a single response
regulator and may regulate the expression of genes through several path-
ways involving different response regulators. This suggests that ArcB may
mediate broad cellular responses to a variety of environmental signals and
stimuli. However, very little information is currently available about the
specific response regulator(s) that mediates the A. actinomycetemcomitans
response to AI-2. Identifying these transcriptional factors and demonstrat-
ing that they accept the transfer of phosphate fromArcB will be an essential
step to confirm the role of this sensor kinase in the AI-2-dependent cellular
response.
191
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
AI-2 signal transduction in P. gingivalis
The components that comprise the AI-2 signal transduction pathway of
P. gingivalis are largely uncharacterized. In silico analysis of the P. gingivalis
genome identified homologs of LuxQ and LuxO, but not of LuxP or LuxU.
Interestingly, the gene encoding the LuxQ homolog, designated gppX, has
recently been shown to encode a sensor kinase that regulates the expression of
P. gingivalis proteases (19). Because protease expression is modulated in the
P. gingivalis luxS mutant, it is possible that GppX represents the sensor kinase
that is responsible for AI-2-mediated regulation of the protease genes.
In summary, the AI-2 signal transduction cascade of Vibrio species is
not wholly conserved in other organisms that produce and respond to a
LuxS-dependent signal. Indeed, recent studies suggest that different struc-
tural forms of the signal molecule exist. It is clear that additional work will
be required to clarify the similarities and differences in the signal transduc-
tion pathways contributing to the cellular responses among these diverse
organisms.
Crosstalk and implications for AI-2 signal specificity
The widespread distribution and conservation of luxS has led to spec-
ulation that AI-2 functions as a universal quorum sensing signal that may
report total bacterial biomass in a polymicrobial community, or the meta-
bolic status of that community. Support for the universal recognition of
AI-2 comes mostly from numerous studies that show that conditioned
medium from luxS-containing bacteria stimulates bioluminescence of the
V. harveyi reporter strain BB170. Indeed, V. harveyi BB170 is able to respond
to AI-2 from many diverse and unrelated organisms. However, is this
promiscuity a trait that is unique to V. harveyi, or do other organisms also
respond to heterologous signals? Our studies have shown that conditioned
medium from A. actinomycetemcomitans complements a luxS mutation in
P. gingivalis, suggesting that P. gingivalis is able to recognize and respond to
the A. actinomycetemcomitans signal (13). Furthermore, McNab et al. (33)
showed that a P. gingivalis luxS mutant was complemented in trans by AI-2
produced by Streptococcus gordonii. Normal biofilms formed when the
mutant organism adhered to wild-type S. gordonii cells, whereas no bio-
films developed when a LuxS-deficient P. gingivalis adhered to a luxS
mutant of S. gordonii (33). Thus, the ability of AI-2 to communicate between
bacterial species is not limited solely to the Vibrio bioluminescence
response. Although these data are suggestive that AI-2 may function as a
universally recognized signal, one cannot yet exclude the possibility that
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
192
LuxS-dependent signaling may display some degree of specificity. The AI-2
regulon has not been fully characterized for any single bacterial species and
it is not currently known whether all of the genes that respond to the
cognate signal of a given organism also exhibit a comparable response to
AI-2 from other bacteria. Thus it is possible that some genes may be
promiscuous in their response to signal (i.e. respond to both cognate and
heterologous signals) whereas other genes may respond preferentially to
the cognate signal. Furthermore, from an ecological perspective, the desir-
ability of a universally detected signal is not obvious. In complex mixed-
species communities such as the dental biofilm, constituent organisms
(many, if not all, expressing luxS) will be present at different cell densities,
at different metabolic states, and in different local environments. Given the
intricate and complex interrelationships that occur among the different
organisms in mature multispecies biofilms, it would be beneficial if these
organisms could sense not only how many other bacteria were present, but
sense specific other organisms that are cohabiting the local environment.
In this scenario, LuxS-dependent signaling could serve to define specific
niches in complex microbial communities.
Several mechanisms can be envisioned that would support specificity
of LuxS-dependent cell-to-cell communication. One possibility is that the
fine structure of AI-2 varies between organisms. This possibility has
already been acknowledged since the precursor of AI-2, 4,5-dihydroxy-2,
3- pentanedione, is unstable and can potentially rearrange to form different
products (49). It is also possible that different bacteria may uniquely
modify the core structure of AI-2, which in turn may impart specificity to
the signaling process. The best evidence that supports the existence of
structural variation of AI-2 signals is the recent finding that AI-2 from
V. harveyi and Salmonella typhimurium are structurally distinct (6, 37).
Thus, it will be essential to identify the structure of AI-2 from additional
bacteria in order to clarify the extent of structural variation that exists with
AI-2, and to determine whether structural variants of the signal induce
differential responses from organisms.
Specificity of LuxS-dependent signaling might also occur at the level of
signal detection and/or signal transduction. Multiple sensor kinases and/or
AI-2 receptors, or the presence of a hierarchy of response elements, can
define specific responses of bacteria to AI-2 and contribute to organism-
specific responses. In this respect, it is interesting that multiple two-
component systems have been shown to control LuxS-regulated genes in
E. coli (see Chapter 7). Elucidation of these signal transduction pathways
will be necessary to determine the overall distribution and the degree of
193
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
conservation of the AI-2 signal transduction cascades. Only after this infor-
mation is obtained will we have a more accurate picture of the distribution,
specificity, and outcomes of LuxS-dependent cell-to-cell communication.
ACKNOWLEDGEMENTS
Research was supported by Public Health Service Grant DE14605 from
the National Institute of Dental and Craniofacial Research.
REFERENCES
1 Beare, P. A., R. J. For, L. W. Martin and I. L. Lamont 2003. Siderophore-mediated
cell signaling in Pseudomonas aeruginosa: divergent pathways regulate virulence
factor production and siderophore receptor synthesis. Molec. Microbiol. 47:
195207.
2 Blehert, D. S., R. J. Palmer, J. B. Xavier, J. S. Almeida and P. E. Kolenbrander 2003.
Autoinducer 2 production by Streptococcus gordonii DL1 and the biofilm phenotype
of a luxS mutant are influenced by nutritional conditions. J. Bacteriol. 185: 485160.
3 Bramanti, T. E., S. C. Holt, J. L. Ebersole and T. Van Dyke 1993. Regulation of
Porphyromonas gingivalis virulence: hemin limitation effects on the outer mem-
brane protein(OMP) expressionand biological activity. J. Periodont. Res. 28: 4646.
4 Burgess, N. A., D. F. Kirke, P. Williams et al. 2002. LuxS-dependent quorum
sensing in Porphyromonas gingivalis modulates protease and haemagglutinin
activities but is not essential for virulence. Microbiology. 148: 76372.
5 Champagne, C. M., S. C. Holt, T. E. van Dyke, B. J. Gordon and L. Shapira 1996.
Lipopolysaccharide isolated from Porphyromonas gingivalis grown in hemin-
limited chemostat conditions has a reduced capacity for human neutrophil
priming. Oral Microbiol. Immunol. 5: 31925.
quorum sensing signal containing boron. Nature 415: 5459.
7 Chung, W. O., Y. Park, R. J. Lamont et al. 2001. Signaling system in
Porphyromonas gingivalis based on a LuxS protein. J. Bacteriol. 183: 39039.
8 Cook, G. S., Costeton, J. W. and J. J. Lamont 1998. Biofilm formation by
Porphyromonas gingivalis and Streptococcus gordonii. J. Periodont. Res. 33: 3237.
9 Demuth, D. R., D. C. Irvine, J. W. Costerton, G. S. Cook and R. J. Lamont 2001.
Discrete protein determinant directs the species-specific adherence of
Porphyromonas gingivalis to oral streptococci. Infect. Immun. 69: 573641.
10 Demuth, D. R. and H. F. Jenkinson 1997. Structure, function and antigenicity of
strepotococcal antigen I/II polypeptides. Molec. Microbiol. 23: 18390.
11 Du, L. D. and P. E. Kolenbrander 2000. Identification of saliva-regulated genes of
Streptococcus gordonii DL1 by differential display using random arbitrarily primed
PCR. Infect. Immun. 68: 48347.
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
194
12 Egland, P. G., L. D. Du and P. E. Kolenbrander 2001. Identification of indepen-
dent Streptococcus gordonii SspA and SspB functions in coaggregation with
Actinomyces naeslundii. Infect. Immun. 69: 751216.
13 Fong, K. P., W. O. Chung, R. J. Lamont and D. R. Demuth 2001. Intra- and
interspecies regulation of gene expression by Actinobacillus actinomycetemcomitans
LuxS. Infect. Immun. 69: 762534.
14 Fong, K. P., L. Gao and D. R. Demuth 2003. luxS and arcB control aerobic growth
of Actinobacillus actinomycetemcomitans under iron limitation. Infect. Immun.
71: 298308.
15 Frias, J., E. Olle and M. Alsina 2001. Periodontal pathogens produce quorum
sensing signal molecules. Infect. Immun. 69: 34314.
16 Gomez, J. A., M. T. Criado and C. M. Ferreiros 1998. Cooperation between the
components of meningococcal transferrin receptor, TbpAand TbpB, in the uptake of
transferriniron by the 37kDa ferric binding protein(FbpA). Res. Microbiol. 149: 3817.
17 Grenier, D. and D. Mayrand 1986. Nutritional relationships between oral bac-
teria. Infect. Immun. 53: 61620.
18 Guggenheim, M., S. Shapiro, R. Gmur and B. Guggenheim 2001. Spatial
arrangements and associative behavior of species in an in vitro oral biofilm
model. Appl. Environ. Microbiol. 67: 134350.
19 Hasegawa, Y., S. Nishiyama, K. Nishikawa et al. 2003. A novel type of two-
component regulatory system affecting gingipains in Porphyromonas gingivalis.
Microbiol. Immunol. 47: 84958.
20 Haubek, D., O. K. Ennibi, K. Poulsen, N. Benzarti and V. Baelum 2004. The
highly leukotoxic JP2 clone of Actinobacillus actinomycetemcomitans and progres-
sion of periodontal attachment loss. J. Dent. Res. 83: 76770.
21 Hayashi, J., K. Nishikawa, R. Hirano, T. Noguchi and F. Yoshimura 2000.
Identification of a two-component signal transduction system involved in fim-
briation of Porphyromonas gingivalis. Microbiol. Immunol. 44: 27982.
22 Kolenbrander, P. E. 2000. Oral microbial communities: biofilms, interactions,
and genetic systems. A. Rev. Microbiol. 54: 41337.
23 Lamont, R. J. and H. F. Jenkinson 1998. Life below the gum line: pathogenic
mechanisms of Porphyromonas gingivalis. Microbiol. Molec. Biol. Rev. 62: 124463.
24 Lamont, R. J., A. El-Sabaeny, Y. Park et al. 2002. Role of the Streptococcus gordonii.
SspB protein in the development of Porphyromonas gingivalis biofilms on strepto-
coccal substrates. Microbiology 148: 162736.
25 Lenz, D. H., K. C. Mok, B. N. Lilley et al. 2004. The small RNA chaperone Hfq
and multiple small RNAs control quorum sensing in Vibrio harveyi and Vibrio
chloerae. Cell 118: 6982.
26 Liljemark, W. F., C. G. Bloomquist, B. E. Reilly et al. 1997. Growth dynamics in
a natural biofilm and its impact on oral disease management. Adv. Dent. Res.
11: 1423.
27 Lilley, B. N. and B. L. Bassler 2000. Regulation of quorum sensing in Vibrio
harveyi by LuxO and sigma-54. Molec. Microbiol. 36: 94054.
195
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
28 Liu, Y. and H. M. Fletcher 2001. Environmental regulation of recA gene expres-
sion in Porphyromonas gingivalis. Oral Microbiol. Immunol. 16: 13643.
29 Maeda, K., H. Nagata, Y. Yamamoto et al. 2004. Glyceraldehyde-3-phosphate
dehydrogenase of Streptococcus oralis functions as a coadhesin for Porphyromonas
gingivalis major fimbriae. Infect. Immun. 72: 134148.
30 Malpica, R., B. Franco, C. Rodriguez, O. Kwon and D. Georgellis 2004.
Identification of a quinine-sensitive redox switch in the ArcB kinase. Proc.
31 Matsubara, M., S. I. Kitaoka, S. I. Takeda and T. Mizuno 2000. Tuning of porin
expression under anaerobic growth conditions by his-to-asp cross phosphorelay
through both the EnvZ-osmosensor and ArcB anaerosensor in Escherichia coli.
Genes Cells 5: 55569.
32 Matsushika, A. and T. Mizuno 2000. Characterization of three putative sub-
domains in the signal input domain of the ArcB hybrid sensor in Escherichia
coli. J. Biochem. 127: 85560.
33 McNab, R., S. K. Ford, A. El-Sabaeny et al. 2003. LuxS-based signaling in
Streptococcus gordonii: autoinducer 2 controls carbohydrate metabolism and bio-
film formation with Porphyromonas gingivalis. J. Bacteriol. 185: 27484.
34 Merritt, J., F. Qi, S. D. Goodman, M. H. Anderson and W. Shi 2003. Mutation
of luxSaffects biofilmformationin Streptococcus mutans. Infect. Immun. 71: 19729.
35 Mikx, F. H. and J. S. van der Hoeven 1975. Symbiosis of Streptococcus mutans and
Veillonella alcalescens in mixed continuous cultures. Arch. Oral Biol. 20: 40710.
recognizes a chemically distinct form of the bacterial quorum sensing signal
AI-2. Molec. Cell. 15: 67787.
38 Paster, B. J., S. K. Boches, J. L. Galvin et al. 2001. Bacterial diversity in human
subgingival plaque. J. Bacteriol. 183: 377083.
39 Rodreiguez, C., O. Kwon and D. Georgellis 2004. Effect of D-lactate on the
physiological activity of the ArcB sensor kinase in Escherichia coli. J. Bacteriol.
186: 208590.
40 Smalley, J. W., A. J. Birss, A. S. McKee and P. D. Marsh 1991. Haemin-restriction
influences haemin-binding, haemagglutination and protease activity of cells and
extracellular membrane vesicles of Porphyromonas gingivalis W50. FEMS
Microbiol. Lett. 61: 637.
41 Socransky, S. S. and A. D. Haffajee 1992. The bacterial etiology of destructive
periodontal disease: current concepts. J. Periodontol. 63: 32231.
42 Socransky, S. S., A. D. Haffajee, M. A. Cugini, C. Smith and R. L. Kent 1998.
Microbial complexes in subgingival plaque. J. Clin. Periodontol. 25: 13444.
43 Sroka, A., M. Sztukowska, J. Potempa, J. Travis and C. A. Genco 2001.
Degradation of host heme proteins by lysine- and arginine-specific cysteine
proteinases (gingipains) of Porphyromonas gingivalis. J. Bacteriol. 183: 560916.
D
.
R
.
D
E
M
U
T
H
A
N
D
R
.
J
.
L
A
M
O
N
T
196
44 Surette, M. G., M. B. Miller and B. L. Bassler 1999. Quorum sensing
in Escherichia coli, Salmonella typhimurium and Vibrio harveyi: a new family
of genes responsible for auto-inducer production. Proc. Natn. Acad. Sci. USA
96: 163944.
46 Wen, Z. T. and R. A. Burne 2004. LuxS-mediated signaling in Streptococcus
mutans is involved in regulation of acid and oxidative stress tolerance and biofilm
47 Whittaker, C. J., C. M. Klier and P. E. Kolenbrander 1996. Mechanisms of
adhesion by oral bacteria. A. Rev. Microbiol. 50: 51350.
48 Winzer, K., K. R. Hardie and P. Williams 2002. Bacterial cell-to-cell communica-
tion: sorry, cant talk now gone to lunch. Curr. Opin. Microbiol. 5: 221522.
50 Xie, H., S. Cai and R. J. Lamont 1997. Environmental regulation of fimbrial gene
expression in Porphyromonas gingivalis. Infect. Immun. 65: 226571.
51 Xie, H., G. Cook, J. W. Costerton et al. 2000. Intergeneric communication in
dental plaque biofilms. J. Bacteriol. 182: 70679.
52 Xie, H., N. Kozlova and R. J. Lamont 2004. Porphyromonas gingivalis genes
involved in fimA regulation. Infect. Immun. 72: 6518.
53 Yaku, H., M. Kato, T. Hakoshima, M. Tsuzuki and T. Mizuno 1997. Interaction
between CheY response regulator and the histidine containing phosphotransfer
(HPt) domain of the ArcB sensor kinase in Escherichia coli. FEBS Lett. 408:
33740.
54 Zhang, Y., Y. Lei, A. Khammanivong and M. C. Herzberg 2004. Identification of
a novel two-component system in Streptococcus gordonii V288 involved in biofilm
formation. Infect. Immun. 72: 348994.
197
Q
U
O
R
U
M
S
E
N
S
I
N
G
I
N
T
H
E
D
E
N
T
A
L
B
I
O
F
I
L
M
CHAPTER 9
Quorum-sensing-dependent regulation
of staphylococcal virulence and biofilm
development
Jeremy M. Yarwood
Department of Microbiology, Carver College of Medicine,
University of Iowa, IA, USA
INTRODUCTION
Staphylococci are a genus of bacteria remarkably adept at causing a
variety of human and animal diseases. These range from relatively benign
skin infections, such as impetigo, to much more serious ones, including
endocarditis, osteomyelitis, toxic shock syndrome, and those associated
with implanted medical devices. In fact, the staphylococci are a leading
cause of nosocomial infections worldwide, and the continuing emergence
of highly drug-resistant strains has created an immediate need for the
development of new antimicrobial therapies and strategies. Since the iden-
tification of the accessory gene regulator (Agr) quorum sensing system in
Staphylococcus aureus, and subsequently in other staphylococcal species, it
has been assigned a central role in the regulation of staphylococcal viru-
lence. As such, it has attracted substantial attention as a potential target for
controlling staphylococcal disease.
Although recent studies have shown that virulence-gene regulation by
Agr is considerably more complex in vivo than initially understood from
studies in vitro, it remains clear that expression of Agr, or even lack thereof,
is an important determinant in staphylococcal disease development.
agr mutants have been shown to be attenuated for virulence in some animal
models of infection, including a murine arthritis model, an osteomyelitis
model, and a skin abscess model (reviewed in (34)). It has also been shown
that expression of Agr, and of Agr-regulated exotoxins, facilitates escape of
S. aureus internalized by epithelial cells (49). Furthermore, evidence is
beginning to emerge that the Agr system may play a major role in biofilm
formation, development, and behavior, with important implications for
human disease.
199
This chapter will discuss the nature of the agr locus, potential interac-
tion of Agr with other virulence gene regulators, the emergence of Agr
variants, the impact of Agr expression on biofilm formation, and the
viability of therapeutic strategies targeting the Agr system. It will focus
primarily on the Agr system of S. aureus, in which Agr, and virulence in
general, has been best studied, and to a lesser extent on that of
Staphylococcus epidermidis. Agr homologs have been identified in many
additional staphylococcal species, but little or no investigation has been
conducted into regulation of virulence by Agr in these staphylococci.
THE Agr QUORUM SENSING SYSTEM
The agr locus consists of two divergent operons (Figure 9.1) (reviewed
in (34)). The P2 operon (agrACDB) encodes the proteins necessary
for signal synthesis, processing, secretion, and recognition, whereas the
transcript of the P3 operon, RNAIII, mediates the regulatory effects of Agr
expression. The autoinducing peptide (AIP) signal, which can consist of
seven to nine residues, is formed by cleavage and processing of the AgrD
protein. A characteristic thiolactone ring is formed between a generally
conserved central cysteine and the peptides C-terminal carboxyl group; this
cyclical structure is generally required for the activity of AIP. Both the
cleavage of AgrD and the secretion of AIP are thought to be mediated by
the membrane protein AgrB, although other proteins may be involved as
well. In a comparison of AIP sequences from multiple Staphylococcus
species, only the central cysteine and the five-membered thiolactone ring
are generally conserved (reviewed in (34, 38)). One interesting exception is
found in some strains of Staphylococcus intermedius, which produce an
active nonapeptide AIP, which contains a serine instead of a cysteine
group and forms a cyclic lactone. The length of the N-terminal tail varies
fromtwo to four amino acids; this results in pheromones with seven to nine
residues in all. Lengthening or shortening the tail of an S. epidermidis AIP
by a single amino acid residue results in a lack of biological activity in this
bacterium, suggesting that AIP length is important for biological activity.
In flask cultures, the amount of AIP in the mediumgenerally increases
in correlation with increasing cell density. Upon reaching sufficient AIP
concentration (this has been reported to be during the mid-log phase,
although the timing may vary from strain to strain) signaling via the
AgrAAgrC system leads to increased transcription of both the P2 and P3
operons. AgrA and AgrC form the response regulator and histidine kinase
receptor, respectively, of a two-component regulatory system that responds
J
.
M
.
Y
A
R
W
O
O
D
200
to the secreted AIP. AgrA is thought to be constitutively phosphorylated,
thus its activation may require dephosphorylation. Binding of AgrA to the
agr promoters has not been demonstrated; other regulatory proteins such
as staphylococcal accessory regulator A (SarA) likely are a critical compo-
nent of Agr autoinduction and response to the AIP.
The transcript of the P3 operon, RNAIII, is considered to be the effector
of the agr locus in mediating the repression or induction of quorum-
controlled genes. Levels and timing of RNAIII transcription vary from strain
to strain and can be correlated with the relative levels of Agr-regulated
secreted or surface factors (29). d-Toxin (hld) is also translated from the
RNAIII molecule, although disruption of d-toxin translation does not appear
to impair the regulatory capabilities of RNAIII. Although the RNAIII
nucleotide sequence is not well conserved, the secondary structure is,
including several stemloop motifs. When RNAIII from S. epidermidis,
agrA agrC agrD agrB
P2 P3
RNAII RNAIII
hld
AgrC
AgrA
A
g
r
B
AIP =
SarA
P
Figure 9.1. Model of the Agr system(see text for additional description). The agrDproduct is
processed by AgrB into AIP and secreted into the extracellular environment. Recognition
of AIP by AgrC triggers transfer of a phosphate group between AgrC and AgrA. AgrA,
together with the transcriptional regulator SarA, acts to increase transcription of both the
P2 and P3 operons. The transcript of the P3 operon, RNAIII, is the effector molecule of
the agr locus and encodes d-toxin (hld) as well.
201
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
S. simulans, and S. warneri were expressed in S. aureus, they completely
repressed expression of the Protein A (spa) gene, similar to native S. aureus
RNAIII, and they stimulated expression of a-toxin (hla) and serine protease,
suggesting conservation of some important regulatory function among
species (51). However, the ability to stimulate a-toxin and serine protease
was impaired compared with S. aureus RNAIII, indicating that repression
and activation functions might be independently encoded within the
RNAIII molecule.
Much remains unknown of how RNAIII exerts its regulatory effects,
although RNAIII is capable of regulation both at the transcriptional and
translational levels (reviewed in (34)). For instance, the transcription termi-
nation loop of RNAIII is necessary for repression of spa transcription,
whereas the 3
0
end of RNAIII is complementary to the translation initiation
site of spa mRNA and reportedly blocks its translation. In regulation of hla
translation, the 5
0
region of RNAIII is complementary to the hla mRNA
leader sequence. The hla mRNA leader folds into an untranslatable form
unless prevented from doing so by RNAIII. Benito et al. (2) proposed that
several structurally different populations of RNAIII might coexist in vivo,
and that RNAIII undergoes conformational changes necessary for specific
functions. For example, two functions of RNAIII, translation of hld
and translation regulation of hla, are thought to be mutually exclusive
from a structural standpoint. In the in vitro structure of RNAIII, part of
the hla mRNA binding sequence is hidden, whereas the RBS sequence
and initiation codon of hld remain accessible. Presumably, then, some con-
formational change must occur for the hla mRNARNAIII interaction to
occur. There is also some functional redundancy in the S. aureus RNAIII
molecule, as non-overlapping 5
0
and 3
0
regions of the molecule are indepen-
dently active in stimulating hla transcription. This may be related to the
presence of nearly identical sequences in two distant stem-loops, which are
complementary to the canonical ShineDelgarno sequence. It has been
proposed that RNAIII could thus act by interfering with translation of
other regulatory proteins via these anti-S-D sequences or that the C-rich
loops could serve as decoy binding sites for regulatory proteins, potentially
titrating them out of circulation and/or altering their DNA-binding
characteristics (34).
In fact, many of the regulatory effects exerted by RNAIII are likely to be
indirect. Recent analysis of Agr regulation of sed expression suggests that
the late log growth phase increase in sed transcription occurs via the Agr-
mediated reduction in Rot (repressor of toxin) activity rather than via a
direct effect of Agr (54). This was supported by the observation that the sed
J
.
M
.
Y
A
R
W
O
O
D
202
promoter is not regulated by the Agr system in a rot mutant background.
Agr does not appear to affect rot transcription. However, it has been shown
to downregulate Rot activity by an as yet undefined mechanism. One
explanation is that Rot might be an RNAIII-binding protein, and activation
of agr results in titration of Rot from its gene targets (32).
Agr-REGULATED GENES
Virulence determinants regulated by Agr are summarized in Table 9.1.
In general, agr expression in batch cultures leads to the increased expres-
sion of secreted virulence factors (exoenzymes and exotoxins) and the
decreased expression of several surface-associated adhesins and virulence
factors. For most strains this occurs during the transition from late-log
growth to early stationary phase, also known as postexponential phase.
A model for how this might affect pathogenesis in vivo has been proposed
Table 9.1. Virulence factors regulated by Agr
Virulence Factor Agr effect
aureolysin (metalloprotease)
TSST-1
enterotoxin B
enterotoxin C
a-, b-, d-, g-toxins
exfoliatins A and B
fatty-acid-modifying enzyme (FAME)
hyaluronate lyase
lipase
phospholipase C
proteases (Spl A, B, D, F; V8 protease)
staphylokinase
cell-surface associated capsular polysaccharide (type 5)
collagen-binding protein /
coagulase
fibronectin-binding proteins A and B
protein A
vitronectin-binding protein
Source: Adapted from (5, 41).
203
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
for abscess formation; similar phenomena may be important in biofilms as
well (Figure 9.2). During early growth, expression of surface adhesins and
immuno-evasive factors is important for colonization. However, as bacter-
ial cell density becomes high and cells exhaust local nutrient sources,
expression of tissue-degrading exoenzymes and immunostimulatory toxins
facilitates the spread of staphylococci from this localized infection to new
potential sites of infection. Exoenzyme activity may also provide nutrient
release from the surrounding tissues.
With the exception of capsular polysaccharides, factors upregulated by the
Agr system with particular significance for virulence can generally be divided
into two classes, exoenzymes and exotoxins (reviewed in (12)). Exoenzymes,
such as lipase and the proteases, contribute to host tissue degradation, perhaps
to create a source of nutrients or to facilitate escape from the localized infec-
tion. Proteases also degrade host proteins important for the immune response,
Log Lag
GROWTH PHASE
Stationary
Surface-
associated
adhesins
Secreted
virulence
factors
Spread to
new sites
PMNs
F
L
A
S
K

C
U
L
T
U
R
E

L
o
g

c
e
l
l

n
o
.
A
B
S
C
E
S
S
B
I
O
F
I
L
M
Figure 9.2. Model of Agr regulation of virulence genes in vitro and in vivo. Expression of
cell-surface-associated adhesins enhances colonization and proliferation of staphylococci
at the site of infection. Growth and maturation of the staphylococcal colony leads to an
abscess, a mature biofilm, or both. Expression of extracellular enzymes and toxins facilitates
escape of staphylococci from the localized infection and subsequent colonization of
secondary sites or hosts. PMNs: polymorphonuclear cells. Adapted from (41).
J
.
M
.
Y
A
R
W
O
O
D
204
such as the neutrophil defensins, platelet microbiocidal proteins, and anti-
bodies, thus providing some protection from host immunity. Regulation of
lipase production itself may be due to the upregulation by Agr of proteolytic
activity, enhancing conversion of the pro-form of the lipase to mature lipase,
rather than regulation at the transcriptional level. Fatty-acid modifying
enzyme (FAME) is also strongly activated by Agr. FAME can inactivate bacteri-
cidal lipids often found in staphylococcal abscesses through esterification of
the lipids into alcohols. These lipids are frequently released fromglycerides in
the abscess, perhaps by the action of staphylococcal lipase. Accordingly, most
strains that produce lipase also produce FAME.
Exotoxins upregulated by Agr include the hemolysins (a-, b-, d-, g-toxin),
which have general lytic activity against a broad range of host cells, and the
pyrogenic toxin superantigens (SAgs), which have broad immunostimulatory
activity. The a- and d-toxins are both thought to be pore-forming agents
strongly regulated by Agr. a-Toxin is positively regulated at both the transcrip-
tional and translational levels by RNAIII, whereas d-toxin is directly encoded
via RNAIII. a-Toxin is a highly potent toxin, killing erythrocytes, mononuclear
immune cells, and epithelial and endothelial cells; d-toxin is a small, surfactive
protein and is active against many types of membranes; b-toxin is likely a
sphingomyelinase, causing hydrolysis of sphingomyelin in the membrane
outer leaflet patches of erythrocytes and eventual collapse of the lipid bilayer;
g-toxin is a member of a family of bi-component toxins of S. aureus in which
the pore-forming activity is mediated by two synergistically acting proteins,
and is strongly hemolytic but much less leukotoxic.
SAgs generally exert their effects through non-antigen-specific binding
of professional antigen-presenting cells and T-cells. Typical antigens might
stimulate one out of 10,000 T-cells that specifically recognize that particu-
lar antigen; superantigens may stimulate and cause polyclonal prolifera-
tion of 20% or more of all circulating T-cells. This results in the release of
high levels of cytokines, leading to the symptoms of toxic shock such as
vascular dilation, loss of blood pressure, and subsequent organ damage and
failure. In general, most of the SAgs are activated by the Agr system,
although there are some exceptions. Staphylococcal enteroxin A (SEA),
for instance, is produced throughout growth in an Agr-independent man-
ner. It is not clear what the significance of this Agr-independent expression
and the relative contribution of Agr-independent toxins versus Agr-activated
toxins in a particular infection type might be. Many of the SAgs are also
enterotoxins, exhibiting emetic activity, which is separable within the protein
structure from their superantigenic activity, and are responsible for the domin-
ance of S. aureus as a leading cause of food poisoning.
205
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
Several capsular serotypes have been identified in S. aureus; more than
90% of clinical isolates produce capsular polysaccharide (reviewed in (12)).
Agr has been described as a positive regulator of serotype 5 capsule poly-
saccharide production, both in vitro and in vivo in a rabbit endocarditis
model (63). Although there are some conflicting data, the preponderance of
evidence suggests that capsule production enhances virulence and may
resist phagocytosis and clearance by host immune cells. Immunization
with certain capsule serotypes also is protective against challenge by
S. aureus in animal models of infection.
There are numerous factors important for virulence that are down-
regulated by Agr as well. These include several surface-associated adhesins
such as protein A, fibronectin-binding proteins (fnbA, fnbB), vitronectin-
binding protein, and coagulase (reviewed in (12)). Many of these proteins
can also be found in substantial quantities in the growth medium, suggest-
ing that their release from the cell may have importance as well. Protein
A was the first staphylococcal surface protein to be characterized and is
noted for its ability to bind the Fc region of mammalian IgG. By binding
IgG, Protein A may interfere with the phagocytosis of opsonized bacteria.
Protein A can also mediate staphylococcal adherence to von Willebrand
factor, a host extracellular matrix protein, suggesting that Protein A may
influence several aspects of the colonization and infectious processes. Two
structurally similar proteins, FnbpA and FnbpB, have been shown to
mediate S. aureus binding to fibronectin. Fibronectin is a ubiquitous pro-
tein found in the extracellular matrix of most tissues, as well as in soluble
form in many body fluids, and is necessary for the adhesion of almost all
cell types. Fibronectin is one of the host proteins that rapidly coat foreign
objects, such as an intravascular catheter, thus facilitating adherence of
staphylococci to this de facto biological surface. The fibronectin-binding
proteins may also play a role in invasion of host cells by binding soluble
fibronectin, which is then recognized by integrins on the host cell. This
results in phagocytosis of the host-protein-coated bacteria. Vitronectin is
an adhesive glycoprotein found in circulation at several extracellular
matrix sites, particular during tissue or vascular remodeling. Similar to fibro-
nectin- or fibrinogen-binding proteins, vitronectin binding likely facilitates
colonization by staphylococci of host tissues or host-protein-coated
implanted devices. Coagulase production is the primary criterion used to
distinguish S. aureus from other staphylococcal species in a clinical micro-
biology setting. Coagulase binds soluble fibrinogen and also binds human
prothrombin to form a complex which converts soluble fibrinogen to
insoluble fibrin. Coagulase is cell-wall-associated, although it does not
J
.
M
.
Y
A
R
W
O
O
D
206
have a cell-wall-anchoring sequence. The role of coagulase in staphylococcal
pathogenesis is not well understood. It could be that fibrin clotting around
infection foci protects the bacteria from elements of host immunity.
Using a S. aureus gene array, Dunman et al. (11) identified 104 genes
induced and 34 genes repressed in an Agr-dependent manner. This study
supports in general the idea that extracellular virulence factors are activated
by Agr and surface adhesins repressed. However, the majority of genes
identified as being Agr-regulated were in fact not known virulence factors,
but were instead involved in such cellular processes as amino-acid meta-
bolism and nutrient transport. Considering this evidence, as well as the
identification of Agr homologs in other, less virulent staphylococci, one can
speculate that the Agr system may not have evolved originally to facilitate
virulence, but perhaps for the coordination of more basic biological
functions.
QUORUM SENSING IN OTHER STAPHYLOCOCCI
Staphylococcus epidermidis
In general, the Agr systemin S. epidermidis appears to be highly similar
to that in S. aureus (55, 60). It is growth-phase-dependent and, with a few
exceptions, upregulates exoprotein production while downregulating sev-
eral surface-associated proteins. In particular, both lipase and protease
activity are greatly downregulated in a S. epidermidis agr mutant. Overall,
homology of the agr loci between S. aureus and S. epidermidis is 68%. The
d-toxin presumably encoded by the S. epidermidis RNAIII molecule differs
in three amino acids from that produced by S. aureus, and is upregulated in
post-exponential phase, as is RNAIII. d-Toxin activity was found in 21 of 23
S. epidermidis strains tested. Agr was also shown to be indirectly involved in
production of the antibiotic epidermin by S. epidermidis via regulation of
EpiP, a protease involved in the formation of mature epidermin (27).
Most infections caused by S. epidermidis are chronic and generally less
severe than those caused by S. aureus (reviewed in (12)) owing to the
absence in S. epidermidis of most of the immunostimulatory toxins found
in S. aureus. However, severe disease can still result from colonization by
S. epidermidis, particularly in immunocompromised patients; S. epidermidis
is perhaps the most frequent contaminant of implanted medical devices.
Potential consequences of S. epidermidis infection include abscess forma-
tion, endocarditis, peritonitis, ventriculitis, and sepsis. Much of the devel-
opment of these diseases is dependent in large part upon the response of
207
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
the human innate immune systemto specific microbial products, or immune
modulins. The lipopolysaccharide (LPS) of Gram-negative bacteria is an
example of a particularly potent immune modulin. The peptidoglycan and
lipoteichoic acid (LTA) of Gram-positive bacteria are also immune modulators,
but to a much lesser extent than LPS. However, S. epidermidis was also found to
produce the phenol-soluble modulins (PSM, reviewed in (58)). These modu-
lins are a pro-inflammatory complex of peptides that have been shown to
induce cytokine production by monocytes, cause degranulation, and inhibit
spontaneous apoptosis in human neutrophils, and serve as a chemoattractant
for neutrophils and monocytes. S. epidermidis has been shown to produce at
least three of these peptides, PSMa, PSMb, and PSMg. PSMg is in fact d-toxin
and encoded by the agr locus, PSMa is similar to d-toxin, and PSMb is similar
to the SLUSH peptides from S. lugdunensis and the gonococcal growth
inhibitor peptides from S. haemolyticus. Recently, it was found that agr regu-
lates the productionof these modulins inS. epidermidis (58). This study was the
first to ascribe a major role to quorum sensing in hostpathogen interaction
during S. epidermidis infection. PSM production was completely ablated in an
agr mutant; the agr mutant did not induce production of TNFa by human
myeloid cells nor did it induce chemotaxis of neutrophils to the extent that a
wild-type strain did. The authors propose that quorum-sensing regulation
allows PSM production only at the appropriate stage of infection, enabling
S. epidermidis growth and survival in the host. For instance, production of
immunostimulatory factors early in infection when only a few bacterial cells
were present would stimulate neutrophil activity and attraction would likely
result in clearance from the host. Alternatively, production of PSM and extra-
cellular enzymes at the later stages of infection when bacterial cell density is
high would result in tissue degradation associated with a strong immune
response and the enzymatic activity, both providing nutrients and facilitating
escape from the localized abscess (similar to the model of quorum sensing in
S. aureus infections).
Staphylococcus lugdunensis
An agr-related sequence was originally described in this organism by
Vandenesch et al. (57). Atranscript with homology to RNAIII was identified
in some of the strains, and these same strains had hemolysin activity
similar to that caused by S. aureus d-toxin, but the RNAIII molecule from
S. lugdunensis does not encode any peptide homologous to d-toxin. More
recently, it was found that this hemolytic activity was actually mediated by
three closely related SLUSH peptides encoded by a non-agr locus (9).
J
.
M
.
Y
A
R
W
O
O
D
208
Staphylococcus saprophyticus
Recently, an agr homolog was cloned fromS. saprophyticus (44), a cause
of urinary tract infections. Although no genes have yet been confirmed to
be regulated by Agr in this organism, RNAIII levels do appear to increase
with time and the agr locus was found in all clinical isolates tested. The
S. saprophyticus RNAIII does not appear to encode any peptide with homol-
ogy to d-toxin, consistent with the observation that S. saprophyticus is not
hemolytic.
Agr homologs have also been identified by PCR amplification and
sequencing in several additional staphylococcal species by Dufour et al.
(10) in a study described below.
Agr SPECIFICITY GROUPS
Four distinct Agr groups, based on the identity of the AIP produced,
have been described in S. aureus (reviewed in (34)). AIP produced by one
group generally inhibits signaling by staphylococci from a different Agr
group by competitive binding to the AgrC receptor. In addition, one AIP
produced by S. epidermidis inhibits signaling by three of the four S. aureus
agr groups. Conversely, S. aureus Agr group IV is the only one capable of
inhibiting S. epidermidis signaling.
In a survey of numerous staphylococcal species, Dufour et al. (10)
found the presence of agr homologs in 12 other staphylococcal species
besides those already described in S. aureus {4}, S. epidermidis {3}, and
S. lugdunensis {2} (numbers in brackets represent the number of Agr
groups identified for each species). These included S. auricularis {2},
S. capitis subsp. capitis {2}, S. caprae {2}, S. simulans {2}, S. arlettae {1},
S. carnosus {1}, S. cohnii subsp. cohnii {1}, S. cohnii subsp. urealyticum
{1}, S. gallinarum {1}, S. intermedius {1}, and S. xylosus {1}. The authors
found extreme divergence in the agr locus sequence among species, with
only 10% of the agrBCD nucleotides being absolutely conserved. The most
extensive divergence appears to be within the regions thought to be impor-
tant for signal synthesis (agrD), processing (C-terminal portion of agrB),
and recognition (portion of the N-terminal half of agrC). AgrA, which does
not interact directly with the AIP, shows much less sequence divergence.
This study also found significant covariance between these genes, as one
might expect in order for a strain to properly secrete and recognize a unique
AIP. In general, phylogenetic relationships established by using 16S rDNA
strongly resembled relationships determined by using agrC or agrB.
209
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
However, phylogenicity established by using agrD was very different from
that determined by using 16S, suggesting that agrDsequences are the most
divergent of all. Thus, the authors propose that the driving force in Agr
evolution might be mutation of the agrD sequence coding for the AIP.
Under the appropriate environment, the strain carrying the mutated agrD
might accumulate compensatory mutations in agrB and agrC to allow
recovery of a functional Agr quorum sensing system. A model for Agr
evolution in the context of a biofilm is discussed later in this chapter.
There is some correspondence as well between phylogenetic trees, Agr
groups, and disease type (24). For instance, most menstrual TSS strains
belong to Agr group III, whereas most exfoliatin-producing strains belong
to Agr group IV. Interestingly, a strong correlation was found between
vancomycin treatment failure and infection due to MRSA encoding Agr
group II. It may simply be that this Agr group is associated with select
S. aureus clones, known as glycopeptide intermediate-resistance S. aureus
(GISA), that have intrinsic advantages under vancomycin selection (33).
Li et al. (28) did find that genetic polymorphism of Agr was linked to
pathogenicity of S. epidermidis in China. Group I isolates were predominant
in pathogenic isolates, whereas isolates from healthy patients tended to be
group II. It was also shown that some Agr alleles were more highly
associated with infection of a specific host (human vs. bovine) (17).
It has been proposed that Agr polymorphism might contribute to exclu-
sion of other strains of the same species fromthe colonization or infection site
or to isolate certain staphylococcal populations (25, 34), but there is no clear
evidence for this phenomenon. From an epidemiological standpoint, it is
difficult to untangle whether Agr plays an active or a passive role in coloniza-
tion and disease, owing to the covariance of many other virulence-associated
genes among Agr groups. Further complicating the studies are questions
about whether or not the agr system is even expressed in many environments
in vivo, particularly in areas of lowcell density. Indeed, RNAIII expression was
frequently not detected in the lungs of cystic fibrosis (CF) patients (20).
Nevertheless, Goerke et al. (22) addressed this issue of Agr interference in a
study of cystic fibrosis patients and healthy controls. They found that strain
replacement in CF patients was accompanied by a change in the Agr group
80%of the time. However, in 10 cases where more than one strain of S. aureus
was detected in the patients, 6 consisted of co-colonization by strains from
interfering Agr groups, suggesting that Agr-group-based interference was not
an active phenomenon in these cases. This was a much more difficult issue to
assess in healthy individuals. Although the authors found several cases where
healthy carriers became sequentially colonized by strains belonging to an
J
.
M
.
Y
A
R
W
O
O
D
210
interfering Agr group and were rarely co-colonized by strains frominterfering
Agr groups, they were unable to rule out the possibility of a time gap between
the loss of one strain and colonization by a second. However, in three cases
where a person was colonized by more than one strain, two cases were of
strains belonging to the same Agr group and the third consisted of isolates
fromAgr groups I and IV, which are only partly inhibitory for each other. Kahl
et al. (26) found that the success of isolates from CF patients, as measured by
their prevalence, persistence, or ability to co-colonize with other S. aureus
strains, was independent of Agr group specificity. In addition, the Agr speci-
ficity group distribution did not differ significantly when compared with
isolates from other infection types and from healthy carriers. Lina et al. (30)
did show that, among healthy volunteers, S. aureus colonization rates were
negatively correlated with the rate of colonization by Corynebacteriumspp. and
S. epidermidis. Only one type of S. aureus agr allele was detected in each carrier,
but the results did not generally support in vitro Agr-specific cross-inhibition
experiments. For example, the relatively frequent co-isolation of Agr groups
not mutually inhibitory, such as Agr groups I and IV, should have been
observed if cross-inhibition was an important determinant in colonization,
but this was not the case. Jarraud et al. (24) concluded that the data do not
necessarily support a direct role for Agr group in the type of human disease
caused, but may instead reflect an ancient evolutionary division of S. aureus in
which other determinants developed in subsequent staphylococcal lineages
facilitate pathogenesis and disease type. Regardless, additional studies are
needed to address the potential functions of various Agr groups in staphylo-
coccal colonization or disease pathogenesis, as the studies conducted so far
have been limited to very few disease types.
It has also been proposed that the ability of S. epidermidis to generally
inhibit quorum sensing by most S. aureus strains may help to explain why
S. epidermidis predominates on the skin and in infections of indwelling
medical devices (38). This assumes that the agr locus (i) is expressed in
these loci, and (ii) facilitates colonization, ideas for which little either
supporting or conflicting evidence exists.
Agr REGULATION OF VIRULENCE IN THE CONTEXT OF OTHER
REGULATORS AND THE ENVIRONMENT IN VIVO
Unarguably, agr expression must occur within the context of the extra-
cellular environment, growth phase, and overall cell metabolism. Early on
in the study of Agr regulation, it was recognized that there is a temporal
211
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
program of exotoxin expression that is unaltered by the timing of RNAIII
expression. For instance, Vandenesch et al. (56) found that hla transcription
could be delayed as much as six hours after RNAIII reached its maximum
level. In the case of Agr-activated toxins, several other appropriate environ-
mental conditions are required for full expression of these proteins, at
least in vitro. These conditions include neutral pH, sufficient oxygen and
carbon dioxide concentrations, and elevated protein. Thus, interaction of
Agr with other regulatory systems that respond to diverse signal inputs is
essential for a complete understanding of the staphylococcal quorum
response.
SrrAB
Yarwood et al. (69) proposed one potential link between cellular meta-
bolism and quorum sensing: the two-component system SrrAB (also
described independently by Throup et al. (52)). This homolog of the
Bacillus subtilis ResDE two-component system regulates certain virulence
factors in response to oxygen concentration; srrAB mutants are growth-
defective in the absence of oxygen, presumably owing to the role of srrAB in
activating enzymes necessary for synthesis of alternative electron accep-
tors. It is possible that SrrAB responds to an intermediate of the electron-
transport pathway in staphylococci, such as reduced menaquinone, thus
providing a direct link to cellular energy metabolism.
The sae locus
A second regulatory locus, sae, encoding a two-component regulatory
system, SaeRS, and two additional genes (saeP and saeQ) was identified
with effects on multiple staphylococcal virulence factors (18, 19, 34, 35).
Nuclease and coagulase in particular were repressed at the transcriptional
level in an saeRS mutant. SaeRS expression does not appear to affect
RNAIII production and instead is thought to act together with or down-
stream of Agr. Coincident with the onset of RNAIII synthesis, one early sae
locus transcript disappears and three new ones appear. This switch is
blocked in an agr mutant as well as by diverse environmental conditions,
including NaCl, pH<6, and subinhibitory levels of clindamycin, condi-
tions that were previously shown to affect exoprotein synthesis. Thus, it has
been proposed that sae coordinates quorum sensing with some environ-
mental signals (35).
J
.
M
.
Y
A
R
W
O
O
D
212
ArlRS
ArlRS is a two-component regulatory system that regulates autolysis
and expression of the NorA multidrug efflux pump (13). Increased expres-
sion of agr was observed in an arl mutant (14). Correspondingly, the
expression of several Agr-activated genes were increased in arl mutants,
such as a-toxin, lipase, and serine protease. Interestingly, spa (Protein A)
expression was also increased in arl mutants, an effect apparently mediated
in part by SarA (discussed later in this chapter) as well as by agr. Expression
of arlRS was itself reduced in an agr mutant, so ArlRS and Agr apparently
form an autorepression circuit in that arlRS counters agr induction and
vice versa.
The s
B
factor
The staphylococcal alternative sigma factor, s
B
, is activated by environ-
mental stresses (such as ethanol treatment) and energy depletion;
s
B
appears to act generally opposite of Agr as it represses many secreted
virulence factors while upregulating some surface adhesins (3). In certain
conditions, s
B
may contribute to biofilm development by staphylococci.
However, despite its apparent role in regulation of virulence factors and
biofilm formation, a clear role for s
B
in virulence has not yet been estab-
lished. Perhaps this is due to the limited number of studies addressing
s
B
-controlled virulence, particularly in models of long-term, chronic infec-
tions. No direct link between s
B
and agr expression has been identified,
although lower levels of the RNAIII transcript have been found in strains
with an intact sigB operon (4, 23).
ClpX and ClpP
Clp proteolytic complexes have been shown to be important for viru-
lence and survival by several pathogenic bacteria under stress conditions.
A ClpX or ClpP mutant of S. aureus is attenuated in virulence, and both
ClpX and ClpP are important for growth under oxidative stress or at low
temperature (16). Interestingly, absence of either ClpX or ClpP reduces the
transcription of RNAIII and AIP activity. Correspondingly, expression of
several extracellular proteins, including a-toxin, are reduced in Clp
mutants. No mechanism has yet been identified whereby the Clp proteases
exert this regulation on the Agr system, though it may serve as an impor-
tant regulatory pathway for virulence regulation under certain conditions.
213
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
The SarA family
No consensus sequence has been identified as a regulatory target for
Agr among Agr-regulated genes. Presumably, then, RNAIII must fre-
quently exert its regulatory activity via intermediary proteins (37). Some
of these intermediary regulators likely belong to the SarA family, an
expanding group of homologous winged helixturnhelix transcription
factors (reviewed in (5)). This family includes SarA, SarR, SarT, SarS,
SarU, SarY, and Rot. Many of these proteins are an integral part of viru-
lence gene regulation, often influencing expression of the agr operon itself
and sometimes even opposing Agr regulation of certain genes. SarA, the
most extensively studied member of this family, is thought to regulate agr
expression directly via binding to the agr promoter region. The agr tran-
scripts RNAII and RNAIII are both diminished in sarA mutants; a 29 bp
sequence located between the agr P2 and P3 promoters, identified as a SarA
binding site, is required for the transcription of RNAII in S. aureus (6). In
addition, hemolysin production can be restored to wild-type levels in sarA
mutants by induction of RNAIII supplied in trans. However, SarA also
mediates virulence gene regulation independently of Agr, activating some
genes also activated by Agr (e.g. hemolysins) while repressing others also
activated by Agr (e.g. proteases). Many of these effects are likely to be direct
regulatory actions, as several SarA-regulated genes have a consensus SarA
binding site upstream. The relationship to Agr of other members of the
SarA family is diverse, and likely helps to fine-tune the quorumresponse in
response to other signals. SarR downregulates SarA expression through
binding of the sarA promoter, in effect repressing Agr expression, whereas
SarU activates RNAIII transcription. SarS activates protein A synthesis;
SarT represses a-toxin production. Experimental evidence suggests that
Agr likely represses spa by suppressing sarS transcription. In contrast,
Agr does not appear to upregulate hla expression via repression of sarT.
The global regulator of virulence, Rot, generally acts to counter Agr activity
in repressing some secreted proteins and activating surface-associated
virulence genes (43).
REGULATION BY Agr IN VIVO
The regulation of virulence genes by Agr appears to be considerably
more complex in vivo than was understood from experimentation in vitro.
The contribution of Agr to virulence may depend heavily upon the site of
infection; and one animal model of infection will imply a role for Agr very
J
.
M
.
Y
A
R
W
O
O
D
214
different from that found by using another animal model. For instance,
growth and virulence gene regulation in the kidney, where both pH and
osmolarity are high, is vastly different from growth in an endocarditis
vegetation. Even within an endocarditis vegetation, exposure of cells to
nutrients is much greater on the surface of the vegetation than in the
depths of the vegetation.
When agr expression has been examined in vivo, a complex picture of
Agr regulation has emerged. Goerke et al. (20) found that RNAIII expres-
sion by staphylococci in the sputum of cystic fibrosis patients was highly
variable and did not correlate with expression of hla, which is strongly
regulated by Agr in vitro. The same group also determined that mutation
of agr did not affect hla expression in a guinea pig model of device-related
infection (21). They found that, in the two strains examined, RNAIII
expression was significantly lower in vivo than during growth in vitro and
negatively correlated with bacterial densities. Yarwood et al. (68) used
microarray analysis of virulence gene expression in vivo in a rabbit model
of TSS. Surprisingly, despite the repression of agr expression in serum
cultures and in vivo, several secreted virulence factors were upregulated,
sufficient to cause acute disease in the animals. There was also little
difference between an agr wild type and an agr mutant in expression of
exotoxins and causation of disease. This lack of correlation between agr
expression and toxin expression in vivo suggests the presence of other
regulatory mechanisms that can up- or downregulate many virulence fac-
tors independently of Agr activity.
Xiong et al. (65) did confirm several aspects of the in vitro model of the
Agr regulatory circuit in an experimental endocarditis model, as well as
identifying an intriguing exception. As might be expected, maximal RNAII
activation in vegetations occurred early, followed by increasing RNAIII
expression. This correlated with increased bacterial densities within the
vegetations, which were higher in the vegetations than in kidney or spleen
tissues. The authors concluded that RNAIII activation in vivo is time- and
cell-density-dependent, and perhaps also tissue-specific. Surprisingly,
RNAIII activation was also observed in vegetations formed by using Agr
signaling mutants (though to a lesser extent than in the wild type), suggest-
ing that an RNAII-independent mechanism of RNAIII activation may exist
in vivo. In addition, there was no correlation between RNAII promoter
activity and vegetation densities.
Rothfork et al. (42) hypothesized that bacterial clumping mediated by
fibrinogen might create a microenvironment within the host that promotes
density-dependent activation of agr expression independently of overall
215
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
bacterial burden. The plasma protein fibrinogen is an important compo-
nent of the acute inflammatory response. It helps to promote neutrophil
migration and adhesion, induction of cytokine synthesis, coating of foreign
bodies, walling off of infection sites, and initiation of wound healing.
S. aureus possesses several cell-associated and secreted factors that directly
interact with fibrinogen or its soluble precursor, fibrin. Rothfork et al.
showed that, in a murine abscess model, transient depletion of the animal
of fibrinogen significantly reduced the bacterial burden and overall mor-
bidity and mortality in the animals. This was not observed in infection by an
agr mutant. Fibrinogen depletion also inhibited activation in vivo of RNAIII
transcription, as well as expression of the quorum-activated virulence
factors a-toxin and capsule. The data suggest that fibrinogen-mediated
clumping is sufficient to concentrate the autoinducer and promote quorum
sensing. The same effects could also be mediated by fibronectin. This study
provides an important mechanistic link between the innate immune
response and pathogenesis of S. aureus, as well as insight into regulation
of agr expression in vivo.
Agr AND STAPHYLOCOCCAL BIOFILMS
The concept of biofilm formation and its relevance to human disease
has been well reviewed elsewhere (8, 39). In general, biofilms are a complex
community of bacteria enclosed in a matrix that is usually self-produced.
They are of particular clinical relevance, as biofilm-associated bacteria are
more resistant to, or tolerant of, antimicrobial treatment and are also
resistant to clearance by the host immune system. Many staphylococcal
infections appear to be associated with, or resemble, biofilms: these include
endocarditis, osteomyelitis, and even some skin infections. However, the
most common staphylococcal biofilm-associated infections are associated
with implanted medical devices, such as intravascular catheters. These
infections are most commonly caused by S. epidermidis; in fact, the ability
to formbiofilms is considered the primary virulence factor of S. epidermidis.
Quorumsensing in staphylococcal biofilms is an important, emerging area
of investigation, as many factors thought to be important in adherence of
staphylococci to biological surfaces are under the control of the Agr system.
Furthermore, emerging evidence suggests that the Agr phenotype of
staphylococci might influence the development and outcome of long-
term, chronic staphylococcal infections.
There are at least three important stages in staphylococcal biofilm
development and behavior. The first is the initial attachment of cells to a
J
.
M
.
Y
A
R
W
O
O
D
216
biotic or abiotic surface, usually mediated by surface adhesins. The second,
or maturation, stage involves the accumulation of cells into multilayered
clusters enclosed in a self-produced matrix. The matrix usually consists of
the polysaccharide intercellular adhesion (PIA). In the host milieu, how-
ever, it is not entirely clear whether PIAis in fact required to forma biofilm-
like community. One might imagine that, through intracellular binding
mediated by host cell matrix components, a biofilm-like structure could be
achieved together with the important characteristics of a biofilm (nutri-
tional gradients, protection from host immune factors and predation, etc.)
without the presence of PIA. The third stage of biofilm development
involves detachment of cells from the biofilm; this may facilitate the
colonization of sites distant from the original infection site. The factors
contributing to detachment are both external and internal to the biofilm.
Physical factors such as shear and physical disruption of the biofilminduce
large-scale detachment; emerging evidence suggests that biofilm-
associated bacteria may also actively promote their own detachment.
There are potential mechanisms whereby Agr expression might impact
each of these stages of biofilm development, based on a very limited
number of studies in vitro.
Initial attachment
There appear to be two general mechanisms by whichstaphylococci attach
to a surface, as illustrated by colonization of an intravascular catheter. During
insertion of the catheter, attachment to the naked polymer surface occurs
through non-specific, physiochemical interactions, such as hydrophobic inter-
actions. Subsequent to implantation, the catheter surface becomes coated with
components of the host matrix, such as fibrinogen, fibronectin, and collagen.
This facilitates more specific interactions between the staphylococci and what
is nowa biological surface mediated by specific receptors on the staphylococci,
such as the fibrinogen- and fibronectin-binding proteins. Several of these
specific staphylococcal receptors are negatively regulated by Agr. In some
staphylococcal species large proteins that might mediate non-specific, hydro-
phobic interactions with the uncoated polymer surface are also regulated by
Agr (e.g. the autolysin AtlE in S. epidermidis).
Maturation
Little evidence exists for or against a contribution of the Agr system to
the maturation of biofilms. In particular, the expression of PIA is not
217
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
regulated by Agr (59). However, it is conceivable that accumulation in the
host of staphylococcal cells in a biofilm, particularly through mutual bind-
ing by adjacent bacteria of host factors, would be enhanced by the contin-
ued expression of surface adhesins.
Detachment
Vuong et al. (62) have proposed that expression of d-toxin, a protein
with surfactant properties and encoded by the agr locus, might contribute to
detachment of cells from a biofilm. Thus, in combination with the down-
regulation of surface adhesins, Agr may well play an important role in
facilitating release of staphylococcal cells fromthe biofilm. Indeed, we have
observed enhanced detachment of agr-expressing cells from a biofilm (67),
but have not yet been able to confirm the contribution of Agr to this
phenomenon. Large-scale detachment events would also be expected to
influence mature biofilm structure, at least temporarily, thus potentially
influencing the maturation stage of biofilm development as well.
At first glance, the various existing literature regarding the role of the
Agr system in staphylococcal biofilm formation and behavior appear some-
what inconsistent. Several investigators have grown different strains under
different growth conditions and, not surprisingly, have obtained different
results. In a survey of 105 S. aureus strains, Vuong et al. (62) found a strong
correlation between lack of Agr activity (as measured by lack of d-toxin
production) and ability to adhere to polystyrene. The authors attributed
this, at least in part, to the surfactant properties of d-toxin, as addition of
increasing concentrations of d-toxin decreased attachment of S. aureus to
polystyrene. In two studies with somewhat conflicting results, Shenkman
et al. first found that agr mutants showed (compared with wild-type
S. aureus) increased adherence to immobilized fibrinogen, increased induc-
tion of platelet aggregation, and had little impact on adherence to immo-
bilized fibronectin, von Willebrand factors, bovine corneal extracellular
matrix, and endothelial cells (47). The difference in adherence properties
developed primarily under flow conditions, suggesting different adhesion
mechanisms under static and flow conditions. In the second study, it was
concluded that RNAIII downregulated S. aureus adherence to fibrinogen
under static conditions while upregulating S. aureus adherence to fibronec-
tin and endothelial cells under both static and flow conditions (48). The
authors also found that the contribution of activated platelets in S. aureus
adherence to endothelial cells was downregulated by RNAIII, likely due to
decreased adherence to fibrinogen, a plasma protein thought to bridge
J
.
M
.
Y
A
R
W
O
O
D
218
S. aureus, platelets and endothelial cells. Pratten et al. (40) showed pleo-
tropic effects of both the agr and sar operons on expression of surface
molecules responsible for binding to substrata.
To address whether the variable results found in the literature were the
result of different strains or different growth conditions, Yarwood et al. (67)
grewthe same isogenic pair (wild-type versus agr mutant) in biofilms under
several conditions. In this study, the contribution of Agr to biofilm devel-
opment was found to be dependent on growth conditions. In some cases,
agr expression decreased bacterial attachment and biofilm formation.
Under other conditions, it enhanced biofilm formation or, in the case of
flow-cell biofilms, appeared to have no effect at all, even when clearly
expressed (Figure 9.3; for additional images see (67)). Not surprisingly,
the nature of the growth surface appears to be especially important in
detecting a contribution of Agr to biofilm development, for example,
whether or not the surface is coated with biologically relevant proteins
Figure 9.3. Three-dimensional reconstruction of a Staphylococcus aureus biofilm. Cells
expressing a quorum-controlled green fluorescent protein reporter are green; the
remaining biofilm is red from staining with propidium iodide. Each side of the grid
represents about 600mm. (See also color plate section.)
219
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
with which staphylococci can interact and receptors for which might be
regulated by agr.
In the first study of its kind to address directly the biofilm-forming
capabilities of agr mutants in vivo, Vuong et al. (61) found that a
S. epidermidis agr mutant showed increased binding to epithelial cells and a
higher colonization rate in a rabbit model of an indwelling medical-device-
related infection. They also confirmed that deletion of agr or inhibition of Agr
activity led to thicker biofilms in vitro. These results were consistent with those
of a study conducted earlier by the same laboratory group, in which a
S. epidermidis agr mutant showed increased primary attachment and biofilm
formation, as well as expression of the cell-surface-associated autolysin AtlE
(59). (Repetitive sequences inAtlEare thought to interact hydrophobically with
abiotic surfaces.) As in S. aureus, production of PIA by the S. epidermidis agr
mutant was similar to that of the wild type. As expected, the agr mutant lacked
d-hemolysin production. Addition of increasing concentrations of d-toxin
resulted in decreased attachment of S. epidermidis cells to polystyrene, where
10mg ml
1
d-toxin was sufficient to reduce biofilm formation of the agr
mutant strain to the same levels found in the agr wild-type strain.
Interestingly, there may also be some role for agr expression in the
resistance of staphylococcal biofilms to antibiotic exposure. Under condi-
tions where an agr mutant formed a smaller biofilm than its wild-type
parent did, the mutant was also more sensitive to rifampicin treatment,
but not to oxacillin (67). The basis for this variation in sensitivity is
unknown, although there is precedent for the regulation of other antibiotic
resistance mechanisms by Agr. Regulation of NorA, a multidrug efflux
pump involved in resistance to quinolones, by the DNA-binding protein
NorR, was found to require an intact Agr system (53).
Thus, the role of Agr in biofilm development and behavior varies from
species to species and from one environment to another, although some
consistent themes are emerging. It will be critical in future investigations to
identify those growth conditions that best mimic the environment in vivo in
order to most effectively study Agr regulation of virulence factors in bio-
films. Even more desirable are studies examining the contribution of Agr to
biofilm formation in vivo, the first of which is described in the next section.
Agr VARIANTS AND THEIR ROLE IN STAPHYLOCOCCAL
PATHOGENESIS
Agr variants (cells either overexpressing or underexpressing Agr com-
pared with the parental strain) have been frequently isolated from cultures
J
.
M
.
Y
A
R
W
O
O
D
220
in vitro, suggesting that staphylococci maintain some capacity to alter their
Agr phenotype or maintain Agr-negative subpopulations. Somerville et al.
(50) found that repeated passage of S. aureus in vitro resulted in the loss
of Agr function in a large percentage of the population, along with corres-
ponding hemolytic and aconistase activity. The authors hypothesized
that frequent mutations of agr create a mixed population of bacteria, with
some cells expressing colonization factors while others tend to express
secreted exotoxins. Under a particular environment with specific ecological
and/or immunological selection, the Agr variant best able to adapt would
emerge.
Agr mutants are frequently found among clinical isolates. Vuong et al.
(61) showed that the percentage of strains with defective quorum-sensing
systems was significantly higher among isolates from patients with infec-
tions of joint prostheses than among isolates from the skin of healthy
controls (36% and 5%, respectively). This same group also found that
26% of 105 S. aureus isolates failed to produce d-toxin, indicating that
they were deficient in quorum-sensing-mediated regulation (62). When
Goerke et al. (20) isolated staphylococci from the lungs of cystic fibrosis
patients, not only did the strains generally express lowlevels of RNAIII, but
several isolates were also found to be Agr-negative. Fowler et al. (15) showed
that the percentage of S. aureus isolates recovered from patients with
persistent bacteremia with defective d-toxin production (a consistent indi-
cator of Agr activity) was higher than in isolates from patients with resolv-
ing bacteremia (71% and 39%, respectively). The authors postulated that
lack of agr expression might contribute to persistent bacteremia through
the increased expression of the S. aureus surface adhesion gene, fnbA, in
these mutants. The fibronectin-binding protein encoded by fnbA has been
shown to enhance S. aureus adhesion to, invasion of, and persistence within
endothelial cells. Intracellular invasion may contribute then to resistance to
antibiotics, as vancomycin penetrates poorly into endothelial cells. Thus,
lack of agr expression may facilitate a protected intracellular reservoir for
S. aureus. Indeed, an agr mutant was incapable of escape from the endo-
some (49) or of inducing apoptosis (64), suggesting a prominent role for
Agr in invasion of and persistence in host cells.
One area of particular concern in staphylococcal pathogenesis is the
emergence of staphylococci with intermediate resistance to glycopeptide
antibiotics (GISA). Interestingly, GISA are frequently isolated from
biomedical-device-related infections, which are also likely to be biofilm-
associated; these same GISA have been shown to be predominantly Agr-
negative (45). The same study also suggested that loss of Agr function
221
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
might in fact contribute to the development of vancomycin tolerance, an
intriguing idea yet to be confirmed.
Schwan et al. (46) examined the behavior of mixed populations of
hyperhemolytic, hemolytic, and non-hemolytic variants in a murine
abscess model of infection. They found that the percentage of non-
hemolytic variants, likely representing Agr-negative bacteria, recovered
from the wound increased over time, whereas the number of hyperhemolytic
variants (Agr overexpressors) decreased dramatically over the same time
period. A wound infection model demonstrated the same trend, though to
a lesser degree. In contrast, in a model of systemic infection, hemolytic
variants seemed to be favored in isolates recovered from murine livers and
spleens. Thus, Agr activity likely facilitates survival and pathogenesis in
some host environments, but not in others. Additional studies will be of
great importance in monitoring Agr phenotypes of clinical isolates from
various infection types (preferably multiple isolates from each patient) and
corresponding this to disease progression and outcome.
Although there is only scattered evidence for this idea, one can imagine
that Agr-negative variants are better suited to biofilm formation and long-
term, chronic infection as (i) they tend to express the surface adhesins that
mediate cell-to-cell and cell-to-surface interactions, while downregulating
factors that may facilitate detachment, such as d-toxin, and (ii) they tend to
express more immuno-evasive factors, such as protein A, than immuno-
stimulatory ones (such as the superantigens). In addressing the first idea,
our laboratory has found that Agr-negative variants become the predomi-
nant form in biofilms grown in a serum-based medium (66). It is not yet
clear whether this is due to a selective pressure against Agr-positive cells,
increased generation of Agr variants in the biofilm, active detachment of
cells expressing Agr, or some combination of all three. Preliminary data
also indicate that the Agr-positive population is not completely lost from
the biofilm (J. M. Yarwood, unpublished data), suggesting that there may
be some mechanism to retain the capability to express invasive factors at an
appropriate stage of infection. Indeed, we have detected the frequent
detachment of cells expressing agr from the biofilm (67). This may have
important clinical implications, as detaching cells expressing agr are also
likely to be expressing extracellular virulence factors important in causing
acute infection.
One potential model of S. aureus Agr evolution in the context of a
chronically infected host is presented in Figure 9.4. Upon establishment
of infection, mutations (usually point mutations) accumulate in the agr loci
of S. aureus cells. These mutations result in the conversion of a significant
J
.
M
.
Y
A
R
W
O
O
D
222
part of the population to a quorum-sensing negative phenotype. The Agr-
negative phenotype confers some protection on the staphylococcal popula-
tion as a whole, due to increased expression of immuno-evasive factors, and
facilitates attachment and accumulation through increased expression of
host protein-binding factors. This protected environment is conducive to
the continued growth of staphylococci and additional accumulation of
mutations in the agr locus. On very rare occasions, appropriate mutations
are acquired by the agr-negative variant to return functionality of the agr
locus, such as alteration of the AgrC receptor to recognize an AIP variant.
In an extremely rare case, it is also possible that mutations in all appro-
priate areas of the agr locus are gained simultaneously to create a function-
ally unique Agr specificity group. Along with other Agr-positive cells, these
1
2
3
4
5
Figure 9.4. Model of Agr variant generation in Staphylococcus (see text for additional
description). Colonization of a surface by an Agr-positive strain (gray cells) leads to
microcolony development (step 1). Agr-negative variants (white cells) arise in the biofilm
(step 2) through point mutation, genetic rearrangement, transposon insertion, etc.
Occasionally, an Agr-negative variant acquires compensatory mutations to recognize a
variant AIP and regains a functional Agr system (banded cells), albeit one distinct from the
original strain (step 3). This new Agr specificity group strain detaches, along with other
Agr-positive cells (step 4). Ecological pressures are then exerted over an extended time
period (step 5). These pressures may select for certain virulence characteristics of the
new strain, which is likely also divergent from the original strain in other virulence
factors, and the strain emerges as a major Agr specificity group.
223
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
new Agr specificity-group cells detach from the biofilm through expression of
invasive virulence factors or production of d-toxin and establish infection
elsewhere in the host or, alternatively, colonize a secondary host. In some
cases, appropriate ecological pressures are present to allow emergence of this
new Agr specificity group from among the established groups. The combined
rarity of these events accumulation of several, eventually positive mutations,
and selection for any emergent Agr specificity group would only give rise to a
major new S. aureus Agr group very infrequently. This would be consistent
with the identification of only four distinct S. aureus agr groups thus far,
despite frequent mutation of the agr locus. The driving force behind this
cycle is, in part, the advantage conferred by maintaining a mixed population
of cells, where Agr-negative variants prevent recognition by immune surveil-
lance and cells expressing agr provide additional nutrient sources through host
tissue degradation or facilitate escape from the localized infection at appro-
priate times. Thus, the emergence of distinct Agr groups may be a by-product
of this mode of Agr evolution in which the generation of variants itself is
important. However, it is noteworthy that this Agr-negative phenotype is often
generated through non-reversible mutation of the agr locus, rather than by a
reversible, conditional switching of agr expression on and off. This is consis-
tent withsome evolutionary advantage for generation of distinct Agr specificity
groups.
INHIBITION OF STAPHYLOCOCCAL QUORUM SENSING
AS A THERAPEUTIC TOOL
The use of inhibitors of quorum sensing has been proposed as one
mechanism for controlling staphylococcal infections (25; see also Chapter 4).
In a skin abscess model it was shown that co-administration of the synthetic
Agr group II AIP together with the bacterial inoculation significantly attenu-
ated aninfectioncaused by anAgr group I strain(31). However, it has also been
found that cross-inhibiting pheromones mimic agr mutations inboth S. aureus
and S. epidermidis and enhance biofilmformation (59, 61, 62). This result calls
for extreme caution in the use of quorum-sensing inhibitors, as it is conceiv-
able that such treatments, while mitigating the acute phase of infections,
might facilitate chronic, biofilm-associated infections.
RIP/RAP
With some degree of controversy, a second quorum-sensing systemhas
been described in S. aureus that is proposed to regulate Agr activity. This
J
.
M
.
Y
A
R
W
O
O
D
224
system (see (1, 7, 34) and other studies by N. Balaban and colleagues)
consists of the autoinducer RNAIII-activating protein (RAP) and its target
molecule TRAP. RAP is described as an ortholog of the ribosomal protein
L2 and as being synthesized early in growth. Immunization against RAP
was shown to mitigate pathology in a murine cutaneous S. aureus infection
model. Reportedly, when RAP reaches a threshold concentration, it induces
the histidine phosphorylation of the membrane protein TRAP. This event
leads to the upregulation of agr transcription through an undescribed
mechanism. Once AIP is made, it has been reported to lead to the down-
regulation of TRAP phosphorylation. A protein produced by S. xylosus and
resembling, in the N-terminal sequence of RAP, RNAIII inhibiting peptide
(RIP) has been reported to act as an agonist of TRAP, inhibiting its phos-
phorylation and, consequently, agr expression. Treatment with synthetic
RIP inhibits several types of S. aureus and S. epidermidis infections, includ-
ing those that are biofilm-related or caused by multiple drug-resistant
staphylococci. Apparently, these therapeutic effects can be observed when
RIP is applied locally or systemically.
The RAP/TRAP/RIP story thus far is somewhat unsatisfying. Little has
been described regarding the properties of RIP, and even its amino-acid
sequence and structure remain questionable. It is also conceivable that,
when used at high concentrations, RIP has sufficient amphipathic, or
surfactant, properties to prevent bacterial attachment, not unlike many
other proteins. All of the experiments in vivo have administered RIP before
or coincident with bacterial challenge and it is clear that, in these cases, RIP
inhibits bacterial adherence to surfaces. However, there is no evidence that
RIP has any effect against established biofilm, and few data are available as
to the overall effect of RIP on bacterial physiology or virulence. It is also not
clear whether a RIP-impregnated intravascular catheter would continue to
inhibit staphylococcal adhesion against the more or less continual chal-
lenge that likely occurs in vivo, from either the epithelium or transient
bacteremias. Implanted devices are soon coated with host matrix proteins, a
fact that limits the efficacy of many device surface treatments as the now
biotic surface provides several protein-receptor-specific targets for staphy-
lococci to bind to. Furthermore, very little is known as to the regulatory
targets of this system. Besides being reported to inhibit Agr activation, no
other gene targets have been conclusively identified. This is particularly
unsatisfying, as inhibition of Agr activity would be expected to increase
bacterial adhesion, yet, in fact, the RIP-treated cells are less likely to adhere;
this result suggests that RIP may in fact simply be acting as a surfactant-
like molecule. Finally, other laboratory groups have been unable to detect
225
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
the RNAIII-activating activity in supernatants of agr-null strains, despite
the presumed presence of RAP, contributing to the controversy as to the
true nature of this molecule (34, 36).
CONCLUSION
As more studies regarding quorum sensing emerge, it is becoming clear
that the role and expression of Agr is intimately tied to environmental condi-
tions, infection type, and the input of other virulence regulators. As with many
biological processes, there are few rules regarding quorum sensing to which
there are not important exceptions. However, the data also suggest that, under
certain conditions, quorum sensing may have an important part to play in
everything from influencing the chronic nature of disease to staphylococcal
metabolism. Thus, additional studies are sorely needed to flesh out the role of
Agr in staphylococcal virulence, particularly through animal models of infec-
tion and epidemiological monitoring of Agr variants. Even given the caveats
associated with the use of Agr inhibitors in staphylococcal infections, inter-
ference with quorum signaling may yet prove to be beneficial in treatment of
certain staphylococcal infections, particularly those in which inhibition of
toxin production would be of immediate benefit.
ACKNOWLEDGEMENTS
J. M. Y. was supported by a Ruth L. Kirschstein National Research
Service Award from the National Institute of General Medical Sciences and
a gift fromthe Procter &Gamble Company. E. Peter Greenberg is gratefully
acknowledged for helpful discussions.
REFERENCES
1 Balaban, N., T. Goldkorn, R. T. Nhan et al. 1998. Autoinducer of virulence as a
target for vaccine and therapy against Staphylococcus aureus. Science 280: 43840.
2 Benito, Y., F. A. Kolb, P. Romby et al. 2000. Probing the structure of RNAIII, the
Staphylococcus aureus agr regulatory RNA, and identification of the RNA domain
involved in repression of protein A expression. RNA 6: 66879.
3 Bischoff, M., P. Dunman, J. Kormanec, et al. 2004. Microarray-based analysis of
the Staphylococcus aureus s
B
regulon. J. Bacteriol. 186: 408599.
4 Bischoff, M., J. M. Entenza and P. Giachino 2001. Influence of a functional sigB
operon on the global regulators sar and agr in Staphylococcus aureus. J. Bacteriol.
183: 51719.
J
.
M
.
Y
A
R
W
O
O
D
226
5 Cheung, A. L. and G. Zhang 2002. Global regulation of virulence determinants
in Staphylococcus aureus by the SarA protein family. Front. Biosci. 7: 182542.
6 Chien, Y., A. C. Manna, S. J. Projan and A. L. Cheung 1999. SarA, a
global regulator of virulence determinants in Staphylococcus aureus, binds to a
conserved motif essential for sar-dependent gene regulation. J. Biol. Chem. 274:
3716976.
7 DellAcqua, G., A. Giacometti, O. Cirioni et al. 2004. Suppression of drug-
resistant staphylococcal infections by the quorum-sensing inhibitor RNAIII-
inhibiting peptide. J. Infect. Dis. 190: 31820.
8 Donlan, R. M. and J. W. Costerton 2002. Biofilms: survival mechanisms of
clinically relevant microorganisms. Clin. Microbiol. Rev. 15: 16793.
9 Donvito, B., J. Etienne, L. Denoroy et al. 1997. Synergistic hemolytic activity of
Staphylococcus lugdunensis is mediated by three peptides encoded by a non-agr
genetic locus. Infect. Immun. 65: 95100.
10 Dufour, P., S. Jarraud, F. Vandenesch et al. 2002. High genetic variability of the
agr locus in Staphylococcus species. J. Bacteriol. 184: 11806.
11 Dunman, P. M., E. Murphy, S. Haney et al. 2001. Transcription profiling-based
identification of Staphylococcus aureus genes regulated by the agr and/or sarA loci.
J. Bacteriol. 183: 734153.
12 Fischetti, V. A., R. P. Novick, J. J. Ferretti, D. A. Portnoy and E. J. I. Rood 2000.
Gram-Positive Pathogens. Washington, DC: ASM Press.
13 Fournier, B., R. Aras and D. C. Hooper 2000. Expression of the multidrug
resistance transporter NorA from Staphylococcus aureus is modified by a two-
component regulatory system. J. Bacteriol. 182: 66471.
14 Fournier, B., A. Klier and G. Rapoport 2001. The two-component system ArlS-
ArlR is a regulator of virulence gene expression in Staphylococcus aureus. Mol.
15 Fowler, V. G., Jr., G. Sakoulas, L. M. McIntyre et al. 2004. Persistent bacteremia
due to methicillin-resistant Staphylococcus aureus infection is associated with agr
dysfunction and low-level in vitro resistance to thrombin-induced platelet micro-
biocidal protein. J. Infect. Dis. 190: 11409.
16 Frees, D., S. N. Qazi, P. J. Hill and H. Ingmer 2003. Alternative roles of ClpX and
ClpP in Staphylococcus aureus stress tolerance and virulence. Mol. Microbiol.
48: 156578.
17 Gilot, P. and W. van Leeuwen 2004. Comparative analysis of agr locus diversifi-
cation and overall genetic variability among bovine and human Staphylococcus
aureus isolates. J. Clin. Microbiol. 42: 12659.
18 Giraudo, A. T., A. Calzolari, A. A. Cataldi, C. Bogni and R. Nagel 1999. The sae
locus of Staphylococcus aureus encodes a two-component regulatory system.
FEMS Microbiol. Lett. 177: 1522.
19 Giraudo, A. T., A. L. Cheung and R. Nagel 1997. The sae locus of Staphylococcus
aureus controls exoprotein synthesis at the transcriptional level. Arch. Microbiol.
168: 538.
227
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
20 Goerke, C., S. Campana, M. G. Bayer et al. 2000. Direct quantitative transcript
analysis of the agr regulon of Staphylococcus aureus during human infection in
comparison to the expression profile in vitro. Infect. Immun. 68: 130411.
21 Goerke, C., U. Fluckiger, A. Steinhuber, W. Zimmerli and C. Wolz 2001. Impact
of the regulatory loci agr, sarA and sae of Staphylococcus aureus on the induction of
alpha-toxin during device-related infection resolved by direct quantitative tran-
script analysis. Mol. Microbiol. 40: 143947.
22 Goerke, C., M. Kummel, K. Dietz and C. Wolz 2003. Evaluation of intraspecies
interference due to agr polymorphism in Staphylococcus aureus during infection
and colonization. J. Infect. Dis. 188: 2506.
23 Horsburgh, M. J., J. L. Aish, I. J. White et al. 2002. s
B
modulates virulence
determinant expression and stress resistance: characterization of a functional
rsbU strain derived from Staphylococcus aureus 83254. J. Bacteriol. 184:
545767.
24 Jarraud, S., C. Mougel, J. Thioulouse et al. 2002. Relationships between
Staphylococcus aureus genetic background, virulence factors, agr groups (alleles),
and human disease. Infect. Immun. 70: 63141.
25 Ji, G., R. Beavis and R. P. Novick 1997. Bacterial interference caused by auto-
inducing peptide variants. Science 276: 202730.
26 Kahl, B. C., K. Becker, A. W. Friedrich et al. 2003. agr-dependent bacterial
interference has no impact on long-term colonization of Staphylococcus aureus
during persistent airway infection of cystic fibrosis patients. J. Clin. Microbiol. 41:
5199201.
27 Kies, S., C. Vuong, M. Hille et al. 2003. Control of antimicrobial peptide synthesis
by the agr quorum sensing system in Staphylococcus epidermidis: activity of the
antibiotic epidermin is regulated at the level of precursor peptide processing.
Peptides 24: 32938.
28 Li, M., M. Guan, X. F. Jiang et al. 2004. Genetic polymorphism of the accessory
gene regulator (agr) locus in Staphylococcus epidermidis and its association with
pathogenicity. J. Med. Microbiol. 53: 5459.
29 Li, S., S. Arvidson and R. Mollby 1997. Variation in the agr-dependent expression
of alpha-toxin and protein A among clinical isolates of Staphylococcus aureus from
patients with septicaemia. FEMS Microbiol. Lett. 152: 15561.
30 Lina, G., F. Boutite, A. Tristan et al. 2003. Bacterial competition for human nasal
cavity colonization: role of staphylococcal agr alleles. Appl. Environ. Microbiol. 69:
1823.
31 Mayville, P., G. Ji, R. Beavis et al. 1999. Structure-activity analysis of synthetic
autoinducing thiolactone peptides from Staphylococcus aureus responsible for
virulence. Proc. Natn. Acad. Sci. USA 96: 121823.
32 McNamara, P. J., K. C. Milligan-Monroe, S. Khalili and R. A. Proctor 2000.
Identification, cloning, and initial characterization of rot, a locus encoding a
regulator of virulence factor expression in Staphylococcus aureus. J. Bacteriol.
182: 3197203.
J
.
M
.
Y
A
R
W
O
O
D
228
33 Moise-Broder, P. A., G. Sakoulas, G. M. Eliopoulos et al. 2004. Accessory gene
regulator group II polymorphism in methicillin-resistant Staphylococcus aureus is
predictive of failure of vancomycin therapy. Clin. Infect. Dis. 38: 17005.
34 Novick, R. P. 2003. Autoinduction and signal transduction in the regulation of
staphylococcal virulence. Mol. Microbiol. 48: 142949.
35 Novick, R. P. and D. Jiang 2003. The staphylococcal saeRS system coordinates
environmental signals with agr quorum sensing. Microbiology 149: 270917.
36 Novick, R. P., H. F. Ross, A. M. N. S. Figueiredo et al. 2000. Activation and
inhibition of the staphylococcal AGR system. Science 287: 391.
37 Novick, R. P., H. F. Ross, S. J. Projan et al. 1993. Synthesis of staphylococcal
virulence factors is controlled by a regulatory RNA molecule. EMBO J. 12:
396775.
38 Otto, M. 2001. Staphylococcus aureus and Staphylococcus epidermidis peptide
pheromones produced by the accessory gene regulator system. Peptides 22:
16038.
39 Parsek, M. R. and P. K. Singh 2003. Bacterial biofilms: an emerging link to
disease pathogenesis. Annu. Rev. Microbiol. 57: 677701.
40 Pratten, J., S. J. Foster, P. F. Chan, M. Wilson and S. P. Nair 2001. Staphylococcus
aureus accessory regulators: expression within biofilms and effect on adhesion.
Microbes Infect. 3: 6337.
41 Projan, S. J. and R. P. Novick 1997. The molecular basis of pathogenicity. In K. B.
Crossley and G. L. Archer (eds.), The Staphylococci in Human Disease, pp. 5581.
New York: Churchill Livingstone.
42 Rothfork, J. M., S. Dessus-Babus, W. J. Van Wamel, A. L. Cheung and H. D.
Gresham 2003. Fibrinogen depletion attenuates Staphyloccocus aureus infection
by preventing density-dependent virulence gene up-regulation. J. Immunol. 171:
538995.
43 Said-Salim, B., P. M. Dunman, F. M. McAleese et al. 2003. Global regulation of
Staphylococcus aureus genes by Rot. J. Bacteriol. 185: 61019.
44 Sakinc, T., P. Kulczak, K. Henne and S. G. Gatermann 2004. Cloning of
an agr homologue of Staphylococcus saprophyticus. FEMS Microbiol. Lett. 237:
15761.
45 Sakoulas, G., G. M. Eliopoulos, R. C. Moellering, Jr. et al. 2002. Accessory gene
regulator (agr) locus in geographically diverse Staphylococcus aureus isolates
with reduced susceptibility to vancomycin. Antimicrob. Agents Chemother. 46:
1492502.
46 Schwan, W. R., M. H. Langhorne, H. D. Ritchie and C. K. Stover 2003. Loss of
hemolysin expression in Staphylococcus aureus agr mutants correlates with selec-
tive survival during mixed infections in murine abscesses and wounds. FEMS
Immunol. Med. Microbiol. 38: 238.
47 Shenkman, B., E. Rubinstein, A. L. Cheung et al. 2001. Adherence properties of
Staphylococcus aureus under static and flow conditions: roles of agr and sar loci,
platelets, and plasma ligands. Infect. Immun. 69: 44738.
229
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
48 Shenkman, B., D. Varon, I. Tamarin et al. 2002. Role of agr (RNAIII) in
Staphylococcus aureus adherence to fibrinogen, fibronectin, platelets and endothe-
lial cells under static and flow conditions. J. Med. Microbiol. 51: 74754.
49 Shompole, S., K. T. Henon, L. E. Liou et al. 2003. Biphasic intracellular expres-
sion of Staphylococcus aureus virulence factors and evidence for Agr-mediated
diffusion sensing. Mol. Microbiol. 49: 91927.
50 Somerville, G. A., S. B. Beres, J. R. Fitzgerald et al. 2002. In vitro serial passage of
Staphylococcus aureus: changes in physiology, virulence factor production, and agr
nucleotide sequence. J. Bacteriol. 184: 14307.
51 Tegmark, K., E. Morfeldt and S. Arvidson 1998. Regulation of agr-dependent
virulence genes in Staphylococcus aureus by RNAIII from coagulase-negative
staphylococci. J. Bacteriol. 180: 31816.
52 Throup, J. P., F. Zappacosta, R. D. Lunsford et al. 2001. The srhSR gene pair from
Staphylococcus aureus: genomic and proteomic approaches to the identification
and characterization of gene function. Biochemistry 40: 10392401.
53 Truong-Bolduc, Q. C., X. Zhang and D. C. Hooper 2003. Characterization of
NorR protein, a multifunctional regulator of norA expression in Staphylococcus
aureus. J. Bacteriol. 185: 312738.
54 Tseng, C. W., S. Zhang and G. C. Stewart 2004. Accessory gene regulator control
of staphylococcal enterotoxin d gene expression. J. Bacteriol. 186: 1793801.
55 Van Wamel, W. J., G. van Rossum, J. Verhoef, C. M. Vandenbroucke-Grauls and
A. C. Fluit 1998. Cloning and characterization of an accessory gene regulator
(agr)-like locus from Staphylococcus epidermidis. FEMS Microbiol. Lett. 163: 19.
56 Vandenesch, F., J. Kornblum and R. P. Novick 1991. A temporal signal, indepen-
dent of agr, is required for hla but not spa transcription in Staphylococcus aureus.
J. Bacteriol. 173: 631320.
57 Vandenesch, F., S. J. Projan, B. Kreiswirth, J. Etienne and R. P. Novick 1993. Agr-
related sequences in Staphylococcus lugdunensis. FEMS Microbiol. Lett. 111: 11522.
58 Vuong, C., M. Durr, A. B. Carmody et al. 2004. Regulated expression of patho-
gen-associated molecular pattern molecules in Staphylococcus epidermidis:
quorum-sensing determines pro-inflammatory capacity and production of
phenol-soluble modulins. Cell. Microbiol. 6: 7539.
59 Vuong, C., C. Gerke, G. A. Somerville, E. R. Fischer and M. Otto 2003. Quorum-
sensing control of biofilm factors in Staphylococcus epidermidis. J. Infect. Dis. 188:
70618.
60 Vuong, C., F. Gotz and M. Otto 2000.Construction and characterization of an agr
deletion mutant of Staphylococcus epidermidis. Infect. Immun. 68: 104853.
61 Vuong, C., S. Kocianova, Y. Yao, A. B. Carmody and M. Otto 2004. Increased
colonization of indwelling medical devices by quorum-sensing mutants of
Staphylococcus epidermidis in vivo. J. Infect. Dis. 190: 14981505.
62 Vuong, C., H. L. Saenz, F. Gotz and M. Otto 2000. Impact of the agr quorum-
sensing systemon adherence to polystyrene in Staphylococcus aureus. J. Infect. Dis.
182: 168893.
J
.
M
.
Y
A
R
W
O
O
D
230
63 Wamel, W. V., Y. -Q. Xiong, A. Bayer et al. 2002. Regulation of Staphylococcus
aureus type 5 capsular polysaccharides by agr and sarA in vitro and in an experi-
mental endocarditis model. Microb. Pathogen. 33: 739.
64 Wesson, C. A., L. E. Liou, K. M. Todd et al. 1998. Staphylococcus aureus Agr and
Sar global regulators influence internalization and induction of apoptosis. Infect.
Immun. 66: 523843.
65 Xiong, Y. Q., W. Van Wamel, C. C. Nast et al. 2002. Activation and transcriptional
interaction between agr RNAII and RNAIII in Staphylococcus aureus in vitro and
in an experimental endocarditis model. J. Infect. Dis. 186: 66877.
66 Yarwood, J. M. 2004. Quorum sensing in Staphylococcus aureus biofilms.
Presented at the 11th International Symposium on Staphylococci and
Staphylococcal Infections, Charleston, SC, October 2004.
67 Yarwood, J. M., D. J. Bartels, E. M. Volper and E. P. Greenberg 2004. Quorum
sensing in Staphylococcus aureus biofilms. J. Bacteriol. 186: 183850.
68 Yarwood, J. M., J. K. McCormick, M. L. Paustian, V. Kapur and P. M. Schlievert
2002. Repression of the Staphylococcus aureus accessory gene regulator in serum
and in vivo. J. Bacteriol. 184: 1095101.
69 Yarwood, J. M., J. K. McCormick and P. M. Schlievert 2001. Identification of a
novel two-component regulatory systemthat acts in global regulation of virulence
factors of Staphylococcus aureus. J. Bacteriol. 183: 111323.
231
S
T
A
P
H
Y
L
O
C
O
C
C
A
L
V
I
R
U
L
E
N
C
E
A
N
D
B
I
O
F
I
L
M
S
CHAPTER 10
Cell-density-dependent regulation
of streptococcal competence
M. Dilani Senadheera, Celine Levesque and Dennis G. Cvitkovitch
Dental Research Institute, University of Toronto,
Toronto, Canada
INTRODUCTION
A brief history
In the 1920s Frederick Griffith, a medical officer at the Ministry of
Health in Britain, made a significant discovery regarding Streptococcus
pneumoniae, a bacterium that caused a pneumonia epidemic in London.
While examining the strain variability within different groups of pneumo-
cocci, Griffith noted that an avirulent strain of the bacteriumcould revert to
the virulent type or remain unchanged following subculture (37). Because
this phenomenon enabled the bacterium to acquire a novel heritable
phenotype, Griffith coined the term transformation principle to describe
the phenotypic changes he observed.
In his classic experiment, Griffith studied a highly infective, encapsu-
lated S strain that formed smooth colonies, and an avirulent R strain, which
had no capsule and formed rough colonies when grown on blood agar (37).
When healthy mice were injected with the S strain, they died of septice-
mia, whereas separate admission of the R strain or the heat-killed
S strain appeared to be harmless. However, when the live R strain and
the heat-killed S strain were injected simultaneously, the mice died.
Surprisingly, when blood samples drawn from these dead animals were
analyzed, both R and S live strains were detected. Based on these results,
Griffith concluded that a transforming factor, present in the heat-killed
S strain, was able to transform an avirulent R strain into a capsulated,
virulent S strain.
Over the next few decades, Griffiths inspiring work on transformation
was followed up by a number of scientists. In 1944, by isolating various
233
components of a dead encapsulated virulent pneumococcal strain
and testing each component for its ability to transform, Oswald
T. Avery, Colin M. MacLeod, and Maclyn J. McCarty concluded that a
nucleic acid of the deoxyribonucleic type was the factor involved in trans-
formation (6). However, it was not until 1952 that Alfred Hershey and
Martha Chase conclusively demonstrated that DNA and not proteins are
the transforming factor by using bacteriophage infecting Escherichia coli
cells (46). They observed that, when bacteriophage containing
32
P-labeled
DNA and
35
S-labeled proteins were used to infect E. coli, only
32
P-DNA
entered the cell, whereas the
35
S-proteins stayed outside. Hence, they con-
cluded that DNA carried the hereditary information that coded for the
replication of the bacteriophage.
These pioneering experiments were aimed at elucidating the mech-
anism of genetic exchange, responsible for horizontal gene transfer. We
now know that bacteria can acquire foreign DNA through three distinct
mechanisms: transduction, which requires an intermediate bacteriophage;
conjugation, which involves cell-to-cell contact via an organelle called a
pilus; and by transformation, which is simply the uptake and integration
of free foreign DNA into the cells chromosome. Understanding these
mechanisms of genetic exchange, responsible for horizontal gene transfer,
is indispensable to understanding the genetic plasticity of bacteria and
to re-evaluating some of our therapeutic strategies against bacterial
infections.
In this chapter we examine how natural competence for genetic trans-
formation in streptococci is controlled by a cell-density-dependent quorum-
sensing system. Genetic competence is a transient physiological state that
enables a bacterium to recognize, process, and integrate foreign DNA into
its genome, leading to transformation (16, 23, 39, 69). We will begin by
focusing on S. pneumoniae, whose transformation process has been studied
for decades, and which has one of the best-characterized systems among
Gram-positive cocci. Consequently, we will also discuss the control of gene
transfer in oral streptococci, whose transformation mechanisms have also
been extensively studied. Interestingly, the oral cavity harbours over 400
different species, many of which can act as DNA donors and provide access
to a communal gene pool (26). As organisms colonizing the primary
portals of entry to the human body, the ability of oral microbes to uptake
DNA and transform is potentially of great significance to human health.
S. pneumoniae, whose ecological niche is the human nasopharynx, has been
previously shown to acquire B-lactam (e.g. penicillin) resistance from oral
streptococci such as Streptococcus mitis and S. oralis via transformation
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
234
(21, 22). With this in mind, we will discuss how genetic competence is
regulated in response to cell population density in streptococci.
Cell-density-dependent competence development
Bacterial transfer of genes via natural transformation depends on a
number of factors. The recipient cells have to be in a metabolically active,
competent state; it is also essential that free DNA (chromosomal or plas-
mid) be available for active uptake (16, 23, 39, 69). Because a cell utilizes a
considerable amount of its resources and energy for the processing, uptake
and incorporation of such DNA, it makes intuitive sense that competence is
synchronized with cell population density, which would be an indication of
the free homologous DNA that is available for uptake.
In the 1960s Pakula and Walczak (89) and Tomasz and Hotchkiss
(116) discovered that in streptococcal cultures the induction of genetic
competence took place only at a certain cell population density. In their
report, Tomasz and Hotchkiss describe two phases of competence: an
initial lag phase yielding a low number of competent cells, and a second
phase displaying the fast onset and explosive spread of the competent state
throughout the majority of cells in the culture. To test whether this latter
phase was activated by an external cell product, they conditioned a non-
competent cell culture with cell-free, filtered supernatant that was derived
from a competent culture. Interestingly, an immediate induction of com-
petence was observed in the treated, initially non-competent, culture.
Further testing revealed that a protein-like macromolecular compound
present in the supernatant of competent cultures acted as an activator
to induce competence in otherwise non-competent cells (89, 114116).
Following these pioneering experiments, the competence-activating
peptide molecule was successfully purified from S. pneumoniae strain
CP1200 (40, 41). Sequence analysis of this secreted molecule revealed a
17-residue cationic peptide that retained its full biological activity when
chemically synthesized (40). Because this synthetic heptadecapeptide
could stimulate competence in streptococcal cultures, it was designated
CSP, for competence stimulating peptide (reviewed below).
The ability to control cell physiology in concert with cell population
density, a widespread phenomenon termed quorumsensing, is observed in
both Gram-negative and Gram-positive organisms (8, 55, 112, 122). It is well
established that, by density-dependent cell-to-cell communication, bacteria
are able to respond to changes in the environment by altering gene expres-
sion. Such coordinated expression of genes facilitates group behavior
235
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
among cells that is typically observed in multicellular organisms. In addi-
tion to inducing competence, quorum sensing is known to control a variety
of physiological activities in bacteria, including antibiotic production, spor-
ulation, biofilm differentiation, conjugation, and bioluminescence (18, 30,
40, 64, 76, 90, 107, 111). In Gram-positive bacteria, quorum-sensing sys-
tems often consist of three components: a signal peptide messenger mole-
cule and a two-component signal transduction system(TCSTS) comprising
a histidine kinase (HK) and a response regulator (RR) (8, 112). As bacteria
multiply, corresponding secreted peptide molecules accumulate proportio-
nately in the surrounding environment. These molecules are then detected
by a membrane-associated HK protein once a minimal stimulatory con-
centration is reached (110). As a result, the sensor kinase is autophosphory-
lated on a conserved histidine residue, and subsequently the phosphate
group is transferred to a cognate response regulator (RR) protein.
Phosphorylation of the RR results in its activation, thereby controlling
the transcription of various gene(s) by altering the binding affinity of
RNA polymerase to one or more of the promoter regions belonging to the
target genes.
Among streptococcal species, S. pneumoniae has one of the best-
characterized density-dependent signal transduction systems. In this bac-
terium, two independent quorum-sensing mechanisms control compe-
tence and bacteriocin production. In addition, the quorum-sensing
systems of S. mutans and S. gordonii have also been well studied. In the
proceeding sections, we will discuss the competence-inducing quorum-
sensing system of S. pneumoniae and how it triggers the expression of
genes involved in DNA uptake.
GENETIC COMPETENCE REGULATION
IN S. PNEUMONIAE
The ComCDE competence regulon in S. pneumoniae
In S. pneumoniae, competence for genetic transformation is regulated
by a quorum-sensing system encoded by two genetic loci, comCDE and
comAB (13, 45). The comC, comD, and comE genes encode the CSP peptide-
precursor, the histidine kinase (HK), and the response regulator (RR),
respectively (12, 40, 42, 90). The secretion apparatus necessary for CSP
maturation and export is encoded by the comA and comB genes. When the
CSP reaches a critical extracellular concentration it is detected by the ComD
receptor, which undergoes autophosphorylation at a conserved histidine
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
236
residue. The phosphorylated ComD then transfers its phosphate group to
ComE, which is activated and, in turn, controls the transcription of its own
production, as well as that of the so-called late genes that are essential for
DNA processing, uptake, and recombination. The induction of this latter
pathway, which induces competence-related genes, is dependent upon the
activation of an alternative sigma factor designated ComX (60). In strepto-
coccal cultures, the accumulation of CSP and the development of compe-
tence is not passive, but a tightly regulated property. In the following
sections, the ComABCDE regulon and the induction of the early and late
competence genes will be discussed.
The competence-stimulating peptide (CSP)
The discovery of CSP was significant, not only as a valuable research
ingredient in pneumococcal genetics, but also as an important tool that has
been utilized to enhance our understanding of the quorum-sensing
mechanisms of Gram-positive bacteria in general. The gene, comC, that
encodes CSP was identified by searching the S. pneumoniae genome for the
reverse-translated amino acid sequence that was derived fromCSP (40, 41).
Sequence analysis of comC revealed that its deduced primary translational
product (the precursor peptide of CSP) was a 41 amino-acid peptide, of
which 17 amino acids of the C-terminus formed the biologically active CSP
that was secreted. Activation of CSP was found to be dependent on cleavage
of the pro-peptide at a GlyGly residue present in the 1 and 2 location
(relative to the processing site) in its 24 amino-acid N-terminal leader
sequence during post-translational modification (Figure 10.1) (41, 90).
Further analysis of this N-terminal sequence revealed that it belonged to
the double-glycine-type leader peptide family and was unrelated to the
N-terminal signal sequences that direct proteins across the cytoplasmic
membrane via the sec-dependent pathway (44). Analysis of leader peptides
Figure 10.1. Multiple sequence alignment of the deduced amino acid sequences of the CSP
precursors from different strains belonging to Streptococcus gordonii, S. pneumoniae and
S. mutans. A vertical box indicates the processing site that is followed by the GlyGly motif
in each sequence.
237
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
revealed the following consensus: hydrophobic residue (15), Leu (12),
Ser (1), Glu (8), Leu (7), Ile (4), Gly (2), Gly (1). The residues in
4, 7, 12 and 15 positions almost always occupied hydrophobic
residues, but only the Gly at 2 was universally conserved among these
peptides.
GlyGly leader peptides were first discovered in the precursors of lacto-
coccin A, a bacteriocin secreted by Lactobacillus lactis (48). Bacteriocins are
ribosomally synthesized, proteinaceous compounds that have antimicrobial
activity against closely related bacteria and are found in both Gram-negative
and Gram-positive bacteria (102). In addition to bacteriocins, these double-
glycine-type leaders were later discovered in pheromone peptide precursors
(such as in the CSP precursor) of Gram-positive bacteria that are involved in
cell-to-cell communication (40). Very often, their precursor peptides are
encoded by genes that are cotranscribed with one or more neighboring
genes whose products act as ATP-binding cassette (ABC) transporters and
their accessory proteins. It has been shown that disruption of the structural
genes that code for ABC transporters will prevent the secretion of their
substrate peptides (49, 86), thereby providing evidence that the export of
these activators requires a dedicated ABC transporter (41) (Figure 10.2).
The comAB locus in S. pneumoniae encodes the CSP secretion appara-
tus, which includes an ABC-transporter (ComA) and its accessory protein
(ComB) (49, 50). The indispensable role of comA in the development
of genetic competence in S. pneumoniae was discovered by virtue of an
insertionduplication mutant with a disrupted comA, which was non-
transformable and unable to produce the competence factor (86). The
mutant cells were, however, capable of becoming competent when supple-
mented with exogenous competence factor. It was therefore hypothesized
that the defect in competence resulted from interruption of the synthesis,
release, or activity of the competence factor. It was later shown that
ComAB was the secretion apparatus of CSP (41). In ComA, the N-terminal
domain has two conserved sequence motifs associated with proteolytic
activity that are responsible for cleaving the leader peptide from the
precursor CSP at the GlyGly residues (13). Interestingly, ComA falls
into a unique family of ABC transporters that carry out proteolytic proces-
sing of their substrates concomitant with export, and are dedicated to
exporting bacteriocins (35, 81). As in the case of CSP, bacteriocins exported
by these ABC-transporters comprise a double-glycine leader in their
N-terminal leader sequence. Because CSP has no apparent bacteriocin-
like activity, it is probable that these transporters may serve other
functions as well.
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
238
P
r
e
-
C
S
P
E
a
r
l
y

C
S
P
-
i
n
d
u
c
e
d

g
e
n
e
s
C
S
P
B
M
e
m
b
r
a
n
e
C
o
m
D
O
u
t
s
i
d
e
C
y
t
o
p
l
a
s
m
A
C
o
m
E
C
o
m
E
~
P
c
o
m
A
B

c
o
m
C
D
E
c
o
m
W
c
o
m
X
1

a
n
d

c
o
m
X
2
+
1
6
r
p
l
B
C
D
E
F
r
p
s
C
E
H
s
e
c
Y
+
5
5
r
e
c
A
s
s
b
B
c
g
l
A
d
p
r
A
+
7
7
D
N
A

u
p
t
a
k
e

a
n
d

r
e
c
o
m
b
i
n
a
t
i
o
n
g
r
o
E
S
g
r
o
E
L
d
n
a
K
c
i
a
R
+
1
5
L
a
t
e

C
S
P
-
i
n
d
u
c
e
d

g
e
n
e
s
D
e
l
a
y
e
d

C
S
P
-
i
n
d
u
c
e
d

g
e
n
e
s
C
S
P
-
r
e
p
r
e
s
s
e
d

g
e
n
e
s
F
i
g
u
r
e
1
0
.
2
.
T
h
e
d
y
n
a
m
i
c
n
e
t
w
o
r
k
o
f
g
e
n
e
s
r
e
g
u
l
a
t
e
d
b
y
t
h
e
C
S
P
-
d
e
p
e
n
d
e
n
t
q
u
o
r
u
m
s
e
n
s
i
n
g
m
e
c
h
a
n
i
s
m
t
h
a
t
r
e
g
u
l
a
t
e
s
c
o
m
p
e
t
e
n
c
e
i
n
S
t
r
e
p
t
o
c
o
c
c
u
s
p
n
e
u
m
o
n
i
a
e
.
T
h
i
s
f
i
g
u
r
e
i
s
m
o
d
i
f
i
e
d
f
r
o
m
t
h
a
t
p
u
b
l
i
s
h
e
d
b
y
P
e
t
e
r
s
o
n
e
t
a
l
.
(
9
5
)
b
a
s
e
d
o
n
r
e
s
u
l
t
s
f
r
o
m
t
h
e
i
r
m
i
c
r
o
a
r
r
a
y
a
n
a
l
y
s
i
s
o
f
C
S
P
-
m
o
d
u
l
a
t
e
d
g
e
n
e
s
.
B
l
a
c
k
s
o
l
i
d
a
r
r
o
w
s
:
r
e
g
u
l
a
t
i
o
n
o
f
t
h
e
a
u
t
o
c
a
t
a
l
y
t
i
c
q
u
o
r
u
m
-
s
e
n
s
i
n
g
c
i
r
c
u
i
t
.
B
l
a
c
k
d
o
t
t
e
d
a
r
r
o
w
s
:
k
n
o
w
n
l
i
n
k
s
t
o
l
a
t
e
g
e
n
e
s
a
n
d
D
N
A
p
r
o
c
e
s
s
i
n
g
p
r
o
t
e
i
n
s
.
B
l
a
c
k
s
t
r
i
p
e
d
a
r
r
o
w
s
:
h
y
p
o
t
h
e
t
i
c
a
l
l
i
n
k
s
o
f
e
a
r
l
y
g
e
n
e
p
r
o
d
u
c
t
s
t
o
l
a
t
e
,
r
e
p
r
e
s
s
e
d
,
o
r
d
e
l
a
y
e
d
g
e
n
e
s
.
B
l
a
c
k
d
a
s
h
e
d
a
r
r
o
w
:
h
y
p
o
t
h
e
t
i
c
a
l
r
e
t
r
o
-
r
e
g
u
l
a
t
i
o
n
(
9
5
)
.
The ComD, CSP receptor
During quorum sensing, when the cell population density reaches a
critical threshold, CSP is detected by a membrane-associated HK protein.
The CSP receptor, ComD, was first identified by using two Streptococcus
gordonii strains, Challis and NCTC 7865 (42). It had been observed that the
streptococcal competence pheromones exhibit strain specificity such that a
signaling peptide from one strain may or may not be able to induce
competence in another non-competent strain. In S. gordonii, the Challis
and NCTC 7865 strains represent two distinct pherotypes or transforma-
tion groups that only respond to their own CSPs. For instance, the com-
petence factor present in the supernatant of a Challis culture is unable to
induce competence in NCTC7865 and vice versa. On the other hand, Challis
is able to induce competence in a non-competent S. gordonii strain (Wicky)
that belongs to the same pherotype. To identify the competence peptide
receptor, this strain-specificity of CSP was exploited. Specifically, the comD
genes belonging to Challis and NCTC 7865 were expressed in strain Wicky
by using a Campbell-type mutagenesis strategy (42). Subsequently, the
ability of the recombinant strain to induce competence when supplemen-
ted with CSP derived from Challis and NCTC 7865 strains was assayed.
In contrast to the parent Challis strain that responded only to its own CSP,
the recombinant Wicky strain became competent in the presence of CSP
derived from NCTC 7865, thereby proving that ComD is the receptor
for CSP.
The ComD receptor consists of an N-terminal membrane-spanning
domain and a C-terminal cytoplasmic kinase domain, which contains the
histidine residue that is phosphorylated (100, 110). By conducting multiple
sequence alignment of 348 kinase domains, more than 90%of the proteins
belonging to the histidine kinase superfamily were divided into distinct
subfamilies (36). As a result, most of the peptide pheromone-activated
sensor kinases, including ComD in S. pneumoniae, were incorporated in
an HPK10 subfamily. Members of this HPK10 subfamily belong to a group
of proteins called the orthodox kinases, which comprise a membrane-
spanning N-terminal domain and a C-terminal cytoplasmic kinase domain
(36, 110). Interestingly, this group differs from other surface-located histi-
dine kinases in their regions of homology and in their membrane-spanning
domains. These homology boxes are characteristic highly conserved
(fingerprint) amino-acid sequences that are believed to play crucial roles
in substrate binding, catalysis, and/or structure (e.g. the H-, N-, D-, F-, and
G-boxes). Members of the HPK10 subfamily have no proline residue in the
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
240
H-box that acts as the site of histidine phosphorylation, and contain a
characteristic tyrosine that is located two residues downstream from the
conserved histidine residue. In addition, HPK10 HKs lack a D-box, which is
part of the nucleotide-binding domain, and their N-box contains only one
asparagine residue. In addition, in contrast to most prokaryotic histidine
kinases, which consist of two such transmembrane segments, orthodox
kinases usually possess six or seven transmembrane segments in the
N-terminal membrane-associated domain (20, 42, 67). Sequence analysis
of ComD of Streptococcus pneumoniae Rx and S. pneumoniae A66 revealed a
high level of sequence similarity in the receptor domain of these kinases
(90). Despite this similarity, the corresponding peptide pheromones were
not able to cross-activate these kinases (96), thereby highlighting the
importance of this region in the ComD receptor for CSP specificity. With
reference to peptide molecules that activate the HPK10 subfamily of
kinases, almost all peptide pheromones that have been characterized are
unmodified peptides (54). As with CSP in S. pneumoniae, these signaling
molecules are synthesized as precursor peptides containing a double gly-
cine leader, which serves as a secretion signal that is cleaved concomitant
with export.
The ComE response regulator
The ComE comprises the RR belonging to the ComDE TCSTS that is
activated when CSP reaches a critical concentration. Previously, it was
shown that the disruption of comE abolished two processes: response to
synthetic CSP and endogenous competence induction (90). Like other
response regulators, ComE consists of an N-terminal regulatory domain
and a C-terminal DNA-binding effector domain (110). Following the phos-
phorylation of ComD, the phosphate group is transferred to a conserved
aspartate residue present in the regulatory domain in ComE. The phos-
phorylation of ComE, designated ComE$P, enhances its binding to various
promoter regions throughout the genome, thereby regulating gene expres-
sion at the transcriptional level. ComE$P-regulated genes belong to the
early phase of the competence development pathway, which sets the stage
for the synthesis of proteins that are directly involved in the transport and
integration of extracellular DNA.
Notably, it was demonstrated that, after exposure to CSP, the transcrip-
tion rates of the comCDE and comAB operons were significantly increased
during competence development; this was consequently followed by
a burst in CSP secretion (2, 90, 119). It was also observed that this response
did not occur in ComE-deficient mutants (90), thereby suggesting a
241
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
ComE-mediated role in the transcriptional activation of comCDE. Later, by
using electrophoretic mobility shift assays, it was demonstrated that ComE
specifically binds to the comC and comA promoters, strongly supporting its
role as a transcriptional factor that activates the comAB and comCDE
promoters (119).
The ComE binding-site was found to possess two 9 bp imperfect direct
repeats separated by 12 bp (119). More specifically, it was identified as the
following sequence: aCAtTTct(a/g)G12 bpACA(t/g)TtgAG,
where the most conserved bases are capitalized. The presence of the
ComE binding-site in the comA and comC promoters and their resultant
increase in expression generates an autocatalytic regulatory loop represen-
tative of a positive feedback mechanism (i.e. during competence induction,
comC can stimulate its own synthesis and secretion; phosphorylation of
ComE increases its response to CSP and also amplifies the production and
secretion rate of this signal molecule) (45). This abrupt rise in CSP probably
ensures that the entire population of cells reaches its competence state
simultaneously. However, this self-inducing mechanism is transient.
Based on Northern hybridization studies and lacZ reporter gene assays,
the comAB and comCDE transcripts are highest only 5 min after CSP
addition, and then decrease and disappear almost entirely at 20min
(3, 119). These genes whose transcripts appear the earliest when cells are
exposed to CSP are designated the early competence genes and include at
least 8 genes (3, 11, 60, 73, 91, 94, 95, 90).
ComX: the link to late competence genes
After the ComE binding-site had been identified, it was compared
with a common regulatory sequence motif in the promoters belonging to
coordinately expressed late competence genes previously identified by
Campbell et al. (11) and Pestova and Morrison (91). Because these
sequences did not share a common sequence motif, it was believed that
ComE activated the late genes via one or more intermediate factors, later
identified as ComX (60). ComX, also called Sigma X (s
x
), is synthesized by
duplicate genes, comX1 and comX2, and constitutes an alternative sigma
factor that can be derived by co-purification with RNA polymerase isolated
from competent streptococcal cultures (60, 75). The ComX shows high
homology to the s
H
in Bacillus subtilis, a member of the s
70
RNA polymer-
ase sigma subunit family, and is essential for the transcription of the late
genes involved in the processing, binding, uptake, and recombination of
DNA during transformation (73, 94, 95, 98). ComX acts by directing an
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
242
RNA polymerase to a recognition sequence, TACGAATA, referred to as the
cin-box or com-box, that is present in the late-gene promoters (59).
Interestingly, the regulation of ComX is dependent on CSP, the phosphor-
ylation of ComE, and the activity of both comX1 and comX2 to obtain a
normal level of competence (60). It was shown that, in a mutant lacking
both comX genes, transformation ability was abolished altogether. More
recently, Luo et al. identified a second early gene product, ComW, shown to
be important for the optimal activity of ComX (74). Although its mode of
action is yet to be characterized, the ectopic coexpression of comX and comW
demonstrated that, together, just the products of these early genes were
sufficient to induce competence.
Before discussing the late competence-induced genes, it is important
to describe the actual transformation process. Transformation is initiated
by the binding of exogenous double-stranded DNA molecules in linear or
circular form (plasmids) to the bacterial cell surface. Consequently, these
double-stranded molecules are split into single strands by the activity of
EndA (125). One strand is actively transported into the cell in a 3
0
to 5
0
direction, whereas the complementary strand is degraded by nucleases
(84). Depending on the presence of homologous sequences, the newly
translocated DNA strands can be fully or partly integrated into the host
genome via recombination. In a competent state, this recombination event
is favored by the presence of single-stranded gaps throughout the host
chromosome (23). In plasmid DNA the re-circularization process is
less efficient because a second strand, complementary to the first, must
be present.
With the recent advances in molecular technology that includes the use
of high-density DNA microarrays, researchers have been able to monitor
the global changes in gene expression during competence development
(94, 95, 98). Based on these studies, it has been demonstrated that the
transcriptional kinetics of at least 188 genes are transiently affected during
this process. Most recently, in a DNA microarray study conducted by
Peterson et al. (95), CSP-induced genes exhibited four temporally distinct
expression waves: early, late, and delayed gene induction, and gene repres-
sion. Notably, the majority of the CSP-responsive genes belonged to the late
phase, whose expression was optimal at approximately 10min following
CSP addition. These included 81 genes in 21 clusters, whose products have
a direct role in DNA uptake and homologous recombination during trans-
formation. These include ssbB (cilA), dalA, ccl, cglABCD, celAB, and cflAB
(11, 14, 60, 91). Once imported into the cell, the DNA molecules are
integrated into the host chromosome via homologous recombination
243
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
assisted by RecA (79, 87). Interestingly, the recA, cinA, dinF, and lytA genes
that comprise the recA operon were induced in the late phase by CSP (94,
95, 98). The genes that displayed a delayed expression pattern showed the
accumulation of mRNA starting from the first minute after exposure to
CSP until after the optimal peak was obtained for the late genes. Although
the expression of certain delayed genes appeared to be unaffected by the
inactivation of ComX (95), the majority of these genes were associated
with bacterial stress response. For instance, the entire dnaK heat-shock
operon, which included hrcA, grpE, dnaK, and dnaJ (103) was induced in
response to CSP.
Competence-induced cell lysis
Any discussion of competence would naturally evoke the question
regarding the source of donor DNA that is used for transformation.
Previously, it was believed that these DNA molecules were derived from
dead bacteria that had lysed by natural causes. In contrast to this traditional
view, emerging investigations have demonstrated that, in pneumococci,
the release of donor DNA is triggered by the addition of CSP (88, 108, 109).
More specifically, during co-cultivation, competent streptococci can actively
acquire DNA by killing 5%20% of their non-competent neighbors of the
same strain in a planktonic population when exposed to CSP (108).
Moreover, it was observed that cell lysis did not occur in a ComE-deficient
mutant, thereby demonstrating that the ComDE signal transduction sys-
tem was involved in DNA release as well as in its uptake. At the present, the
mechanism of lysis and how it is related to competence induction remains
to be investigated. However, based on current evidence, DNA release is
accomplished by heterolysis (i.e. lysis of one bacterium that is caused by
another) coordinated by proteins such as autolytic amidase, LytA, and an
autolytic lysozyme, LytC (88, 108, 109). From an evolutionary perspective,
the discovery of competence-induced lysis is of enormous significance in
increasing the genetic plasticity of pneumococci. Previously, Steinmoen
et al. (109) indicated that, in S. pneumoniae, cell lysis and DNA uptake were
coordinated to ensure the presence of a sufficient amount of homologous
DNA during competence development, which would increase the chances
for homologous recombination leading to the emergence of novel geno-
types. In contrast to this DNA release-and-uptake model, a more recent
investigation into CSP-induced lysis by Moscoso and Claverys (88) pro-
vided evidence that would question the validity of cell lysis as a method
evolved for maximized genetic exchange. They showed that, although
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
244
competence decreased after its maximumat 20min after CSP addition, the
amount of liberated DNAcontinued to increase and reached a maximumin
the stationary phase, when cells were no longer capable of DNA uptake.
Supporting their view is the appearance of nuclease activity (EndA and at
least one other nuclease) when competence is triggered. Hence, Moscoso
and Claverys suggest that competence-induced DNA release serves a
role different from that previously thought and may possibly have a role
in nutrient acquisition, biofilm formation, or the release of toxins (e.g.
pneumolysin, teichoic and lipoteichoic acids).
Competence shutoff
In S. pneumoniae, genetic competence is a transient phenomenon that
takes place at a certain cell population density. Based on emerging evi-
dence, this process is tightly regulated; this makes sense, because consti-
tutive uptake of foreign DNA would generate havoc in the cell by
overloading the DNA repair systems, causing damage to the host chromo-
some and also by generating lethal mutations. Despite the importance of
competence shutdown for cell viability, there has been little insight into its
mechanism in the past. However, several hypotheses describing the
mechanism of competence shutoff have been described.
One such hypothesis, proposed by Lee and Morrison, states that the
complete shutoff of competence followed by its continued suppression is
ComX-mediated (60). Consistent with this is the rapid degradation of
ComX as competence declines (60, 75). This is likely caused by a proteolytic
event that prevents perpetual competence. In addition to degradation of
ComX, it was assumed that an additional negative control mechanism was
acting to shut off competence, because a rapid decrease of late competent
gene transcripts was observed even when the ComX protein was still
present. In addition, Lee and Morrison (60) suggested that competence
shutdown was likely mediated by the inhibitory activity of a putative late
competent gene product, which was designated ComI. To identify such
inhibitory candidates, Peterson et al. (95) performed targeted gene dele-
tions to screen previously uncharacterized late competence genes, which
were identified by global genome analysis, whose temporal profile would
significantly alter competence decay. Based on their results, none of the
candidate genes selected for mutagenesis was involved in the competence
shutdown mechanism.
Previously, Alloing et al. (3) had suggested that a ComE-specific
phosphatase, CEP, could manipulate competence activation as well as its
245
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
shutdown by altering the unphosphorylated to phosphorylation ratio of
ComE (ComE : ComE$P). More specifically, optimal transcription of
comCDE would occur in a partly phosphorylated ComE/ComE$P state,
whereas a fully phosphorylated ComE would increase the transcription of
late competence genes, including the gene responsible for CEP production.
Increased production of CEP would then deplete ComE$P, which, in turn,
would reduce the expression of late competence genes, thereby shutting
down competence. However, there was no evidence of a putative CEP,
based on microarrays conducted by different researchers (95, 98), thereby
warranting the requirement for more studies to elucidate the molecular
pathway(s) that lead to competence shutdown.
The CiaRH and VicRK two-component signal transduction
systems (TCSTS)
The S. pneumoniae CiaRH TCSTS was linked to competence by
Guenzi et al. (38) based on an observation that an amino-acid change in
CiaH (ciaH
T230P
) caused complete inhibition of competence. Because sup-
plementing these ciaH
T230P
mutant cultures with exogenous competence
factor did not restore competence, these authors argued that the CiaRH
signal transduction system was involved in the early steps of competence
regulation. A few years later, it was observed that a mutant deficient in the
ciaR gene became competent under conditions that inhibited the develop-
ment of competence in the wild-type strain (33). However, by examining
ciaR mutants that were derived from spontaneous revertants of ciaH
T230P
,
that had a restored ability to transform, it became apparent that competence
was derepressed in these ciaR mutants. A link between the CiaRH and
ComCDE signal transduction systems was first obtained by Echenique et al.
(24). In their studies, inactivation of ciaRH resulted in the overexpression
of comCDE, and it was proposed that CiaR negatively regulated the compe-
tence regulon. The work by Martin et al. (80) provided supporting evidence
by using transposon mutagenesis to identify comCDE-upregulated mutants
(called ComCDE
UP
or CUP) that included independent insertions in the
ciaR gene. Additional evidence for the CiaR-mediated negative regulation
of the competence regulon was provided recently based on two indepen-
dent transcriptome studies (82, 104). Despite accumulating knowledge
suggesting some type of cross-regulation between these signal trans-
duction systems, it remains to be established whether CiaR is directly
or indirectly involved in comCDE repression. With reference to this,
the ciaRH and comCDE systems exhibited 83% identity over a fragment
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
246
24 bp long that included their 10 promoter regions (80). Hence, it is
possible that comCDE would be targeted for regulation by harboring a CiaR
binding site in this segment.
In addition to CiaRH, the VicRK TCSTS (58) has also been associated
with competence development (25, 117). The S. pneumoniae vicRK, also
called TCS02 (58), 492hk/rr (113), and micAB (25), is the ortholog of the
essential yycFG signal transduction system in B. subtilis (27). In contrast to
B. subtilis, which requires both the HK and the RR pair for its viability, only
the RR is essential for survival in S. pneumoniae (58). Based on the findings
of Echenique and Trombe (25), the phosphorylated VicR acts upstream of
ComDE to repress competence when oxygen is limited. Interestingly, VicK
is an atypical kinase that harbours a PAS domain. Usually, PAS domains
are involved in monitoring intracellular signals such as redox potential
(124), suggesting that the response to oxygen limitation might be sensed
intracellularly. Interestingly, in a VicK-deficient strain, transformability
was decreased by three orders of magnitude.
Based on accumulating evidence, competence development in S. pneu-
moniae involves many dynamic interactions within the cell that not only
involve a mechanismthat monitors the population density, but also seemto
connect DNA uptake to other important physiological processes. For
instance, competence is influenced by oxygen availability, temperature,
pH, and Ca
2
and Mg
2
concentrations (45). Also importantly, the
comCDE competence regulon is located near the putative origin of chromo-
some replication, ori (31). Hence, the comori co-location enables the bac-
terium to sense the rate of replication (which is dependent on nutrient
availability) and develop competence in concert with the environmental
conditions (15). Previously, Alloing et al. (3) discovered that quorumsensing
is affected in an obl-deficient mutant (obl for oligopeptide-binding lipo-
protein) suggesting that transformation is affected by an oligopeptide
permease in response to nutrient availability. However, more studies are
needed to elucidate the molecular and genetic mechanisms that connect
these signals to the competence regulon.
The Blp quorum-sensing system
The BlpABCSRH regulon forms a second quorum-sensing system in
S. pneumoniae that shares common features with the ComABCDE compe-
tence regulon (19). The Blp regulon (blp for bacteriocin-like peptide) con-
trols the production of class II bacteriocins and their immunity proteins in
response to a threshold cell population density (19, 97). In this system, the
247
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
blpC encodes the BIP (bacteriocin-inducing peptide) pheromone, whereas
the blpRH encodes the histidine kinase sensor (BlpH) and response reg-
ulator (BlpR), respectively. Although the function of BlpS is unknown, it
shares high similarity with the C-terminal DNA-binding domain of BlpR,
but lacks the N-terminal domain. Similar to CSP, BlpC contains a char-
acteristic GlyGly motif that is believed to be cleaved by the blpAB-encoded
secretion apparatus, prior to export of the mature BIP signaling peptide.
Sequence analysis of BlpC from different strains demonstrated allelic
variation resulting in at least four different BIPs, thereby leading to
strain-specific activity of this signaling molecule reminiscent of the phero-
type variation for CSP.
Previously, based on microarray analyses, it was believed that cross-
communication was unlikely between the Com and Blp quorum-sensing
systems at the pheromone or histidine kinase level (15). It was argued that
genes affected by crosstalk would possibly have been detected during global
genome analyses conducted to identify genes regulated by either CSP or
BIP addition. More recently, the addition of BIP and CSP signaling pep-
tides were demonstrated to activate the transcription of an operon, encod-
ing an ABC transporter (QsrAB) of unknown function in S. pneumoniae Rx
strain (56). Moreover, cross-induction was achieved by a hybrid-direct-
repeat motif present in the target promoter that responded to both ComE
and BlpR. Because homology searches suggested a putative role for QsrAB
as a sodium pump, it is possible that it protects the bacterium against
osmotic stress (56).
GENETIC COMPETENCE IN ORAL STREPTOCOCCI
In the oral cavity, there is a diverse range of habitats including both soft
and hard tissues. Consequently, to be able to survive and prosper, many oral
bacteria have evolved highly specialized mechanisms to overcome fluctua-
tions in the composition of nutrient supply, local oxygen availability, pH,
shear forces due to saliva flow and mastication, and a range of host defense
mechanisms (77). Oral streptococci, originally called viridans streptococci,
are the species that predominantly inhabit the mouth and upper respiratory
tracts of humans as commensal bacteria and constitute approximately 20%
of the normal human oral flora (57). Oral streptococci can cause opportu-
nistic infections at oral or non-oral sites when environmental conditions
favor an overgrowth of particular species or when they escape their normal
ecologic niches and establish at another site, usually as a consequence of
some breakdown in normal host defenses (101).
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
248
The classification of oral streptococci has undergone significant
changes in the past several years. A comprehensive review (28) divided
the oral streptococci into five distinct species groups based on phenotypic
characteristics, DNADNA re-association, and 16S rRNA gene sequencing:
(i) the Streptococcus anginosus group, (ii) the Streptococcus mitis group,
(iii) the Streptococcus mutans group, (iv) the Streptococcus salivarius group,
and (v) the Streptococcus sanguinis group.
Oral streptococci respond to signals resulting from the proximity,
density, and identity of microbial neighbors. Through the process of
quorum sensing, bacteria can indirectly determine population density by
sensing concentration of a signal molecule (8). As discussed earlier, the
work on pneumococcal competence had pioneered a research area on
quorum-sensing regulation of streptococcal competence. Also importantly,
this gene-exchange phenomenon has long been described as a character-
istic of species within the mitis group (62). To map the incidence of natural
competence in the genus Streptococcus (43) a number of streptococcal
strains were screened by PCR for the presence of the competence operon
(comCDE). By using primers complementary to the Arg- and Glu-tRNA
genes that flank the comCDE operon in S. pneumoniae, it was observed that
most streptococcal species belonging to the mitis and anginosus groups
possessed the comCDE operon. Recently, natural genetic competence has
been reported in other oral streptococcal species belonging to the sanguinis
and mutans groups (Table 10.1) (for review see 16).
Table 10.1. Naturally competent human oral streptococci
Group Species
Anginosus S. anginosus
S. constellatus
S. intermedius
S. mitis
Mitis S. cristatus
S. oralis
Mutans S. mutans
S. ratti
Sanguinis S. sanguinis
S. gordonii
Source: Adapted from (16).
249
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
Horizontal gene transfer exerts a strong selective force on a bacterial
population, leading to the evolution of prokaryotic genomes (10). As natural
competence for genetic transformation is involved in horizontal gene
transfer, it plays a significant role in gene acquisition/loss and strain
heterogeneity, thereby leading to overall evolution (29). The potential for
gene transfer is high among streptococci in oral biofilms; emerging evi-
dence suggests that quorum sensing may be coupled with expression of
competence for acquisition of foreign DNA via transformation. In the
following sections, specific examples of quorum-sensing regulation of
natural competence in two important oral streptococci, namely
Streptococcus gordonii and S. mutans, will be discussed.
Transformation in Streptococcus gordonii
Genetics of competence development
S. gordonii is a human commensal bacterium classified into the san-
guinis group (28). It participates in the formation of the dental biofilm as a
primary colonizer of the tooth pellicle, creating a template for the sub-
sequent attachment of other bacteria and leading to the establishment of
the complex oral biofilm (121). By producing hydrogen peroxide, this spe-
cies may also play a protective role in relation to periodontal diseases, as
hydrogen peroxide inhibits the growth of Gram-negative periodontopatho-
gens (47). Although S. gordonii is considered to be relatively non-patho-
genic, this bacterium is among the most common oral bacteria associated
with infective endocarditis (32).
S. gordonii Challis (formerly Streptococcus sanguis Challis) was the first
oral streptococcus shown to be capable of transformation (62). Competence
in S. gordonii develops in the early to mid exponential growth phase and is
modulated by environmental stimuli that trigger the production of an
extracellular competence factor (CF). Although all of the stimuli are still
unknown, it has been demonstrated that a slightly basic pH and the pre-
sence of albumin or heat-inactivated serum can trigger competence devel-
opment in S. gordonii (70). CF acts in the same manner as S. pneumoniae
CSP by functioning as a population density indicator for competence
induction. Two CFs have been isolated from S. gordonii species; both are
19 amino-acid peptides that are processed from a 50 amino-acid precursor
peptide containing the characteristic N-terminal double-glycine type lea-
der, as seen in S. pneumoniae (42, 72). Amino-acid sequence alignment
showed that these two peptides have 74% amino-acid identity. The comC
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
250
gene encoding CF is located immediately upstream of the comD and comE
genes which encode the HK and RR of a TCSTS, respectively (42). As with
S. pneumoniae, the S. gordonii comCDE operon is located between the Arg-
tRNA and Glu-tRNA genes (42, 43).
At a different genomic locus, there is a comABoperonencoding the ComAB
transporter, a member of the ABC-type transporters present in S. pneumoniae.
Amino-acid sequences of the ComAproteins fromS. pneumoniae and S. gordonii
displayed 82% identity (71); the ComB sequences revealed that the peptides
were 55% identical. Analysis of the S. gordonii genome located 3
0
proximal
to comB revealed the presence of an additional locus, called orfX, that
had no similarity to known sequences (71). The S. gordonii orfX locus
contains three small putative open reading frames, designated comX,
orfM, and orfO. The comX gene encodes a putative 52 amino-acid basic
peptide with a predicted isoelectric point (pI) of 10.2. Complementation
experiments demonstrated that ComX is able to complement transform-
ation deficiency in the non-transformable S. gordonii strain Wicky, a non-
strain that had lost the ability to synthesize the CF. As ComX is responsible
for competence-stimulating activity in S. gordonii Wicky, it has been
proposed that CF may not be the only signal molecule; consequently, a
CF independent pathway for competence induction may exist in S. gordonii
(96, 98).
Regulation of competence
Experimental evidence suggested that the mechanisms involved in cell-
density-dependent regulation of competence development in S. pneumoniae
and S. gordonii were fundamentally similar. A model depicting the pathway
involved in the regulation of competence in S. gordonii has been proposed by
Jenkinson (51) based on the data obtained for S. pneumoniae (Figure 10.2).
Competence develops through a quorum-sensing mechanism in which CF
serves as the signal molecule. The CF precursor is most likely processed
and exported outside the cell by the ComAB transporter system. When the
extracellular CF reaches a critical threshold concentration, it is sensed by
the N-terminal domain of the membrane-bound ComD receptor (70).
This interaction results in the autophosphorylation of ComD and the subse-
quent activation of the response regulator ComE by phosphorylation
(ComE$P). In general, phosphorylation enhances the affinity of response
regulators for their DNA binding sites, leading to increased expression of
target genes.
In addition to controlling the expression of the comCDE operon,
ComE$P also regulates transcription of late competence genes (comY
251
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
operon, cipA, and recA) via a regulatory protein (Figure 10.3). The comY
operon contains four genes, designated comYA, comYB, comYC, and
comYD, and encodes a putative dedicated secretion system thought to be
involved in DNAuptake. It has been demonstrated that a S. gordonii comYA-
defective mutant was completely deficient in transformation and exhibited
decreased levels of DNA binding and hydrolysis (72). Interestingly, the
comYA gene product shares significant similarity (57% homology) to the
ComGAtransporter of Bacillus subtilis (72). The S. gordonii comYAgene also
represents the first competence gene conserved between the transformable
streptococci and the well-studied transformable B. subtilis (70, 72). The cipA
and recA genes encode proteins necessary for DNA recombination. The
product encoded by the cipA (competence-inducible peptide) gene is pre-
dicted to be a putative 51 kDa protein believed to be involved in chromo-
somal integration and plasmid reassembly. As expected, S. gordonii cipA
mutants are defective in chromosomal transformation (70).
Another ABC-type transporter, designated Hpp, has been shown to be
involved in competence development by S. gordonii (Figure 10.3). Hpp is a
lipoprotein-dependent oligopeptide permease that preferentially transports
hexa- and heptapeptides (52). The Hpp permease is composed of cytoplas-
mic membrane-bound lipoproteins that are highly similar (60% amino
acid identity) to the substrate-binding components of the S. pneumoniae
Ami oligopeptide permease (4). It has been demonstrated that mutations in
genes encoding components of the S. gordonii Hpp oligopeptide permease
transporter affect competence development by reducing the efficiency of
transformation (52). A possible explanation is that the Hpp-transported
peptides may modulate the expression or activity of a ComE-specific phos-
phatase, which, in turn, controls the level of ComE$P (51). The activity of
Hpp also affects the expression of cshA, a gene encoding a high-molecular-
mass cell-wall-anchored polypeptide essential for colonization of the oral
cavity by S. gordonii (83). CshA mediates the binding of S. gordonii to
human fibronectin, Actinomyces naeslundii, Streptococcus oralis, and
Candida albicans (53).
In Gram-positive bacteria, it has been demonstrated that peptide trans-
port systems are not only essential for the uptake of peptides as a source of
nutrients, but also involved in affecting other cellular functions and proper-
ties, such as the development of competence, sporulation, and bacterial
adherence. The Hpp system may influence the development of S. gordonii
genetic competence by functioning in sensing environmental changes
during growth. As proposed by Jenkinson (51), the utility of this would be
to provide an additional level of control on the ComE regulatory circuit and
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
252
O
u
t
s
i
d
e
C
y
t
o
p
l
a
s
m
A B
C
o
m
D
C
o
m
D
C
o
m
C
P
c
o
m
C
o
m
E
R
P
P
e
p
t
i
d
e
s
H
p
p
c
s
h
A
C
o
a
g
g
r
e
g
a
t
i
o
n
C
o
l
o
n
i
z
a
t
i
o
n
c
o
m
Y

o
p
e
r
o
n
c
i
p
A
r
e
c
A
C
o
m
E
C
F
C
o
m
E
~
P
M
e
m
b
r
a
n
e
F
i
g
u
r
e
1
0
.
3
.
S
c
h
e
m
a
t
i
c
r
e
p
r
e
s
e
n
t
a
t
i
o
n
o
f
q
u
o
r
u
m
-
s
e
n
s
i
n
g
r
e
g
u
l
a
t
i
o
n
o
f
g
e
n
e
t
i
c
c
o
m
p
e
t
e
n
c
e
i
n
S
t
r
e
p
t
o
c
o
c
c
u
s
g
o
r
d
o
n
i
i
.
M
o
d
e
l
p
r
o
p
o
s
e
d
b
y
J
e
n
k
i
n
s
o
n
(
5
1
)
.
S
e
e
t
e
x
t
f
o
r
d
e
t
a
i
l
s
.
R
:
r
e
g
u
l
a
t
o
r
y
p
r
o
t
e
i
n
.
P
:
p
h
o
s
p
h
a
t
a
s
e
.
a means by which alternative environmental signals might be integrated
into the S. gordonii quorum-sensing-dependent competence pathway.
Transformation in Streptococcus mutans
Genetic regulation of competence development
S. mutans is considered a major causative agent of dental caries, one of
the most common human infectious diseases. In the United States, billions
of dollars are spent annually to treat dental caries or caries-related infec-
tions. S. mutans initiates dental caries while living in the biofilm environ-
ment of dental plaque, which comprises a diverse community of
microorganisms found on tooth surfaces and embedded in an extracellular
matrix of bacterial and host origin (78). The main virulence factors asso-
ciated with S. mutans cariogenicity are its ability to adhere and coaggregate
on teeth, its acidogenicity, and its aciduricity (7, 85). Consequently, the
tooth surface is an indispensable natural habitat for S. mutans. As demon-
strated for S. gordonii, S. mutans was also found to possess a quorum
sensing system that closely resembles the com regulon of S. pneumoniae.
One genomic locus contains the comC, comD, and comE genes encoding the
precursor to the competence-stimulating peptide (CSP), a transmembrane
HK, and an RR, respectively, with the comD-encoded protein being the CSP
receptor (65). In contrast to what was found in S. pneumoniae and other oral
streptococci, in S. mutans the comC gene is encoded divergently proximal to
the comDE genes. Moreover, in S. mutans, the comCDE region is not
flanked by the Arg- and Glu-tRNA genes (64) and therefore cannot be
detected by PCR using primers complementary to the tRNA-encoding
genes (43). The S. mutans CSP was found to be a 21 amino-acid peptide
derived from a 46 amino-acid peptide precursor. The peptide precursor is
believed to be produced constitutively and cleaved during its export via the
ComAB transporter system (93). As seen in S. pneumoniae and S. gordonii,
the S. mutans comCDE region is not closely linked to the comAB operon (1).
Interestingly, deletion mutations in any of the S. mutans comgenes resulted
in mutants with a residual level of transformation (64). This is in contrast
to the transformability of S. pneumoniae, in which the com mutants had no
detectable transformation (12). This is one clue that the mechanisms of the
quorum-sensing-dependent competence pathway may be regulated differ-
ently among various streptococcal species. Indeed, another fundamental
difference between S. pneumoniae and S. mutans is the optimal level of
competence achieved under saturating CSP concentrations. In this condi-
tion, nearly all of the S. pneumoniae cells become competent, whereas this
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
254
percentage is typically between 0.1% and 1% in an S. mutans population.
While the reasons for this are unclear, it is probable that, in S. mutans,
alternate signals and/or shutoff mechanisms may hinder the spread of
competence throughout the whole population.
A model describing the general pathway of quorum-sensing-mediated
regulation of genetic competence in S. mutans is illustrated in Figure 10.4.
When the mature CSP reaches a critical threshold concentration at cell
densities typical of the early to mid exponential growth phase, it interacts
with the membrane-associated HK, ComD. This interaction results in the
phosphorylation of ComD, which in turn stimulates the phosphorylation of
its cognate RR, ComE. The phosphorylated form of ComE acts at promoter
sites (Pcom) for genes whose expression is upregulated during the devel-
opment of competence. Based on the work done in pneumococci, the
following consensus sequence for the ComE-binding site has been pro-
posed: aCAtTTca/gG-N
12
-ACAt/gTTgAG (119). In S. mutans, at least six
genes are known to be upregulated by the ComCDE two-component signal
transduction system, comC, comDE, comAB, and comX (66). By using a
transcriptional fusion of the comX promoter to a promoterless reporter
gene, Aspiras et al. (5) demonstrated a positive correlation between the
pcomX transcription in the presence of CSP and competence development.
Similar to S. pneumoniae, it is believed that comX encodes an alternative
sigma factor that directs the transcription of RNApolymerase by altering its
binding affinity to the promoters of several late competence-specific genes
involved in DNA uptake and processing (73). Indeed, bioinformatic analy-
sis of the S. mutans UA159 genome revealed that many putative open
reading frames, including late-competence gene homologs, possess a
sequence in their promoter regions (TTTTT-N
9
-TACGAATA, where N
9
represents a 9bp linker region) that resembles the com box of S. pneumoniae
(11). Interestingly, several of these genes are unrelated to genetic competence.
These findings suggest a complex network of genes that are regulated by the
quorum-sensing systemand that control other cellular functions in addition to
competence induction in S. mutans.
Study of competence in model biofilms
Dental biofilmformation is initiated by the interactions between plank-
tonic cells and tooth surfaces in response to appropriate environmental
signals. Bacteria in a biofilm mode of growth are distinct from planktonic
cells in their gene expression and cellular physiology (106). Interestingly, it
has been demonstrated that S. mutans cells growing in biofilms can be
transformed more efficiently (at a rate 10- to 600-fold higher) than their
255
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
O
u
t
s
i
d
e
C
y
t
o
p
l
a
s
m
A B
C
o
m
D
H
K
1
1
V
i
c
K
V
i
c
R
R
R
1
1
~
P
R
R
1
1
C
e
l
l

s
e
g
r
e
g
a
t
i
o
n
?
B
i
o
f
i
l
m

f
o
r
m
a
t
i
o
n
?
B
i
o
f
i
l
m

f
o
r
m
a
t
i
o
n
?
A
l
t
e
r
n
a
t
e

s
i
g
m
a

f
a
c
t
o
r
L
a
t
e

c
o
m
p
e
t
e
n
c
e

g
e
n
e
s
C
o
m
p
e
t
e
n
c
e
O
t
h
e
r

g
e
n
e
s
g
t
f
B
g
t
f
C
V
i
c
R
~
P
c
o
m
D
c
o
m
C
c
o
m
A
c
o
m
B
P
c
o
m
P
c
o
m
P
c
o
m
P
c
o
m
C
o
m
E
c
o
m
X
c
o
m
E
C
S
P
C
o
m
E
~
P
M
e
m
b
r
a
n
e
?
?
F
i
g
u
r
e
1
0
.
4
.
S
c
h
e
m
a
t
i
c
r
e
p
r
e
s
e
n
t
a
t
i
o
n
o
f
q
u
o
r
u
m
-
s
e
n
s
i
n
g
r
e
g
u
l
a
t
i
o
n
o
f
g
e
n
e
t
i
c
c
o
m
p
e
t
e
n
c
e
i
n
S
t
r
e
p
t
o
c
o
c
c
u
s
m
u
t
a
n
s
.
S
e
e
t
e
x
t
f
o
r
d
e
t
a
i
l
s
.
free-living counterparts (64). It is likely that the high density of bacterial
cells growing in biofilms can facilitate the exchange of intraspecies genetic
material by providing optimal conditions for the activation of the CSP-
dependent quorum-sensing system. In addition, natural transformation
has also been shown to occur in mixed-species biofilms in vitro. One of
the best examples was the intergeneric transfer of tetracycline resistance
in a mixed-species biofilm from B. subtilis to a Streptococcus sp. grown
in a constant-depth biofilm fermentor (99). Moreover, genetic exchange
between a Gram-negative (Treponema denticola) and a Gram-positive
(S. gordonii) bacterium has also been demonstrated in an artificial oral
biofilm (118). Biofilms may thus function as potential genetic reservoirs,
allowing horizontal gene transfer of virulence factors and the dissemina-
tion of antibiotic resistance genes.
The first report linking the cell-density-dependent quorum-sensing
system to biofilm formation was described in S. gordonii. By using trans-
poson mutagenesis, Loo et al. (68) found that inactivation of comD resulted
in a biofilm-defective phenotype. In a recent study, Gilmore et al. (34) used
real-time PCR to quantify the expression of S. gordonii genes known or
assumed to be involved in biofilm formation. They demonstrated that the
comD and comE genes encoding the TCSTS required for competence in
S. gordonii were both upregulated in the biofilm phase. In S. mutans,
biofilm formation has also been linked to the competence regulon.
Li et al. (66) demonstrated that inactivation of any of the genes encoding
the CSP signaling systemresulted in the formation of an abnormal biofilm.
In particular, they found that the comC null and comAB null mutants
formed biofilms that lacked the wild-type architecture, whereas the comD,
comE and comX null mutants formed biofilms with reduced biomass. The
fact that the biofilms formed by the comC mutant differed from the comD,
comE, and comX mutant biofilms suggested that S. mutans may possess at
least one alternate CSP and/or receptor (17). Further support for multiple
CSP receptors includes the observation that exogenous addition of syn-
thetic CSP into the comCDE mutant culture only partly restored the wild-
type biofilm architecture (65, 66). Further studies demonstrated a positive
correlation between pcomX activity in high cell density areas within the
biofilm population and competence development (5). Because S. mutans
relies on a biofilm lifestyle for its survival in the oral cavity, it is not
surprising that this bacterium has evolved a quorum-sensing signaling
system that helps it to adapt and survive while living in biofilms.
More recently, Petersen et al. (92) have demonstrated that CSP pro-
motes the biofilm mode of growth of Streptococcus intermedius, thereby
257
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
enhancing the role of CSP as a biofilm sensor. S. intermedius, a member
of the anginosus group, is a bacterium found in the gingival crevice and
is associated with abscesses (101). It has been shown that CSP-treated
S. intermedius cells formed more biofilm than untreated cells, whereas
competence development occurred exclusively in the presence of CSP (92).
Other CSP-regulated phenotypes
Within the oral cavity, S. mutans is often exposed to transient environ-
mental conditions. For example, during the ingestion of dietary sugars by
the host, carbohydrate concentrations can range from 100mm to as low as
1 mmduring sleep. It also encounters pHshifts ranging fromabove 7.0 to as
low as 3.0 (in carious sites). Consequently, the ability of S. mutans to
tolerate acidic pH is crucial to its survival and pathogenicity. To determine
whether cell density can modulate adaptation to acid in S. mutans, Li et al.
(63) examined the ability of mutants with defects in the ComCDE signaling
system to withstand acid challenge. They found that inactivation of
the comCDE genes resulted in a diminished acid tolerance response,
which could be restored to wild-type levels by supplementing with synthetic
CSP. These results clearly demonstrated a connection between the CSP-
dependent competence signaling pathway and the acid-tolerance response
of S. mutans.
Another TCSTS, comprising the hk11 and rr11 genes, has also been
shown to be important for the survival of S. mutans at an acidic pH(65, 66).
Interestingly, a com-box consensus sequence was found in the promoter
region of rr11, while the deletion of the hk11 or rr11 resulted in defects in
biofilm formation. A closer examination of the mutant biofilms by scan-
ning electron microscopy revealed that they have a sponge-like architecture
composed of cells organized in very long chains, a feature that was pre-
viously observed in the biofilm formed by the comC null mutant. Hence, it
was suggested that hk11 may act as another CSP receptor, in which inter-
action could activate another pathway related to cell segregation and ulti-
mately affect the S. mutans biofilm architecture (Figure 10.4). However,
further studies will be necessary to assign a role to HK11 as a CSP receptor.
In addition to hk11/rr11, characterization of the VicK/VicR TCSTS in
S. mutans revealed that it controls several important physiological proper-
ties, including competence development and biofilm formation, in this
bacterium (105).The vicK and vicR (also named covR/S (61)) genes encode
a putative HK and putative RR, respectively (9). Based on transformation
efficiency assays that were conducted with a vicR-overexpressing mutant, it
was observed that its natural transformability (without added CSP) was
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
258
reduced by approximately 100-fold compared with its UA159 wild-type
parent. Interestingly, although the addition of synthetic CSP increased
the transformation efficiency of the wild-type by approximately 1,000-
fold, the transformability of a vicK-null mutant and the vicR-overexpressing
strain were not restored to wild-type levels. Especially, when CSP was
added, the vicK-deficient mutant showed a 60-fold reduction in transform-
ation efficiency relative to the wild type, thereby suggesting that VicK is
responsive to CSP. However, as in the case of HK11, more studies are
warranted to assign a definite role for VicK as another receptor for CSP.
Recently, Wen et al. (120) examined the S. mutans homolog of the
bacterial trigger factor, designated RopA, which is a molecular chaperone
highly conserved in most bacteria. Interestingly, they found that inactiva-
tion of ropA affected multiple S. mutans phenotypic properties that
included decreased stress tolerance, reduced transformation efficiency,
and the formation of altered biofilms. Given the established role of the
trigger factor in proper folding, processing, and secretion of membrane-
associated and extracellular proteins, the authors suggested that RopA may
be involved in the processing and/or secretion of proteins, which can affect
competence development, biofilm formation, and bacterial stress response
in S. mutans.
CONCLUSIONS AND FUTURE PERSPECTIVES
Although the process of transformation has been known in strepto-
cocci for over 80 years, it is only in the past 10 years that we have under-
stood its molecular underpinnings. In this chapter, we have discussed how
quorum sensing in streptococci leads to the development of competence.
Interestingly, accumulating evidence suggests that genetic competence is
one of many important physiological properties regulated by this elegant
mechanism that has been linked to virulence in a plethora of bacterial
species. Because many of these quorum-sensing pathways have been asso-
ciated with critical cell functions, they have provided us with enticing
targets for antibacterial therapy.
Synthetic signaling molecules can be used to artificially induce intra-
species communication in streptococci; this can be exploited to design
therapeutics to treat or prevent streptococcal infections. Emerging litera-
ture discusses the possibility of interfering with quorum sensing as a novel
approach to control bacterial pathogenicity (30, 55, 123). The use of quorum-
sensing blockers to attenuate virulence, as opposed to conventional thera-
peutics such as antibiotics, can offer various advantages. Because these
259
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
blockers can inhibit virulence without necessarily killing bacterial cells, the
development of drug-resistant strains can be minimized.
REFERENCES
1 Ajdic, D., W. M. McShan et al. 2002. Genome sequence of Streptococcus mutans
UA159, a cariogenic dental pathogen. Proc. Natn. Acad. Sci. USA 99(22):
144349.
2 Alloing, G., C. Granadel et al. 1996. Competence pheromone, oligopeptide
permease, and induction of competence in Streptococcus pneumoniae. Molec.
Microbiol. 21(3): 4718.
3 Alloing, G., B. Martin et al. 1998. Development of competence in Streptococcus
pneumoniae: pheromone autoinduction and control of quorum-sensing by the
oligopeptide permease. Molec. Microbiol. 29(1): 7583.
4 Alloing, G., M. C. Trombe et al. 1990. The ami locus of the gram-
positive bacterium Streptococcus pneumoniae is similar to binding protein-
dependent transport operons of gram-negative bacteria. Molec. Microbiol. 4(4):
63344.
5 Aspiras, M. B., R. P. Ellen et al. 2004. ComX activity of Streptococcus mutans
growing in biofilms. FEMS Microbiol. Lett. 238(1): 16774.
6 Avery, O. T., C. M. MacLeod et al. 1944. Studies on the chemical nature of the
substance inducing transformation of pneumococcal types: induction of trans-
formation by a desoxyribonucleic acid fraction isolated from pneumococcus
Type III. J. Exp. Med. 79: 13758.
7 Banas, J. A. 2004. Virulence properties of Streptococcus mutans. Front. Biosci. 9:
126777.
8 Bassler, B. L. 2002. Small talk. Cell-to-cell communication in bacteria. Cell
109(4): 4214.
9 Bhagwat, S. P., J. Nary et al. 2001. Effects of mutating putative two-component
systems on biofilm formation by Streptococcus mutans UA159. FEMS Microbiol.
Lett. 205(2): 22530.
10 Boucher, Y., C. J. Douady et al. 2003. Lateral gene transfer and the origins of
prokaryotic groups. A. Rev. Genet. 37: 283328.
11 Campbell, E. A., S. Y. Choi et al. 1998. A competence regulon in Streptococcus
pneumoniae revealed by genomic analysis. Molec. Microbiol. 27(5): 92939.
12 Cheng, Q., E. A. Campbell et al. 1997. The com locus controls genetic transform-
ation in Streptococcus pneumoniae. Molec. Microbiol. 23(4): 68392.
13 Claverys, J. P. and L. S. Havarstein 2002. Extracellular-peptide control of compe-
tence for genetic transformation in Streptococcus pneumoniae. Front. Biosci. 7:
d1798814.
14 Claverys, J. P. and B. Martin 1998. Competence regulons, genomics and strepto-
cocci. Molec. Microbiol. 29(4): 11267.
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
260
15 Claverys, J. P., M. Prudhomme et al. 2000. Adaptation to the environment:
Streptococcus pneumoniae, a paradigm for recombination-mediated genetic plas-
ticity? Molec. Microbiol. 35(2): 2519.
16 Cvitkovitch, D. G. 2001. Genetic competence and transformation in oral strepto-
cocci. Crit. Rev. Oral Biol. Med. 12(3): 21743.
17 Cvitkovitch, D. G., Y. H. Li et al. 2003. Quorum-sensing and biofilmformation in
Streptococcal infections. J. Clin. Invest. 112(11): 162632.
18 Davies, D. G., M. R. Parsek et al. 1998. The involvement of cell-to-cell signals in
the development of a bacterial biofilm. Science 280(5361): 2958.
19 de Saizieu, A., C. Gardes et al. 2000. Microarray-based identification of a novel
Streptococcus pneumoniae regulon controlled by an autoinduced peptide.
J. Bacteriol. 182(17): 4696703.
20 Diep, D. B., L. S. Havarstein et al. 1994. The gene encoding plantaricin A,
a bacteriocin from Lactobacillus plantarum C11, is located on the same trans-
cription unit as an agr-like regulatory system. Appl. Environ. Microbiol. 60(1):
1606.
21 Dowson, C. G., V. Barcus et al. 1997. Horizontal gene transfer and the evolution
of resistance and virulence determinants in Streptococcus. Soc. Appl. Bacteriol.
Symp. Ser. 26: 42S51S.
22 Dowson, C. G., T. J. Coffey et al. 1993. Evolution of penicillin resistance in
Streptococcus pneumoniae; the role of Streptococcus mitis in the formation of a
low affinity PBP2B in S. pneumoniae. Molec. Microbiol. 9(3): 63543.
23 Dubnau, D. 1999. DNA uptake in bacteria. A. Rev. Microbiol. 53: 21744.
24 Echenique, J. R., S. Chapuy-Regaud et al. 2000. Competence regulation by
oxygen in Streptococcus pneumoniae: involvement of ciaRH and comCDE. Molec.
Microbiol. 36(3): 68896.
25 Echenique, J. R. and M. C. Trombe 2001. Competence repression under oxygen
limitation through the two-component MicAB signal-transducing system in
Streptococcus pneumoniae and involvement of the PAS domain of MicB.
J. Bacteriol. 183(15): 4599608.
26 Erdos, G., S. Sayeed et al. 2003. Development and characterization of a pooled
Haemophilus influenzae genomic library for the evaluation of gene expression
changes associated with mucosal biofilmformation in otitis media. Int. J. Pediatr.
Otorhinolaryngol. 67(7): 74955.
27 Fabret, C. and J. A. Hoch 1998. A two-component signal transduction system
essential for growth of Bacillus subtilis: implications for anti-infective therapy.
J. Bacteriol. 180(23): 637583.
28 Facklam, R. 2000. What happened to the Streptococci: overview of taxonomic
and nomenclature changes. Clin. Microbiol. Rev. 15: 61330.
29 Ferretti, J. J., D. Ajdic et al. 2004. Comparative genomics of streptococcal species.
Ind. J. Med. Res. 119 (Suppl.): 16.
30 Finch, R. G., D. I. Pritchard et al. 1998. Quorum-sensing: a novel target for anti-
infective therapy. J. Antimicrob. Chemother. 42(5): 56971.
261
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
31 Gasc, A. M., P. Giammarinaro et al. 1998. Organization around the dnaA gene of
Streptococcus pneumoniae. Microbiology 144 (2): 4339.
32 Gendron, R., D. Grenier et al. 2000. The oral cavity as a reservoir of bacterial
pathogens for focal infections. Microbes Infect. 2(8): 897906.
33 Giammarinaro, P., M. Sicard et al. 1999. Genetic and physiological studies of the
CiaH-CiaR two-component signal-transducing system involved in cefotaxime resist-
ance and competence of Streptococcus pneumoniae. Microbiology 145 (8): 185969.
34 Gilmore, K. S., P. Srinivas et al. (2003). Growth, development, and gene
expression in a persistent Streptococcus gordonii biofilm. Infect. Immun. 71(8):
475966.
35 Gilson, L., H. K. Mahanty et al. 1990. Genetic analysis of an MDR-like export
system: the secretion of colicin V. EMBO J. 9(12): 387594.
36 Grebe, T. W. and J. B. Stock 1999. The histidine protein kinase superfamily. Adv.
Microb. Physiol. 41: 139227.
37 Griffith, F. 1928. The significance of pneumococcal types. J. Hyg. 27: 11359.
38 Guenzi, E., A. M. Gasc et al. 1994. Atwo-component signal-transducing systemis
involved in competence and penicillin susceptibility in laboratory mutants of
Streptococcus pneumoniae. Molec. Microbiol. 12(3): 50515.
39 Havarstein, L. S. 1998. Bacterial gene transfer by natural genetic transformation.
APMIS (Suppl.) 84: 436.
40 Havarstein, L. S., G. Coomaraswamy et al. 1995. An unmodified heptadecapep-
tide pheromone induces competence for genetic transformation in Streptococcus
pneumoniae. Proc. Natn. Acad. Sci. USA 92(24): 111404.
41 Havarstein, L. S., D. B. Diep et al. 1995. A family of bacteriocin ABC transporters
carry out proteolytic processing of their substrates concomitant with export.
42 Havarstein, L. S., P. Gaustad et al. 1996. Identification of the streptococcal
competence-pheromone receptor. Molec. Microbiol. 21(4): 8639.
43 Havarstein, L. S., R. Hakenbeck et al. 1997. Natural competence in the genus
Streptococcus: evidence that streptococci can change phenotype by interspecies
recombinational exchanges. J. Bacteriol. 179(21): 658994.
44 Havarstein, L. S., H. Holo et al. 1994. The leader peptide of colicin V shares
consensus sequences with leader peptides that are common among peptide
bacteriocins produced by gram-positive bacteria. Microbiology 140 (9): 23839.
45 Havarstein, L. S., and D. A. Morrison 1999. Quorum-sensing and peptide phero-
mones in streptococcal competence for genetic transformation. In G. M. Dunny &
S. C. Winans (eds), Cell-Cell Signaling in Bacteria, pp. 926. Washington, DC:
ASM Press.
46 Hershey, A. D. and M. Chase 1952. Independent functions of viral protein and
nucleic acid in growth of bacteriophage. J. Gen. Physiol. 36: 3956.
47 Hillman, J. D., S. S. Socransky et al. 1985. The relationships between strepto-
coccal species and periodontopathic bacteria in human dental plaque. Arch. Oral.
Biol. 30(1112): 7915.
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
262
48 Holo, H., O. Nilssen et al. 1991. Lactococcin A, a newbacteriocin fromLactococcus
lactis subsp. cremoris: isolation and characterization of the protein and its gene.
J. Bacteriol. 173(12): 387987.
49 Hui, F. M. and D. A. Morrison 1991. Genetic transformation in Streptococcus
pneumoniae: nucleotide sequence analysis shows comA, a gene required for
competence induction, to be a member of the bacterial ATP-dependent transport
protein family. J. Bacteriol. 173(1): 37281.
50 Hui, F. M., L. Zhou et al. 1995. Competence for genetic transformation in
Streptococcus pneumoniae: organization of a regulatory locus with homology to
two lactococcin A secretion genes. Gene 153(1): 2531.
51 Jenkinson, H. F. 2000. Genetics of Streptococcus sanguis. In J. I. Rood (ed.), Gram-
Positive Pathogens, pp. 28794. Washington, DC: American Society for Microbiology.
52 Jenkinson, H. F., R. A. Baker et al. 1996. A binding-lipoprotein-dependent
oligopeptide transport system in Streptococcus gordonii essential for uptake of
hexa- and heptapeptides. J. Bacteriol. 178(1): 6877.
53 Jenkinson, H. F. and R. J. Lamont 1997. Streptococcal adhesion and colonization.
Crit. Rev. Oral. Biol. Med. 8(2): 175200.
54 Kleerebezem, M. and L. E. Quadri 2001. Peptide pheromone-dependent regula-
tion of antimicrobial peptide production in Gram-positive bacteria: a case of
multicellular behavior. Peptides 22(10): 157996.
55 Kleerebezem, M., L. E. Quadri et al. 1997. Quorum-sensing by peptide phero-
mones and two-component signal-transduction systems in Gram-positive bac-
teria. Molec. Microbiol. 24(5): 895904.
56 Knutsen, E., O. Ween et al. 2004. Two separate quorum-sensing systems upre-
gulate transcription of the same ABC transporter in Streptococcus pneumoniae.
J. Bacteriol. 186(10): 307885.
57 Kolenbrander, P. E. 2000. Oral microbial communities: biofilms, interactions,
and genetic systems. A. Rev. Microbiol. 54: 41337.
58 Lange, R., C. Wagner et al. 1999. Domain organization and molecular character-
ization of 13 two-component systems identified by genome sequencing of
Streptococcus pneumoniae. Gene 237(1): 22334.
59 Lee, M. S., B. A. Dougherty et al. 1999. Construction and analysis of a library for
random insertional mutagenesis in Streptococcus pneumoniae: use for recovery of
mutants defective in genetic transformation and for identification of essential
genes. Appl. Environ. Microbiol. 65(5): 188390.
60 Lee, M. S. and D. A. Morrison 1999. Identification of a new regulator in
Streptococcus pneumoniae linking quorum-sensing to competence for genetic
transformation. J. Bacteriol. 181(16): 500416.
61 Lee, S. F., G. D. Delaney et al. 2004. A two-component covRS regulatory system
regulates expression of fructosyltransferase and a novel extracellular carbohy-
drate in Streptococcus mutans. Infect. Immun. 72(7): 396873.
62 Leonard, C. G. 1973. Early events in development of streptococcal competence.
J. Bacteriol. 114(3): 1198205.
263
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
63 Li, Y. H., M. N. Hanna et al. 2001. Cell density modulates acid adaptation in
Streptococcus mutans: implications for survival in biofilms. J. Bacteriol. 183(23):
687584.
64 Li, Y. H., P. C. Lau et al. 2001. Natural genetic transformation of Streptococcus
mutans growing in biofilms. J. Bacteriol. 183(3): 897908.
65 Li, Y. H., P. C. Lau et al. 2002. Novel two-component regulatory system involved
in biofilm formation and acid resistance in Streptococcus mutans. J. Bacteriol.
184(22): 633342.
66 Li, Y. H., N. Tang et al. 2002. A quorum-sensing signaling system essential for
genetic competence in Streptococcus mutans is involved in biofilm formation.
J. Bacteriol. 184(10): 2699708.
67 Lina, G., S. Jarraud et al. 1998. Transmembrane topology and histidine protein
kinase activity of AgrC, the agr signal receptor in Staphylococcus aureus. Molec.
Microbiol. 28(3): 65562.
68 Loo, C. Y., D. A. Corliss et al. 2000. Streptococcus gordonii biofilm formation:
identification of genes that code for biofilm phenotypes. J. Bacteriol. 182(5):
137482.
69 Lorenz, M. G. and W. Wackernagel 1994. Bacterial gene transfer by
natural genetic transformation in the environment. Microbiol. Rev. 58(3):
563602.
70 Lunsford, R. D. 1998. Streptococcal transformation: essential features and appli-
cations of a natural gene exchange system. Plasmid 39(1): 1020.
71 Lunsford, R. D. and J. London 1996. Natural genetic transformation in
Streptococcus gordonii: comX imparts spontaneous competence on strain wicky.
J. Bacteriol. 178(19): 58315.
72 Lunsford, R. D. and A. G. Roble 1997. comYA, a gene similar to comGA of
Bacillus subtilis, is essential for competence-factor-dependent DNA transform-
ation in Streptococcus gordonii. J. Bacteriol. 179(10): 31226.
73 Luo, P., H. Li et al. 2003. ComX is a unique link between multiple quorum-
sensing outputs and competence in Streptococcus pneumoniae. Molec. Microbiol.
50(2): 62333.
74 Luo, P., H. Li et al. 2004. Identification of ComW as a new component in
the regulation of genetic transformation in Streptococcus pneumoniae. Molec.
Microbiol. 54(1): 17283.
75 Luo, P. and D. A. Morrison 2003. Transient association of an alternative sigma
factor, ComX, with RNA polymerase during the period of competence for genetic
transformation in Streptococcus pneumoniae. J. Bacteriol. 185(1): 34958.
76 Magnuson, R., J. Solomon et al. 1994. Biochemical and genetic characterization
of a competence pheromone from B. subtilis. Cell 77(2): 20716.
77 Marsh, P. D. 2000. Oral ecology and its impact on oral microbial diversity.
In Ellen, R. P. (ed.), Oral Bacterial Ecology: the Molecular Basis, pp. 1165.
Wymondham, UK: Horizon Scientific Press.
78 Marsh, P. D. 2004. Dental plaque as a microbial biofilm. Caries Res. 38(3): 20411.
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
264
79 Martin, B., P. Garcia et al. 1995. The recA gene of Streptococcus pneumoniae is part
of a competence-induced operon and controls an SOS regulon. Dev. Biol. Stand.
85: 293300.
80 Martin, B., M. Prudhomme et al. 2000. Cross-regulation of competence
pheromone production and export in the early control of transformation in
81 Marugg, J. D., C. F. Gonzalez et al. 1992. Cloning, expression, and nucleotide
sequence of genes involved in production of pediocin PA-1, and bacteriocin from
Pediococcus acidilactici PAC1.0. Appl. Environ. Microbiol. 58(8): 23607.
82 Mascher, T., D. Zahner et al. 2003. The Streptococcus pneumoniae cia regulon:
CiaR target sites and transcription profile analysis. J. Bacteriol. 185(1): 6070.
83 McNab, R. and H. F. Jenkinson 1998. Altered adherence properties of
a Streptococcus gordonii hppA (oligopeptide permease) mutant result from
transcriptional effects on cshA adhesin gene expression. Microbiology 144 (1):
12736.
84 Mejean, V. and J. P. Claverys 1993. DNA processing during entry in transform-
ation of Streptococcus pneumoniae. J. Biol. Chem. 268(8): 55949.
85 Mitchell, T. J. 2003. The pathogenesis of streptococcal infections: from tooth
decay to meningitis. Nat. Rev. Microbiol. 1(3): 21930.
86 Morrison, D. A., M. C. Trombe et al. 1984. Isolation of transformation-deficient
Streptococcus pneumoniae mutants defective in control of competence, using
insertion-duplication mutagenesis with the erythromycin resistance determinant
of pAM beta 1. J. Bacteriol. 159(3): 8706.
87 Mortier-Barriere, I., A. de Saizieu et al. 1998. Competence-specific induction of
recA is required for full recombination proficiency during transformation in
88 Moscoso, M. and J.-P. Claverys 2004. Release of DNA into the medium by
competent Streptococcus pneumoniae: kinetics, mechanism and stability of the
liberated DNA. Molec. Microbiol. 54(3): 78394.
89 Pakula, R. and W. Walczak 1963. On the nature of competence of transformable
streptococci. J. Gen. Microbiol. 31: 12533.
90 Pestova, E. V., L. S. Havarstein et al. 1996. Regulation of competence for genetic
transformation in Streptococcus pneumoniae by an auto-induced peptide phero-
mone and a two-component regulatory system. Molec. Microbiol. 21(4): 85362.
91 Pestova, E. V. and D. A. Morrison 1998. Isolation and characterization of three
Streptococcus pneumoniae transformation-specific loci by use of a lacZ reporter
insertion vector. J. Bacteriol. 180(10): 270110.
92 Petersen, F. C., D. Pecharki et al. 2004. Biofilm mode of growth of Streptococcus
intermedius favored by a competence-stimulating signaling peptide. J. Bacteriol.
186(18): 632731.
93 Petersen, F. C. and A. A. Scheie 2000. Genetic transformation in Streptococcus
mutans requires a peptide secretion-like apparatus. Oral Microbiol. Immunol.
15(5): 32934.
265
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
94 Peterson, S., R. T. Cline et al. 2000. Gene expression analysis of the Streptococcus
pneumoniae competence regulons by use of DNA microarrays. J. Bacteriol.
182(21): 6192202.
95 Peterson, S. N., C. K. Sung et al. 2004. Identification of competence pheromone
responsive genes in Streptococcus pneumoniae by use of DNA microarrays. Molec.
Microbiol. 51(4): 105170.
96 Pozzi, G., L. Masala et al. 1996. Competence for genetic transformation in
encapsulated strains of Streptococcus pneumoniae: two allelic variants of the
peptide pheromone. J. Bacteriol. 178(20): 608790.
97 Reichmann, P. and R. Hakenbeck 2000. Allelic variation in a peptide-inducible
two-component system of Streptococcus pneumoniae. FEMS Microbiol. Lett.
190(2): 2316.
98 Rimini, R., B. Jansson et al. 2000. Global analysis of transcription kinetics
during competence development in Streptococcus pneumoniae using high density
DNA arrays. Molec. Microbiol. 36(6): 127992.
99 Roberts, A. P., J. Pratten et al. 1999. Transfer of a conjugative transposon,
Tn5397 in a model oral biofilm. FEMS Microbiol. Lett. 177(1): 636.
100 Robinson, V. L., D. R. Buckler et al. 2000. A tale of two components: a novel
kinase and a regulatory switch. Nat. Struct. Biol. 7(8): 62633.
101 Russell, R. R. B. 2000. Pathogenesis of oral streptococci. In J. I. Rood et al. (eds),
Gram-Positive Pathogens, pp. 2729. Washington, DC: American Society for
Microbiology Press.
102 Sablon, E., B. Contreras et al. 2000. Antimicrobial peptides of lactic acid bac-
teria: mode of action, genetics and biosynthesis. Adv. Biochem. Eng. Biotechnol.
68: 2160.
103 Schroder, H., T. Langer et al. 1993. DnaK, DnaJ and GrpE form a cellular
chaperone machinery capable of repairing heat-induced protein damage.
EMBO J. 12(11): 413744.
104 Sebert, M. E., L. M. Palmer et al. 2002. Microarray-based identification of htrA, a
Streptococcus pneumoniae gene that is regulated by the CiaRH two-component
system and contributes to nasopharyngeal colonization. Infect. Immun. 70(8):
405967.
105 Senadheera, D., C. Huang et al. 2004. Streptococcus mutans covR/S genes
control adhesion, biofilm formation and competence development. J. Dent.
Res. Special Issue (Proceedings from the International Association for Dental
Research Annual Meeting), Abstr. 3001.
106 Socransky, S. S. and A. D. Haffajee 2002. Dental biofilms: difficult therapeutic
targets. Periodontol. 2000 28: 1255.
107 Stein, T., S. Borchert et al. 2002. Dual control of subtilin biosynthesis and
immunity in Bacillus subtilis. Molec. Microbiol. 44(2): 40316.
108 Steinmoen, H., E. Knutsen et al. 2002. Induction of natural competence in
Streptococcus pneumoniae triggers lysis and DNA release from a subfraction of
the cell population. Proc. Natn. Acad. Sci. USA 99(11): 76816.
M
.
D
.
S
E
N
A
D
H
E
E
R
A
E
T
A
L
.
266
109 Steinmoen, H., A. Teigen et al. 2003. Competence-induced cells of Streptococcus
pneumoniae lyse competence-deficient cells of the same strain during cocultiva-
tion. J. Bacteriol. 185(24): 717683.
110 Stock, A. M., V. L. Robinson et al. 2000. Two-component signal transduction.
A. Rev. Biochem. 69: 183215.
111 Strauch, M. A. and J. A. Hoch 1993. Signal transduction in Bacillus subtilis
sporulation. Curr. Opin. Genet. Dev. 3(2): 20312.
112 Sturme, M. H., M. Kleerebezem et al. 2002. Cell to cell communication by
autoinducing peptides in gram-positive bacteria. Antonie Van Leeuwenhoek
81(14): 23343.
113 Throup, J. P., K. K. Koretke et al. 2000. A genomic analysis of two-component
signal transduction in Streptococcus pneumoniae. Molec. Microbiol. 35(3): 56676.
114 Tomasz, A. 1965. Control of the competent state in Pneumococcus by a hor-
mone-like cell product: an example for a new type of regulatory mechanism in
bacteria. Nature 208(6): 1559.
115 Tomasz, A. and S. M. Beiser 1965. Relationship between the competence anti-
gen and the competence-activator substance in pneumococci. J. Bacteriol. 90(5):
122632.
116 Tomasz, A. and R. D. Hotchkiss 1964. Regulation of the transformability of
pneumococcal cultures by macromolecular cell products. Proc. Natn. Acad. Sci.
USA 51: 4807.
117 Wagner, C., A. Saizieu Ad et al. 2002. Genetic analysis and functional character-
ization of the Streptococcus pneumoniae vic operon. Infect. Immun. 70(11): 61218.
118 Wang, B. Y., B. Chi et al. 2002. Genetic exchange between Treponema denticola
and Streptococcus gordonii in biofilms. Oral Microbiol. Immunol. 17(2): 10812.
119 Ween, O., P. Gaustad et al. 1999. Identification of DNA binding sites for ComE,
a key regulator of natural competence in Streptococcus pneumoniae. Molec.
Microbiol. 33(4): 81727.
120 Wen, Z. T., D. G. Suntharaligham et al. 2005. Trigger factor in Streptococcus
mutans is involved in stress tolerance, competence development, and biofilm
formation. Infect. Immunol. 73: 21925.
121 Whittaker, C. J., C. M. Klier et al. 1996. Mechanisms of adhesion by oral bacteria.
A. Rev. Microbiol. 50: 51352.
122 Winans, G. M. D. a. S. C. (ed.) 1999. Cell-Cell Signaling in Bacteria. Washington,
DC: American Society of Microbiology Press.
123 Zhang, L. H. and Y. H. Dong 2004. Quorum-sensing and signal interference:
diverse implications. Molec. Microbiol. 53(6): 156371.
124 Zhulin, I. B., B. L. Taylor et al. 1997. PAS domain S-boxes in Archaea, Bacteria
and sensors for oxygen and redox. Trends Biochem. Sci. 22(9): 3313.
125 Puyet, A., B. Greenberg and S. A. Lacks 1990. Genetic and structural character-
ization of end A, a membrane-bound nuclease required for transformation of
Streptococcus pneumoniae. J. Molec. Biol. 213: 72738.
267
R
E
G
U
L
A
T
I
O
N
O
F
S
T
R
E
P
T
O
C
O
C
C
A
L
C
O
M
P
E
T
E
N
C
E
CHAPTER 11
Signaling by a cell-surface-associated signal
during fruiting-body morphogenesis in
Myxococcus xanthus
Lotte Sgaard-Andersen
Max Planck Institute for Terrestrial Microbiology,
Marburg, Germany
INTRODUCTION
Over the past decade the perception of bacterial cells as autonomous
individuals, each following their own agenda and not interacting with each
other, has been replaced by the view that bacteria interact extensively both
withinand betweenspecies by means of intercellular signal molecules. Eachof
these signal molecules constitutes part of an information system that is
constructed of four parts: the donor cell synthesizing the signal, the signal
molecule, the recipient cell, and the output response. As in any other
information system, the signal must be tailored to the talents of the recipient.
A clear example of a tailor-made signal molecule is the C-signal molecule in
Myxococcus xanthus. Most intercellular signals identified in bacteria are small
(i.e. with a molecular mass of less than 1000Da), freely diffusible molecules
that are part of quorum sensing systems, which help bacterial cells to assess
population size (90). However, that is not the case for the C-signal in
Myxococcus xanthus. The C-signal is a 17 kDa cell-surface-associated protein
and is thus non-diffusible, and it helps to guide M. xanthus cells into nascent
fruiting bodies and to assess their position in a field of cells.
C-signal transmission occurs by a contact-dependent mechanism,
i.e. it depends on direct contact between the donor and the receiving cell.
The C-signal is used repeatedly during the starvation-induced formation of
spore-filled fruiting bodies in M. xanthus. Early during fruiting-body for-
mation, the C-signal induces the aggregation of cells into the nascent
fruiting bodies. In the late stage of fruiting body formation, the C-signal
induces sporulation of cells that have accumulated inside the fruiting
bodies. During the same time interval, the C-signal induces the express-
ion of a large number of genes. Why, then, has M. xanthus adopted a
269
cell-surface-associated signal molecule rather than a freely diffusible molecule
to induce the two major morphogenetic events in fruiting body formation?
M. xanthus moves by gliding motility and is faced with the problem that the
average gliding speed is only 25 mmmin
1
. It has been argued that this speed,
which is only 50100 times higher than the rate of continental drift, is so low
that the directive properties of a diffusible aggregation signal would disperse
before M. xanthus cells could reorient in a gradient of such a signal.
The solution adopted by M. xanthus cells to overcome this problem is to use
a cell-surface-associated signal. The advantage of this type of signal is that it
moves at the same speed as the cells and therefore allows cells to reorient
without losing their sense of direction. A second advantage conferred by a
signal molecule that hinges on a contact-dependent signal transmission
mechanism is that it confers information to the recipient about its position
relative to that of other cells in its immediate neighbourhood, i.e. signaling
levels are proportional to cell density. Thus, the output response of a cell can
be tied in with the position of that cell relative to that of other cells. This
property of the C-signaling system is the key to an understanding of one of
the hallmarks in fruiting-body formation, the position-specific sporulation of
those cells that have accumulated inside the fruiting bodies.
The complete understanding of an intercellular signaling system
entails an understanding of how the signal molecule is synthesized, how
synthesis is regulated, how the signal is received, and how the reception of
the signal is transformed into an output response. In this chapter, these
questions are discussed in the context of the C-signaling system in
M. xanthus and its role in fruiting-body formation.
MULTICELLULARITY AS A SURVIVAL STRATEGY
Myxobacteria are the gold standard for social bacteria, although in
strong competition with Actinomycetes. Myxobacteria have adopted a
social lifestyle or multicellularity as a survival strategy. Myxobacteria
are found in the topsoil, where they feed on organic matter and prey on
other microorganisms by secreting hydrolytic enzymes and antimicrobials
(68). M. xanthus forms two morphologically distinct types of biofilm,
depending on the nutritional status of the cells (Figure 11.1). In the pre-
sence of nutrients, the motile, rod-shaped cells grow and divide.
Myxobacteria move by gliding (31, 81); if cells are present on a solid surface
they formspreading, cooperatively feeding colonies (13). Cells at the edge of
a colony spread coordinately over the surface, forming a thin, film-like
structure (Figure 11.1). The developmental program that culminates in
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
270
the formation of the spore-filled fruiting bodies is initiated in response to
starvation of cells at a high density on a solid surface (13). Fruiting-body
formation proceeds in a relatively stereotyped pattern of morphological
events, which are separated in time and space (Figures 11.1, 11.2).
The starting point is an initially unstructured lawn of randomly oriented
cells on a solid surface. The first explicit signs of fruiting-body formation
are evident after 46h with changes in cell behaviour (29) as the cells begin
to aggregate to form small aggregation centers (Figure 11.2). A character-
istic of the aggregation process is that cells move into the aggregation
centers organized as streams of cells, rather than entering the centers
as single cells from all directions (61) (Figure 11.2). As more cells enter
the aggregation centers, these centers increase in size and eventually
become symmetric mounds. By 24 h, the aggregation process is complete
and the nascent fruiting bodies each contain c. 10
5
densely packed cells
(Figures 11.1, 11.2). Inside the nascent fruiting bodies, the rod-shaped,
motile cells undergo morphological and physiological differentiation
into spherical, non-motile, dormant spores, resulting in mature fruiting
bodies (Figure 11.1). Although the aggregation process takes c. 24h to com-
plete, spore maturation is finished c. 72 h after the onset of starvation.
Although all cells exposed to starvation have the potential to develop into
spores (14, 52) only 1020% of cells differentiate into spores. These cells are
specifically those which have accumulated inside the mounds. Up to 30%
nutrients
+ nutrients
0 h 6 h 9 h
12 h 18 h 24 h
72 h
Figure 11.1. The life cycles of Myxococcus xanthus. In the panel to the left a vegetative,
spreading colony is shown. The diameter of the colony is c. 2 cm. In the scanning EM
pictures to the right, the different stages of fruiting body formation are shown. In the 72 h
panel, a fruiting body has been opened to visualize the spores. Scale bars in 24 and 72 h
panels: 5 and 10mm, respectively. The scanning EM pictures are adapted from Kuner and
Kaiser (43).
271
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
remain outside the fruiting bodies. These cells remain rod-shaped and
differentiate to a cell type called peripheral rods. Even after extended periods
of starvation, peripheral rods do not differentiate into spores (60). Finally,
the remaining cells undergo lysis (69).
Fruiting-body formation involves temporally coordinated changes in
gene expression in which genes are turned on at specific time points during
development (27, 41) (Figure 11.2). Moreover, developmental gene expression
is spatially controlled (Figure 11.2). Genes turned on after 6h are preferen-
tially expressed in aggregating and sporulating cells, whereas genes activated
prior to 6h are expressed in all cells including peripheral rods (30). Finally,
genetic and biochemical evidence suggests that fruiting-body formation
depends on the exchange of intercellular signals (71) (Figure 11.2).
Myxobacteria are the only bacteria that cope with starvation by forming
spore-filled fruiting bodies. This survival strategy is optimally suited to the
multicellular life-style of myxobacteria. Each fruiting body consists of c. 10
5
0
C
Cellular organization
Developmental gene
expression
Hours of development
Intercellular signals
6 12 18 24 72
A
0 h 6 h 15 h 24 h
Figure 11.2. Morphogenesis of fruiting bodies in Myxococcus xanthus. On the developmental
timeline, each triangle represents a gene, which is turned on during fruiting-body
formation. Black triangles represent genes that are expressed in all cells; grey triangles
represent genes that are expressed only in aggregating and sporulating cells. The times of
action of the A and C signals are indicated by the arrows below the timeline. Cellular
arrangements during the different stages of fruiting-body formation are shown above the
timeline. Cell arrangements are visualized by fluorescence microscopy of green fluorescent
protein (GFP)-labeled cells. GFP-labeled wild-type cells were co-developed with non-
fluorescent wild-type cells at a ratio of 1 : 40. Images were acquired at the time points
indicated. White circles and arrows in the 6h image indicate aggregation centers and
streams, respectively. White arrows in the 15 h image indicate streams of cells entering a
nascent fruiting body. Scale bar: 50mm. The pictures of the cellular arrangements during
the different stages of fruiting-body formation are adapted from (28).
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
272
spores and is essentially a spore-filled sac. Under the appropriate condi-
tions, the spores in fruiting bodies germinate to give rise to motile, meta-
bolically active vegetative cells. Germination immediately gives rise to a
new, spreading, cooperatively feeding colony. Thus, fruiting bodies are
optimally designed to ensure that a new vegetative cycle is initiated by a
community of cells rather than by a single cell.
GLIDING MOTILITY IN M. XANTHUS
The multicellular lifestyle of M. xanthus crucially depends on the
ability of cells to display active movement. M. xanthus cells move by gliding
motility, which is the movement of a rod-shaped cell in the direction of
the cells long axis on a surface in the absence of a flagellum (22). Gliding
and its regulation constitutes the basis for the formation of spreading
colonies and f ruiting-bo dy fo rma tio n. M. xanthus has two mech anisms
for gliding, adventurous (A)- and social (S)-motility (24). Mutations in both
systems (A
, S
mutants) result in a non-motile phenotype; mutations that

inactivate only one system leave the cells motile by means of the remaining,
intact system. The speed of gliding is highly variable for a given cell as it moves
across the substrate (Figure 11 .3 b). Periodically, cells stop and then either
resume gliding in the same direction or undergo a reversal in which the
head becomes the tail and vice versa (5, 28, 29, 82 ) (Figure 11.3 a). S-motility
is generally only operational when cells are within contact distance of each
other (23), whereas A-motility is operational in single cells (23).
The motility engine that provides the force for S-motility is generated
by the retraction of type IV pili (Tfp) (32, 92); for a review on Tfp, see (55).
Tfp are polarly localized structures and are normally only present at one
pole at a time ( 32) (Figure 11.3c ). Extension of T fp from a cell pole, followed
by retraction of Tfp, provides the force for cell movement (58, 76, 85).
In addition to Tfp, S-motility depends on LPS O-antigen (6) and on
extracellular matrix fibrils (1). Extracellular matrix fibrils are fibers
30nm thick that coat the cell surface and form an extracellular matrix
in which neighboring cells are interconnected (1, 4). The fibrils are
made from equal amounts of polysaccharide and protein (3). The contact-
dependence of Tfp-dependent motility in M. xanthus has been rationalized
by the observation that the polysaccharide portion of the extracellular
matrix fibrils triggers the retraction of Tfp (51). According to this scheme,
S-motility relies on the extension of Tfp from the pole of a cell and
attachment to the extracellular matrix fibrils on a nearby cell, followed
273
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
(a)
(b)
G
l
i
d
i
n
g

s
p
e
e
d

(
m

m
i
n
1
)
Time (seconds)
0 s 60 s 120 s 180 s 240 s 300 s 360 s 420 s
480 s 540 s 600 s
120 240 360 480 720
660 s 720 s 780 s 840 s 900 s
2 4
1
3
2
1
0
2
1
3
3 5
Direction of gliding
S-engine
Type IV pili
Pulling
A-engine
Active nozzle-like structures
Pushing
(c)
600 840
Figure 11.3. Characteristics of Myxococcus xanthus gliding motility. (a) A sequence of
phase-contrast images of a single cell gliding on a solid surface. Images were recorded
every 15 s for 900s. Slime deposited by the cell is evident as a refractile trail; white
arrows indicate the direction of gliding. (b) The speed profile of the cell in (a). The
gliding speed was calculated for each 15 s interval between recordings and plotted as a
function of recording time. The dashed horizontal lines indicate the detection limit for
active movement (0.35 mmmin
1
). The dotted vertical lines indicate intervals where the
cell showed active movement (intervals 2, 3, and 4) and periods of no active movement
(intervals 1 and 5). A change in gliding speed froma negative to a positive value or vice versa
indicates a reversal. This cell reversed its direction of gliding at t 75, 345, 570 and 870s.
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
274
by retraction of Tfp and thus the forward movement of the Tfp
containing cell. Thus, retraction of Tfp essentially pulls cells forward. Tfp
are only observed at one pole at a time (32). Nevertheless, A
cells, which
move only by means of their S-engine, reverse direction at regular intervals
(82), suggesting that the pole at which the Tfp are located changes
regularly.
M. xanthus cells moving by means of A-motility leave behind a slime
trail. Currently, the secreted slime in these slime trails is thought to be
the key to an understanding of A-motility. Wolgemuth et al. (91) identified
(by electron microscopy) nozzle-like structures, which are clustered at
both poles of A- m otile cells in M. xanthus (Figure 11.3c ). These s tructures
are similar to the junctional pore complexes used in gliding cyanobacteria
to secrete the slime that generates gliding motility in these bacteria (25).
Accordingly, it was proposed that the nozzle-like structures in M. xanthus
are also involved in slime secretion and that slime extrusion generates
the force for movement. To account for how slime extrusion could
produce sufficient thrust to propel a bacterial cell, Wolgemuth et al. (91)
proposed that slime is introduced into the nozzle-like structures in a
dehydrated form. The subsequent hydration of the slime would cause
it to swell and be extruded from the nozzle. Adherence of the hydrated
slime to the substrate would provide the slime with a footing that would
allow it to push the cell forward. Slime extrusion thus essentially pushes
cells forward. Even though the nozzle-like structures are observed at
both poles, slime extrusion is only observed at one pole at a time, implying
that only one polar cluster of the nozzle-like structures is actively
se creting slime at a t ime ( Figure 11.3c ). Nevertheless, A
cells that
move as single cells, and hence only translocate by means of their A-engine,
reverse direction regularly (5). This, in turn, implies that the polarity of
the active nozzles also switches at regular intervals. The molecular compo-
sition of the nozzle-like structures in M. xanthus is so far unknown.
Figure 11.3. (cont.)
(c) Longitudinal cross-section of M. xanthus cells on a solid surface, showing the arrangement
and location of the two engines involved in gliding motility. The dashed and solid lines
indicate the outer and inner membrane, respectively. Type IV pili project from the left
pole. Black cylinders indicate proteins involved in biogenesis and retraction of type IV pili
spanning the cell envelope. White barrels indicate the slime-extruding nozzle-like structures
implicated in A-motility. These structures are present at both poles; however, only one
cluster at a time is actively secreting slime. In this cell, the cluster at the right pole secretes
slime as indicated in grey. Parts (a) and (b) were adapted from(28); (c) was adapted from(78).
275
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
Interestingly, homologs of the TolR, TolB, and TolQ proteins have been
shown to be important for A-motility (89, 93). In Gram-negative bacteria
these proteins, together with the TolA and Pal proteins, form a complex,
which is involved in transport processes across the inner and outer mem-
branes (53).
INTERCELLULAR SIGNALING DURING FRUITING-BODY
MORPHOGENESIS
Whereas gliding motility and its regulation constitutes the basis for the
formation of fruiting bodies, the role of the intercellular signals involved
in fruiting-body formation seems to be to coordinate and synchronize the
efforts of thousands of cells during the construction process. Genetic
evidence suggests that there may be at least five intercellular signals,
known as the A- to E-signals, which are involved in fruiting-body formation
(71). Initially, mutants deficient in the synthesis of an intercellular signal
required for fruiting-body formation were identified in a collection
of mutants that were unable to complete fruiting body formation and
sporulation (18). A specific characteristic of a group of these mutants
was that their development was rescued by co-development with wild-type
cells. This rescue is referred to as extracellular complementation (18)
to emphasize that it does not involve transfer of genetic material from
wild-type cells to mutant cells. In addition it was found that none of the
non-autonomous mutants were auxotrophs, suggesting that extracellular
complementation did not involve cross-feeding (18). Extracellular comple-
mentation experiments involving pairs of non-autonomous mutants lead
to the classification of the mutants into four classes referred to as the
asg (A-signal), bsg (B-signal), csg (C-signal) and dsg (D-signal) mutants,
respectively (18). In these co-development experiments, mutants from one
class were able to rescue development of a mutant from a different class, i.e.
an asg mutant could rescue sporulation of a bsg mutant and vice versa,
whereas mutants belonging to the same class did not result in rescue of
development when co-developed (18). Later, a fifth class of non-autonomous
mutants referred to as the esg (E-signal) mutant was identified (12).
Extracellular complementation was explained by suggesting that the inability
of a non-autonomous mutant to complete development was caused by the
inability to produce an intercellular signal required for development. In the
co-development experiments, this missing signal would be provided either
by wild-type cells or by a mutant from a different class of non-autonomous
mutants (18).
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
276
Despite intensive work, only two of the intercellular signals the A- and
C-signals have been characterized biochemically. Likewise, the A- and
C-signals are the only signals whose function has been clarified. The
A-signal becomes important for development after 2 h of starvation (44)
(Figure 11.2). The A-signal consists of a subset of amino acids, which are
produced by the action of extracellular proteases that are activated in
response to starvation (45, 65). The A-signal functions as part of a system
that measures the density of starving cells: each starving cell produces a
constant amount of A-signal and once a threshold concentration of A-signal
is reached development proceeds (46). The function of the A-signaling
system therefore seems to be to ascertain that fruiting-body morphogenesis
is not initiated unless a sufficiently high number of cells are starving. The
C-signal comes into action after 6h of starvation, coinciding with the first
signs of morphogenesis (40).
THE C-SIGNAL INDUCES THREE RESPONSES THAT ARE
SEPARATED IN TIME AND SPACE
Mutants that cannot synthesize the C-signal are unable to aggregate
and sporulate (73) and the expression of the genes that are normally turned
on from 6h is reduced or abolished (40). The C-signal is encoded by the
csgA gene (73). Consequently, in a csgA mutant the developmental program
is arrested after 6h. Development of csgA mutant cells can be restored in
either of two ways: by co-development with wild-type cells or with cells from
either the asg, bsg, dsg, or esg classes of non-autonomous mutants (18), or by
exogenous C-signal (36).
Formally, the developmental defects in a csgA mutant and the rescue
of these defects by extracellular complementation or by exogenous C-signal
only provide evidence that the C-signal is required for fruiting-body formation.
These experiments do not provide evidence that the C-signal is a true signal
that induces a specific response. To distinguish between these possibilities,
Kruse et al. (42) overproduced the C-signal early during starvation, the idea
being that if the C-signal was a true signal that induced a response(s), then
this/these response(s) would be induced earlier in the strain overproducing
the signal. In the strain overproducing the C-signal early during development
it was observed that aggregation, sporulation, and C-signal-dependent gene
expression were indeed induced earlier than in wild-type cells. Thus, this
experiment provided the crucial piece of evidence demonstrating that the
C-signal induces aggregation, sporulation, and full expressionof developmental
genes, which are turned on after 6h.
277
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
THE MOLECULAR NATURE OF THE C-SIGNAL
The molecular nature of the C-signal has been controversial. Kim and
Kaiser (36, 37) originally purified the C-signal by detergent extraction and
biochemical fractionation from starving M. xanthus cells on the basis of
its ability to rescue development of csgA cells. Kim and Kaiser identified the
C-signal as a protein with molecular mass c. 17 kDa, which is encoded by
the csgA gene. The understanding that this 17 kDa protein was the C-signal
was questioned by several observations that suggested that the CsgA
protein might act as an enzyme to produce the C-signal. Firstly, from the
sequence of the csgA gene, the size of the CsgA protein was predicted to be
25 kDa (47). Secondly, the CsgA protein is homologous to members of
the short-chain alcohol dehydrogenase (SCAD) family of intracellular
proteins (2, 47). SCAD enzymes contain two conserved sequence motifs,
both of which are present in the CsgA protein: an N-terminal motif corres-
ponding to the NAD(P)-coenzyme binding pocket and a more C-terminal
motif involved in the catalytic mechanism (63). Consistently, the CsgA
protein was observed to bind NAD
in vitro (47). Thirdly, exogenous

CsgA protein purified from E. coli restored development of csgA cells; how-
ever, exogenous CsgAproteins carrying substitutions in either the coenzyme-
binding pocket or the catalytic site failed to restore development of csgA cells
(47). Finally, overproduction of the SocA protein, which is homologous to
SCAD enzymes, restored development in a csgA mutant in vivo (48, 49).
However, enzymatic activity of full-length CsgA protein has not been
demonstrated.
Recently, these conflicting data were at least partly reconciled. First it
was shown that the CsgA protein exists in two forms, one with a molecular
mass of c. 25 kDa (designated p25), which corresponds to full-length CsgA
protein, and one with a molecular mass of c. 17 kDa (designated p17) (42),
which is similar in size to the C-signal protein purified by Kim and Kaiser
(36, 37). The p25 protein is present in vegetative cells and accumulates
during fruiting-body formation, whereas p17 is only detected in starving
cells (42). By partially purifying the C-signal from starving cells, Lobedanz
and Sgaard-Andersen showed that the C-signal co-purifies with p17 (54).
In biochemical fractionation experiments, p17 and p25 were both found to
be localized to the outer membrane (54). Moreover, by using antibodies
directed against the N- and C-terminal of p25, evidence was provided
that p17 corresponds to the C-terminal c. 17 kDa of p25 (54). It has been
found consistently that recombinant p17 proteins, which corresponded
to the C-terminal c. 17kDa of p25 and which were purified from E. coli, have
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
278
C-signal activity (54). Importantly, these recombinant p17 proteins lack the
NAD
coenzyme-binding pocket and are unable to bind NAD
in vitro (54).
Thus, p17 does not depend on SCAD activity to engage in C-signaling. Rather
these data strongly support the idea that p17 is the bona fide C-signal.
The precise N-terminus of p17 remains to be determined. Another
unsolved question regarding the exact biochemical nature of the C-signal
is how p25 and p17 are anchored in the outer membrane. A homology
model of p25 revealed that p25 is not an amphiphilic protein and does not
have the b-barrel structure typical of outer membrane proteins (54).
However, in Triton X-114 phase separation experiments, p17 and p25 both
partition to the detergent phase (54). As this partitioning pattern is a
characteristic of amphiphilic membrane proteins and lipoproteins, this
led to the speculation that p25 and p17 are post-translationally modified
with a hydrophobic molecule and that this hydrophobic moiety serves to
anchor the proteins in the outer membrane. Neither the chemical nature
of this modification nor the position in p17 and p25 that carries the
modification have been determined. It should be pointed out that secreted
lipoproteins are typically modified on an N-terminal Cys residue. However,
p25 does not contain a lipoprotein signal peptide.
SYNTHESIS OF THE C-SIGNAL
To understand a signaling pathway, it is essential to define the mechanisms
by which cells produce this signal. Shimkets and co-workers have provided
evidence that the csgA gene is transcribed from the same promoter during
vegetative growth and development (50) and that the translation start
codon for p25 synthesis is essential for synthesis of the C-signal (47). These
observations argued against the idea that p17 was synthesized from an
alternative start codon. A synthesis mechanism for p17 involving translational
frameshifts also appeared unlikely, as this mechanismwould involve at least two
frameshifts. Synthesis mechanism involving transcriptional or translational
mechanism having been ruled out by Shimkets and co-workers, Lobedanz
and Sgaard-Andersentested whether p17 synthesis involved proteolytic proces-
sing of p25 (54). This hypothesis was tested experimentally by using an in vitro
protease assay in which total cell extracts were prepared from starving
M. xanthus cells. When this extract was added to a recombinant p25 protein
purified fromE. coli, a protein with a molecular mass of c. 17 kDa was produced.
This protein was recognized by the antibodies against the C-terminus of p25,
whereas the antibodies against the N-terminus of p25 did not recognize it.
Control experiments provided evidence that the recombinant p25 protein
279
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
does not harbor the protease activity involved in the N-terminal proteolytic
processing of p25. By adding protease inhibitors specific for different types of
proteases to the cell extract, evidence was obtained that the protease involved in
p25 processing is a serine protease. Consistent with the observation that p17 is
only detected in developing cells (42), it was observed that the activity of the
protease is developmentally regulated. The protease and the corresponding
gene remain to be identified.
Several observations in addition to those already mentioned still need
to be explained before the molecular nature of the C-signal can be declared
solved. In particular, the inability of mutant p25 proteins, which carry sub-
stitutions in the coenzyme binding pocket or in the catalytic site, to rescue
development of csgAcells in the C-signal bioassay (47) is thought-provoking. It
has been speculated that the substitutions in the coenzyme binding pocket
may interfere with proteolytic processing of p25 and that the substitutions in
the catalytic site could interfere with recognition by the C-signal receptor (54).
An unresolved issue is the potential enzymatic activity of p25. Is p25
an enzymatic fossil, which no longer has SCAD activity and which only
functions as a precursor for p17? Or does p25 still have enzymatic activity?
Clearly, this potential enzymatic activity is not required for fruiting body
formation, as development of csgA cells is rescued by exogenous p17.
Irrespective of whether p25 has SCAD activity, it is interesting to speculate
about why p25 was, at some point in its evolution, adopted to become the
precursor for the 17 kDa C-signal protein. An enzyme is optimized to recognize
a substrate. This interaction could potentially be exploited to generate a
signalreceptor pair in which the signal molecule is the substrate-binding
part of the enzyme and the receptor is the original substrate or a protein that
carries the original substrate as a secondary modification. It will be interesting
to see whether this is the case for the C-signal receptor (cf. below).
C-SIGNAL TRANSMISSION RELIES ON A CONTACT-DEPENDENT
MECHANISM
Early on, Kaiser and co-workers observed a conspicuous relationship
between cell motility and C-signal transmission. First, it was found that
non-motile cells have an altered pattern of developmental gene expres-
sion (39). This altered pattern matches exactly the pattern observed in
csgA mutants. The similarity in the pattern of developmental expression
suggested that motility is required for C-signal transmission. Secondly,
it was found that both the donor cell of the C-signal as well as the
receiver cell of the C-signal need to be motile in order for C-signal
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
280
transmission to occur (34). The interpretation of these experiments was
that specific cellcell contacts are needed for C-signal transmission to
occur and that motility is required to establish these contacts.
This hypothesis was tested in a simple and elegant experimental setup
developed by Kim and Kaiser (33). Briefly, an agar plate containing
starvation medium was scored with a finely grained sandpaper to
create grooves, which were approximately 0.5 to 1 cell length in width.
Subsequently, non-motile cells were positioned in the grooves. In
the grooves, the rod-shaped cells generally aligned with extensive
side-to-side contacts and end-to-end contacts. Amazingly, this simple
manipulation of cell position restored C-signal-dependent gene expression
and sporulation in the non-motile mutants. Interestingly, cells outside the
grooves also aligned in a pattern that established extensive side-to-side
contacts. However, these cells neither expressed C-signal dependent genes
nor did they sporulate. The defining difference between cells in the grooves
and outside the grooves is the frequency of end-to-end contacts (33). Thus,
Kim and Kaiser suggested that C-signal transmission depends on direct
contacts between adjacent cells and that these contacts were end-to-end
contacts (33). Cell motility would be required to establish these contacts (33).
Later, the observation that wild-type cells and csgA cells need to be
in direct contact in order for extracellular complementation to occur has
offered further support to the contact-dependent C-signal transmission
mechanism (36). Finally, the findings that CsgA antibodies recognize
epitopes, located on the surface of developing wild-type cells (74) and
that the 17 kDa C-signal protein is localized to the outer membrane
showed that the C-signal is non-diffusible, and thus further supported
the idea that C-signal transmission would rely on a contact-dependent
mechanism (54). Definitive evidence that C-signal transmission depends
on specific end-to-end contacts is still lacking. However, analyses of cell
behaviour during rippling, which is a specialized type of organized
cell movement that is occasionally observed during the early stages
of aggregation, lend support to the notion that C-signal transmission
involves end-to-end contacts (70).
THE C-SIGNAL TRANSDUCTION PATHWAY
In order to understand how a single signal can induce three responses,
i.e. aggregation, sporulation, and developmental gene expression, which
are separated in time and space, it is essential to elucidate the structure of
the signal transduction pathway. Several of the components in the C-signal
281
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
transduction have been identified (Figure 11.4). Most of these components
have been identified by using genetic approaches in which developmental
mutants were specifically screened for those deficient in C-signal-
dependent responses (52, 80).
The receptor of the C-signal has yet to be identified. The first
recognized component in the C-signal transduction pathway is the DNA
binding response regulator protein FruA, which consists of an N-terminal
receiver domain and a C-terminal DNA binding domain (15). fruA mutants
are unable to aggregate and sporulate and are deficient in the expression of
C-signal-dependent genes (15, 26, 62, 78). FruA activity is regulated at the
level of transcription as well as at a post-translational level (Figure 11.4).
The regulatory mechanism acting at the level of fruA transcription ensures
the timely induction of fruA transcription after 36h of starvation. Several
pathways converge to stimulate fruA transcription: the early-acting A-signal
induces fruA transcription by an unknown mechanism (15, 62). The MrpC
protein, which is a homolog of the cyclic AMP receptor protein in E. coli
(84), binds to the fruA promoter and induces fruA transcription. Finally, the
DevT protein directly or indirectly stimulates transcription of fruA (7).
The regulatory mechanism acting at the post-translational level results in
the activation of FruA. Genetic evidence suggests that this activation
involves the phosphorylation of a conserved Asp residue in the N-terminal
receiver domain (15). In other response regulator proteins, the corresponding
Asp residue is phosphorylated by a cognate histidine protein kinase (64).
Moreover, genetic evidence suggests that FruA phosphorylation is induced
by the C-signal (15, 78). The cognate FruA histidine protein kinase has yet to
be identified.
Downstream from phosphorylated FruA the C-signal transduction
pathway branches (Figure 11.4). One branch leads to aggregation; the
proteins in the Frz chemosensory system act in this branch (78, 80).
The Frz proteins constitute a cytoplasmic signal transduction system that
controls several gliding-motility parameters including the frequency of
gliding reversals (5, 28). The Frz proteins are similar to proteins involved
in chemotaxis responses in other bacteria (88). In this branch, the C-signal
is an input signal to the Frz system and induces methylation of the FrzCD
protein (78), a methyl-accepting chemotaxis protein (57). C-signal-induced
methylation of FrzCD occurs via phosphorylation of FruA and depends on
the FrzF methyltransferase (15, 78) (Figure 11.5). Presumably C-signaling
alters the activity of the Frz system in such a way that aggregation is
induced. Genetic evidence suggests that C-signaling inhibits the activity
of the Frz system (77). This in turn, alleviates the Frz-dependent inhibition
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
282
S
t
a
r
v
a
t
i
o
n
1
F
r
u
A
f
r
ff
u
rr
A
uu
d
o
f
A
R
e
l
A
(
p
)
p
p
G
p
p
D
e
v
T
M
r
p
C
A
-
s
i
g
n
a
l
s
d
e
K
t
o
d
K
F
r
u
A
-
P
A
g
g
r
e
g
a
t
i
oo
n S
p
o
r
u
l
a
t
i
o
n
p
2
5
p
1
7
1
S
d
e
K
T
o
d
K
c
s
g
A
gg
a
c
t
A
tt
B
C
D
H
P
K
2
F
r
z
d
e
dd
v
T
R
S
7
5
3
6
c
s
g
A
gg
F
i
g
u
r
e
1
1
.
4
.
M
u
l
t
i
p
l
e
s
i
g
n
a
l
t
r
a
n
s
d
u
c
t
i
o
n
p
a
t
h
w
a
y
s
c
o
n
v
e
r
g
e
t
o
c
o
n
t
r
o
l
m
o
r
p
h
o
g
e
n
e
s
i
s
.
T
h
e
s
c
h
e
m
a
t
i
c
i
l
l
u
s
t
r
a
t
e
s
C
-
s
i
g
n
a
l
t
r
a
n
s
m
i
s
s
i
o
n
b
e
t
w
e
e
n
t
w
o
c
e
l
l
s
.
T
h
e
d
a
s
h
e
d
a
n
d
c
o
n
t
i
n
u
o
u
s
l
i
n
e
s
i
n
d
i
c
a
t
e
t
h
e
o
u
t
e
r
a
n
d
i
n
n
e
r
m
e
m
b
r
a
n
e
,
r
e
s
p
e
c
t
i
v
e
l
y
.
F
o
r
s
i
m
p
l
i
c
i
t
y
,
t
h
e
c
o
m
p
o
n
e
n
t
s
i
n
t
h
e
p
a
t
h
w
a
y
a
r
e
o
n
l
y
s
h
o
w
n
i
n
t
h
e
c
e
l
l
t
o
t
h
e
r
i
g
h
t
.
T
h
e
s
e
c
o
m
p
o
n
e
n
t
s
a
r
e
a
l
s
o
p
r
e
s
e
n
t
i
n
t
h
e
c
e
l
l
t
o
t
h
e
l
e
f
t
.
I
n
t
h
i
s
c
e
l
l
,
o
n
l
y
c
s
g
A
i
s
s
h
o
w
n
t
o
i
l
l
u
s
t
r
a
t
e
t
h
e
s
i
g
n
a
l
a
m
p
l
i
f
i
c
a
t
i
o
n
l
o
o
p
l
a
b
e
l
l
e
d
.
T
h
e
s
e
c
o
n
d
s
i
g
n
a
l
a
m
p
l
i
f
i
c
a
t
i
o
n
l
o
o
p
i
s
l
a
b
e
l
l
e
d
.
T
h
e
s
y
m
b
o
l
i
n
d
i
c
a
t
e
s
t
h
e
h
y
p
o
t
h
e
t
i
c
a
l
C
-
s
i
g
n
a
l
r
e
c
e
p
t
o
r
;
H
P
K
i
n
d
i
c
a
t
e
s
t
h
e
h
y
p
o
t
h
e
t
i
c
a
l
F
r
u
A
h
i
s
t
i
d
i
n
e
p
r
o
t
e
i
n
k
i
n
a
s
e
;
a
n
d
7
5
3
6
s
y
m
b
o
l
i
z
e
s
t
h
e
s
p
o
r
u
l
a
t
i
o
n
g
e
n
e
m
o
n
i
t
o
r
e
d
b
y
t
h
e
T
n
5
l
a
c
7
5
3
6
i
n
s
e
r
t
i
o
n
.
T
h
e
p
o
s
s
i
b
l
e
c
o
n
v
e
r
g
e
n
c
e
p
o
i
n
t
o
f
t
h
e
s
i
g
n
a
l
t
r
a
n
s
d
u
c
t
i
o
n
p
a
t
h
w
a
y
s
d
e
f
i
n
e
d
b
y
t
h
e
S
d
e
K
a
n
d
T
o
d
K
k
i
n
a
s
e
s
w
i
t
h
t
h
e
C
-
s
i
g
n
a
l
t
r
a
n
s
d
u
c
t
i
o
n
p
a
t
h
w
a
y
i
s
i
n
d
i
c
a
t
e
d
b
y
t
h
e
s
h
a
d
e
d
b
o
x
.
S
e
e
t
h
e
t
e
x
t
f
o
r
a
d
e
t
a
i
l
e
d
e
x
p
l
a
n
a
t
i
o
n
o
f
t
h
e
p
a
t
h
w
a
y
s
.
of the MglA protein, a member of the RasRabRho superfamily of small
eukaryotic GTPases (21), which controls polarity switching of the A- and
S-engines (83). Consequently, MglA inhibits polarity switching of Tfp and
the active nozzle-like structures (Figure 11.5). The connection between the
C-signal and the Frz system, its interaction with the gliding machinery, and
the connection to cell behavior is discussed further below.
The second branch downstream from phosphorylated FruA leads to
C-signal-dependent gene expression and sporulation (15, 26) (Figure 11.4).
The C-signal and FruA jointly regulate the expression of at least 50
genes (26) including the devTRS operon (15). The DevTRS proteins, in
turn, are required for the expression of the sporulation gene tagged by
the Tn5lac 7536 insertion (52). Moreover, DevT is required for full expres-
sion of the fruA gene (7). In addition to regulating the expression of
C-signal-dependent genes, FruA regulates the expression of at least eight
genes, including the dofA gene, in a C-signal independent manner (26)
(Figure 11.4). Presumably these genes are induced by unphosphorylated
FruA.
A third branch in the C-signal transduction pathway is located
upstream from FruA and leads to increased transcription of the csgA gene
(35) (Figure 11.4). csgA is transcribed in vegetative cells and transcription
increases approximately four-fold during development (50). This increase
in csgA transcription is paralleled in the accumulation of p25, which also
FrzCD
FrzE
MglA
FrzF
FrzG
C-signal
Direction of gliding
FrzCD-CH
3
FruA FruA-P
Figure 11.5. The C-signal inhibits polarity switching of the gliding engines in Myxococcus
xanthus. Longitudinal cross-section of a M. xanthus cell on a solid surface, showing the
arrangement and location of the two engines involved in gliding motility. The interactions
that result in C-signal-dependent inhibition of polarity switching of the two gliding engines
during fruiting-body formation are indicated in black. C-signaling results in increased
methylation of the FrzCDprotein. Genetic evidence suggests that this leads to an inhibition
of the FrzE histidine protein kinase. This in turn results in an alleviation of the Frz-
dependent inhibition of the MglA protein. Consequently, MglA inhibits polarity switching
of the two gliding engines. The figure was adapted from (78).
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
284
increases approximately four-fold in response to starvation (42). The
increase in csgA transcription in response to starvation involves RelA and
the stringent response (9) and the four proteins encoded by the act operon
(17). Specifically, ActA and ActB, both of which are response regulators, are
required for full transcription of csgA during development whereas ActC
and ActD are important for the correct timing of csgA transcription. It
has been suggested that the Act proteins are specifically involved in
the C-signal-dependent increase in csgA transcription (17). However, an
interaction between C-signal transmission and act transcription or the
Act protein has yet to be reported. Following csgA transcription, p25 accu-
mulates, is exported, and is then processed to produce the 17 kDa C-signal
protein. The p17 protein is exposed on the cell surface and may interact with
a C-signal receptor on a different cell.
The C-signal transduction pathway is only active during starvation.
Starvation sets the pathway in motion by inducing the stringent response
(75). The stringent response, in turn, induces csgA transcription (9) and
A-signal accumulation (20) (Figure 11.4).
The C-signal transduction pathway contains two signal-amplification
loops, which ensure that cells engaged in C-signaling are exposed
to increased levels of C-signaling during development. In the first loop,
C-signaling induces aggregation and thus the accumulation of cells at a
higher cell density. As C-signal transmission involves a contact-dependent
mechanism, the prediction is that during the aggregation process the
level of C-signaling to which an aggregating cell is exposed increases
(Figure 11.4, loop 1). In the second amplification loop, C-signaling results
in increased csgA transcription; this in turn, results in p25 accumulation,
which is subsequently processed to p17, whichmay thenengage inC-signaling
with a neighboring cell (Figure 11.4, loop 2).
HOW DOES A SINGLE SIGNAL INDUCE THREE RESPONSES
SEPARATED IN TIME AND SPACE?
A hallmark in fruiting-body formation is the temporal and spatial
coordination of aggregation and sporulation: aggregation precedes sporu-
lation and cells do not initiate sporulation until the aggregation process is
complete and cells have accumulated at a high cell density inside the fruiting
bodies. With the C-signal inducing aggregation as well as sporulation, the
question becomes: how is the C-signal transduction pathway structured
to ensure the spatial and temporal coordination of aggregation and
sporulation?
285
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
Three independent lines of evidence converge to suggest that
the three C-signal-dependent responses are induced at different thresholds
of C-signaling. Kim and Kaiser added different amounts of exogenous
C-signal to starving csgA cells and observed that aggregation and early
C-signal-dependent gene expression were induced at an intermediate
level of C-signaling whereas sporulation and late C-signal-dependent
gene expression were induced at a higher level (35). By manipulating
expression of the csgA gene in vivo, Li et al. (50) made similar observations.
Finally, by systematically varying the accumulation of the C-signal in vivo,
Kruse et al. (42) observed that increased synthesis of the C-signal early
during development resulted in early aggregation, sporulation and C-signal-
dependent gene expression. Vice versa, decreased accumulation of the C-signal
resulted in delayed aggregation and sporulation. Importantly, it was
also observed that overexpression of the C-signal results in uncoupling of
aggregation and sporulation, with spores being formed outside fruiting
bodies. Thus, the regulated increase in C-signaling levels is a key parameter
in the temporal and spatial coordination of aggregation and sporulation.
The current view of how the C-signal ensures the temporal coordination
of aggregation and sporulation is as follows (79). The C-signal thresholds in
combination with the ordered increase in the level of C-signaling ensure that
the C-signal first induces aggregation and early gene expression and then late
gene expression and sporulation. According to this model, the C-signal is a
timer of developmental events. Moreover, the C-signal is a non-diffusible
morphogen as it induces distinct morphogenetic events at distinct
thresholds.
The spatial coordination of aggregation, sporulation and gene expression
is likely to be a direct consequence of the contact-dependent C-signal transmis-
sion mechanism (79). This signal transmission mechanism ensures that the
level of C-signaling to which a cell is exposed reflects cell density and thus the
position of a cell. The high level of C-signaling that induces late gene expres-
sion and sporulation is only obtained in cells that are closely packed inside the
nascent fruiting bodies. Consequently, only cells that have accumulated inside
the nascent fruiting bodies express late genes and undergo sporulation. The
observation that overexpression of csgA results in uncoupling of aggregation
and sporulation suggests that in wild-type cells the balance between the two
C-signal amplificationloops is carefully regulated to ensure that the sporulation
threshold is explicitly obtained by cells inside the nascent fruiting bodies.
Thus, the mechanism of the C-signal essentially allows cells to decode their
position with respect to that of other cells and in that way match gene expres-
sion, and ultimately sporulation, to their position.
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
286
This model also provides an explanation for the spatial control of
C-signal-dependent gene expression. C-signal-dependent genes are prefer-
entially expressed in aggregating and sporulating cells whereas they are not
expressed in the peripheral rods (30). Peripheral rods are present at a lowcell
density outside the nascent fruiting bodies. Consequently, they will only
infrequently engage in direct contacts with other cells with C-signal trans-
mission. Therefore, they experience a low level of C-signaling that allows
neither C-signal-dependent gene expression nor sporulation.
MULTIPLE SIGNAL TRANSDUCTION PATHWAYS CONTROL
MORPHOGENESIS
Two additional signal transduction pathways converge with the C-signal
transduction pathway to regulate aggregation, sporulation, and gene
expression after 6h (Figure 11.4). Each of the pathways is defined by
a specific histidine protein kinase. The SdeK histidine protein kinase is
synthesized in a RelA-dependent manner immediately after the initiation
of starvation; the SdeK pathway converges with the C-signal transduction
pathway to stimulate aggregation, sporulation, and gene expression (16, 66).
The pathway defined by the TodK histidine protein kinase inhibits aggrega-
tion, sporulation, and gene expression (67). Synthesis of TodKis inhibited by
starvation in a RelA-independent manner (67). Genetic evidence suggests
that the TodK-dependent inhibition of aggregation, sporulation and gene
expression is alleviated after 69h of starvation (67). Interestingly, SdeK and
TodK are both predicted to be located in the cytoplasm and they both contain
PAS domains in their sensor part. PAS domains have been implicated in
sensing changes in redox potential, oxygen, light, small ligands, and overall
energy level of a cell and also mediate proteinprotein interactions (86).
Therefore, it has been proposed that the kinase activities of TodK and SdeK
are controlled by intracellular signals, which are informative about the
metabolic state of individual cells and which are indicative of continued
starvation (67). According to this model, continued starvation of cells results
in the accumulation of these signals. This would subsequently trigger an
alteration in the activity of the TodK and SdeK kinases, resulting in the
alleviation of the inhibitory effect of TodK and stimulating the activating
effect of SdeK on the C-signal transduction pathway. In this model, the
C-signal transduction pathway is anintegrationpoint at whichthe intercellular
signals needed to coordinate the efforts of thousands of cells during fruiting-
body formation and intracellular signals reflecting the energy status of
individual cells are integrated. The advantage of this kind of integration
287
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
would be that productive C-signaling, and thus morphogenesis, is strictly
coordinated with the energy status of individual cells and only occurs when
the energy status of individual cells is appropriate. The SdeK and TodK
pathways may also help to make sure that, if nutrients become available
during fruiting body formation, then fruiting-body formation is arrested and
vegetative growth resumes.
THE C-SIGNAL-DEPENDENT MOTILITY RESPONSE
As fruiting-body formation depends on changes in cell behavior from
spreading to aggregation, a complete understanding of the morphogenetic
properties of the C-signal entails a description of how this signal molecule
alters cell behavior. Using fluorescent time-lapse video microscopy, Jelsbak
and Sgaard-Andersen (29) found that, during the aggregation stage of
fruiting body formation, the C-signal induces a motility response that
includes increases in the gliding speed and in the duration of gliding
intervals and decreases in the stop and reversal frequencies. The combined
effect of these changes is a switch in the motility behaviour of individual
cells from an oscillatory to a unidirectional type of behaviour (Figure 11.6).
In order for M. xanthus cells to display directional movement such
as that seen during aggregation, they need to be able to regulate their
frequency of direction changing. A priori, a stop as well as a reversal
could result in a change in the direction of movement. However, as
M. xanthus cells adhere to each other and to the substratum, they are
not buffeted by Brownian motion. Consequently, among the motility
parameters that characterize gliding in M. xanthus, only a reversal results
in a change in the direction of movement. Therefore, among the motility
parameters regulated by the C-signal, the key parameter is the reversal
frequency (29). The remaining motility parameters regulated by the C-signal
may contribute to aggregation by increasing the net distance travelled by a
cell per minute (29).
The C-signal dependent decrease in the reversal frequency fits with the
genetic and biochemical evidence demonstrating that the C-signal is an
input signal to the Frz chemosensory system (78) (Figures 11.5, 11.6).
Consistent with the observation that the C-signal induces a decrease in
the reversal frequency, methylation of FrzCD correlates with a low reversal
frequency (56).
The C-signal-dependent decrease in the reversal frequency suggests
that the C-signal inhibits polarity switching of the Tfp and nozzle-like
structures involved in S- and A-motility, respectively (77). The net
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
288
result of this inhibition is that the two motility engines are locked at
their respective poles, and consequently C-signaling cells are locked in a
unidirectional mode of behavior.
C-SIGNAL-INDUCED AGGREGATION: A MODEL
The ability of C-signaling cells to move a long distance is clearly
beneficial during the aggregation stage of fruiting-body formation.
However, in order for cells to aggregate they need to move with a sense
of direction. The identification of the motility parameters controlled by the
C-signal in combination with the contact-dependent C-signal transmission
mechanism has allowed the generation of a model for C-signal-induced
aggregation (Figure 11.7). The model is composed of three discrete events.
The basic event is an end-to-end contact between two cells with C-signal
transmission. Cells engaged in end-to-end contact with C-signal transmis-
sion gain the ability to travel over longer net distances for as long as they
0 10 20 30 40 50
0
0 10 20 30 40 50
10
20
30
40
50
wild type
wild type
csgA
csgA
0
10
20
30
40
50
wild type
wild type
csgA
csgA
wild type
csgA
y

p
o
s
i
t
i
o
n

(
m
)
y

p
o
s
i
t
i
o
n

(
m
)
3 h 15 h
x position (m)
x position (m)
Figure 11.6. Effect of the C-signal on the motility behaviour of M. xanthus cells. By means of
fluorescent time-lapse video-microscopy the x, y positions of wild-type cells, which express
GFP and which were co-developed with wild-type cells that do not express GFP, and of csgA
cells, which express GFP and which were co-developed with csgA cells that do not express
GFP, were recorded every 15 s for a period of 900s. The trajectories in the left panel show
the behaviour of two wild-type cells and of two csgA cells starved for 3 h (i.e. 3 h prior to the
initiation of C-signaling); the trajectories in the right panel show the behavior of three wild-
type cells and of three csgA cells starved for 15 h (i.e. 9h after the initiation of C-signaling).
Before the initiation of C-signaling, wild-type and csgA cells behave similarly. After the
initiation of C-signaling in wild-type cells, these cells display a unidirectional type of
behavior. The figure was adapted from (30).
289
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
End-to-end contact
Chain formation
Stream formation
Change in behaviour
induced by C-signal
transmission
Figure 11.7. Amodel for C-signal-induced aggregation.The model consists of three elements.
The basic element is an end-to-end contact between two cells with C-signal transmission
followed by a change in cell behavior. The second element is chain formation, which is
predicted to be a consequence of repeated end-to-end contacts with C-signal transmission
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
290
maintain the end-to-end contact. In a population of starving cells, this basic
event is predicted to result in the second discrete event, chain formation
(Figure 11.7). End-to-end contacts will occur frequently at the high cell
density required for normal development. This will sequentially result in
the recruitment of cells to pairs of cells that are C-signaling and, thus, in the
formation of chains of cells. Cells in a chain are moving at the same speed
and in the same direction, which is determined by the direction of move-
ment of the cell at the leading end of the chain. The information about
which direction to move in is relayed from the leading cell in a chain to the
following cell by the direct cellcell contacts. A reversal of direction of a cell
positioned inside a chain would break the chain. However, cell reversals in
chains are unlikely based on two observations. First, the cohesiveness of
the cells increases during development (72). Second, cells are exposed to
increasing levels of C-signaling during aggregation (see above).
The arrangement and movement of cells in chains is predicted to result
in the third discrete event, streamformation (Figure 11.7). Movement of cell
in a chain may create alignment of neighboring cells. This will result in the
formation of secondary chains that are associated with the initiating chain
by lateral interactions. This could lead to the formation of the streams of
cells that have been observed experimentally (Figure 11.2). Aggregation
Figure 11.7. (cont.)
between cells in a field. Consider cell A in scenario 1, which for simplicity is located at the
edge of an aggregate and moving towards the aggregate. This cell is moving at a high speed
in long gliding intervals and with low stop and reversal frequencies, as it is engaged in
C-signaling with cells inside the aggregate. The bold, single-headed arrow indicates this
type of behavior. Cells not engaged in C-signaling move at low speeds and with high stop
and reversal frequencies, as marked by the double-headed arrows. If cell B establishes an
end-to-end contact with cell A (scenario 2), then the two cells will transmit the C-signal to
each other. As a consequence, cell B is induced to move with a high speed and low stop and
reversal frequencies as long as it maintains the end-to-end contact. As cell A is moving
towards the aggregation center, cell B can only move with a high speed and low stop and
reversal frequencies if it also moves towards the aggregation center. This sequence of
events could be repeated sequentially, adding cell C, etc., thus creating a chain in which cells
are moving in the direction determined by the leading cell A (scenarios 34). The third
element in the model is stream formation. Cell movement in chains is predicted to create
alignment of neighboring cells. This will result in the formation of secondary chains of cells
(marked by dark gray colour). The cells in these secondary chains are associated with the
initiating chain by lateral cellcell contacts and with other cells in the secondary chain by
end-to-end contacts. Together, an initiating chain and its associated secondary chains will
make up a stream. The figure was adapted from (28).
291
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
centers could form by collisions of streams or by a single stream turning on
itself in a spiral movement. Once the leading cell of a stream is trapped
inside an aggregation center, the remaining cells in the stream will follow
into that center, as the direction of movement of a cell in a stream is deter-
mined by the leading cell. In this model C-signal transmission is a local event
between two cell ends, which occurs without reference to the global pattern,
and the result is a global organization of cells. Therefore, according to this
model C-signal-induced aggregation is a self-organizing process.
SIGNAL INTEGRATION DURING FRUITING-BODY
MORPHOGENESIS
Formation of the spore-filled fruiting bodies is clearly an effective survival
strategy in response to starvation. However, it is also costly, as only 10%20%
of cells differentiate into spores. The starvation-induced stringent response and
the intercellular A-signal systemconstitute two checkpoints, which ensure that
cells only embark on fruiting-body formation when starvation is severe and
gives the impression of being long-lasting. The C-signal comes into action
when the cells have passed the tests of the stringent response and the A-signal.
The primary functions of the C-signal are to induce and coordinate aggregation,
sporulation and full gene expression after 6h. Thus, the C-signaling system
cannot be regarded as a system whose primary function is to monitor starva-
tion. Rather the C-signal is the temporal and spatial coordinator of the different
morphogenetic events involved in fruiting-body formation. Regarded in this
way, the stringent response and the A- and C-signaling systems essentially act
to stimulate fruiting-body formation. Predictably, it is becoming evident that
fruiting-body formation is also subject to intensive negative control.
The negative regulators of fruiting-body formation include two cyto-
plasmic histidine protein kinases (EspA and TodK) (8, 67), a member of
the NtrC family of two-component transcriptional regulators (SpdR) (19),
the Che3 chemosensory system (38), two serine/threonine protein kinases
(Pkn1 and Pkn2) (59, 87), a protein that is homologous to flavin-containing
monooxygenases (BcsA) (11), and a protein of unknown biochemical function
that modulates the stringent response (SocE) (10). Although the presence of a
PAS domain in the sensor part of the TodK kinase hints that it could be
involved inmonitoring the metabolic state of cells (67), the specific parameters
monitored by these systems are still unknown. However, the multitude of
negative regulators of fruiting-body formation indicate that cells are continu-
ously monitoring the conditions to evaluate whether or not fruiting-body
formation should continue.
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
292
From a design point of view, a regulatory circuit that includes the
integration of positively as well as negatively acting pathways is highly
robust. It endows the system with the capacity to tailor the final decision,
i.e. to continue or not to continue fruiting-body formation, to the specific
conditions to which the cells are exposed.
CONCLUDING REMARKS
The C-signal is a remarkably versatile signal and we nowhave a general
framework for understanding how a single signal induces three responses
that are separated in time and space. However, many questions are still
unanswered. So far, the identification of components important for devel-
opment has to a large extent depended on the often painstaking identifica-
tion of relevant genes and proteins on a step-by-step basis using genetic
approaches. Recently, the M. xanthus genome sequence was finished
(W. C. Niermann, personal communication). This together with proteome
analysis (26) and the construction of an M. xanthus DNA microarray
(H. Kaplan and D. Kaiser, personal communication) promise to pave the
way for the systematic identification of additional important players in the
regulatory pathways that drive and coordinate fruiting-body formation.
ACKNOWLEDGEMENTS
I thank Jimmy Jakobsen for carefully reading the manuscript, and Martin
Overgaard, Anders Aa. Rasmussen, Sune Lobedanz, Eva Ellehauge and Lars
Jelsbak for many helpful discussions.
REFERENCES
1 Arnold, J. W. and L. J. Shimkets 1988. Cell surface properties correlated with
cohesion in Myxococcus xanthus. J. Bacteriol. 170: 57717.
2 Baker, M. E. 1994. Myxococcus xanthus C-factor, a morphogenetic paracrine
signal, is similar to Escherichia coli 3-oxoacyl-[acyl-carrier-protein] reductase and
human 17 beta-hydroxysteroid dehydrogenase. Biochem. J. 301: 31112.
3 Behmlander, R. M. and M. Dworkin 1994. Biochemical and structural analyses of
the extracellular matrix fibrils of Myxococcus xanthus. J. Bacteriol. 176: 6295303.
4 Behmlander, R. M. and M. Dworkin 1991. Extracellular fibrils and contact-
mediated cell interactions in Myxococcus xanthus. J. Bacteriol. 173: 781021.
5 Blackhart, B. D. and D. R. Zusman 1985. Frizzy genes of Myxococcus xanthus are
involved in control of frequency of reversal of gliding motility Proc. Natn. Acad.
Sci. USA 82: 87714.
293
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
6 Bowden, M. G. and H. B. Kaplan 1998. The Myxococcus xanthus lipopolysac-
charide O-antigen is required for social motility and multicellular development.
7 Boysen, A., E. Ellehauge, B. Julien and L. Sgaard-Andersen 2002. The DevT
protein stimulates synthesis of FruA, a signal transduction protein required
for fruiting body morphogenesis in Myxococcus xanthus. J. Bacteriol. 184:
15406.
8 Cho, K. and D. R. Zusman 1999. Sporulation timing in Myxococcus xanthus is
controlled by the espAB locus. Molec. Microbiol. 34: 71425.
9 Crawford, E. W. and L. J. Shimkets 2000. The Myxococcus xanthus socE and csgA
genes are regulated by the stringent response. Molec. Microbiol. 37: 78899.
10 Crawford, E. W. and L. J. Shimkets 2000. The stringent response in Myxococcus
xanthus is regulated by SocE and the CgsA C-signaling protein. Genes Dev.
14: 48392.
11 Cusick, J. K., E. Hager and R. E. Gill 2002. Characterization of bcsA mutations
that bypass two distinct signaling requirements for Myxococcus xanthus develop-
ment. J. Bacteriol. 184: 514150.
12 Downard, J., S. V. Ramaswamy and K. S. Kil 1993. Identification of esg, a genetic
locus involved in cell-cell signaling during Myxococcus xanthus development.
J. Bacteriol. 175: 776270.
13 Dworkin, M. 1996. Recent advances in the social and developmental biology of
the Myxobacteria. Microbiol. Rev. 60: 70102.
14 Dworkin, M. and S. M. Gibson 1964. A system for studying microbial
morphogenesis: rapid formation of microcysts in Myxococcus xanthus.
Science 146: 2434.
15 Ellehauge, E., M. Nrregaard-Madsen and L. Sgaard-Andersen 1998. The FruA
signal transduction protein provides a checkpoint for the temporal co-ordination
of intercellular signals in M. xanthus development. Molec. Microbiol. 30: 80717.
16 Garza, A. G., J. S. Pollack, B. Z. Harris et al. 1998. SdeK is required for early
fruiting body development in Myxococcus xanthus. J. Bacteriol. 180: 462837.
17 Gronewold, T. M. A. and D. Kaiser 2001. The act operon controls the level and
time of C-signal production for Myxococcus xanthus development. Molec.
18 Hagen, D. C., A. P. Bretscher and D. Kaiser 1978. Synergism between morpho-
genetic mutants of Myxococcus xanthus. Dev. Biol. 64: 28496.
19 Hager, E., H. Tse and R. E. Gill 2001. Identification and characterization of spdR
mutations that bypass the BsgA protease-dependent regulation of developmental
gene expression in Myxococcus xanthus. Molec. Microbiol. 39: 76580.
20 Harris, B. Z., D. Kaiser and M. Singer 1998. The guanosine nucleotide (p)ppGpp
initiates development and A-factor production in Myxococcus xanthus. Genes Dev.
12: 102235.
21 Hartzell, P. L. 1997. Complementation of sporulation and motility defects in a
prokaryote by a eukaryotic GTPase. Proc. Natn. Acad. Sci. USA 94: 98816.
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
294
22 Henrichsen, J. 1972. Bacterial surface translocation: a survey and a classification.
Bacteriol. Rev. 36: 478503.
23 Hodgkin, J. and D. Kaiser 1979. Genetics of gliding motility in Myxococcus
xanthus (Myxobacterales): genes controlling movement of single cells. Molec.
Gen. Genet. 171: 16776.
24 Hodgkin, J. and D. Kaiser 1979. Genetics of gliding motility in Myxococcus
xanthus (Myxobacteriales): two gene systems control movement. Molec. Gen.
Genet. 171: 17791.
25 Hoiczyk, E. and W. Baumeister 1998. The junctional pore complex, a prokaryotic
secretion organelle, is the molecular motor underlying gliding motility in
cyanobacteria. Curr. Biol. 8: 11618.
26 Horiuchi, T., M. Taoka, T. Isobe, T. Komano and S. Inouye 2002. Role of
fruA and csgA genes in gene expression during development of Myxococcus
xanthus. Analysis by two-dimensional gel electrophoresis. J. Biol. Chem. 277:
2675360.
27 Inouye, M., S. Inouye and D. R. Zusman 1979. Gene expression during develop-
ment of Myxococcus xanthus: pattern of protein synthesis. Dev. Biol. 68: 57991.
28 Jelsbak, L. and L. Sgaard-Andersen 1999. The cell surface-associated intercel-
lular C-signal induces behavioral changes in individual Myxococcus xanthus cells
during fruiting body morphogenesis. Proc. Natn. Acad. Sci. USA 96: 50316.
29 Jelsbak, L. and L. Sgaard-Andersen 2002. Pattern formation by a cell surface-
associated morphogen in Myxococcus xanthus. Proc. Natn. Acad. Sci. USA
99: 20327.
30 Julien, B., A. D. Kaiser and A. Garza 2000. Spatial control of cell differentiation in
Myxococcus xanthus. Proc. Natn. Acad. Sci. USA 97: 9098103.
31 Kaiser, D. 2003. Coupling cell movement to multicellular development in myxo-
bacteria. Nature Rev. Microbiol. 1: 4554.
32 Kaiser, D. 1979. Social gliding is correlated with the presence of pili in
Myxococcus xanthus. Proc. Natn. Acad. Sci. USA 76: 59526.
33 Kim, S. K. and D. Kaiser 1990. Cell alignment required in differentiation of
Myxococcus xanthus. Science 249: 9268.
34 Kim, S. K. and D. Kaiser 1990. Cell motility is required for the transmission of
C-factor, an intercellular signal that coordinates fruiting body morphogenesis of
Myxococcus xanthus. Genes Dev. 4: 896904.
35 Kim, S. K. and D. Kaiser 1991. C-factor has distinct aggregation and sporulation
thresholds during Myxococcus development. J. Bacteriol. 173: 17228.
36 Kim, S. K. and D. Kaiser 1990. C-factor: a cell-cell signaling protein required for
fruiting body morphogenesis of M. xanthus. Cell 61: 1926.
37 Kim, S. K. and D. Kaiser 1990. Purification and properties of Myxococcus xanthus
C-factor, an intercellular signaling protein. Proc. Natn. Acad. Sci. USA87: 36359.
38 Kirby, J. R. and D. R. Zusman 2003. Chemosensory regulation of developmental
gene expression in Myxococcus xanthus. Proc. Natn. Acad. Sci. USA 100:
200813.
295
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
39 Kroos, L., P. Hartzell, K. Stephens and D. Kaiser 1988. A link between cell
movement and gene expression argues that motility is required for cell-cell
signaling during fruiting body development. Genes Dev. 2: 167785.
40 Kroos, L. and D. Kaiser 1987. Expression of many developmentally regulated genes
in Myxococcus depends on a sequence of cell interactions. Genes Dev. 1: 84054.
41 Kroos, L., A. Kuspa and D. Kaiser 1986. A global analysis of developmentally
regulated genes in Myxococcus xanthus. Dev. Biol. 117: 25266.
42 Kruse, T., S. Lobedanz, N. M. S. Berthelsen and L. Sgaard-Andersen 2001.
C-signal: A cell surface-associated morphogen that induces and coordinates
multicellular fruiting body morphogenesis and sporulation in M. xanthus.
43 Kuner, J. M. and D. Kaiser 1982. Fruiting body morphogenesis in submerged
cultures of Myxococcus xanthus. J. Bacteriol. 151: 45861.
44 Kuspa, A., L. Kroos and D. Kaiser 1986. Intercellular signaling is required for
developmental gene expression in Myxococcus xanthus. Dev. Biol. 117: 26776.
45 Kuspa, A., L. Plamann and D. Kaiser 1992. Identification of heat-stable A-factor
from Myxococcus xanthus. J. Bacteriol. 174: 331926.
46 Kuspa, A., L. Plamann and D. Kaiser 1992. A-signaling and the cell density
requirement for Myxococcus xanthus development. J. Bacteriol. 174: 73609.
47 Lee, B.-U., K. Lee, J. Mendez and L. J. Shimkets 1995. A tactile sensory system of
Myxococcus xanthus involves an extracellular NAD(P)
-containing protein. Genes

Dev. 9: 296473.
48 Lee, K. and L. J. Shimkets 1994. Cloning and characterization of the socA locus
which restores development to Myxococcus xanthus C-signaling mutants.
49 Lee, K. and L. J. Shimkets 1996. Suppression of a signaling defect during
Myxococcus xanthus development. J. Bacteriol. 178: 97784.
50 Li, S., B.-U. Lee and L. J. Shimkets 1992. csgA expression entrains Myxococcus
xanthus development. Genes Dev. 6: 40110.
51 Li, Y., H. Sun, X. Ma et al. 2003. Extracellular polysaccharides mediate pilus
retraction during social motility of Myxococcus xanthus. Proc. Natn. Acad. Sci.
USA 100: 54438.
52 Licking, E., L. Gorski and D. Kaiser 2000. Acommon step for changing cell shape
in fruiting body and starvation-independent sporulation in Myxococcus xanthus.
53 Llamas, M. A., J. J. Rodriguez-Herva, R. E. W. Hancock et al. 2003. Role of
Pseudomonas putida tol-oprL gene products in uptake of solutes through the
cytoplasmic membrane. J. Bacteriol. 185: 470716.
54 Lobedanz, S. and L. Sgaard-Andersen 2003. Identification of the C-signal, a
contact-dependent morphogen coordinating multiple developmental responses
in Myxococcus xanthus. Genes Dev. 17: 215161.
55 Mattick, J. S. 2002. Type IV pili and twitching motility. A. Rev. Microbiol. 56:
289314.
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
296
56 McBride, M. J., T. Kohler and D. R. Zusman 1992. Methylation of FrzCD, a
methyl-accepting taxis protein of Myxococcus xanthus, is correlated with factors
affecting cell behaviour. J. Bacteriol. 174: 424657.
57 McBride, M. J., R. A. Weinberg and D. R. Zusman 1989. Frizzy aggregation
genes of the gliding bacterium Myxococcus xanthus show sequence similarities
to the chemotaxis genes of enteric bacteria. Proc. Natn. Acad. Sci. USA
86: 4248.
58 Merz, A. J., M. So and M. P. Sheetz 2000. Pilus retraction powers bacterial
twitching motility. Nature 407: 98102.
59 Munoz, D. J., S. Inouye and M. Inouye 1991. A gene encoding a protein serine/
threonine kinase is required for normal development of M. xanthus, a gram-
negative bacterium. Cell 67: 9951006.
60 OConnor, K. A. and D. R. Zusman 1991. Development in Myxococcus xanthus
involves differentiation into two cell types, peripheral rods and spores. J. Bacteriol.
173: 331833.
61 OConnor, K. A. and D. R. Zusman 1989. Patterns of cellular interactions during
fruiting-body formation in Myxococcus xanthus. J. Bacteriol. 171: 601324.
62 Ogawa, M., S. Fujitani, X. Mao, S. Inouye and T. Komano 1996. FruA, a putative
transcription factor essential for the development of Myxococcus xanthus. Molec.
63 Oppermann, U., C. Filling, M. Hult et al. 2003. Short-chain dehydrogenases/
reductases (SDR): the 2002 update. Chem. Biol. Interact. 1434: 24753.
64 Parkinson, J. S. 1993. Signal transduction schemes of bacteria. Cell 73: 85771.
65 Plamann, L., A. Kuspa and D. Kaiser 1992. Proteins that rescue A-signal-defective
mutants of Myxococcus xanthus. J. Bacteriol. 174: 331118.
66 Pollack, J. S. and M. Singer 2001. SdeK, a histidine kinase required for
67 Rasmussen, A. A. and L. Sgaard-Andersen 2003. TodK, a putative histidine
protein kinase, regulates timing of fruiting body morphogenesis in Myxococcus
xanthus. J. Bacteriol. 185: 545264.
68 Reichenbach, H. 1999. The ecology of the myxobacteria. Environ. Microbiol. 1: 1521.
69 Rosenbluh, A., R. Nir, E. Sahar and E. Rosenberg 1989. Cell-density-dependent lysis
and sporulation of Myxococcus xanthus in agarose beads. J. Bacteriol. 171: 49239.
70 Sager, B. and D. Kaiser 1994. Intercellular C-signaling and the traveling waves of
Myxococcus. Genes Dev. 8: 2793804.
71 Shimkets, L. J. 1999. Intercellular signaling during fruiting-body development of
Myxococcus xanthus. Au. Rev. Microbiol. 53: 52549.
72 Shimkets, L. J. 1986. Role of cell cohesion in Myxococcus xanthus fruiting body
73 Shimkets, L. J., R. E. Gill and D. Kaiser 1983. Developmental cell interactions in
Myxococcus xanthus and the spoC locus. Proc. Natn. Acad. Sci. USA 80: 140610.
74 Shimkets, L. J. and H. Rafiee 1990. CsgA, an extracellular protein essential for
297
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
75 Singer, M. and D. Kaiser 1995. Ectopic production of guanosine penta- and
tetraphosphate can initiate early developmental gene expression in Myxococcus
xanthus. Genes Dev. 9: 163344.
76 Skerker, J. M. and H. C. Berg 2001. Direct observation of extension and retraction
of type IV pili. Proc. Natn. Acad. Sci. USA 98: 69014.
77 Sgaard-Andersen, L. 2004. Cell polarity, intercellular signaling and mor-
phogenetic cell movements in Myxococcus xanthus. Curr. Opin. Microbiol.
7: 58793.
78 Sgaard-Andersen, L. and D. Kaiser 1996. C factor, a cell-surface-associated
intercellular signaling protein, stimulates the cytoplasmic Frz signal trans-
duction system in Myxococcus xanthus. Proc. Natn. Acad. Sci. USA 93:
26759.
79 Sgaard-Andersen, L., M. Overgaard, S. Lobedanz et al. 2003. Coupling gene
expression and multicellular morphogenesis during fruiting body formation in
Myxococcus xanthus. Molec. Microbiol. 48: 18.
80 Sgaard-Andersen, L., F. J. Slack, H. Kimsey and D. Kaiser 1996. Intercellular
C-signaling in Myxococcus xanthus involves a branched signal transduction
pathway. Genes Dev. 10: 74054.
81 Spormann, A. M. 1999. Gliding motility in bacteria: insights from studies of
Myxococcus xanthus. Microbiol. Molec. Biol. Rev. 63: 62141.
82 Spormann, A. M. and A. D. Kaiser 1995. Gliding movements in Myxococcus
83 Spormann, A. M. and D. Kaiser 1999. Gliding mutants of Myxococcus xanthus
with high reversal frequencies and small displacements. J. Bacteriol. 181:
2593601.
84 Sun, H. and W. Shi 2001. Analyses of mrp genes during Myxococcus xanthus
development. J. Bacteriol. 183: 67339.
85 Sun, H., D. R. Zusman and W. Shi 2000. Type IV pilus of Myxococcus xanthus is
a motility apparatus controlled by the frz chemosensory system. Curr. Biol. 10:
11436.
86 Taylor, B. L. and I. B. Zhulin 1999. PAS domains: internal sensors of oxygen,
redox potential, and light. Microbiol. Molec. Biol. Rev. 63: 479506.
87 Udo, H., M. Inouye and S. Inouye 1996. Effects of overexpression of Pkn2, a
transmembrane protein serine/threonine kinase, on development of Myxococcus
88 Ward, M. J. and D. R. Zusman 1999. Motility in Myxococcus xanthus and its role
in developmental aggregation. Curr. Opin. Microbiol. 2: 6249.
89 White, D. J. and P. L. Hartzell 2000. AglU, a protein required for gliding motility
and spore maturation of Myxococcus xanthus, is related to WD-repeat proteins.
90 Winans, S. C. and B. L. Bassler 2002. Mob psychology. J. Bacteriol. 184: 87383.
91 Wolgemuth, C., E. Hoiczyk, D. Kaiser and G. Oster 2002. How myxobacteria
glide. Curr. Biol. 12: 36977.
L
.
S
G
A
A
R
D
-
A
N
D
E
R
S
E
N
298
92 Wu, S. S. and D. Kaiser 1995. Genetic and functional evidence that Type IV pili
are required for social gliding motility in Myxococcus xanthus. Molec. Microbiol. 18:
54758.
93 Youderian, P., N. Burke, D. White and P. L. Hartzell 2003. Identification of genes
required for adventurous gliding motility in Myxococcus xanthus with the trans-
posable element mariner. Molec. Microbiol. 49: 55570.
299
C
O
N
T
A
C
T
-
D
E
P
E
N
D
E
N
T
S
I
G
N
A
L
I
N
G
I
N
M
.
X
A
N
T
H
U
S
Index
Note: page numbers in italics refer to figures and tables.
A to E signals 276
ABC transporters 238
Lsr 154
QsrAB 248
Streptococcus gordonii 252
accessory gene regulator see Agr
ActA response regulator 2845
ActB response regulator 2845
Actinobacillus actinomycetemcomitans
AI-2 183
iron acquisition regulation 184
Lsr transport system 189
receptor 18890
response 189
sensor kinase 1901
signal structure 190
signal transduction 1878
signal transduction cascade 188
Hfq protein 187
Lux S 182
luxS 131
inactivation 1367
Porphyromonas gingivalis recognition of
signal 192 3
RbsB 1889
response to AI-2 188
RTX leukotoxin 186
Actinomyces, lactate production in dental
biofilm 177
N-acylhomoserine lactones (AHLs) 104
absence from oral bacteria 181
bacterial crosstalk 135
cognate signals 74
LuxR interaction 11718
QS 135
biofilm growth 180
system dependency 67
virulence 180
receptors 43, 523
recycling 49
signals 72, 768
reception blockade 736
sulfonamide derivatives 75
synthetic analogs 736
categories 745 see also AHL synthase
S-adenosylhomocysteine (SAH) 1213, 1268
S-adenosylmethionine (SAM) 44, 121,
1523, 155
adhesins, surface-associated 2067
Aeromonas hydrophila, biofilmformation 70
Affymetrix Pseudomonas aeruginosa
GeneChip
1
824
aggregation 2857
aggregation centers 271, 2912
Agr
correspondence with disease types 210
evolution in Staphylococcus aureus 2224
group II AIP 224
model 201
polymorphism 21011
QS system 199203
specificity groups 20911, 2234
staphylococcal biofilms 21617, 21819
staphylococcal virulence factors 2067
Staphylococcus epidermidis 207
Staphylococcus saprophyticus 209
variants 2204
generation in Staphylococcus 223
negative 222, 224
virulence factor activation 207
virulence gene regulation 199
in vivo 21416
301
Agr (cont.)
virulence regulation 21112, 213
agr gene
expression 213, 219 20
locus 200
mutations 220, 221, 2224
AgrA 201
agrBCD nucleotides 20910
AgrC 201
agrD 20910
agrobacteria, pathogenic 40
Agrobacterium tumefaciens 3940
QS 42
reporter 76
Ti plasmids 403
conjugation 41 3
Agr-regulated genes 2037
AHL synthase 43
AI-1 11921
knockout of luxS effects on synthesis 155
Vibrio harveyi 1345
AI-2 1523
Actinobacillus actinomycetemcomitans 183
Lsr transport system 189
sensor kinase 1901
signal structure 190
bacteria with ambiguous roles 132 4
Campylobacter jejuni 133
Clostridium perfringens 131
crosstalk 1924
dental plaque 1357
Escherichia coli 130, 153
Helicobacter pylori 1323
induction by LuxS 119
LuxS role 126 8
metabolic pathway of production 1213
microbial intestinal flora production 155
mixed-species signaling in cystic
fibrosis 1378
Photorhabdus luminescens 133
Porphyromonas gingivalis 183
virulence 186
production 127, 139
Pseudomonas aeruginosa 138
QS
interspecies signal 1345
oral bacteria 1825
signal 12832 , 134 5
Vibrio 188
QS-mediated virulence repression 107
receptor in A. actinomycetemcomitans
18890
reporter of metabolic status 13841
Salmonella 153
Serratia marcescens 1301
signal specificity 192 4
signal transduction
in Actinobacillus
actinomycetemcomitans 1878
cascade 188
in Porphyromonas gingivalis 1878
signaling 1535
pathway 1289
Streptococcus gordonii secretion 187
structural forms 190
structures 123 6, 125 , 153
Vibrio harveyi 134 5
AI-3 130
bacterialhost communication 157, 158
knockout of luxS effects on synthesis
155
microbial intestinal flora production 155
QS signaling cascade 165
QseBC quorum sensing through 163
receptor recognition 164
signaling 1535
AI-3-dependent quorum sensing signaling
cascade 164
AiiA enzyme 74
AinS in Vibrio fischeri 129
AIP signal 2001
Allium sativum (garlic) 81
A-motility 273 6, 2889
anthranilate 28, 34
antibiotics
QS inhibitor combinations 8990
QS inhibitor synergy 90
resistance 65, 8990
antigen-presenting cells 204 5
antimicrobials
myxobacterial secretion 270
P. aeruginosa tolerance 84
Arabidopsis thaliana 910
arc operon 45
ArcB 131
arcB inactivation 191
ArcB sensor kinase 1901
ArlRS 213
asg mutants 276
A-signal 276, 277
attaching and effacing (AE) lesions 152,
157, 158
attKLM operon 489
attM gene 48
autoinducers 23, 139
antibodies 1415
immunomodulatory effects 1112
synthesis 8
Vibrio cholerae 112
Vibrio fischeri 117 see also AI-2
I
N
D
E
X
302
autoinducing peptide see AIP signal
autoinduction 43
Bacillus subtilis oligopeptide signaling 118
bacterial colonization inhibition 723
bacterialhost communication,
AI-3 157, 158
bacteriocins 238
BcsA 292
bfrAB genes 177
biofilms 65 6
antimicrobial tolerance 845
architecture 69
bacterial gene expression 70
CSP signaling 257
detachment 218
development 176
QSI effects 845
EPEC 167
eukaryotic defense strategy 713
filamentous 70
formation limitation 90
genetic competence 2558
growth
conditions 6970
mode 856
hydrogen peroxide treatment 85
infections 656, 89
maturation 176
microcolony formation 167
mixed 178
models 846
Myxococcus xanthus 270
natural blockers 713
QS
in formation 70
formation inhibition 107 11
QS-controlled genes 71
QSI drug effects 846
sloughing 84 5
staphylococcal 21617
Agr 21819
agr gene mutants 220
antibiotic exposure 220
behavior 21617
detachment 21820
development 21617, 219 20
Staphylococcus aureus 219
maturation 217 18
Streptococcus 1312
Vibrio cholerae 1078 , 110
formation regulation 108
host infection 109 see also dental
biofilm; Pseudomonas aeruginosa,
biofilms
biofouling, eukaryotic defense strategy 713
bioluminescence expression 76
Blp quorum-sensing system 2478
BlpABCSRH regulon 2478
BlpC 248
BlpR 248
cross-communication 248
BlpS 248
Borrelia burgdorferi, luxS system 1334
bsg mutants 276
B-signal 276
bundle-forming pili (BFP) 166 7
Burkholderia cepacia
biofilm formation 70
cystic fibrosis 137
N-butyryl homoserine lactone see C
4
-HSL
butyryl lactone 119
C
4
-HSL 23, 5, 89, 234
cystic fibrosis sputa 13
mammalian cell transcription regulator
activation 12
pqsABCDE induction inhibition 30
C
4
-HSL synthase 25
C-30
drug specificity 82
Pseudomonas aeruginosa QS
suppression 88
QS inhibitor pulmonary infection
treatment model 889
QS system inhibition 14, 845
C-56 845
Caenorhabditis elegans 910
CAI-1
signal pathways 112
signaling 108
Campylobacter jejuni 133
capsular polysaccharide 2056
cell density, quorum sensing 140
cell lysis, competence-induced 2445
cell-to-cell adherence, interbacterial in
dental biofilm 179
cell-to-cell signaling 117
bacterial host 156, 1579
co-ordinated behavior 141
dental biofilm 17594
EHEC 1557
LuxS-dependent 1934
Vibrio fischeri 1401 see also quorum
sensing
Che3 chemosensory system 292
CheY 191
cholera 102
pathogenesis 102
cholera toxin (CT) 102, 109
Chromobacterium violaceum 76
303
I
N
D
E
X
ciaR gene mutations 2467
CiaRH 2467
Clostridium perfringens, AI-2 131
ClpP 213
ClpX 213
coagulase 2067
pro duction by Stap hylococcus aureus 20 6 7
colony expansion 67
com regulon, Streptococcus pneumoniae
2545
ComA 238, 251
comAB gene 255
comAB locus 238
comAB operon 251
ComAB transporter 251
ComB 238, 251
com-box 255
comC gene 2378, 255
ComCDE competence regulon
Streptococcus pneumoniae 23642
signal 2467
comCDE operon 249
comCDE region 2545
ComD 2401
CSP interaction 255
receptor 2401
comD genes 240
comDE gene 255
ComDE signal transduction system 244
ComE
binding site 2412
cross-communication 248
response regulator 241 2
ComE phosphorylation (ComEP) 2412 ,
251 2, 255
ComE-specific phosphatase (CEP) 2456
ComI 245
competence factors (CF) 2501
competence-stimulating peptide (CSP) 235 ,
2368, 2435
cell lysis 2445
ComD interaction 255
QS regulation 239
receptor 2401, 257
signaling in biofilm formation 257
Streptococcus intermedius biofilm
growth 2578
competence-stimulating peptide (CSP)
precursor 2368, 2545
amino acid sequences 237
competence-induced cell lysis 2445
competitive inhibitors 74
ComX 2424, 245
comX gene 255
comY operon 252
conjugation factor (CF) 41 3
CqsAS system 1046
crown gall tumors 40, 41
csg mutants 276
csgA gene 2845
aggregation 286
extracellular complementation 281
mutants 277
sporulation 286
transcription 2845
CsgA protein 278
CshA 252
C-signal molecule in Myxococcus xanthus
269, 276, 277
aggregation induction 2857 , 28992,
291
cell motility relationship 2801
contact-dependent mechanisms 2801
end-to-end contacts 281
molecular nature 2789
motility
behavior 289
gliding engines 284
response 2889
responses 277
sporulation induction 2857
synthesis 27980
transduction pathway 2815, 283
morphogenesis control 2878
transmission 269 70, 2801, 283
mechanism 286
CTX genetic element 102
CVO26 76
cystic fibrosis 23
AI-2 in mixed-species signaling 1378
antibiotic-resistant bacteria 8990
autoinducers in sputa 13
chronic endobronchiolitis 89
Pseudomonas aeruginosa
infections 68
mucoid variants 89
RNAIII 210, 215
sputum samples 13, 334
CytR protein 108
Delisea pulchra (red macroalga) 73
furanones 713, 756
delta-toxin, see d -toxin
dental biofilm
bacterial community 1756
bacterial interactions 1779
contact-dependent bacterial signaling
17680
dental plaque 135
interbacterial cell-to-cell-adherence 179
lactate production by bacteria 177
I
N
D
E
X
304
LuxS 1823
LuxS-dependent virulence regulation
1858
mixed 178
oral bacteria growth 181
populational shifts 176
QS-dependent communication 1801
quorum sensing 17594
dental caries 1756
Streptococcus mutans 254
dental plaque 1357
Streptococcus 136
DevT protein 282
Dictyostelium discoideum 910
4,5-dihydroxy 2,3-pentanedione (DPD)
1213, 1523
derivatives 1256
DNA
bacterial acquisition 234
transforming factor 234
DNA microarray transcriptomics 812
Drosophila melanogaster 910
dsg mutants 276
D-signal 276
EAF plasmid 166
elastase 89, 1112
endobronchiolitis, chronic 89
endocarditis 216
enterobactin-like siderophores 185
Enterococcus faecalis, oligopeptide
signaling 118
enterotoxins 2045
EnvZ 191
epinephrine 1589
Escherichia coli
AI-2 130, 153, 183
activity 1201
communication 188
as QS signal 12930
diffusely adherent (DAEC) 151
enteroaggregative (EAEC) 151
enterohemorrhagic (EHEC) 1512
AI-3 QS signaling cascade 165
attaching and effacing lesions 152
cell-to-cell signaling 1557, 160
flagellar expression 163
flagellar regulon expression
activation 164
LEE genes 157, 158
LuxS-dependent autoinducer 1535
luxS mutant 156, 1579
luxS QS system 1523
motility expression 163
QS regulation 1667
QS signaling cascade 15965
virulence 157, 158, 160
enteroinvasive (EIEC) 151
enteropathogenic (EPEC) 151
adherence factor plasmid 166
biofilm development 167
pathogenesis 166
QS regulation 1667
quorum sensing 1667
enterotoxigenic (ETEC) 151
pSB401-containing 76
QS 139
SdiA 1645
virulence and LuxS-dependent regulation
15168
esg mutants 276
E-signal 276
EspA histidine protein kinase 292
exoenzymes 204
exopolymeric substances (EPS) 69
exotoxins
expression 21112
staphylococci 204
extracellular complementation 276
csgA gene 281
fatty-acid modifying enzyme (FAME) 205
fibrinogen 21516
fibronectin 206
fibronectin-binding proteins 2067, 221
fimA transcription 17980
flagellar regulon 1623, 166
expression 163, 1667
transcription 1623
FlgM 1623
FlhDC flagellar regulon 1623
flhDC gene
transcription 157, 158
transcriptional start sites 163
fnbA 2067, 221
fnbB 2067
FruA 282, 284
fruiting-body morphogenesis
intracellular signaling 2767
Myxococcus xanthus 26993, 272
aggregation and sporulation
coordination 2857
negative regulators 292
signal integration 2923
spore-filled 2712
Frz proteins 2824, 288
fungal extracts 801
furanones 845
bacterial colonization inhibition 723
Delisea pulchra 713
drug specificity 82
efficacy 867
305
I
N
D
E
X
furanones (cont.)
food content 80
halogenated 713
Pseudomonas aeruginosa QS
suppression 88
QS inhibitor pulmonary infection
treatment model 889
Serratia liquefaciens inhibition 723
furanosyl-borate-diester 153
Fusobacterium nucleatum
AI-2 quorum sensing 182
Lux S 182
gacA gene 89
GadX 1667
Galleria melonella 910
garlic 81, 845
gene transfer, horizontal 234, 250
GeneChips 812
genetic competence
biofilms 2558
cell-density-dependent development 2356
development 2434
genetic regulation 2545
genetics 2501
induction 2356
oral bacteria 24859
phases 235
regulation
in Streptococcus gordonii 251
in Streptococcus pneumoniae 23648,
251
shut-off 2456
streptococci 2345
genetic exchange 234
GISA 2212
GlyGly leader peptides 238
gppX 192
green fluorescent protein 3, 768, 77, 86
growth, sessile mode 656
Haemophilus influenzae, cystic fibrosis 137
HapR 112
hapR mutations 1067, 108, 109
hcnABC operon 33
hemolysins 204
2-heptyl-3-hydroxy-4-quinolone 234
2-heptyl-4-hydroxyquinolone oxide 32
Hfq protein 187
histidine kinase (HK) 236, 248, 2545
histidine protein kinase 2878
hk11 gene 258
hla gene 2013
transcription 202, 21112
hla RNA leader 202
homocysteine 1213
homoserine lactone ring 75
HPK10 subfamily 2401
Hpp permease 252
hydrogen peroxide 85
b-hydroxybutyryl homoserine lactone 119
immunocompromised patients 1
infections, biofilm 656, 89
infective endocarditis 250
interleukin-8 (IL-8) 1112
lactate production by dental biofilm
bacteria 177
lactonases 1415, 74
las boxes 23, 6
las system 19, 23, 31, 68
Pseudomonas aeruginosa 1, 4
PQS relationship 245
lasA gene transcripts 1213
LasB elastase 234, 33
lasB gene 234
transcripts 1213
lasI mutant 1011, 69
lasIrhll mutant 1012
LasR 2, 3, 234, 77
lasR gene
mutant 1011, 234
transcripts 1213
LasRLasI system 68
lecA expression 32
LEE (locus of enterocyte effacement) 152
EHEC virulence 157, 158
EPEC 1667
expression regulation 164
LEE1 152
leukotoxins 186
ligand binding 4950
lipB transporter 67
localized adherence (LA) 166
Lsr transporter system 153, 154, 189
LsrB structure 124
luxCDABE 101
LuxI protein 11718
Vibrio fischeri 129
LuxI-family proteins 104
LuxIR pathway 129
LuxLM proteins 11718
LuxO 1056, 112, 192
luxO mutations 1067
LuxP 124, 153
LuxQ 191, 192
LuxR 3, 55, 77
activation antagonism 75
AHL interaction 11718
I
N
D
E
X
306
OHHL displacement 73
overexpression 768
receptor 745
LuxR-type proteins 104
LuxS 11742
AI-2 production 119
cell-to-cell signaling in bacteria 1525
characterization 123
Escherichia coli virulence regulation
15168
identification 11921
oral bacteria 1823
phylogenetic tree 122
role 1268
signal molecule role 138
Vibrio fischeri 129
Vibrio harveyi 187
Vibrio vulnificus 129
virulence regulation 1858
luxS gene 1201, 182
Actinobacillus actinomycetemcomitans 131
expression by S. gordonii 187
inactivation 1312, 1367, 191
knockout 155
location 128
mutant 129, 1301, 157, 158, 1667
A-12 signaling 1535
Porphyromonas gingivalis 1845, 187
Shigella flexneri 131
operon 128
luxS system, Borrelia burgdorferi 1334
LuxS/AI-2, Escherichia coli 12930
luxS/AI-2, Neisseria meningitidis 134
LuxS-dependent signaling 1845
cell-to-cell communication 1934
LuxSPQ system 1046
LuxU 191
LytA (autolytic amidase) 244
LytC (autolytic lysozyme) 244
mannopine 48
medical devices, implanted 216
methionine 127
recycling 1334
(2R,4S)-2-methyl-2,3,3,4-
tetrahydroxytetrahydrofuran
(R-THMF) 124
metK 134
microbial activity 656
modulins, Staphylococcus epidermidis 208
mot operon 157, 158
MrcpC protein 282
multicellularity, survival strategy 2703
mvaT gene 89
myxobacteria
antimicrobial secretion 270
social lifestyle 270
spore-filled fruiting bodies 2723
starvation 2723
Myxococcus xanthus
aggregation 288
signal 26970
aggregation centers 271, 2912
A-signal 119
biofilms 270
colonies 270
C-signal 119, 269, 277, 2789
aggregation induction 2857,
28992, 291
contact-dependent mechanism 2801
end-to-end contacts between cells
281, 28991
molecular nature 2789
motility behavior 289
motility response 2889
sporulation induction 2857
synthesis 27980
transmission 26970, 2801, 283
transmission mechanism 286
C-signal transduction pathway 2815, 283
branches 2824
morphogenesis control 2878
signal amplification loops 285
fruiting-body morphogenesis 26993, 272
aggregation and sporulation
coordination 2857
negative regulators 292
signal integration 2923
gliding motility 26970, 2736, 275
adventurous 2736
social 2736
intracellular signaling 2767
life cycles 271
motility 26970, 2736, 275, 2889
C-signal 289
C-signal dependent response 2889
C-signal transmission 2801
directional 288
gliding engines 284
stream formation 2912
nozzle-like structures 2756, 2889
peripheral rods 287
reorientation 26970
slime trail 2756
spore development 2712
spore-filled fruiting bodies 2712, 292
starvation 2712, 292
stringent response 2712, 292
Tfp 2735
307
I
N
D
E
X
natural blockers, biofilm 713
Neisseria meningitidis, luxS/AI-2 134
NFkB induction 12
norepinephrine 1589
siderophore activity 185
NtrC transcriptional regulators 292
3-O-C
12
-HSL 2 3, 5, 89
cystic fibrosis sputa 13
immunomodulatory effects 11 12
mammalian cell transcription regulator
activation 12
prostaglandin E
2
link 12
OccR 45
OdDHL 68, 74
OHHL 42
analogs 745
bioluminescence studies 768
displacement from LuxR 73
furanone efficacy studies 867
OOHL 42, 434
degradation 56
structure 2, 51
TraR
binding 50
maturation 4950
polar interactions 523
protection against proteolysis 567
opines 41, 43, 44 5
oral bacteria
genetic competence 248 59
specializations 248
oropharyngeal flora 137
osteomyelitis 216
3-oxo-C6-homoserine lactone see OHHL
3-oxo-C8-homoserine lactone see OOHL
3-oxo-C
12
-HSL 234 , 30
N-(3-oxododecanoyl) homoserine lactone
see 3O-C
12
-HSL
N-[3-oxo-dodecanoyl]-L-homoserine lactone
see OdDHL
p17 protein 27880, 2845
p25 protein 27880, 2845
mutant 280
pathogenicity of Pseudomonas aeruginosa 8
pATK84b 45
patulin 84 5
pea exudates 79
Penicillium, QSI activity 81
periodontal disease 1756
Pfs 1213
phenol-soluble modulins 208
phnAB genes 29, 30
expression 32
Photorhabdus luminescens 133
pili, type IV 2735
Pisum sativum (pea) exudates 79
Pkn1 292
Pkn2 292
plant extracts 80 1
plant-associated bacteria 3940
polymorphonuclear leucocytes (PMNs),
biofilm mode of growth 856
polyphosphate kinase (PPK) 89
polysaccharide intercellular adhesion
(PIA) 21617
Actinobacillus actinomycetemcomitans
signal recognition 1923
adherence to viridans streptococci 1867
AI-2 183
quorum sensing 182
signal transduction 1878 , 1924
virulence modulation 186
hemin acquisition 184 5
LuxS 182, 183
luxS
inactivation 1367
mutant 1845, 187
LuxS-dependent signaling 1867
mixed biofilms 178
proteases 192
Streptococcus cristatus adherence
17980
Streptococcus gordonii adherence
179, 1867
Treponema denticola association 177
virulence 136 7, 1845
PQS see Pseudomonas quinolone
signal (PQS)
pqsABCD 2930
pqsABCDE operon 31
pqsH gene 258
pqsR mutant 33
PqsR regulation 30
Prevotella intermedia, AI-2 quorum
sensing 182
prostaglandin E
2
12
Protein A 2067, 213
proteins, intrinsically unstructured 4950
AI-2 138
biofilms 8, 33
QS role in development 6971
cell-to-cell signals 24
furanone effects 824
mucoid variants 89
PAO1 genome sequencing 56
pathogenicity 8, 910, 1378
pyoveridine 184, 185
I
N
D
E
X
308
QS 68
activity in vivo 1213
behavior regulation 14
circuitry 245
genes 4, 83
influence on virulence 912, 11
las 1, 4
rhl 1, 4
suppression 88
as therapeutic target 1315
QS factors 19
QS inhibitors 824
in garlic 81
target genes 4, 83
QS regulon 5
activation 89
QS-regulated genes 7
expression 7, 8, 78
4-quinolones 35
tobramycin treatment 84
virulence 912
PQS 334
virulence factors 1, 8, 234, 68
control by PQS 33
QS inhibitors 824
virulence genes 910, 11
Pseudomonas putida, biofilm formation 70
Pseudomonas quinolone signal (PQS)
35, 2335
cystic fibrosis sputa 13, 334
discovery 234
drug target potential 34
genes for production 289
Pseudomonas aeruginosa
QS circuitry relationship 245
virulence 334
production
regulation 31
timing 312
synthesis 27
genetics 259
regulation 2930
pTiBo542 45
pulmonary doseresponse models 5,
868, 87
pulmonary infection, QS inhibitor
treatment model 889
pyocyanin 28
pyoveridine 184, 185
qscR system 3
QseA 15961
qseA mutation 1667
QseB 1612
QseBC 162
flagellar regulon transcription 1623
quorum sensing 163
through AI-3 163
QseC 1612
qseD 164
qseE 164
qseF 164
QsrAB 248
4-quinolones 35
quorum sensing (QS)
Agr in Staphylococcus aureus 199200
Agrobacterium tumefaciens 42
AHL-dependent 86, 135, 180
AI-2 as signal 12832
interspecies 1345
oral bacteria 1825
Vibrio 188
AI-3-dependent signaling cascade 164
ArlRS 213
biofilms 70
formation inhibition 10711
gene control 71
cell density 140
cell-density-dependent 257
circuitry 245
colony expansion 67
CSP-dependent 239
deficiency 845
definition 138, 140
dental biofilm 17594
bacteria communication 1801
Escherichia coli 139
EHEC signaling cascade 15965
EPEC 1667
luxS in EHEC 1523
gene regulation in Pseudomonas
aeruginosa 7, 8, 78
green fluorescent monitors 3, 77
green fluorescent sensors 768
homoserine lactone systems 11718
intracellular signals 66
las system 31
luxS in EHEC 1523
oligopeptide pathways 11819
oral streptococci 249
activity in vivo 1213
behavior regulation 14
circuitry and PQS 245
las system 1, 4
rhl system 1, 4
therapeutic target 1315
virulence influence 912, 11
QseBC system through AI-3 163
regulation
in EPEC/EHEC 1667
in Vibrio 101
309
I
N
D
E
X
quorum sensing (QS) (cont.)
regulatory mutants 845
rhl system 31
Staphylococcus epidermidis 207 8
sae locus 212
SrrAB 212
staphylococcal 199 226
therapeutic tool 2246
streptococcal 2356
surface-associated behavior 67
systems
function 67 8
inhibition by C-30 14
through QseBC 163
TraR/TraI systems 47
Vibrio cholerae 101, 110
apparatus 106
infectious cycle 109
negative regulation of virulence genes
1067 , 111
regulation 104 6, 105
sensing systems 1046
targets 106 7
Vibrio harveyi 120
pathways 185
quorum sensing factors 19
quorum sensing inhibitor selectors 79,
80, 81
quorum sensing inhibitors (QSI) 66 8, 72,
75 6
antibiotic combination 8990
antibiotic synergism 90
anti-pathogenic therapy 678
biofilm effects 846
novel natural 7981
Pseudomonas aeruginosa QS suppression
88
pulmonary dose-response model 5 , 87
screens 768
novel 789
treatment model 889
quorum sensing regulon 7, 71
Pseudomonas aeruginosa 5, 89, 10
RbsB 1889, 190
rbsDACBK 189
RbsK 1889
red fluorescent protein 78
relA gene 8 9
repABC operon 53
response regulator (RR) 236, 2545
BlpR 248
rhamnolipid 89
Rhizobiaceae 3940
rhizosphere 39
rhl system 1 9, 23 4, 31, 68
Pseudomonas aeruginosa 1 , 4
PQS relationship 245
Rhl1RRh1I system 68
rhlI gene 25
mutant 1011
S-ribosyl homocysteine (SRH) 1213
RIP/RAP QS system 224 6
RNAII 214
RNAIII 201 3, 204
cystic fibrosis 210, 215
fibrinogen depletion 21516
sarA mutants 214
RopA 259
Rot virulence regulator 214
rpoN gene 89
rpoS gene 8 9
rr11 gene 258
rsmA gene 89
sae locus 212
salivary pellicle, colonizing bacteria 176
Salmonella
AI-2 153
PhoB/PhoR 162
SdiA 1645
Salmonella typhimurium
ABC transporter 153
AI-2 183
activity 1201
binding protein structure 124
Lsr transporter 153
PmrB 161
SarA family 214
SdeK histidine protein kinase 2878
SdiA 1645
sed gene expression 2023
Serratia liquefaciens 70
inhibition by furanones 723
Serratia marcescens, luxS mutants 1301
Shiga toxin (Stx) 152
Shigella flexneri, luxS mutation 131
short-chain alcohol dehydrogenase (SCAD)
278
siderophores 185
enterobactin-like 185
sigma factor s
B
213
sigma X see ComX
skin infections 216
slime trail 2756
SLUSHpeptides, Staphylococcus lugdunensis
208
SmcR system 129
S-motility 2736, 2889
I
N
D
E
X
310
SocE 292
spa gene
expression 213
transcription inhibition 202
SpdR 292
sporulation 2857
SrrAB 212
sspAB genes 1767
staphylococcal enterotoxin A (SEA) 205
staphylococci 199226
agr homologs 20910
Agr specificity groups 20911
Agr variants
generation 223
in pathogenesis 2204
Agr-regulated genes 2037
AIP signal 2001
attachment 217
biofilms 21617
Agr 21819
agr gene mutants 220
antibiotic exposure 220
behavior 21617
detachment 21820
development 21617, 21920
maturation 21718
ClpP 213
ClpX 213
diseases 199
enterotoxins 2045
exoenzymes 204
exotoxins 204
expression 21112
fibrinogen 21516
hemolytic variants 222
with intermediate resistance to
glycopeptide antibiotics (GISA)
2212
QS 199226
therapeutic tool 2246
RIP/RAP QS system 2246
RNAII 214
RNAIII 2013, 204, 214
cystic fibrosis 215
SarA family 214
sigma factor s
B
213
surface adhesion 217
virulence 203, 2067
regulation by Agr 21112
Staphylococcus aureus
Agr evolution 2224
Agr QS system 199200
biofilm 216
capsular polysaccharide 2056
cystic fibrosis 137
gene array 207
oligopeptide signaling 118
Staphylococcus epidermidis
agr mutants in biofilm formation 220
biofilms 216
infections 2078, 211
QS 2078
Staphylococcus lugdunensis 208
SLUSH peptides 208
starvation
myxobacteria 2723
Myxococcus xanthus 2712, 292
streptococci
biofilms 1312
cell density-dependent competence
development 2356
regulation 23360
ComE response regulator 2412
ComX 242
dental plaque 136
genetic transformation 2345
lactate production in dental biofilm 177
luxS/AI-2 QS systems 1312
oral 234, 24859
Porphyromonas gingivalis adherence
1867
quorum sensing 2356
Streptococcus cristatus, Porphyromonas
gingivalis adherence 17980
AI-2 secretion 187
cell density-dependent QS 257
ComD 240
infective endocarditis 250
interaction with saliva 1767
mixed biofilms 178
Porphyromonas gingivalis adherence
179, 1867
transformation 2504
Treponema denticola genetic exchange
257
Streptococcus intermedius biofilm 2578
dental caries 254
luxS 13940
oral cavity 2589
pathogenicity 258
relA 13940
RopA 259
survival 258
transformation 2549
virulence factors 254
Streptococcus pneumoniae
avirulent strain 2334
CiaRH TCSTS 2467
311
I
N
D
E
X
Streptococcus pneumoniae (cont.)
com AB locus 238
com regulon 2545
ComCDE competence regulon 23642
ComD sequence analysis 241
CSP-dependent QS 239
genetic competence
regulation 23648
shut off 2456
luxS mutant 140
oligopeptide signaling 118
R strain 2334
S strain 2334
Streptomyces, butyryl lactone signals 119
survival, multicellularity 2703
swarming 67
swr system 67
T cells 2045
tatC gene 11
TCP pathogenicity island 102, 109
TcpPH 103
TCSTS (two-component signal
transduction system) 236, 2467, 258
T-DNA transfer 401
Tfp 2735, 2889
Ti plasmids 46
tobramycin 845
TodK histidine protein kinase 2878, 292
TolB 276
TolQ 276
TolR 276
TonB 1845
Tox R regulon 103
toxA transcripts 1213
toxic shock 2045
d-toxin 2013, 204
cell detachment from biofilms 218
S. epidermidis 207
toxin-coregulated pilus (TCP) 102
ToxR regulon 102, 1067
ToxRS 103
ToxT 103
Tr1R 48, 56
tra box 50, 535
structure 2, 51
TraR-dependent promoters 54
tra genes 41
traI gene 434
TraM 468, 556
transcriptional autoregulation 162
transcriptomics 814
DNA microarray 81
transformation
groups 240
principle 2334
process 243
transforming factor 2334
TRAP 2246
TraR
class I and II promoters 535
C-terminal DNA binding domain 52
cytoplasmic membrane binding 49
dimerization interface 502
function 503
inactivation 556
maturation 4950
models 54
N-terminal pheromone binding domain
52
OOHL polar interactions 523
post-transcriptional regulation of
activity 469
protection against proteolysis 567
structure 2, 503, 51
transcription activator 535
traR gene 434
expression regulation 445
regulation 46
TraR-dependent promoters 54
transcription start sites 535
TraR-TraI quorum-sensing system 42
trb genes 41
Treponema denticola
Porphyromonas gingivalis association 177
Streptococcus gordonii genetic
exchange 257
twin-arginine-translocation (TAT)
pathway 11
Veillonella, lactate production in dental
biofilm 177
vfr gene 89
Vibrio
AI-2 cell signaling 1289
AI-2 quorum sensing signal 188
Vibrio cholerae 10112
AI-2 signaling pathway 1289
autoinducers 112
biofilms 1078, 110
formation regulation 108
host infection 109
environmental fitness 111
environmental phase of life cycle 110
flagellar torque monitoring 108
infections 104
LuxQ 191
LuxU 191
pathogenicity 1014
QS 101, 110
apparatus 106
I
N
D
E
X
312
negative regulation of virulence genes
1067, 111
regulation 1046, 105
sensing systems 1046
targets 1067
QS-deficient hapR mutants 108
toxigenic strains 111
virulence genes 102, 103
QS negative regulation 1067, 111
Vibrio fischeri 43, 101
autoinducer 117
signaling pathways 129
Vibrio harveyi 101, 1046
AI-1 1345
AI-2 1345
luciferase operon 1046
LuxLM proteins 11718
LuxS protein 187
luxS/AI-2 pathway 11921
QS pathways 120, 185
reporter strain generation 11920
Vibrio polysaccharide synthesis (vps)
genes 108
Vibrio vulnificus 129
vicK gene 2589
vicR gene 2589
VicRK TCSTS 247
violacein 76
virulence
AHL-dependent QS 180
Escherichia coli
enterohemorrhagic (EHEC) 157,
158, 160
LuxS-dependent regulation 15168
LuxS-dependent regulation 1858
PQS 334
Porphyromonas gingivalis 1367, 1845
modulation by AI-2 186
staphylococci 203, 2067
virulence factors
Agr regulation 21112
Pseudomonas aeruginosa 1, 8, 234, 68
control by PQS 33
QS inhibitors 824
virulence genes
Agr regulation 204
Pseudomonas aeruginosa 910, 11
QS negative regulation 1067
Vibrio cholerae 102, 103
QS negative regulation 1067, 111
virulence regulator, Rot 214
vitronectin 206
vitronectin-binding protein 2067
VpsR response regulator 108
VpsT response regulator 108
vqsR gene 8
313
I
N
D
E
X
rhl l rhl R
RhlI RhlR
RhlR
RhlR
Target genes:
rhlAB rhlI lasB
Lasl
las l las R
LasR
LasR
LasR
Target genes:
lasB,A lasI toxA
xcpR,P apr rhlR
C
4
-HSL
C
4
-HSL
C
4
-HSL
C
4
-HSL
3-oxo-C
12
-HSL
3-oxo-C
12
-HSL
3-oxo-C
12
-HSL
3-oxo-C
12
-HSL
Figure 1.1. The las and rhl QS systems in P. aeruginosa. Both the las and rhl systems are
composed of a transcriptional regulator protein (LasR, RhlR) and an autoinducer synthase
(LasI, RhlI). The LasI synthase produces 3-oxo-C
12
-HSL, and the RhlI synthase produces
C
4
-HSL. The autoinducer then binds to its respective cognate protein. The LasR:3-oxo-
C
12
-HSL complex activates transcription of genes involved in the production of several
virulence factors, including lasB, aprA, and toxA, as well as upregulating transcription of lasI
and of rhlR. The RhlR:C
4
-HSL complex activates transcription of several genes involved in
virulence, including rhlAB.
(a)
(b)
(c)
G5
T4
210
9
9
N
C
C
N
206
1
3
1
3
Figure 3.3. Ternary structure of TraR, OOHL, and tra box DNA. (a) Just the N-terminal
domains of a dimer. The dimerization interface is highlighted, and is along the length of
-helix 9 of each monomer. The OOHL and water molecule bound in each N-terminal
domain are shown in space-fill and can be seen most clearly in the right-most monomer.
The linker between the N-terminal domains and the C-terminal domains is marked with a
C. (b) Just the C-terminal domains of the TraR dimer bound to DNA. Hydrogen bonds
between the side-chains of arginine 206 and 210 of the recognition helix and bases T4 and
G5 of the binding site are shown. The dimerization interface is also highlighted, along
-helix 13 of each monomer, and the linker between the NTD and CTD of each monomer is
marked with an N. (c) A view of a full dimer down the long axis of the DNA, showing the
asymmetry of the crystallized protein due to the flexible linkers. Coordinates for these
models are from references 81 and 92.
Green fluorescent cells Dark cells
(a)
(b)
No
signal
or
signal
block
+
signal
R
e
s
p
o
n
s
e

t
i
m
e
s
:
1
0

1
5

m
i
n
4

5

h
r
s
lasR
P
lasB
Gfp(ASV)
luxR
P
luxl
Gfp(ASV) luxR
P
luxl
Gfp(ASV)
lasR
P
lasB
Gfp(ASV)
Active receptor Blocked receptor AHL signal QSI
R
R
R
R
R
R
R*
R*
R*
R
Figure 4.2. (a) Molecular details of two green fluorescent QS monitors. The LuxR-based
monitor responds to a variety of AHL signals except for BHL, whereas the LasR-based
monitor primarily responds to OdDHL. (b) Microscopic inspection of Escherichia coli cells
carrying the LuxR-based monitor (right panels show fluorescent microscopy). The cells
respond to the addition of 10 nM OHHL by producing detectable amounts of the unstable
GFP protein. If the OHHL is removed by washing, or if a QSI compound is added which
blocks GFP synthesis, growing cells lose the GFP signal in 45 hours.
Figure 4.4. Gene expression atlas of QS-controlled genes and QSI target genes in
Pseudomonas aeruginosa PAO1. Upregulated genes are shown in green; downregulated
genes are red. QS-controlled gene expression is shown in the outer half of the atlas;
furanone C-30 target genes are indicated in the inner half. QS-controlled genes were
identified by growing a P. aeruginosa PAO1 lasIrhlI mutant with or without exogenous
AHL signals (2 mM OdDHL and 5 mM BHL) and retrieving samples for microarray analysis at
several culture densities. C-30 target genes were identified by growing P. aeruginosa PAO1
with or without 10 mM furanone C-30. P. aeruginosa PAO1 ORFs are shown in red (sense
strand) and blue (antisense strand) in the innermost circle. The outermost scale gives the
gene localization (in base pairs) in the genome. The color scale bar gives the correlation
between color coding and fold-change in gene expression.
R
(a) (c) (b)

6 h after injection
RFP GFP RFP GFP
(d)
R
+
R
R
lasR P
lasB
gfp(ASV) P
tac
dsred P
lasB
gfp(ASV) P
tac
dsred lasR
R
R
OdDHL QSI
Time 0
(e)
Active receptor Blocked receptor
Figure 4.5. The pulmonary doseresponse model. (a) Mice are challenged intratracheally
with alginate beads containing P. aeruginosa. (b) Photomicrographs of mouse lung tissue
infected with P. aeruginosa. The arrow points at an alginate bead surrounded by numerous
PMNs. Adapted from reference (126). (c) QSI drugs can be injected intravenously in the tail
vein. (d, e) Mouse lung tissue infected with P. aeruginosa carrying the LasR-based monitor
PlasBgfp for detection of cell-to-cell signaling (green fluorescence) and a tag for simple
identification in tissue samples (red fluorescence) examined by SCLM. (d) Mice were
administered C-30 via intravenous injection at time zero. (e) Infected animals were
sacrificed in groups of three at the time point indicated and the lung tissue samples were
examined by SCLM according to Hentzer et al. (43). Parts (d) and (e) are adapted from (43).
Figure 9.3. Three-dimensional reconstruction of a Staphylococcus aureus biofilm. Cells
expressing a quorum-controlled green fluorescent protein reporter are green; the
remaining biofilm is red from staining with propidium iodide. Each side of the grid
represents about 600mm.

Bacterial Cell To Cell Communication

Uploaded by

Copyright:

Available Formats

Bacterial Cell To Cell Communication

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

Bacterial Cell To Cell Communication

Uploaded by

Copyright:

Available Formats

This page intentionally left blank

Bacteri al Cel l -to-Cel l Communi cati on

strains exhibit reduced virulence gene expression at high cell density.

), suggesting that AHL-like autoindu-

). This suggests that these organisms produce a signal that

mutants) result in a non-motile phenotype; mutations that

in vitro (47). Thirdly, exogenous

coenzyme-binding pocket and are unable to bind NAD

-containing protein. Genes

(a) (c) (b)

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.