Biochem. J.

(1988) 253, 337-343 (Printed in Great Britain) 337

Effects of ciprofibrate and 2-15-(4-chlorophenyl)pentyljoxirane-
2-carboxylate (POCA) on the distribution of carnitine and CoA
and their acyl-esters and on enzyme activities in rats
Relation between hepatic carnitine concentration and carnitine acetyltransferase activity

A. K. M. Jalaluddin BHUIYAN,* Kim BARTLETT,* H. Stanley A. SHERRATTt and Loranne AGIUSI

Human Metabolism Research Centre, Departments of *Child Health and Clinical Biochemistry, tPharmacological Sciences and
tMedicine, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH, U.K.

The effects of feeding the peroxisome proliferators ciprofibrate (a hypolipidaemic analogue of clofibrate) or
POCA {2-[5-(4-chlorophenyl)pentyl]oxirane-2-carboxylate} (an inhibitor of CPT I) to rats for 5 days on the
distribution of carnitine and acylcarnitine esters between liver, plasma and muscle and on hepatic CoA
concentrations (free and acylated) and activities of carnitine acetyltransferase and acyl-CoA hydrolases were
determined. Ciprofibrate and POCA increased hepatic [total CoA] by 2 and 2.5 times respectively, and [total
carnitine] by 4.4 and 1.9 times respectively, but decreased plasma [carnitine] by 36-46 %. POCA had no effect
on either urinary excretion of acylcarnitine esters or [acylcarnitine] in skeletal muscle. By contrast,
ciprofibrate decreased facylcarnitine] and [total carnitine] in muscle. In liver, ciprofibrate increased the
[carnitine]/[CoA] ratio and caused a larger increase in [acylcarnitine] (7-fold) than in [carnitine] (4-fold),
thereby increasing the [short-chain acylcarnitine]/[carnitine] ratio. POCA did not affect the [carnitine]/
[CoA] and the [short-chain acylcarnitine]/[carnitine] ratios, but it decreased the [long-chain acylcarnitine]/
[carnitine] ratio. Ciprofibrate and POCA increased the activities of acyl-CoA hydrolases, and carnitine
acetyltransferase activity was increased 28-fold and 6-fold by ciprofibrate and POCA respectively. In
cultures of hepatocytes, ciprofibrate caused similar changes in enzyme activity to those observed in vivo,
although [carnitine] decreased with time. The results suggest that: (1) the reactions catalysed by the short-
chain carnitine acyltransferases, but not by the carnitine palmitoyltransferases, are near equilibrium in liver
both before and after modification of metabolism by administration of ciprofibrate or POCA; (2) the
increase in hepatic [carnitine] after ciprofibrate or POCA feeding can be explained by redistribution of
carnitine between tissues; (3) the activity of carnitine acetyltransferase and [total carnitine] in liver are closely
related.

INTRODUCTION chondrial redox state, the capacity for f-oxidation

depends on [carnitine] and [CoA]. Carnitine is essential
Mitochondria oxidize long-chain fatty acids rapidly for the transfer of long-chain fatty-acyl groups into
and completely to acetyl-CoA, although they only oxidize mitochondria, and buffers free [CoA]. Administration of
very-long-chain fatty acids slowly [1-3]. Liver peroxi- hypolipidaemic drugs of the fibrate group including
somes oxidize long-chain and very-long-chain fatty acids clofibrate, ciprofibrate and bezafibrate [4], and the
to medium-chain acyl-CoA esters by successive removal hypoglycaemic compound POCA, the CoA ester of
of acetyl groups as acetyl-CoA. Medium-chain acyl-CoA which powerfully inhibits CPT I [5], induces both
esters formed by partial peroxisomal ,-oxidation transfer proliferation of peroxisomes and peroxisomal fl-oxida-
their acyl groups to carnitine catalysed by COT and CAT tion, and the capacity for mitochondrial f8-oxidation of
in peroxisomes [2,3]. These acylcarnitine esters then palmitoylcarnitine [6,7]. Fibrates also increase [carnitine]
diffuse to the mitochondria, where the acyl groups are [6,8] and [CoA] [9-11] in rat liver, and clofibrate increases
transferred to CoA in the matrix by the action of CPT I, both mitochondrial and cytosolic [CoA] [12].
CPT II, COT and CAT, where their oxidation to acetyl- Although the metabolism of carnitine and CoA is
CoA is completed. Acetyl groups formed by peroxisomal intimately related through the action of the carnitine
fl-oxidation are also transferred to the mitochondrial acyltransferases, [carnitine], [CoA], [short-chain acylcar-
matrix as acetylcarnitine. The capacity of the liver to nitine], [short-chain acyl-CoA], [long-chain acylcarnitine]
oxidize long-chain and very-long-chain fatty acids is and [long-chain acyl-CoA] do not appear to have been
therefore partly determined by the activities of the determined in the same internally consistent experiment.
carnitine acyltransferases. In addition to the activities of In this study we used ciprofibrate {2-[4-(2,2-dichloro-
mitochondrial and peroxisomal enzymes, and the mito- cyclopropyl)phenoxy]-2-methylpropionic acid} [13] and

Abbreviations used: POCA, 2-[5-(4-chlorophenyl)pentyl]oxirane-2-carboxylate; POCA-CoA, 2-[5-(4-chlorophenyl)pentyl]oxirane-2-carbonyl-

CoA; CAT, camitine acetyltransferase; COT, carnitine octanoyltransferase; CPT, CPT I, CPT II, carnitine palmitoyltransferase (EC 2.3.1.21), its
overt and latent forms.

Vol. 253
338 A. K. M. J. Bhuiyan and others

POCA to change fatty acid metabolism so as to (approx. 100 mg) was homogenized with a Polytron homo-
investigate the relation between [carnitine], [CoA], genizer in 600 g1 of 1 M-HC14, to which was added
[acylcarnitine] and [acyl-CoA] and the activities of CAT palmitoyl-L-[methyl-'H]carnitine (73 pmol; 8000 d.p.m.;
and acyl-CoA hydrolases in liver and skeletal muscle, 1.85 x 106 Bq/,umol) as internal standard. The homogen-
and [carnitine] and [acylcarnitine] in plasma. ate was centrifuged (10000g), and free and short-chain
acylcarnitine esters were determined in the neutralized
MATERIALS AND METHODS supernatant. For determination of long-chain acylcarni-
tine esters, the pellet was dissolved in 200 1ul of 1 M-KOH
Materials and incubated at 50 C for 2 h. Protein was precipitated
Ciprofibrate was a gift from Sterling-Winthrop by 300 ,ul of 1 M-HCIO4 and removed by centrifugation
Research, POCA was from Byk Gulden Chemische (10000 g). The supernatant (400 ,1) was neutralized with
Fabrik (Konstanz, Germany), and L-carnitine was from 95 ,ul of 1 M-KOH and 200 1ul of 1 M-Hepes, pH 7.2 and
Sigma-Tau (Rome, Italy). [methyl-3H]Carnitine was precipitated KC104 was removed by centrifugation
from Amersham International (Amersham, Bucks., (10000 g). Free carnitine formed by hydrolysis was
U.K.). Enzymes and confactors were from Sigma Chemi- determined and corrected for incomplete recovery by
cal Co. (St. Louis, MO, U.S.A.) or from Boehringer reference to the internal standard. With plasma and
Corp. (Mannheim, Germany). Sources of culture media urine samples the homogenization step was omitted.
were as in ref. [14].
Determination of free CoA and acyl-CoA esters
Animals Frozen liver (about 200 mg) was homogenized in
Male albino Wistar rats (180-220g bodywt.) were 1.0 ml of 1 M-HC104 containing 2 mM-dithiothreitol,
bred locally, and guinea pigs (250-300 g, body- wt.) were and centrifuged (10000 g). The supernatant was neutral-
obtained from Bantin and Kingman, Grimston, Ald- ized to pH 5-6 with 3 M-KOH (150 lsl), and half was
borough, Hull, U.K. They were fed on standard rat or used directly for determination of free CoA. The rest was
guinea-pig chow (Scientific Diets Services, Stepfield, used for the determination ofacid-soluble acyl-CoA, after
Witham, Essex, U.K.) ad libitum. Ciprofibrate and the hydrolysis of the esters by adding 30 ,ul of 3 M-KOH and
ethyl ester of POCA were dissolved in acetone as 10 % incubating at 45 C for 30 min and then neutralizing with
(v/v) solutions and sprayed on the chow, to give 5 M-HCI, to give total CoA (acid-soluble CoA plus free
concentrations of 0.1 % and 0.2% (w/w) respectively, CoA). The protein precipitate (containing acid-insoluble
which were checked by h.p.l.c. analysis [15]. The rats long-chain acyl-CoA) was hydrolysed by adding 200 ,l
were fed on control diet or diet containing drugs for 5 of 1 M-KOH and 500 ,l of 10 mM-dithiothreitol and
days. In Expt. 1 the rats were caged in groups of five for incubation at 45 C for 20 min. Protein was precipitated
4 days, and on day 5 they were put singly in metabolism with 300 1l of 1 M-HC104 and the supernatant
cages for collection of urine. In Expt. 2 the rats were neutralized.
caged in groups of five for 5 days until used. The animals CoA in the extracts was determined enzymically by
were killed by intraperitoneal injection of pentobarbital using 2-oxoglutarate dehydrogenase, by an automated
(60 mg/kg body wt.) between 09:00 and 10:00 h. Blood fluorimetric assay using a centrifugal analyser (Cobas
was withdrawn from the inferior vena cava, and the Bio 8326; Hofmann-La Roche). The assay medium
liver was rapidly removed and weighed. Part of the liver contained 100 mM-Tris/HCl (pH 7.9), 5 mM-MgCl2,
was immediately freeze-clamped, and part was homo- 2 mM-NADI, 5 mM-2-oxoglutarate, 2 mM-dithiothreitol,
genized in a medium containing 50 mM-Tris/HCl, 0.4 mM-thiamin pyrophosphate and 2-oxoglutarate de-
pH 7.0, 100 mM-KCl, 5 mM-EDTA, 20 mM-KF and hydrogenase (1O munits) in a final volume of 200 1d,
0.05 % (w/v) Lubrol PX, for enzyme assays. Enzyme and 10 1 of either unhydrolysed or hydrolysed extract or
activities were determined the same day. Skeletal muscle CoA standards (25, 50 or 100 /pM). The reaction was
(hind limb, gastrocnemius, quadriceps) was rapidly started by addition of 2-oxoglutarate dehydrogenase
removed and freeze-clamped. after a preincubation spin of SO s at 100g. The increase in
fluorescence (excitation 340 nm, emission 450-460 nm)
Studies in vitro using cultured hepatocytes after addition of enzyme was monitored at 30 s intervals
Parenchymal hepatocytes were isolated from normal for 8 min at 25 'C. The CoA concentration was deter-
rats (180-200 g) by collagenase perfusion of the liver [16] mined from the increase in fluorescence by end-point
and cultured in monolayer in 25 cm2 flasks as described analysis. End-points were reached at 100-150 s, and the
previously [14]. Minimum essential medium supple- increase in fluorescence was linear with [CoA] up to
mented with 5 % (v/v) fetal bovine serum was used for 100 pM.
the initial attachment stage (4 h). Subsequently cells were
maintained in serum-free medium supplemented with Measurement of enzyme activities
10 nM-dexamethasone, 10 nM-insulin and the concentra- The following enzymes were assayed in homogenates
tions of ciprofibrate indicated. Medium was changed of fresh liver or cultured hepatocytes spectrophoto-
daily and enzyme activities were determined after metrically at 30 'C using a centrifugal analyser (Cobas
4 days. The hepatocyte monolayer was washed twice Bio 8362) as described previously: glutamate dehydro-
with 0.14 M-NaCl and then extracted in the medium used genase (EC 1.4.1.2) and citrate synthase (EC 4.1.3.7) [18];
for fresh tissue (above). hydroxyacyl-CoA dehydrogenase (EC 1.1.1.35) [19];
CAT (EC,2.3.1.7), palmitoyl-CoA hydrolase (EC 3.1.2.2)
Concentrations of carnitine and acylcarnitine esters in and acetyl-CoA hydrolase (EC 3.1.2.1) [20]. The last
tissues, plasma and urine enzyme was assayed in the presence of 2 mM-ATP or
Carnitine was determined radioenzymically [17] with 1 mM-ADP, which respectively activate or inhibit the
some modifications. Freeze-clamped liver or muscle cytoplasmic enzyme [21]. All assays were run against
1988
Carnitine, CoA and peroxisomal proliferators 339

Table 1. Concentrations of CoA, carnitine and their acyl esters in livers from control, ciprofibrate-fed and POCA-fed rats
Concentrations are given as nmol/g wet wt., as means + S.E.M.; significances of differences from controls: *P < 0.05; **P < 0.01;
***P < 0.005. The apparent mass-action ratios for the reactions catalysed by the carnitine short-chain acyl- and palmitoyl-
transferases are defined as [carnitine][short-chain acyl-CoA]/[CoA][short-chain acylcarnitine] and [carnitine][long-chain acyl-
CoA]/[CoA][long-chain acylcarnitine] respectively. The apparent mass-action ratios as measured depend on whole-tissue
concentrations of CoA, carnitine and their acyl esters. However, the carnitine/acylcarnitine carrier in the mitochondrial inner
membrane enables equilibration of the acylation state of the mitochondrial and cytosolic pools of CoA and carnitine, so that
these concentrations define approximately the mass-action ratios (see the text).

Control Ciprofibrate-fed POCA-fed Guinea pig

(n = 7) (n = 9) (n = 5) (n = 2 or 3)

Concentrations
CoA 43.0+2.6 71.5+7.4** 84.6+14.0** 33.8, 51.6
Short-chain acyl-CoA 12.2+2.5 35.2 + 6.1 ** 38.7+3.1*** 21.2, 13.5
Long-chain acyl-CoA 68.7+4.1 147.0+ 11.7*** 189.0+16.8*** 62.8, 45.6
Total CoA 124.0+6.2 253.0 + 13.7*** 313.0+13.3*** 118.0, 111.0
Carnitine 167.0+7.41 645.0+ 14.8*** 328.0+ 11.6*** 316.0.+36.0***
Short-chain acyl-carnitine 40.0+4.3 291.0+22.0*** 71.0+6.8*** 86.0+ 17.0***
Long-chain acyl-carnitine 22.0+2.2 67.0+ 3.5*** 29.0 + 1.6* 61.0+7.0***
Total carnitine 229.0+6.6 1003.0+22.2*** 428.0+ 13.5*** 463.0+47.4***
Concentration ratios
Carnitine/CoA 3.70+0.69 9.09+2.48* 4.00+1.44 7.14
Short-chain acyl-CoA 0
_ 0 0.20
Short-chain acyl-carnitine 0 +
Long-chain acyl-CoA 3.03 +0.83 2.08 +0.56 6.25+ 1.95 1.12
Long-chain acyl-carnitine
Apparent mass-action ratios
For short-chain esters 1.32+0.16 0.99 + 0.38t 2.17 + 0.24** 1.12
For long-chain esters 11.6+0.16 20.0+12 13.2+9.7 66.7
t n=7.

blanks with water in place of enzyme extract. Controls chain acyl groups in the controls and 600 in the treated
for-CAT were run with enzyme extract and omission of groups (Table 1).
L-carnitine. Reaction rates were monitored for 210 s and
determined by a Linear Search and Regression program. Concentrations of carnitine and acylcarnitine esters
The rates were corrected by subtraction of control values. in iver
Acyl-CoA oxidase activity with 25 /iM-palmitoyl-CoA as Ciprofibrate caused a 4-fold increase in hepatic
substrate was recorded colorimetrically [22] with a dual- [carnitine] and a 7-fold increase in [short-chain acyl-
wavelength spectrophotometer. carnitine] (Table 1), with doubling of the [short-chain
acylcarnitine]/[carnitine] ratio). POCA, by contrast,
RESULTS increased [carnitine] and [short-chain acylcarnitine]
by 2-fold without affecting the [short-chain acyl-
Body weight and liver weight of drug-treated rats carnitine]/[carnitine] ratio. However, it caused a small
Administration of ciprofibrate or POCA in the diet for but significant increase in the [long-chain acyl-CoA]/
5 days had no significant effect on the body weights of [long-chain acylcarnitine] ratio (Table 1).
the rats: in Expt. 1, controls, 204+9 g, ciprofibrate,
192+3g, POCA, 198+16g; in Expt. 2, controls, Concentrations of carnitine and acylcarnitine esters in
203 + 6 g, ciprofibrate, 212 + 5 g (means + S.E.M.). Cipro- plasma, skeletal muscle and urine
fibrate, but not POCA, increased the liver weights The plasma [total carnitine] decreased in both cipro-
by 600%: in Expt. 1, controls 9.4 + 0.2 g, ciprofibrate, fibrate- and POCA-treated rats (Table 2), largely owing
15.1 + 0.3 g, POCA, 8.7 + 0.3 g; in Expt. 2, controls, to decreases in [carnitine] (51 and 360% respectively).
10.00.3g, ciprofibrate, 16.3+0.9g (means+S.E.M.). POCA also decreased [short-chain acylcarnitine] and
All data for Expts. 1 and 2 were very similar and were [long-chain acylcarnitine].
pooled. Muscle [total carnitine] was 100% lower after admin-
istration of ciprofibrate, owing to lower [acylcarnitine]
Concentrations of CoA and acyl-CoA esters in liver but not [carnitine] (Table 2). POCA did not decrease
Ciprofibrate caused a 2-fold increase in [total CoA] muscle [carnitine], in agreement with previous results
(free plus acylated) in liver, and POCA a 2.5-fold [15]. In control and POCA-treated rats the percentage of
increase (Table 1). Approx. 100 of the CoA was the carnitine acylated was higher in skeletal muscle and
esterified with short-chain acyl groups in the controls, in liver than in plasma (Tables 1 and 2).
and 120% in both treated groups, and 55 % with long- Ciprofibrate increased the urinary excretion of acyl-
Vol. 253
340 A. K. M. J. Bhuiyan and others

Table 2. Concentrations of carnitine and acylcarnitine esters in rat skeletal muscle and plasma
Experimental details are given in the Materials and methods section. Concentrations are expressed as nmol/g wet wt. or
nmol/ml of plasma, and are means+ S.E.M. with the numbers of observations in parentheses. Significances of differences from
controls: *P < 0.05; **P < 0.01; ***P < 0.005.

Concn.
Percentage
Short-chain Long-chain Total of carnitine
Tissue Carnitine acylcarnitine acylcarnitine carnitine acylated

Skeletal muscle
Control (6) 573+ 17 239+ 18 78+10 889+ 19 36+2
Ciprofibrate-fed (6) 582+13 160 + 23* 43 +4** 785 + 19*** 26+3*
POCA-fed (5) 627+28 194+22 83 + 3 904 + 31 31+3
Plasma
Control (6) 28 +2 4.7 +0.7 1.7+0.1 33.5+ 1.5 17+1
Ciprofibrate-fed (9) 13.7 _2.1** 3.4+0.8 1.2+0.1* 18.0+ 1.8** 23 + 3
POCA-fed (5) 18.1 +0.9** 1.8+0.4 1.1 +0.1** 21.6+0.6** 14+2

Table 3. Estimated contents of carnitine and acylcarnitine esters in skeletal muscle, extraceliular fluid and liver in rats given ciprofibrate
or POCA for 5 days
The total carnitine (free and acylated) of liver was determined (Table 1) and that of skeletal muscle and extracellular fluid was
estimated from measured concentrations (Table 1) and body weights, assuming that muscle is 40 % of body weight and that
extracellular fluid is 15 %. The total carnitine content (sum of free plus acylated forms) is expressed in ,umol, as means + S.E.M.
(n = 5 or 6): *P < 0.05; ***P < 0.005.

Tissue carnitine (all forms) Controls Ciprofibrate-fed POCA-fed

Skeletal muscle 73.0+2 61.8+ 1.5 71.6+0.3

Extracellular fluid 1.01 +0.06 0.54+0.05*** 0.70+0.10
Liver 2.24+0.06 15.9 + 0.8*** 3.72+0.21*
Sum 76.2 78.2 76.0
Urinary output (,umol/day per rat) 0.17+0.03 0.34+0.04 0.21 +0.02

Table 4. Activities of enzymes in livers from rats given ciprofibrate or POCA for 5 days
Experimental details are given in the Materials and methods section. The activities are expressed as units/g wet wt., as
means+ S.E.M. for n observations. Significance of differences from means: *P < 0.05; ***P < 0.005.

Activity
Enzyme Control Ciprofibrate-fed POCA-fed

CAT 0.67+0.04 18.7+ 1.0*** 3.67 +0.74***

(n = 9) (n = 10) (n = 5)
Acetyl-CoA hydrolase 1.49 +0.04 3.56_0.1 1*** 2.53 + 0.07***
(n = 9) (n = 10) (n = 5)
plus 1 mM-ADP (1) 0.75 +0.04 1.09+0.03***
(n = 5) (n = 5)
plus 2 M-ATP (2) 3.60+0.10 8.38+0.16***
(n = 5) (n = 5)
(2)-(1) 2.85 +0.07 7.29_0.14***
Octanoyl-CoA hydrolase 1.32+0.09 5.87+0.73*** 2.53 + 0.06***
(n = 9) (n = 10) (n = 5)
Palmitoyl-CoA hydrolase 0.69+0.02 1.76+ 0.17*** 0.93 + 0.03***
(n = 9) (n = 10) (n = 5)
3-Hydroxybutyryl-CoA dehydrogenase 69.3 + 0.02 99.2+0.8***
(n = 5) (n = 5)
Glutamate dehydrogenase 135 + 7 103 + 10* 116+5
(n = 9) (n= 10) (n = S)
Citrate synthase 15.9+0.4 11.3 +0.7* 14.6 +0.8
(n = 9) (n = 10) (n = 5)
1988
Carnitine, CoA and peroxisomal proliferators 341
24 predominantly due to the cytoplasmic ATP-dependent
0
enzyme. The concentrations of acyl-CoA esters were
-
20 :0 increased by ciprofibrate and POCA (Table 1), despite
elevated acyl-CoA hydrolase activity. The activity of
0
B 16 hydroxyacyl-CoA dehydrogenase with acetoacetyl-CoA
en
C"
as substrate, which measures both the peroxisomal and
0 mitochondrial enzymes [23], was greater after ciprofibrate
-' 12 treatment. By contrast, the activities of glutamate
dehydrogenase and citrate synthase, which only occur in
0 8 the mitochondrial matrix, were lower (Table 4). Acyl-
0 CoA oxidase activities, determined by Dr. P. McNamee,
0 were increased 10-fold and 3-fold per g wet wt. after
4 ciprofibrate treatment and POCA treatment respectively
(results not shown), in agreement with earlier studies
[4,7].
0 200 400 600 800 1000 1200
[Total carnitine] (nmol/g wet wt.) Correlation between hepatic CAT and Icarnitinel
Fig. 1. Relation between hepatic carnitine content and CAT
The activity of hepatic CAT strongly correlated
activity
(r = 0.976+0.047; P < 0.00001) with [total carnitine] in
liver for all treatments (Fig. 1, Table 1). The corre-
Hepatic carnitine (free plus acylated) is expressed as nmol/ lation coefficients for CAT activity versus [carnitine]
g wet wt., and CAT activity is expressed as units/g wet wt. and [short-chain acyl-carnitine] were respectively
Data from control rats (A), POCA-fed rats (v), cipro- 0.971 +0.051 and 0.918+0.085 (P < 0.00001, n = 24).
fibrate-fed rats (0) and guinea pigs (0) are from the
experiments in Tables 1 and 4. Effects of ciprofibrate on enzyme activities in cultured
hepatocytes
Hepatic changes caused by administration of cipro-
carnitine esters (controls, 148 + 32, and ciprofibrate-fed, fibrate may be caused by direct effects of the drug or
318 + 40 nmol/24 h; means+ S.E.M., P < 0.01), but not secondary changes in concentrations of circulating
of free carnitine (controls, 23+3, and ciprofibrate-fed hormones or metabolites [24]. In monolayer-cultured
19 + 4 nmol/24 h). POCA had no significant effect on hepatocytes, ciprofibrate caused similar increases in
excretion of either free carnitine (35 + 8 nmol/24 h) or CAT and acyl-CoA hydrolase activities (Table 5) as
acylcarnitine (179 + 22 nmol/24 h) (Table 3). in vivo (Table 4), indicating that changes in hepatic
enzyme activities (Table 4) are caused directly.
Effects of ciprofibrate and POCA on enzyme activities
in liver DISCUSSION
The activity of CAT increased 28-fold and 6-fold Relation between mitochondrial and peroxisomal
(per g wet wt.) after ciprofibrate and POCA treatment /1-oxidation
respectively (Table 4), and the total liver activity increased
46-fold in livers from ciprofibrate-treated rats, after Chronic administration of fibrate drugs causes a
allowance for hypertrophy. The activities of acyl-CoA massive increase (up to 10-fold) in the number of liver
hydrolases assayed with acetyl-CoA, octanoyl-CoA and peroxisomes and the capacity for peroxisomal /-oxida-
palmitoyl-CoA as substrates were also increased by tion in rats [25,26]. The capacity for mitochondrial ,-
ciprofibrate and POCA treatment (Table 4); the increases oxidation is also increased 2-3-fold, so that the absolute
were greater with ciprofibrate than with POCA. In rat increase is comparable with that in the peroxisomes as
liver, acetyl-CoA hydrolases occur in the cytoplasm and assessed by the capacity for acetyl-CoA formation [6,27].
the mitochondrial matrix [20]. The increased acetyl-CoA [CoA] [6,8], [carnitine] [9,10] and the fatty-acid-binding
hydrolase activity in livers from ciprofibrate-fed rats was Z-protein are increased [28-30]. Fibrate drugs may cause

Table 5. Enzyme activities in hepatocytes cultured with ciprofibrate for 4 days

Hepatocytes isolated from rats maintained on standard diet were cultured for 4 days with the concentrations of ciprofibrate
indicated. The medium was changed daily, and enzyme activities are expressed as munits/mg of protein, as means + S.E.M. Of
results from duplicate flasks from four cultures. Significance of differences from controls: **P < 0.01; ***P < 0.005.

Activity
Enzyme [Ciprofibrate] (ng/ml) ... 0 10 100 1000

CAT 1.02+0.27 1.28+0.30 3.26 + 0.47** 14.1 + 0.47***

Acetyl-CoA hydrolase 2.65 + 0.04 2.79+0.09 4.68+0.25*** 8.59 + 0.56***
Octanoyl-CoA hydrolase 4.9+1.6 5.4+1.7 7.1+ 1.8 15.0+2.5**
Palmitoyl-CoA hydrolase 2.4+1.0 2.6+1.2 2.6+ 1.0 4.7+1.4
Glutamate dehydrogenase 1006 + 294 1035 +748 963 + 243 859 + 351
Vol. 253
342 A. K. M. J. Bhuiyan and others

co-ordinated changes in rodent liver, which increase the At equilibrium, therefore, the ratios of the reactants are
capacity for fatty acid oxidation and which may be an independent of their absolute concentrations. It has been
exaggerated physiological response to the ingestion of proposed that a reaction is at near-equilibrium if the
some natural very-long-chain fatty acids such as erucate mass-action ratio does not differ by more than a factor of
(C22: 1) [31,32]. POCA causes a limited (about 3-fold) 5 from the equilibrium constant [41]. The apparent mass-
increase in peroxisomes and in peroxisomal ,-oxidation, action ratios for the reactions catalysed by CAT (and by
similar to that caused by erucate [7]. POCA (as POCA- other carnitine acyltransferases acting on short-chain
CoA), however, is a very potent inhibitor of CPT I [5] esters) were in the range 1.0-2.2 (Table 1), suggesting
and hence of mitochondrial oxidation of long-chain fatty that these were not too far displaced from equilibrium
acids; and induction of peroxisomes may be due to in vivo (allowing for any zonal differences [42]) (Table 1).
POCA/POCA-CoA itself, or because of this inhibition. This explains the high [short-chain acylcarnitine] in liver
Curiously, simultaneous administration of bezafibrate after administration of ciprofibrate. Since [carnitine] is
and POCA to fasted, but not fed, rats limits peroxisomal greater than [CoA], the [CoA]/[carnitine] ratio is low
proliferation to about the extent caused by POCA alone when the [acylcarnitine]/[acyl-CoA] ratio is high at near-
[33]. However, the mechanisms by which ciprofibrate equilibrium. However, even in control animals the
and POCA produce their effects, and why these differ reactions catalysed by CAT appear to be near equili-
quantitatively, are unknown. brium. Tissue [CoA], [carnitine], [short-chain acyl-CoA]
and [short-chain acylcarnitine] (Table 1) are in the range
Effect of ciprofibrate and POCA on hepatic enzyme of, or greater, than the Km values for pigeon breast-
activities muscle CAT: CoA, 37-60,M; short-chain acyl-CoA,
Administration of ciprofibrate or of POCA for only 5 11-44 /sM; carnitine, 100-1 74 #m ; and short-chain acyl-
days caused large increases in the activities of CAT and carnitine, 350-455 fM [40]. By contrast, the mass-action
of acyl-CoA hydrolases in rat liver. Ciprofibrate caused ratio for the reaction catalysed by CPT was strongly
larger changes than POCA (Table 4). Maximum increases displaced from equilibrium (Table 1), and was not
occur after administration of ciprofibrate or POCA for apparently changed by POCA, even though POCA-CoA
10-14 days [34]. strongly inhibits CPT I [5]. This latter equilibrium may
be influenced by binding of long-chain acyl-CoA and
Role of carnitine as an acyl-group buffer carnitine esters to cell proteins.
Hepatic [carnitine] increases during the fed-to-fasted Mechanism of the increase in liver carnitine
transition in the rat [7,35]. Plasma [carnitine] and concentrations
[acylcarnitine] are also higher during fasting and diabetes
[36], suggesting increased export by the liver. Increased Administration of ciprofibrate increases liver [total
tissue [carnitine] increases the buffering capacity for acyl carnitine] from 0.23 to about 1 ,umol/g wet wt. However
groups and may facilitate maintenance of the [CoA]/ these values are only about 2.4% and 18 % of the
[carnitine] ratio within limits compatible with normal estimated skeletal-muscle [total carnitine] respectively
metabolism [37]. (Table 3). Ciprofibrate increased the proportion of
Both ciprofibrate and POCA caused a similar increase carnitine that was acylated in liver and plasma, but
(2-3-fold) in total liver [CoA] (Table 1). However, decreased it in muscle (Table 2). Plasma [total carnitine]
ciprofibrate caused a much larger increase in total liver was lowered by 46 % and by 35 % respectively, by
[carnitine] than did POCA (Table 1). Both drugs administration of ciprofibrate or POCA (Table 2). This
increased the activity of CAT, and there was a highly decrease cannot be explained by increased urinary
significant correlation between CAT activity and [total excretion or uptake by muscle (Table 3). Increased
carnitine] and [acylcarnitine] for all conditions (Fig. 1), hepatic [carnitine] may be due to greater retention.
but not between CAT and [CoA]. This correlation Muscle [total carnitine] was lowered by 15 % by
also holds for guinea-pig liver, which has higher CAT ciprofibrate. The estimated sum of [total carnitine] in
activity [38] and [total carnitine] than control rat liver muscle, liver and the extracellular fluid (Table 3) was
(Table 1). the same in all three groups. Although this may be a
The mitochondrial inner membrane is impermeable to coincidence, it suggests redistribution of carnitine
CoA and acyl-CoA esters, with separate pools of CoA between tissues after administration of ciprofibrate.
plus acyl-CoA esters in the mitochondrial matrix and There was no apparent simple relation between plasma
other compartments. Measured [CoA], [carnitine], [acyl- and tissue [carnitine] and [acylcarnitine]. Ciprofibrate
CoA] and [acyl-carnitine] are the sums of these (Tables 1 increased the amount of carnitine excreted in the urine;
and 2). However, the mitochondrial and cytosolic this was only a tiny fraction (0.2-0.5 %) of the total tissue
compartments are connected by a carnitine/acylcarnitine content (Table 3).
exchange carrier in the mitochondrial inner membrane Carnitine is synthesized from methionine and lysine,
[3] (assumed to have a low control strength), so that at which are converted into y-butyrobetaine in muscle [1].
equilibrium the ratios of [acylcarnitine]/[carnitine] will In the rat only the liver converts y-butyrobetaine into L-
be equal, and hence the [CoA]/[acyl-CoA] ratios will also carnitine, which is then exported to other tissues [1]. Our
be equal. Further, it has been suggested that peroxisomes results differ from those of a previous study on rats given
are permeable to small water-soluble molecules [39], so 30 mg of clofibrate orally for 14-21 days, in which no
that the acylation status of the major pools of CoA and decreases in muscle or plasma [carnitine] were found [43].
carnitine in the hepatocyte may be related. The reactions However, it is difficult to compare results obtained with
catalysed by the carnitine acyltransferases have two different compounds and dose schedules.
substrates and two products, so that the equilibrium There was a strong correlation between [long-chain
constant, defined as [carnitine][acyl-CoA]/[CoA][acyl- acyl-CoA] and the activity of palmitoyl-CoA hydrolase
carnitine], which has a value of 0.6 [40], is dimensionless. in livers of rats treated with various peroxisomal
1988
Carnitine, CoA and peroxisomal proliferators 343

proliferators [44]. The correlations between liver [carni- 15. Koundakjian, P. P., Turnbull, D. M., Bone, A. J., Rogers,
tine] and CAT activity (Fig. 1) may be analogous. M. P., Younan, S. I. M. & Sherratt, H. S. A. (1984)
However, ciprofibrate increased CAT activity in hepato- Biochem. Pharmacol. 33, 465-473
cytes cultured in the absence of the carnitine precursor y- 16. Seglen, P. 0. (1976) Methods Cell Biol. 13, 29-83
butyrobetaine (Table 5). Hepatocyte [carnitine] declines 17. McGarry, J. D. & Foster, D. W. (1976) J. Lipid Res. 17,
in culture [14], in both the presence and the absence of 277-281
ciprofibrate (results not shown), presumably because of 18. Bergmeyer, H. U. (1974) Methods of Enzymatic Analysis,
lack of precursor. Therefore, ciprofibrate may cause pp. 650-655 (Bergmeyer, H. U., ed.), Academic Press,
parallel rather than interdependent changes in [carnitine] New York
and CAT activity. 19. Brdiczka, D., Pette, D., Brunner, G. & Miller, F. (1968)
Eur. J. Biochem. 5, 294-304
Conclusions 20. Agius, L., Wright, P. D. & Alberti, K. G. M. M. (1987)
Clin. Sci. 73, 3-10
Three conclusions are drawn from this study: first, 21. S6ling, H.-D. & Rescher, C. (1985) Eur. J. Biochem. 147,
that the reactions catalysed by the short-chain carnitine 111-117
acyltransferases, but not by the carnitine palmitoyltrans- 22. Osumi, T. & Hashimoto, T. (1978) Biochem. Biophys. Res.
ferases, are not too far displaced from equilibrium in rat Commun. 83, 479-485
liver both before and after modification of liver 23. Hashimoto, T. (1982) Ann. N.Y. Acad. Sci. 386, 5-12
metabolism by administration of ciprofibrate and 24. Paul, H. S. & Adibi, S. A. (1979) J. Clin. Invest. 64,
POCA; second, that retention of carnitine by the liver, 405-412
and hence redistribution of carnitine between tissues, 25. Svoboda, D. J. & Azarnoff, D. L. (1966) J. Cell Biol. 30,
may be an important factor in the increase in hepatic 442-450
carnitine caused by ciprofibrate and by POCA; third, 26. Lazarow, P. B. & de Duve, C. (1976) Proc. Natl. Acad. Sci.
that hepatic CAT activity and [carnitine] may be closely U.S.A. 73, 2043-2046
related. 27. Van Hoof, F., Hue, L., Vamecq, J. & Sherratt, H. S. A.
(1985) Biochem. J. 229, 387-397
A. K. M. J. B. is a Commonwealth Scholar. We are grateful to 28. Renaud, G., Foliat, A. & Infante, R. (1978) Biochem.
Professor K. G. M. M. Alberti and Professor A. Aynsley-Green Biophys. Res. Commun. 80, 327-334
for their support. 29. Fleischner, G., Meyer, D. K. F., Levine, W. G., Gatmaitan,
Z., Gluck, R. & Arias, I. M. (1965) Biochem. Biophys. Res.
Commun. 67, 1401-1407
REFERENCES 30. Foliot, A., Drocourt, J.-L., Etienne, J.-P., Housset, E.,
Fiessenger, J.-N. & Christoroforov, B. (1977) Biochem.
1. Bremer, J. (1983) Physiol. Rev. 63, 1420-1480 Pharmacol. 26, 547-549
2. Bremer, J. & Osmundsen, H. (1984) in Fatty Acid 31. Neat, C. E., Thomassen, M. S. & Osmundsen, H. (1980)
Metabolism and its Regulation (Nunes, S., ed.), pp. Biochem. J. 186, 369-371
113-154, Elsevier, Amsterdam 32. Neat, C. E., Thomassen, M. S. & Osmundsen, H. (1981)
3. Schultz, H. (1985) in Biochemistry of Lipids and Mem- Biochem. J. 196, 149-159
branes (Vance, D. E. & Vance, J. E., eds.), pp. 116-142, 33. Hertz, R. & Bar-Tana, J. (1987) Biochem. J. 245, 387-392
Benjamin Cummings Publishing Co., Menlo Park, CA 34. Reddy, J. K., Warren, J. R., Reddy, M. K. & Lalwani,
4. Reddy, J. K. & Lalwani, N. D. (1983) CRC Crit. Rev. N. D. (1982) Ann. N.Y. Acad. Sci. 386, 81-108
Toxicol. 12, 1-58 35. Brass, E. P. & Hoppel, C. E. (1978) J. Biol. Chem. 253,
5. Turnbull, D. M., Bartlett, K., Younan, S. I. M. & Sherratt, 2688-2693
H. S. A. (1984) Biochem. Pharmacol. 33, 475-481 36. Frohlich, J., Seccombe, D. W., Hahn, P. H. & Hynie, I.
6. Mannaerts, G. P., Thomas, J., Debeer, L. J., McGarry, (1978) Metab. Clin. Exp. 27, 555-561
J. D. & Foster, D. W. (1978) Biochim. Biophys. Acta 529, 37. Sherratt, H. S. A. (1985) Biochem. Soc. Trans. 13, 1985-
201-211 1987
7. Bone, A. J., Sherratt, H. S. A., Turnbull, D. M. & Osmund- 38. Mannaerts, G. P. & Van Veldhoven, P. P. (1987) in
sen, H. (1982) Biochem. Biophys. Res. Commun. 104, Peroxisomes in Biology and Medicine (Fahimi, H. D. &
708-712 Sies, H., eds.), pp. 169-176, Springer Verlag, Berlin
8. Pande, S. V. & Parvin, R. (1980) Biochim. Biophys. Acta 39. Sauer, F. & Erfle, J. D. (1966) Fed. Proc. Fed. Am. Soc.
617, 363-370 Exp. Biol. 25, 769
9. Ball, M. R., Gumaa, K. A. & McLean, P. (1979) Biochem. 40. Chase, J. F. A. (1967) Biochem. J. 104, 510-518
Biophys. Res. Commun. 87, 489-496 41. Newsholme, E. A., Crabtree, B. & Zammit, V. A. (1980)
10. Voltti, H., Savolainen, M. J., Jauhonen, V. P. & Hassinen, Trends in Enzyme Histochemistry and Cytochemistry:
I. E. (1979) Biochem. J. 182, 95-102 CIBA Found. Symp. 73 (New Ser.), 245-258
11. Halvorsen, 0. (1963) Biochem. Pharmacol. 32, 1128-1131 42. Jungermann, K. & Katz, N. (1982) in Metabolic Compart-
12. Skrede, S. & Halvorsen, 0. (1979) Eur. J. Biochem. 98, mentation (Sies, H., ed.), pp. 411-435, Academic Press,
223-229 New York
13. Arnold, A., McAulitt, J. P., Powers, L. G., Philipis, D. K. 43. Paul, H. S., Gleditsch, C. E. & Adibi, S. A. (1986) Am. J.
& Beyler, A. L. (1979) Atherosclerosis 32, 155-163 Physiol. 251, E311-E315
14. Agius, L., Chowdhury, M. H. & Alberti, K. G. M. M. 44. Berge, R. K. & Arsland, A. (1985) Biochim. Biophys. Acta
(1986) Biochem. J. 239, 593-601 837, 141-151

Received 7 December 1987/23 February 1988; accepted 15 March 1988

Vol. 253