MSU

LIBRARIES
.——.

~

 

 

RETURNING MATERIALS:
P1ace in book drop to
remove this checkout from
your record. FINES wi11
be charged if book is
returned after the date
stamped below.

 

 

 

 

 

 

 

 

 

 

A STUDY OF THE KINETICS, SPECIFICITY,
AND REGULATION OF HEART MITOCHONDRIAL
CARNITINE PALMITOYLTRANSFERASE

BY
Carol J. Biol

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

DEPARTMENT OF BIOCHEMISTRY

1986

.. we?

<1:

ABSTRACT
A STUDY OF THE KINETICS, SPECIFICITY,

AND REGULATION OF HEART MITOCHONDRIAL
CARNITINE PALMITOYLTRANSFERASE

BY
Carol J. Fiol

A study of the kinetic properties of heart
mitochondrial carnitine palmitoyltransferase (CPT) was
undertaken using both CPT purified to apparent homogeneity
from beef heart and membrane bound CPT-I (outer form of CPT)
from rat heart. Purified CPT is an aggregate of molecular
weight 660,000 by Fractogel TSK chromatography with a
subunit molecular weight of 67,000 by SDS PAGE. This
aggregate contains 19 moles of phospholipid per mole of
enzyme which are primarily cardiolipin, phosphatidylcholine
and phosphatidylethanolamine. Purified beef heart CPT has
sigmoidal kinetics with both acyl-CoA and L-carnitine. It
has higher affinity for L-carnitine in the presence of
long-chain acyl-CoA derivatives, but it has the highest
absolute catalytic rate with hexanoyl-CoA. Its catalytic
activity is strongly pH dependent with a pH optima of 7.

The K for palmitoyl-CoA is 1.9 pm and 24.2 uM at pH 8 and

0.5

6, respectively. The K for L-carnitine is 0.2 mM and 2.9

0.5
mM at pH 8 and 6, respectively. Malonyl-CoA (20-600 uM) has

no effect on the kinetics of purified CPT with

palmitoyl-CoA. In contrast, TDGA-CoA increased the K for

0.5
palmitoyl-CoA and reduced the Hill coefficient.

Carol J. Fiol
When octylglucoside is substituted for Triton X—100,
the specificity of purified beef heart CPT in the forward
direction shifts towards the long-chain acyl-CoAs and large
changes in the kinetic constants are observed. At pH 8.0
and 200 pM palmitoyl-CoA, the K0.5 for L-carnitine is 4.9 mM
in 12 mM octylglucoside compared to 0.2 mM in 0.1% Triton

X-100. At pH 6.0, the K for palmitoyl-CoA is 24.2 pH in

0.5
0.1% Triton X-100, compared to 3.1 uM in 12 mM
octylglucoside. Octylglucoside is an apparent competitive
inhibitor of CPT reaction with octanoyl-CoA with a Ki of 15
mM.

Membrane bound CPT-I from rat heart mitochondria shows
substrate cooperativity of similar magnitude to that
exhibited by the purified enzyme from beef heart
mitochondria. The K0.5 for decanoyl-CoA is 3 uM with
mitochondria from both fed and fasted rats. Addition of
malonyl—CoA increased the K0.5 for decanoyl-CoA with no
apparent increase in sigmoidicity or vmax' With 20 pH
malonyl-CoA, the K .

0 5
coefficient is 2.1. CPT-I from fed rats had an apparent Ki

for decanoyl-CoA is 185 pM and the Hill

for malonyl-CoA of 0.3 uM while that from fasted rats was 2.5
pH. The kinetics with L—carnitine were variable: the K0.5
ranged from 0.2 mM to 0.7 mM and the Hill coefficient varied
from 1.2 to 1.8.

These data are consistent with the conclusion that

native CPT exhibits different catalytic properties on either

side of the inner membrane of mitochondria due to its

Carol J. Fiol
non-Michaelis-Menten kinetic behavior, which can be affected
by pH differences and differences in membrane environment.
The data suggests that the physiological levels of
L-carnitine may a be determing factor in the specificity of

CPT i vivo. It is concluded that malonyl-CoA is not a

 

competitive inhibitor of CPT but acts like a negative

allosteric modifier of CPT-I.

 

A papi y mami

ii

ACKNOWLEDGEMENTS

The support of the Department of Biochemistry, its
faculty and students, is gratefully acknowledged. The time
and interest of the members of my thesis guidance
committee--Drs. William C. Deal, Shelagh Ferguson—Miller,
William Wells, and Dale Romsos--in following this research
are greatly appreciated. Special thanks go to Professor
Loran Bieber, my Doctoral Thesis Advisor, who has guided and
supported my work, and has made me the scientist that I
always wanted to be. I wish to thank the other members of
this research group, Kim Vaulkner, and Drs. Patrick
Sabourin, Ron Emaus, Janos Kerner, Wieslava Lysiak, and
Shawn Farrell, who have helped me in many ways to complete
this research and are my friends. I would also like to thank
Dr. Dwight Needles, Nicholas Ringo and Peter Toth for
sharing their knowledge of the systems used in this
research.

I am grateful to my dearest friends, Arlyn, Maria,
Shawn, John, Carmen, Mildred, and my husband, Bob, for their

great camaraderie in this venture.

iii

TABLE OF CONTENTS

Page

LIST OF FIGURES ................ . .................... .. vi
LIST OF TABLES ............................ ...... ...... viii
LIST OF ABBREVIATIONS. ......... .... ...... .. .......... ix
INTRODUCTION .......................................... 1
CARNITINE PALMITOYLTRANSFERASE ........................ 5
Assay Methods .............. . ...................... 5
Intracellular Distribution. .......... . ....... ..... 8
Mitochondrial Distribution ........................ 10

CPT Deficiency...... ..... .. .......... ..... ........ 13
Purification of OPT ............................. .. 15
Kinetics and Specificity...... ...... .............. 20
Properties of Membrane Bound CPT..... ............. 25
Role of OPT in Control of Fatty Acid Oxidation.... 30
EXPERIMENTAL PROCEDURES ........ . ..... . ..... .. ..... .... 35
Material .............. . ..................... ...... 35
Methods. ....................................... 36
Purification of Beef Heart Mitochondrial CPT.. 36
Mitochondrial Isolation... ... ........ 36
Solubilization of Mitochondrial CPT ......... 37
Chromatography of Solubilized CPT.. ......... 38

Exchange of Triton x—1oo for Octylglucoside. 38

Physical Characterization of Purified CPT...... 39

SDS-Polyacrylamide Gel Electrophoresis ...... 39
Phospholipid Analyses ....................... 39
Native Molecular Weight Determination ....... 40
Kinetic Characterization of Purified CPT ....... 40
Activity Assay ............................... 40

Kinetic Measurements with purified CPT...... 40

iv

Kinetic Characterization of Membrane Bound Rat

Heart CPT-I .................................... 42

Other Methods... ....... . ........................ 43

CMC Determinations ................... ....... 43

Protein Determinations ....................... 43

Isolation of Rat Heart Mitochondria .......... 43

RESULTS ..................... . .......................... 46

Purification of Carnitine Palmitoyltransferase
from Beef Heart Mitochondria... ..... .......... ..... 46

Physical Characterization of Purified Carnitine
Palmitoyltransferase ....... .......... ..... ......... 52

Kinetic Characterization of Purified Carnitine

Palmitoyltransferase ........... ............ ........ 55
Preliminary Tests ............................ 55
Determination of Nonmicellar Assay
Conditions............ ......... ..... ......... 59
Nonlinear Kinetics of Purified CPT ........... 63
Substrate Specificity ..... . ...... ......... 67
Effect of pH on the Kinetics of CPT .......... 76

Effect of Malonyl- -CoA and TDGA-CoA on CPT.. . 76

Kinetic Characterization of Membrane Bound Rat

Heart Mitochondrial CPT- I. ................... 84
Effect of Acyl-CoA on Mitochondrial Swelling 84
Kinetics of CPT- I and Effect of Malonyl- -CoA. 88
Effect of Feeding and Fasting on the

Ki(aPP) for Malonyl-CoA ..... ... ..... ....... 88
Kinetics of CPT-I with L-carnitine ........... 93
DISCUSSION ........... ... ............................... 96
SUMMARY AND CONCLUSIONS ................................ 111
LIST OF REFERENCES ........................... .......... 113

LIST OF FIGURES

FIGURE Page
1. Current view of the role of CPT in fatty acid

oxidation........ ....... ... ........................ 3
2. Schematic representation of the semi-automated

kinetic analyzer................OOOOOOOOOOOOOO.COOO ‘1
3. Flowchart for the added subroutine to the computer

programTANKINOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO “
4. Separation of solubilized CPT and CAT on

Sephacryl S-300................. ................... 49
5. SDS polyacrylamide gel electrophoresis of

purified beef heart mitochondrial......... ........ . 51
6. Two dimensional thin layer chromatography of

extracted phospholipids from purified CPT .......... 54
7. Fractogel TSK HW-55 chromatography of

purified CPT....OOOOOIOOOOIOOOOOO ........... O 000000 56
8. Effect of octylglucoside on CPT activity... ........ 57
9. Time course of OPT reaction with DTNB assay........ 58
10. Effect of end point deletion on apparent Hill n.... 60
11. Effect of product accumulation during a substrate

additionexperimentOOIO...OOOOOOOOOIOOOOOOOOOOO..00 61
12. The relationship between the carbon chain-length

of the acyl-CoAs and their one ............. . ....... 62
13. Determination of acyl-CoA micelle formation in

octylglucoside ..................................... 64
14. Double reciprocal plots of CPT reaction velocity

versus acyl—CoA concentration... ......... .. ........ 65
15. Double reciprocal plots of CPT reaction velocity

versus L-carnitine concentration.... ............... 66
16. Double reciprocal plot of COT reaction velocity

versus acyl-CoA concentration ...................... 68

vi

17.

18.

19.

20.

21.

22.

23a

24.

25.

26.

27.

Specificity of OPT for acyl-CoAs of varying
Chain-length ooooo soeeeoeeeeeeeeeeeeeeoeeee eeeeeeeee 69

Relationship between the chain—length of the

acyl-CoA and the KO 5 of CPT for L-carnitine ........ '72
Effect of octylglucoside on the K of CPT for

0.5
octanoyl-CoA ........................................ 78
Effect of pH on the velocity versus substrate
concentration curves for CPT—I ...................... 79
Effect of TDGA—CoA on CPT-I............ ............. 85

Effect of decanoyl-CoA and palmitoyl-CoA on
mitochondrial swelling .............................. 87

Double reciprocal plot of mitochondrial CPT
reaction velocity versus decanoyl-CoA

concentration ....................... ....... ......... 89
Effect of malonyl-CoA on membrane bound CPT ......... 90
Effect of feeding and fasting on malonyl-CoA

inhibition of membrane bound CPT-I ...... .. .......... 92
Replot of KO.5 versus malonyl—CoA concentration ..... 94

Membrane bound CPT-I reaction velocity versus
L-carnitine concentration curve..................... 95

vii

TABLE

II.

III.

IV.

VI.

VII.

VIII.

IX.

XI.

XII.

LIST OF TABLES

Summary of the properties of purified
carnitine palmitoyltransferases ...................

Acyl-Group specificity of
carnitine palmitoyltransferases...................

Summary of the purification of CPT from beef heart
mitochondria.... ..................................

K and V values of OPT for L—carnitine at a
constant concentration of acyl-CoAs of varying
carbon chain-length ...............................

values of COT and CAT for L-carnitine at a
constant concentration of acyl-CoAs of varying
carbon chain-length ........ .............. .........

Kinetic parameters of CPT for acyl-CoAs of
varying chain-length at two fixed concentrations
at L-Carnitine ooooooooooooooooo eoeeoeooeeeeeeeeeee

Effect of octylglucoside on the kinetic parameters
of OPT for palmitoyl-CoA........ ........ ..........

Effect of pH on the kinetic parameters of CPT
in Triton x-100....... ..... .......................

Effect of pH on the kinetics of CPT with
L-carnitine in octylglucoside........... ..........

Effect of malonyl-CoA on the kinetics of purified
CPT with palmitoyl-CoA.. ..... .....................

Effect of TDGA-CoA on the kinetics of purified
CPTWithpalmitOYI‘COA eeeeeeeeee eeeeeeeeeeeeeeeeee

Effect of feeding and fasting on malonyl-CoA
inhibition of membrane bound CPT-I ................

viii

Page

17

21

48

70

73

74

77

BO

81

83

86

91

LIST OF ABBREVIATIONS

Bis Tris- 1,3-bis(tris[hydroxymethyl]-methylamino)-

BSA

CAT

CMC

CoA

COASH

COT

CPT

CPT-I

CPT-II

DTBP

DTNB

EDTA

EGTA

HEPES

Hill n

50

7:

0.5

Bovine serum albumin

Carnitine acetyltransferase

Critical micellar concentration

Coenzyme A

Reduced Coenzyme A

Carnitine octanoyltransferase

Carnitine palmitoyltransferase

Outer form of carnitine palmitoyltransferase
Inner form of carnitine palmitoyltransferase
4,4'-dithiobispyridine
5,5'-dithiobis-(2-nitrobenzoic acid)
(Ethylenedinitrilo)-tetra-acetic acid
Ethyleneglycol-bis-( -amino—ethyl ether)N,N'—
tetra-acetic acid
N-2-hydroxyethylpiperazine-N'—2-ethanesulfonic
acid

Hill coefficient

the inhibitor concentration that causes 50%
inhibition under a specific set of assay
conditions

Inhibition constant

Michaelis constant

Hill constant

ix

QAE

SDS-PAGE

TANKIN

T

DGA

Tris

V

max

Diethyl-(2-hydroxypropyl)aminoethyl-
Sodium dodecyl sulfate polyacrylamide
electrophoresis

Tangent slope kinetic analysis
2-tetradecyl-glycidic acid
tris-(hydroxymethyl)aminomethane

Maximum velocity

INTRODUCTION

Carnitine (gamma—trimethylamino-beta-hydroxybutyrate),
discovered as a component of muscle tissue in 1905 (1), is
ubiquitous in the animal kingdom. Most species are capable
of its synthesis, with one notable exception, the larvae of
Tenebrio molitor (2). Its four carbon chain originates from
lysine (3) and its methyl groups from methionine (4). The
concentration of carnitine varies between tissues and
amongst species (5). Heart and skeletal muscle normally
have a concentration of a few millimoles per liter, while
adipose tissue and liver have lower concentrations. The
initial observations of Fritz (6) on the stimulation of
fatty acid oxidation by L-carnitine in liver, and those of
Bremer (7) on the mitochondrial metabolism of
palmitoylcarnitine led to the proposal of a role for
carnitine in the oxidation of palmitoyl-CoA. The discovery
of carnitine palmitoyltransferase (CPT) (8,9) established
carnitine as the carrier of activated fatty acids through
the CoA impermeable barrier of the inner mitochondria. The
identification and isolation of other carnitine
acyltransferases with specificity for short and medium-chain
acyl residues and their multiple organelle distribution

demonstrate other roles for carnitine (10).

Carnitine acyltransferases are a class of enzymes that

catalyze the following reversible reaction:

L-(-)-carnitine + acyl—CoA‘eacyl-L-(-)-carnitine + CoASH.

These enzymes are classified according to their acyl
substrate specificity into short—, medium- and long-chain
acyltransferases. Carnitine acetyltransferase (CAT) is the
predominant acyltransferase in most tissues (11). It is
found in mitochondria and peroxisomes (12) where it appears
to facilitate the shuttle of acetyl units generated from
B-oxidation into the cytosol, thereby maintaining a free
CoASH pool for continuing reactions and possibly providing
acetyl units for use in synthesis elsewhere in the cell
(13). CAT is also associated with the endoplasmic reticulum
in mammalian species where it has been proposed to have a
role in synthesis (10). Carnitine octanoyltransferase (COT)
has been isolated from liver peroxisomes (14,15) where it
forms medium-chain acylcarnitine esters from the chain
shortened acyl-CoAs produced by peroxisomal B-oxidation of
very long-chain acyl—CoAs. COT also exists in mammalian
microsomes (16), whereas CPT is believed to be primarily a
mitochondrial enzyme (17).

The current view of the role of CPT in mitochondrial
fatty acid oxidation is illustrated in Figure 1. Two forms

of OPT (18) are present on the inner mitochondrial membrane.

453.8% BUM 3&8 3%? 5.. g «0 0H9” Eu. no 302w Hg . H charm

{\J

 

 

 

 

 

 

\<oo-;co_oE
mogom \\nl
2..on ~00 052500 V 'flw
./ $486245 x
/ \ A A 05258.34
«8.584 H PW 1960
H .50
cozoExo - Q Imdoo
965522 <8- :94
128-354 6:323:5—
35: 8558
x322
I/.\ .823

The outer form of OPT (CPT-I) is exposed to the cytosolic
components and catalyzes the formation of acylcarnitine
esters from cytosolic acyl-CoAs. The acylcarnitines are
transported into the matrix of the mitochondria by a
translocase system, in a one to one carnitine/acylcarnitine
exchange (19). There they become the substrates of the
inner form of CPT (CPT-II) which regenerates the acyl-CoAs
in the mitochondrial matrix. Current evidence (20-22)
strongly supports the proposal that one of the major
mechanisms for control of fatty acid oxidation and
ketogenesis is by regulation of the activity of CPT-I
through its inhibition by malonyl-CoA. The most significant
difference between the inner and outer forms of CPT appears
to be the sensitivity of CPT-I to inhibition by malonyl-CoA
and the relative insensitivity of CPT—II (23). Though these
differences have suggested that the two forms of OPT are
distinct enzymes (18,23), when CPT is solubilized from the
mitochondrial membrane it loses its capacity to interact
with malonyl-CoA (23) and CPT—I and CPT-II become
indistinguishable (24). Thus it appears that the membrane
environment plays a pivotal role in the regulation of OPT

activity (24,5).

 

CARNITINE PALMITOYLTRANSFERASE

Assay Methods

 

Two general types of assays have been used to measure
the activity of carnitine acyltransferases: one, the
continuous assays which measure directly the formation or
disappearance of acyl-CoA derivatives at 232 nm (25,26), or
indirectly measure the release of CoASH by reaction with
sulfhydryl reagents such as DTNB (27) or DTBP (28); and two,
discontinuous (end point) assays for which three general
approaches have been used, measurement of the formation of a
specific radioactive acylcarnitine (29,30), measurement of
the formation of acylhydroxamates (31), and measurement of
the exchange of radioactive carnitine into the acylcarnitine
fraction (25). Assays measuring the formation of
acylcarnitines or release of CoASH are known as forward
assays and those which measure the formation of acyl-CoA or
free carnitine are known as reverse assays. In addition, it
is possible to estimate CPT activity in mitochondria with
functional assays measuring flavoprotein reduction (32) or
oxygen consumption (33) during the oxidation of fatty acids,
acyl-CoAs, or acylcarnitines.

None of these methods is ideal and the best choice of

 

method will depend on the problem and purity of the system
studied. When measuring acyltransferase activity in impure
preparations or tissue homogenates with assays that measure
the formation or disappeareance of acyl-CoA or depend on
maintenance of saturating amounts of acyl-CoAs, acyl-CoA
hydrolase activity must be quantitated. This is difficult
with exchange assays or assays which measure acyl-CoA
formation. In addition, tissues such as liver contain a
family of carnitine acyltransferases, with overlapping
substrate specificity, thus the activity measured in
homogenates may represent the contribution of more than one
enzyme.

Isotopic exchange methods, which are run at or near
equilibrium, are inappropriate for kinetic measurements.
The isotope forward assay is very sensitive and is
frequently used in kinetic studies (20,34) but special care
must be taken to ensure that initial rate conditions are
maintained throughout the incubation period. This procedure
can be tedious and is often times omitted. In some studies
the use of long incubation times at low substrate
concentrations suggest that the reaction has gone to near
completion (35). Continuous spectrophotometric assays,
though less sensitive than the isotope forward assay, are
more convenient for kinetic measurements since changes in
enzyme activity with time can be monitored to ensure that a

true initial rate of reaction is being measured. Due to a

high background absorbance in homogenates, the 232 assay is
used only with enzymes of a high degree of purity. With the
232 assay the product formation is measured directly with no
further manipulation. However the low extinction
coefficient of the acyl-CoAs at 232 nm requires higher
product accumulation, therefore there is an increased
possibility of product inhibition. In addition, the
reversibility of the reaction is not prevented and may
affect the initial rates. The DTNB assay can be used to
measure CPT activity in homogenates as well as pure
preparations, allows for easy determination of hydrolase
activity, and the reaction is essentially irreversible by
the removal of free CoASH. However, high concentrations of
DTNB (greater than 0.25 M) have been reported to abolish
both the sensitivity of CPT-I to malonyl-CoA and the
apparent sigmoidal kinetics of CPT—I with palmitoyl-CoA
(36). From these studies and other studies (37) on the
inhibitory effects of disulfides, the presence of reactive
thiol groups in the enzyme essential for catalysis has been
proposed.

In order to measure the inhibitory action of
malonyl-CoA on CPT-I activity, as well as to measure CPT-I
independent of CPT-II activity, membrane integrity must be
maintained. Albumin is commonly used in the assays to
minimize the detergent effects of the long-chain acyl-CoA

substrates on the mitochondrial membrane. McCormick and

Notar-Francesco (38) emphasized the need for precise and
standard assay conditions by demonstrating the striking
dependence of the K0.5 for palmitoyl-CoA on albumin
concentration. They also demonstrated that varying assay
conditions can modify substrate binding to albumin thus
changing the effective concentration of the acyl-00A
substrates. This results in an artifactual effect on CPT
activity. Zammit (39) has shown that lags in the time
course of acylcarnitine formation arise from the time
necessary to obtain equilibrium between free, micellar and
albumin-bound forms of the substrate.

Recently, Zierz and Engel (40) have reviewed the
clinical studies on CPT deficiencies. They have shown that
conflicting results on the nature of the disease are due, in
great part, to methodological problems in the CPT assay.
Their results demonstrate that patients can show severe CPT
deficiency with the isotope exchange assay but normal CPT

activities with the forward assay.

Intracellular Distribution

 

The proposed role for L-carnitine as an acylcarrier
across the GOA-impermeable inner mitochondrial membrane
requires that there be two locations for CPT in the cell.
Early studies on the intracellular localization (8,41,42)

showed that CPT was primarily associated with mitochondria,

but suggested an extramitochondrial location as well.

Later, a more careful investigation by Norum and Bremer
(43), using sucrose density fractionation techniques showed
that the low residual activity in the microsomal fractions
was a result of mitochondrial contamination. Subsequently,
evidence in favour of a dual localization of CPT in the
mitochondria itself was presented by Yates and Garland (44).
They showed that rat liver mitochondria contained an overt
CPT activity accessible to inhibition by 2-bromostearoyl—CoA
and a latent activity inaccessible to inhibition by
2-bromostearoyl-CoA, unless it was previously exposed by
ultrasonic disintegration of the mitochondria. Hoppel and
Tomec (45) used a digitonin fractionation technique to
separate the outer and inner mitochondrial membranes in
sucrose density gradients and showed that CPT was
exclusively located on the inner membrane of mitochondria.
Using phospholipase C to strip beef liver mitochondria of
the outer membrane, Brosnan et al. (46) demonstrated that
both the outer and inner surfaces of the mitochondrial inner
membrane contain CPT which is inhibited by antibodies to the
enzyme (47).

Though it is now generally agreed that CPT is an
exclusively mitochondrial enzyme, there have been further
reports of a microsomal location. Van T01 (48) measured CPT
activity in the rat liver microsomal fraction that could not

be attributed to mitochondrial contamination. In fasted

10

rats, the microsomal activity is increased to about 60% of
CPT activity in mitochondria. Fogle and Bieber (49) found
CPT activity associated with rat heart microsomes which had
been clearly separated from mitochondria by sucrose density
gradient centrifugation. However, the finding that rat
liver contains extramitochondrial COT associated with
peroxisomes and microsomes (50) now suggests that
extramitochondrial long-chain acyltransferase activity
results from the overlapping specificity of medium—chain
acyltransferases. The isolation of microsomal medium-chain
acyltransferases has been made difficult by the apparent
lability of the enzymatic activity (50), but peroxisomal COT
has been purified from both mouse (14) and rat liver (15).
Purified COT has low but significant activity with
long-chain acyl-CoAs and has been shown to be a distinct

enzyme from CPT (51).

Mitochondrial Distribution

 

Several different methods have been used to estimate
the proportion of CPT-I to CPT-II activity in mitochondria.
One common procedure is to measure the overt CPT activity
under non-swelling conditions, and then measure the total
CPT activity by disruption of the mitochondria by sonication
or detergent lysis (29,52-55). CPT-II is estimated from the

difference between total and overt activity. Most of these

11

studies estimate that 25-50% of total CPT activity is due to
CPT-I. Though both activities are usually measured with the
same assay, it is not clear what effect the sonication or
detergent have on the measurement of total activity since
the properties of the enzyme may have changed. Selective
inhibition of CPT-I with malonyl-CoA (20) or
bromostearoyl-CoA (44) has also been used to measure the
remaining CPT-II. These methods estimate CPT-I activity to
be 15%-50% of total CPT.

Another method uses digitonin to selectively solubilize
CPT-I. Hoppel and Tomec (45) showed that CPT activity could
be released with digitonin from the inner membrane without
releasing the matrix enzymes. This resulted in a
preparation of mitoplasts that had severely limited capacity
to oxidize palmitoyl-CoA + L-carnitine, but retained the
ability to oxidize palmitoyl-L-carnitine. From these data,
Hoppel and Tomec proposed that CPT released by digitonin was
CPT-I that was loosely associated with the external surface
of the inner membrane of mitochondria and the remaining CPT
activity was tightly membrane bound CPT-II, exposed to the
matrix side of mitochondria. They estimated that only 15%
of total CPT activity is due to CPT-I. Assuming that only
the outer transferase is released, this method is likely to
yield an underestimate of CPT-I activity if the membrane
begins to disintegrate before all of the outer CPT has been

removed. Hoppel and Tomec (45) used sonication after

 

12

digitonin treatment to ensure complete release of CPT-I and
increased the proportion of outer CPT measured from 15% to
25%. Recently, other investigators (20) have reported that
with increasing concentrations of digitonin significant
solubilization of CPT coincides closely with that of citrate
synthase, a matrix enzyme, and have challenged the validity
of this fractionation method.

Others (32,33,53,56) have used indirect methods
involving multienzymatic pathways in an attempt to measure
the "in situ" activity. Such methods are less reliable
since they do not yield a true measurement of the enzymatic
activity. Wood and Chang (53) compared the oxidation rate
of palmitoyl-CoA + L-carnitine with the rate of oxidation of
palmitoylcarnitine by intact mitochondria to estimate the
CPT-I/CPT-II ratio. They observed a 1:2 ratio which is in
agreement with their estimate obtained using detergent
lysis. Bergstrom and Reitz (32) estimated the CPT—I/CPT-II
ratio from measurements of flavoprotein reduction rates
during oxidation of acyl-CoA/acylcarnitine by mitochondria.
Their estimate of CPT-II activity exceeded that of CPT-I by
a factor of 450. They proposed that "in situ" factors
regulate the activity of CPT-I because of the difference in
CPT I/CPT II ratio obtained with the "in situ" versus the
direct digitonin extraction. A more direct approach to
measure the "in situ" activity of CPT-II using inverted

submitochondrial particles estimates that CPT-I is

13

approximately 30% of total CPT activity (57).

Tissue origin, as well as physiological and
developmental factors, seem to affect the CPT-I/CPT-II
ratio. The proportion of latent CPT appears to be higher in
adipocyte and mammary gland than in liver and kidney, while
heart and skeletal muscle may have the largest proportion of
overt activity (58). Bovine fetal heart mitochondria
oxidize palmitoyl-CoA at 20-30% of the rate of calf heart
mitochondria, but can oxidize palmitoyl-L—carnitine at
similar rates (59). In addition, the proportion of CPT-I to
CPT-II appears to be altered in many cases of OPT deficiency
(60).

In summary, between 15 to 50% of the total
mitochondrial CPT is believed to be exposed to the cytosolic
surface of the inner mitochondrial membrane. Low estimates
of CPT-I activity (33,55) may result from the use of low
ionic strength assay medium since CPT-I but not CPT—II has
been shown to be inactivated in ionic strength below 0.06 M

(61).

CPT Deficiency

 

Numerous cases of muscle CPT deficiency have been
reported (40,60,62,63). A benign condition, unlike muscle
carnitine deficiency, the disease is manifested only when

lipid metabolism is stressed. Clinical symptoms include

14

episodes of muscle weakness, pain, and myoglobinuria induced
by prolonged exercise and aggravated by prior fasting. In
most clinical studies, CPT activity measurements are not
initial rate measurements, but measurements at or
approaching equilibrium. The assumption that optimal
conditions for the assay of OPT in normal tissue are optimal
for the assay of CPT in a patient's tissue is questionable,
in light of the documented changes in the kinetic parameters
of CPT as a response to environmental changes (64,65).
Absolute or partial deficiencies have been reported for
a given patient depending on the methodology used. In
several of these studies (60,62,63) a nearly absolute
deficiency is apparent with the hydroxamate assay as well as
with the isotope exchange assay under the usual assay
conditions. In contrast, significant activity can be
measured in the same patient's tissue with the forward
assay (60,62) and with the isotope exchange assay using
lower concentrations of substrates (62,63). Clearly, the
choice of methodology affects the results. The studies of
Zierz et al. (40) show that assays measuring initial
velocity at low substrate concentrations and minimal product
build up can give normal values of CPT activity in patients
manifesting the disease. From these data, Zierz and Engel
(40) have proposed that CPT deficiency is caused by altered
regulatory properties of a mutant enzyme, which is extremely

sensitive to inhibition by its substrate and/or products.

15

This enzyme also appears to be unusually sensitive to
inhibition by detergents and malonyl-CoA. They have offered
two interpretations for the increased malonyl—CoA
sensitivity; the absence of only malonyl-CoA insensitive CPT
(CPT-II), or a mutant enzyme with increased sensitivity to

malonyl-CoA.

Purification of CPT

The first attempt to purify CPT (25) involved a salt
extraction of lyophylized calf liver mitochondria followed
by chromatography on DEAE-cellulose. This yielded a 22 fold
increase in specific activity. Though the extraction from
the membrane was incomplete and the preparation still
contained a small amount of carnitine acetyltransferase,
these studies helped establish the existence of a separate
long-chain acylcarnitine transferase distinct from CAT.
Since CPT is an integral membrane protein, more successful
purification schemes (24,15) have used non-ionic detergents
to solubilize the enzyme and keep it in solution. Attempts
have been made to purify CPT-I and CPT-II separately
(66,67). With this approach it is uncertain whether a given
extraction procedure will yield only CPT-I or CPT-II. It is
also uncertain whether differences in the two final

preparations are due to differences in the intrinsic

a

16

properties of the enzyme or caused by modified environments
which result from the purification techniques used for each
pool. For these reasons efforts have been made to treat
each pool in an identical manner during purification (32) or
to extract total CPT and attempt to separate the putative
distinct enzymes using chromatographic techniques (15,24).
Table I gives a summary of the properties of several
purified forms of OPT.

West et al. reported the separation of two forms of OPT
(66). Using frozen ox liver they isolated a soluble enzyme
‘with carnitine palmitoyltransferase activity and purified it
850 fold. This preparation was termed "outer" CPT and was
found to be inhibited by 2-bromopalmitoyl-CoA. Another
fraction of CPT activity was extracted from the membrane
sediment by either treatment with Triton X-100 or butanol
and purified 10 fold. This fraction was designated as
CPT-II and was insensitive to inhibition by
2-bromopalmitoyl-CoA. Other significant differences in the
kinetic constants and physical properties of these two
enzymes were observed (see Table I) which suggested that
inner and outer CPT are distinct enzymes. Since peroxisomes
are known to contaminate mitochondrial preparations, it is
possible that the soluble enzyme prepared by West et al. is
not CPT, but COT that was leaked by the rupturing of the
fragile peroxisomal membrane with freeze thawing of the

samples. Interestingly, Miyazawa et al. (15) have shown

17

665325.80
Home £88335 5? 38:05 85 ow .293 :03 25: Tern—D Pan 35 no 838.33 030:5. ..
6883.60 casuglugmump N3 «0 3633 ham—Gong 053m: 93 was owe—P .

 

 

STE .352
2 58... 5.5 2. 5 .85.5Su.55« 2. 5.55 5.: .85 LE
85 5.5 .5 3 .5855. 8 p.55 :58
Egan
an 555 5.5 55 2 .5859 2 5.5 7.85 58
18.3 85.2 2.. .. .. 85.5 €85
u8~ 1.. .8558 58" 3 3?: .55 508.
.8558 S .85
58.2. 85:
S on. 55 5.... 5. .5853 82 55.5" H .55 28
.85:
mm 5.8 S 55 5. u «a 3.5 5+: 8 :8
88 5 55 58.8 2 4.5 ~78
hgnfi
8 5: 55.5 n 2 85.55 58 2 T8 .8
=mm8 83%...8
832.85 - H5 :98 .. 5
a 5.833 832:5 5333
due :5. x .3828. 28: 22855 Eon 8.38

Egan EH8; aHhHME ho avg—Nag BF ho hug

Hang

18

that purified rat liver COT is inhibited by
2-bromopalmitoyl-CoA, while purified rat liver CPT is not.
The identity of the soluble fraction isolated by West et al.
is questionable.

Kopec and Fritz (67) extracted CPT from calf liver
mitochondria using the detergent Tween 20 and purified it by
adsorption onto calcium phosphate gels. They obtained
several enzyme preparations of specific activity between
10-38 units/mg of protein. The preparations with the
highest specific activity were homogenous by SDS-PAGE and
were arbitrarily designated as CPT I. Another protein
fraction which did not adsorb to the calcium phosphate gel
was partially purified and designated as CPT II. This
fraction had no activity with acyl-CoAs, required
preincubation with CoASH to manifest activity and was
unstable at 4°C. Later, it was shown that CPT I could be
transformed into CPT II by treatment with urea and guanidium
chloride. In addition antibodies produced against pure CPT
I could inhibit CPT II activity. These data suggested that
CPT II was a partially denatured form of CPT and not a
distinct enzyme.

More convincing evidence that both pools of OPT are the
same enzyme has been presented by Bergstrom and Reitz (32)
and Clarke and Bieber (24). The former group separated the
outer and inner forms of CPT of rat liver mitochondria using

the digitonin fractionation procedure. They partially

19

purified each fraction independently, but using identical
procedures. The partially purified enzymes had identical
elution volumes on a Sephadex 6-200 column and the same
kinetic constants (see Table 1), although significant
kinetic differences in the "in situ" forms of the enzymes
had been observed. Clarke and Bieber extracted all of the
CPT activity of beef heart mitochondria by treatment with
Triton x-ioo and KCl and purified it to homogeneity. Using
the technique of isoelectric focusing in approach to
equilibrium a single protein band was observed, suggesting
that inner and outer forms of CPT are the same enzyme.
Recently, Miyazawa et al. (15) have purified CPT to near
homogeneity from rat liver and have also found a single
polypeptide. Using antibodies to the purified enzyme (68),
they obtained only a single precipitation line with liver
extracts, and a single in 31329 translation product from
total liver RNA. This translation product appears to be a
precursor of CPT with a slightly higher molecular weight of
71,600. In summary, all of the detergent purified forms of
CPT are apparent aggregates of molecular weight between
150,000-700,000 daltons and subunit molecular weight between

65,000-75,000.

20

Kinetics and specificity.

CPT catalyzes reversible reactions involving acyl-CoAs
and acyl—carnitines as substrates with a broad specificity
for acylgroups from C6 to greater than 020. CPT is highly
specific for L-(-)—carnitine, but also reacts with
norcarnitine (3-hydroxydimethylaminobutyrate) (69) and
thiocarnitine (beta-sulfhydryl-gamma-trimethylaminobutyric
acid) (70) with reduced reaction rates. Deoxycarnitine
(beta-hydroxyl group removed) (71) and aminocarnitine
(3-amino-4-trimethylaminobutyrate) (72) are not substrates
for CPT. These carnitine analogs are competitive inhibitors
of the reaction (69,72). Acyl—(+)—carnitines inhibit
membrane bound CPT and to a lesser extent the soluble enzyme
(73).

Table II summarizes the data from several studies on
the acyl-group specificity of CPT from various tissue
sources and varying degrees of purification. Using intact
mitochondria from calf liver and the reverse isotope assay
Solberg (74) selectively assayed the "in situ" acylcarnitine
specificity of the inner and outer pools of CPT in the
absence and presence of external CoASH, respectively. While
the inner transferase pool had an optimal vm with C7-

ax

moieties, the outer transferase pool had an optimal vmax

with 09- and Clo— moieties. He suggested that the two pools

21

. .940 5.3.80 05 5.55655 08 35.5.3 :5 658550555595 .33 :5 0.5555 55.553 3.55550550355— 503.555
.3543 5.. 55855 0.503 £855.83qu 90.8080: 455.563 555 05.055 9553 55838055555 0.5555 350555.555 5

 

 

 

85 555 555 555 55 55 55 5535+ 55555
55 85 555 55 55 55 55 55 553.5... 855 555+: .85 55.5
85 55 555 555 55 55 55 5.5555
5.5 85 55 5.. 55. 55 5 85 555+: 5.55 .553
55 85 55 5.. 55. 5.. 7.5.55 .5955 8:555:58
55 85 555 55 55 5 5 505.58.. 57.55 .8 .5555
EHH
55 85 55 55 55 5 555 5575. 5.55 558
“ya."
55 85 55 55 55 55 5 55 555+: .85 55,8
85 55 55 555 555 555 555 555 57.5.55 55555
55 85 55 55 55. 55 55 55. 55 85 7.5.55 558
85 555 555 555 55 55 55 8537 .5955
55 85 555 55 55 55 55 55 525+ 85 555+: .35 55.5
5.555:
55 85 85 555 555 555 55. 85 :7: 5.55 5555 558.5555
85 55 55 57.5.55 .5955
55 85 55 55 55 55 55. 55 5 7.5.55 .5555
555 .55 555 555 5o 55 55 55 :5: 8555.55.59.55
. 555 5555555555 55555555 55.5.5856 5.55.5 8.5565

ugggm NZHBHZEO ho MBHOHhHg g

Hag

22

contained different acyltransferases with different relative
velocities in different chain regions. However, it is not
clear from these data what are the specific optima and
relative velocities of CPT-I and CPT-II, since mitochondria
contain CAT with overlapping specificity for the
medium—chain acyl-groups.

Other investigators have used partially purified
preparations of OPT which are assumed to represent "outer"
or "inner" pools. Differences in profiles of these
preparations have been advanced as further evidence that
CPT-I and CPT-II are distinct enzymes. Tubbs et al. (75)
characterized the substrate specificity of the soluble
acyltransferase activity from ox liver isolated by West et
al. (66) which had been postulated to be "outer" CPT from
mitochondria. This enzyme has the highest activity with
myristoyl- and palmitoyl- esters in either direction.

Though is has been purified 850 fold it still retains
activity with short-chain acylgroups. Its broad substrate
specificity resembles that of peroxisomal COT (51). In
contrast, the activity purified by butanol extraction from
the mitochondrial membrane by West et al., which is
postulated to be "inner" CPT, has relative velocities of 15
and 100 for octanoyl-CoA and palmitoyl-CoA, respectively and
no short-chain acyltransferase activity.

Calf liver CPT purified 1200 fold after extraction with

detergent by Kopec and Fritz (67), was free of short-chain

23

acyltransferase activity, but showed a very dissimilar
pattern of substrate specificity from that of the membrane
bound enzyme isolated by West et al. The relative activity
with octanoylcarnitine was only 2% compared to 93% observed
by West et al. Clarke and Bieber (65) were able to resolve
the apparent discrepancy in these profiles by studying the
effect of two different acylcarnitine concentrations on the
acylcarnitine specificity of purified beef heart CPT. At
500 mu acylcarnitine, the specificity profile of purified
beef heart mitochondrial CPT agrees with that reported by
Kopec and Fritz for calf liver mitochondrial CPT, with
increased specificity for long-chain acylcarnitines. At 50
um acylcarnitine the substrate specificity pattern shifts to
one similar to that described by West et al. for ox liver
CPT with relative rates of 100 and 118 for
palmitoylcarnitine and laurylcarnitine, respectively. The
acylcarnitine specificity of OPT is also markedly altered by
detergents. Table II shows the effect of Tween 20 on the
substrate specificity of mouse liver CPT observed by
Miyazawa et al. (15).

The kinetic parameters reported for purified CPT from
similar sources can differ by an order of magnitude, see
Table I. These variations can result, in part, from.
differences in the state of the enzyme after several
purification steps. For example, the dissociation of the

membrane bound enzyme with butanol extraction results in an

24

increase in the K. from 2.2 pH to 9 pH (18). In addition,
assay conditions such as pH and choice of detergent
influence the kinetic parameters of CPT (64,65). Bremer and
Norum showed that increasing concentrations of palmitoyl-CoA
raise the apparent Km for L-carnitine. Therefore, they
proposed that palmitoyl-CoA can behave as a competitive
inhibitor to its cosubstrate L-carnitine (30). In a
separate study they showed the effects of several
detergents, in particular d-palmitoylcarnitine and Tween-80,
on their partially pure preparation from calf liver (34).

At a high concentration of palmitoyl-CoA, the detergents
stimulated the forward reaction and the apparent Km for
L-carnitine decreased. At low palmitoyl-CoA, the detergents
were inhibitory and no changes in the Km for L-carnitine
were observed (35). They concluded that the main reason for
this effect was that the detergent prevented the substrate
inhibition at high concentrations of palmitoyl-CoA.
Phospholipids and albumin were later found to have a similar
inhibitory and stimulatory effect on the activity of CPT
with low and high concentrations of palmitoyl-CoA (76),
respectively. However, phospholipids and albumin had these
effects without lowering the Km for L-carnitine. Recently,
Bremer has postulated that increasing substrate inhibition
by acyl-CoAs as the acyl-chain lengthens explains the
profile of maximum relative activity in the forward

direction in the medium-chain region (5).

25

Clarke and Bieber studied the effect of micelles of
Tween-80 on the kinetic parameters of purified beef heart
CPT in the reverse direction (65). In the micellar
environment the enzyme shows a 3-fold increase in vmax for

the various acylcarnitine substrates and a 5 to 6 fold

increase in Km for C —, C -, C

10 12- and Cl‘-carnitine esters.

8

Properties of Membrane gound CPT

The properties of membrane bound CPT differ
significantly from the properties of the enzyme in solution.
Lags in the time course of the reaction of membrane bound
liver and heart CPT-I with palmitoyl-CoA have been observed
(77-79). These lags decrease with increasing concentrations

of K+ or Mg+2

(77), with preincubation of mitochondria with
palmitoyl-CoA (78), and with sonication or detergent
solubilization of CPT from the membrane (77). In addition,
they are absent when the substrate used is octanoyl-CoA
(77). Cook (78) has suggested that these slow changes in
reaction rate after the addition of substrates are due to
hysteretic behavior of the enzyme and therefore can be
prevented by preincubation with the substrates. Similarly,
Bremer et al. (79) showed that CPT remaining in membrane
residues after extraction with Triton x—ioo is subject to an

apparent activation by the substrate. However, Zammit (39)

has shown that these lags are minimal when the reactions are

26

started by addition of palmitoyl-CoA/albumin mixtures. He
has suggested that the lags are not dependent on intrinsic
properties of CPT-I but may represent the time necessary to
obtain an equilibrium between free, micellar and
albumin-bound forms of the substrate. In albumin containing
media, salts may affect the binding of palmitoyl-CoA to
albumin (38). The association of palmitoyl-CoA with the
mitochondrial membrane appears to increase with rising ionic
strength (61). In this context, the non-specific activation
of CPT by various metal ions may be related to an increase
in the concentration of palmitoyl-CoA in the vicinity of the
enzyme's active site.

Due to the diverse assay conditions, particularly the
molar ratio of albumin and substrate, reports in the
literature on the Km for palmitoyl-CoA and I50 values for
malonyl-CoA of membrane bound CPT-I are extremely variable.
In a comparative study (38), the Km reported for
palmitoyl-CoA is 47 pH in the absence of albumin, and 450 pH
in 2% albumin. Substrate sigmoidicity in the kinetics of
membrane bound CPT with palmitoyl-CoA, has been reported in
some studies (21,80), while linear hyperbolic kinetics have
been reported in others (34,81). In some of these studies
hyperbolic kinetics were reported which became sigmoidal
after addition of malonyl-CoA to the assay medium (29,34).
The relationship between CPT activity and L-carnitine

concentration over a limited range of concentrations (20-400

27

uM) has been reported as hyperbolic (80). The KIn for
L-carnitine is apparently not altered by malonyl-CoA or by
fasting, though significant tissue variation in the Km for
L-carnitine has been observed. Interestingly, higher values
of the Km for L-carnitine have been obtained for tissues in
which CPT has a lower sensitivity to malonyl—CoA inhibition
(82). In contrast, the Km for palmitoyl-CoA does not appear
to vary significantly among tissues (82). Mills et a1. (83)
observed that an increase in pH results in a decrease in the
XI for L-carnitine with no change in the Km for
palmitoyl-CoA.

McGarry and Foster (20,84,85) showed that malonyl-CoA
can be a potent inhibitor of membrane bound CPT. They
concluded that only CPT-I is inhibited by malonyl-CoA, since
the oxidation of palmitoylcarnitine is not affected by
malonyl-CoA and in ruptured mitochondria malonyl-CoA
inhibits approximately one—half of the total CPT activity.
Direct measurements of the effect of malonyl-CoA on CPT-II
are not possible because of the impermeability of the inner
membrane to the CoASH esters. A moderate inhibition by
malonyl—CoA of OPT in inverted mitochondrial vesicles
suggests that CPT—II may be sensitive to malonyl—CoA, though
these preparations could be contaminated with exposed CPT-I
(86). Disruption of the membrane with detergents results in
the release of malonyl-CoA insensitive CPT, but the

sensitivity of CPT activity left associated to the membrane

28

fragments is increased (79,84). This observation stresses
the importance of membrane factors. Nevertheless, the
apparent insensitivity of CPT-II has suggested to some that
CPT-I and CPT-II are different proteins. An alternative
explanation is that CPT-II does not interact with a membrane
component which confers sensitivity to CPT-I. The
observation that other acyl derivatives such as succinyl-CoA
which are primarily located in the matrix compartment of
mitochondria also inhibit CPT (87), provides a possible
mechanism by which CPT-II could be regulated.

Several other CoA esters are potent inhibitors of
membrane bound CPT. Overt CPT activity of liver
mitochondria is sensitive to inhibition by
2-bromopalmitoy1-CoA, but latent CPT is not (88). Latent
CPT can generate bromopalmitoyl-CoA within the matrix from
external 2-bromopalmitoylcarnitine in a reversible reaction.
Palmitoyl-CoA analogs in which the carbonyl group is
replaced by methylene groups (89) and the 2-substituted
oxiran-2-carbonyl—CoA esters (90) are potent inhibitors of
CPT-I. Like malonyl-CoA, 2-tetradecylglycidyl-CoA is an
inhibitor of membrane bound CPT-I but it does not appear to
inhibit the solubilized enzyme (91,20). Both malonyl-CoA
and palmitoyl-CoA protect the enzyme from irreversible
inactivation by TDGA-CoA.

Differences in the phospholipid environment of the

matrix and cytosolic faces of the inner mitochondrial

29

membrane (92,93), may result in different properties of the
two forms of CPT. The functional activity of CPT appears to
be altered in galactosamine hepatotoxicity, a condition in
which the phospholipid composition of the mitochondrial
membrane is altered (94). Phospholipids can activate
partially purified CPT with a low specificity for
phosphatidylcholine (76). Brady et al. (86) have
demonstrated that small changes in membrane fluidity affect
CPT activity. They have also measured increased membrane
rigidity upon addition of malonyl-CoA using fluorescent
probes. They suggested that malonyl-CoA may affect both
CPT-I and CPT-II through changes in the membrane fluidity.
However, attempts to modify the lipid environment of
solubilized CPT have not restored malonyl-CoA sensitivity
(76,92).

At present, the study of CPT is directed towards
understanding the biochemical mechanism by which malonyl-CoA
is inhibitory. Initially, the removal of malonyl-CoA
inhibition by increasing the concentration of palmitoyl-CoA
suggested a simple competitive interaction (20,36,96) at the
active site of the enzyme. More recently, it has been
suggested that malonyl—CoA acts by a mechanism involving
cooperative inhibition due to the apparent ability of
malonyl-CoA to induce sigmoidal kinetics of CPT-I (34).
Since solubilized CPT-I is not inhibited by malonyl-CoA

(2,7), the participation of an additional membrane

30

component(s) has been suggested (64,83,20,97). Kiorpes et
al. (30) demonstrated that TDGA-CoA can inhibit membrane
bound CPT-I irreversibly, and malonyl—CoA can protect
against this inhibition. Recently, it was shown (20) that
solubilization of CPT-I "irreversibly" inhibited by TDGA-CoA
restored CPT-I activity similar to the effect of
solubilization on malonyl-CoA inhibition. Kiorpes et al.
detected in SDS gels, the irreversible binding of
14C-TDGA-CoA to a 90,000 molecular weight mitochondrial
protein from rat liver (91). This protein could be a
regulatory subunit of CPT, but it is not the catalytic
subunit since the catalytic subunit of rat liver CPT has a
molecular weight 63,000 (15). Though these recent data
strongly suggest a separate binding site for malonyl-CoA in

the mitochondrial membrane in liver, the nature of a

malonyl-CoA binding component remains unknown.

Role of CPTgin Control of Fatty Acid Oxidation

Initial studies on the control of fatty acid oxidation
by CPT demonstrated enzyme rates far in excess of the
capacity of mitochondria for fatty acid oxidation when high
concentrations of substrates were used (98-100). It was
proposed that substrate availability might control the rate
of oxidation of fatty acids (101-103). However, the

observation was made that when equal amounts of long—chain

31

fatty acids are supplied to the perfused liver, starved
rats produced ketone bodies at a higher rate than fed rats
(104,105). Differences in the rate of ketogenesis from the
oxidation of octanoic acid were less pronounced, suggesting
that the carnitine acyltransferase system was a primary site
for the regulation of hepatic fatty acid oxidation and
ketogenesis (104). Subsequently, McGarry and Foster
demonstrated the inhibitory action of malonyl-CoA on fatty
acid oxidation and CPT-I reaction (23,84). They proposed a
key role for this metabolite (the first committed
intermediate in the conversion of glucose into fat) in the
coordinated control of fatty acid oxidation and synthesis,
with an auxiliary control of CPT activity effected through
the fluctuation in the levels of fatty acid and L-carnitine
(95).

The control of CPT by malonyl-CoA appears to be
modified in two ways which may act to amplify each other
(106). Hepatic levels of malonyl—CoA vary with the
physiological state of the animal (22,106) and hepatic CPT-I
is less sensitive to malonyl—CoA inhibition in the fasted
and ketotic state (96,29,97,106,108-i10). In the fed state
where a high rate of fatty acid synthesis occurs, both
tissue levels of malonyl-CoA and the sensitivity of CPT-I to
malonyl-CoA inhibition are highest. In the starved state, a
decrease in malonyl-CoA concentration is amplified by the

decreased sensitivity of CPT-I to inhibition by this

32

metabolite, resulting in maximal rates of fatty acid
oxidation. Starvation may affect the properties of CPT
through changes in the lipid composition of the membrane.
Alternatively, changes in the regulatory properties of liver
CPT may be under hormonal control by the insulin/glucagon
ratio (94). Changes in the sensitivity of hepatic CPT to
malonyl-CoA inhibition apparently result from changes in the

K values of CPT for malonyl-CoA (34). The Ki value of CPT

i
for malonyl-CoA from liver of diabetic animals is
approximately 10-fold greater than in controls (111).
Treatment with insulin returns the K1 to control levels.
This finding suggests that in addition to facilitating an
increase in the levels of fatty acids available for
oxidation through enhanced lipolysis (112), a decrease in
insulin levels may modulate ketosis through the inhibition
of CPT-I by malonyl-CoA. High glucagon appears to decrease
the activity of acetyl-CoA carboxylase thus controlling the
malonyl-CoA concentration (113). Recent studies suggest a
direct effect of glucagon on CPT activity. After incubation
of rat hepatocytes with glucagon, SDS electrophoresis of
immunoprecipitates obtained using antibodies to purified CPT
show a phosphorylated protein band which has the molecular
weight of the purified CPT subunit (114).

The regulation of fatty acid oxidation by malonyl-CoA
inhibition of CPT-I appears to be a general phenomenon.

Inhibition by malonyl-CoA is more pronounced in heart and

33

skeletal muscle than in liver, kidney, or adipocyte (114).
Under identical assay conditions, the 150 values for
malonyl-CoA of rat liver CPT is 2.7 uM while the I50 for dog
heart CPT is 20 nM. Though heart and skeletal muscle do not
have active fatty acid synthesis they contain considerable
amounts of malonyl-CoA in the fed state (82). Other
hormones, in particular those associated with stress may
have an effect in tissues such as heart. Methylmalonate, an
intermediate in propionate metabolism, may also have an
important role in the regulation of CPT, especially in
tissues with low levels of fatty acid synthesis but high
gluconeogenesis, such as sheep liver, where methylmalonate
has been shown to be an inhibitor of CPT-I (115).

The therapeutic possibilities of specific inhibitors of
fatty acid oxidation in the treatment of hyperglycemia have
been proposed (116). Inhibition of CPT-I appears to be an
ideal target. Octanoyl-D-carnitine and decanoyl-D-carnitine
have antiketogenic and hypoglycemic effects when
administered to fasted or diabetic rats (117), but can be
toxic at relatively low concentrations presumably due their
detergent properties. TDGA is a potent oral hypoglycemic
and hypoketonemic agent in animals under conditions in which
fatty acids are the main fuel (118). Recently the
antiketogenic effects of DL-aminocarnitine
(3-amino-4-trimethylaminobutyrate) and its acyl derivatives

have been demonstrated (119). These inhibitors of CPT

34

require concentrations that are 0.1-13 of those required for
effective inhibition by acyl-D-carnitines and they show a

more pronounced hypoglycemic effect.

EXPERIMENTAL PROCEDURES

Materials

 

Acyl-CoAs were purchased from PL Biochemicals.
L-carnitine was a generous gift from 0tsuka Pharmaceutical
Co., Japan. Blue Sepharose 4B had been previously
synthesized in the laboratory according to (119). Sephacryl
8—300, Fractogel TSK HW—55, and QAE-Sephadex were purchased
from Pharmacia. Immersible-Cx ultrafilters (Polysulfone
membrane with a nominal molecular weight limit of 30,000
daltons) were from Millipore. Fluorescamine, collagenase,
DTNB, DTBP, and pinacyanol chloride, were from Sigma.
Molybdenum Blue spray for phosphorus detection was from
Applied Science Laboratory (2051 Waukegan Road, Deerfield,
IL). Phospholipid standards and Redi-Coat 2D precoated TLC
plates were from Supelco. Triton X-100 was purchased from
Research Products International Corporation. Purified mouse
liver peroxisomal COT was kindly provided by Dr. Shawn
Farrell. Octylglucoside was kindly provided by Dr.
Ferguson-Miller (Michigan State University, East Lansing,
MI). TDGA-CoA was a generous gift from Drs. Linda and Paul
Brady (Washington State University, Pullman, WA). All other

reagents were of analytical grade.

35

36

Methodg

Purification of carnitine4palmitoyltransferase

from beef heart mitochondria

Several modifications to the procedure in (24) were

made in order to obtain a higher yield of pure enzyme.

Mitochondrial Isolation. Beef hearts were obtained from the
MSU abbatoir, sliced and packed on ice for transportation.
The beef heart ventricles were soaked in ice cold sucrose
buffer (0.25 M sucrose, 5.0 mM Hepes, 0.25 mM EDTA, pH 7.7)
while the fat and connective tissue were throughly removed.
The tissue was homogenized for 45 seconds at high speed in a
Waring blender in batches of 240 g in 2 L of buffer,
followed by 30 second Polytron homogenization in 250 ml
centrifuge tubes. Cellular debris was pelleted with a 15
min spin at 500 x g, and the supernatant was collected by
filtering through cheesecloth. The first mitochondrial
pellet was obtained by centrifugation at 15,000 x g for 15
min. The pellet was resuspended with two strokes of a
loosely fitting Potter—Elvehjem teflon-glass homogenizer in
500 ml of isotonic sucrose buffer (0.25 M reagent grade

sucrose, 2.5 mM HEPES, 0.25 mM EDTA, pH 7.5). The

37

mitochondrial suspension was centrifuged again at 500 x g to
remove remaining debris. A second mitochondrial pellet was
obtained by centrifuging the supernatant at 11,000 x g and
resuspended as previously described. The final
mitochondrial preparation was obtained by a third
centrifugation at 7000 x g, and resuspension in a minimal
volume of isotonic sucrose buffer. The mitochondria were

stored at -80°C.

Solubilization of Mitochondrial CPT. Mitochondrial
suspensions were thawed, pooled, and mixed with one-half
volume of 3 M KCl in 6% Triton X-100. The suspension was
homogenized in 15 ml batches with six strokes of a
Potter-Elvehjem homogenizer at 15°C and centrifuged for 90
min. at 89,000 x 9 (29,000 rpm). Lipid in the supernatant
was removed by suction and the clear apricot-colored
supernatant was collected with care not to disturb the
pelleted membranous debris. The solubilized protein was
equilibrated with Blue buffer (1.0% Triton X-100, 2.5 mM
Hepes, 0.25 mM EDTA, 60 mM KCl, pH 7.5) by exhaustive
dialysis. The dialyzed preparation was centrifuged at
15,000 x g for 15 min. to remove protein that was
precipitated by the decrease in ionic strength. The ionic
strength of the supernatant was then increased to 300 mM KCl
by a second dialysis against the Blue buffer containing 0.3

M KCl.

38

Chromatography of Solubilized CPT. The dialysate was applied
in 35 ml batches to a 40 x 5.0 cm column of Sephacryl S-300
equilibrated with Blue buffer containing 300 mM KCl and
eluted at a flow rate of 75 ml/hr with the equilibrating
Blue buffer. Peak fractions of CPT activity were pooled and
dialyzed against QAE buffer (0.5 x Triton X—100, 5.0 mM
Bis-Tris Propane, 0.25 mM EDTA, 20 mM KCl, pH 9.7) and
applied to a 30 cm x 4.1 cm column of QAE-Sephadex Q-25-120.
CPT activity washed through the column, was pooled, and
concentrated to a final volume of 1 ml using the immersible
ultrafilters describe in Materials. The activity was
dialyzed extensively against buffer (0.1 3 Triton x-100, 2.5
mM Hepes, 0.25 mM BDTA, 60 mM KCl, pH 7.6) and loaded on a
12 cm x 1.6 cm column of Cibacron Blue Sepharose. Four bed
volumes of buffer were used to wash off unbound protein
before a 50 ml linear gradient of 60-860 mM KCl in Blue
buffer was used to elute CPT. Fractions containing CPT were
pooled. Cibacron Blue Sepharose was treated regularly with

pronase to maintain a high binding capacity.

Exchange of Triton X-100 for Octylglucoside. The final

 

step in the purification of CPT, Cibacron Blue Sepharose
Chromatography, served to exchange Triton X-100 for
octylglucoside. The concentration of Triton X-100 was

reduced to 0.002% in the equilibrating buffer prior to the

39

loading of the enzyme to the column. The unbound protein
was washed off with Blue buffer in 0.002% Triton X-100
followed by three bed volume washes of Blue buffer in 25 mM
octylglucoside. The enzyme was eluted with a similar salt
gradient as previously described, in which Triton x-1oo was

substituted with 25 mM octylglucoside.

Physical Characterization of Purified CPT

SDS-Polyacrylamide Gel Electrophoresis. Purified CPT was
electrophoresed on 7% polyacrylamide gels (120) and stained
with Coomasie Brilliant Blue. BSA, ovalbumin, myosin, and

phosphorylase b were used as molecular weight standards.

Phospholipid Analyses. The total phosphorus content of
purified CPT was determined by ashing of the protein and
phosphate quantitation (121). Phospholipids were extracted
with 20 volumes of chloroform:methanol (2:1) and the total
organic solvent soluble phosphate was determined by the Ames
procedure (121). The phospholipids were chromatographed in
two dimensions (122) on Supelco Redi Coat 2D precoated TLC
plates, and localized by spraying with Molybdenum Blue
spray. The phospholipids produced blue spots on a white
background immediately after the application of the spray.

The phospholipids were identified by comparison to

40

standards.

Native Molecular Weight Determination. A 50 x 1.5 cm
Fractogel TSK HW-55 column was equilibrated with buffer
containing 2.5 mM Hepes, 0.25 mM EDTA, 300 mM KCl and 12 mM
octylglucoside at pH 8.0. The column was calibrated using
10 mg/ml each of ovalbumin, BSA, aldolase, thyroglobulin and
1mg/ml of ferritin. A 1 ml sample of purified CPT
containing 200 munits was loaded and 0.3 ml fractions were

collected.

Kinetic Characterization of CPT.
Activity Assay. The forward reaction (formation of
acylcarnitines) was monitored at 412 nm following the
release of CoASH with DTNB. The 2.0 ml reaction volume
contained 115 mM Tris Buffer, 1.1 mM BDTA, 0.1% Triton
X-100, 150 um DTNB, with varying concentrations of
L-carnitine, and acyl-CoA at 25°C. When pH effects were
being studied, DTNB was substituted with DTBP and the buffer

used was 50 mM Bis Tris Propane.

Kinetic Measurements with Purified CPT. Kinetic data were
obtained with a semiautomated kinetic analyzer described in
(123,124). A schematic drawing of this system is shown in

Figure 2. Absorbance-time data were obtained by continously

41

 

 

'—-—--‘ Waterbath

 

 

 

 

Syringe Drive
Substrate Addition

 

 

 

 

 

 

 

Enzyme + Cosubstrate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2 .

 

 

 

 

 

 

 

I I
L ' /
’l ‘7’" /
l Pr
. ogrammable
Monochromator beam : v Photode actor Calculator
#67 ,
a/ >le
Output
Magnetic
Stirrer

 

Schematic represmtation of the semi-amounted kinetic
malyzer. It consists of a Gilford 2600 spectrophotometer
eqnmed with a microprocessor, a five significzit digit
absormnoe readout , water circulaticn with tenperature
control, aid mqnetic cuvette stirrer. It is interfaced
with a progranmable desk-top Hewlett Packard calculator
which nrns the progran TAMIL The reaction is started
by aidition of sthstrate with a pulp driven syringe to a
stirred cuvette containing the enzyne ard oosubstrates.
The rat time—absorbance data are converted to abstrate-
velocity data by the method of tmgent-slcpe mlysis,
and the optimal Hill parameters (V ,80 , Bill n) are
obtained fran the best 1m fit Wee ta.

42

increasing the substrate concentration of a stirred enzyme
assay mixture with the use of a precision syringe drive and
an automated Gilford model 2600 spectrophotometer. The
reaction time was 3.6 min. The concentration of added
substrate was adjusted so that the increase in volume during
the assay was limited to less than 5%. The final
concentration of substrate was optimized for each individual
measurement to be within 3-5x the K0.5' Absorbance data
were obtained to five significant digits. The data were
transformed into velocity-substrate data by a tangent slope
procedure, analyzed as linear plots using the TANKIN program
in a Hewlett-packard 9815 calculator, and plotted with a
Hewlett-Packard 9872 A plotter. Time- and product-
dependent effects on the reaction rate were determined by

systematic analysis of the data (123).

Kinetic Characterization of Membrane Bound Rat Heart CPT-I.

 

The activity of rat heart mitochondrial CPT was measured at
pH 8.0 using the DTNB assay and decanoyl-CoA as substrate.
Triton X—100 was omitted and 150 mM KCl was added to the
assay mixture to prevent mitochondrial swelling. When used
as cosubstrates L-carnitine was 6.0 mM and decanoyl-CoA was
100 uM. For each run a background change in absorbance upon
addition of acyl-CoA was measured with a control run which
lacked L-carnitine. This background absorbance was

subtracted from the absorbance data by an added subroutine

43

to the TANKIN program. The flow chart for this subroutine

is shown in Figure 3.

Other Methods

CMC Determinations. The formation of micelles was monitored
by addition of acyl—CoA substrate or octylglucoside to a
cuvette containing the activity assay mix and 4 uM
pinacyanol chloride. Formation of micelles was observed by
a change in the extinction coefficient of the dye at 610 nm

(125).

Protein Determination. Protein was determined by the
fluorescamine method (126) throughout the purification
procedure and the Fractogel Chromatography using BSA as
standard. Mitochondrial protein was determined by
solubilization with deoxycholate with the Coomasie Blue

Method (127).

Isolation of Rat Heart Mitochondria Mitochondria were
isolated from rat hearts by the procedure described in
(128). Three male Sprague-Dawley rats, 21 days old, were
decapitated and the hearts rapidly removed, rinsed, and
minced in an ice cold medium containing 0.225 M mannitol,
0.075 M sucrose, and 0.2% (w/v) fatty acid free BSA. The

minced hearts were rinsed three times and treated with 200

44

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3. Flowchart for the aided stbroutine to the computer program
TANKIN. This stbroutine eliminates absorbance flames due
to mitochondrial shelling.

45

units of collagenase in 15 ml of medium. After 1 min. the
preparation was homogenized using a Potter-Blvehjem motor
driven pestle. After 3 min. on ice, EGTA was added to a
final concentration of 1 mM. The homogenate was centrifuged
at 600 x g for 5 min. and the supernatant fluid was
collected by filtering through cheesecloth. The
mitochondria were pelleted by centrifugation at 8,000 x g
for 10 min. The mitochondria were washed twice by
resuspension and recentrifugation at 8,000 x g for 10 min
and were finally suspended in 0.225 M mannitol/0.075 M
sucrose. Heart mitocondria isolated in this manner had
respiratory control ratio of 10 or higher with 5 mM
pyruvate/2.5 mM malate as substrates. Acyl—CoA hydrolase
activity was not detectable. The specific activity of CPT-I
was 22-27 (munits/mg of protein) measured with the DTNB

assay at 25°C.

RESULTS

Purification of Carnitine Palmitoyltransferase

from Beef Heart Mitochondria

Preliminary experiments using the available Cibacron
Blue Sepharose resin synthesized in this laboratory revealed
that it was no longer effective in the separation of OPT and
CAT under the conditions reported earlier (24). A loss of
CPT activity in the wash, and an overlap in the elution
peaks for CAT and CPT, indicated a reduction in the binding
capacity of the resin which could not be restored by
treatment with either pronase or denaturing agents. The use
of Cibacron Blue Sepharose from Pharmacia also yielded
unsatisfactory results. New sets of conditions for binding
were tested keeping in mind that non—ionic detergents can
interfere with affinity chromatography by encapsulation of
the dye in the detergent micelles which results in a
reduction of the binding capacity of the resins (129). When
a lower concentration of Triton x-1oo (0.1% Triton X-100)
was used, the enzyme bound tightly at low ionic strength.
However, it was deemed necessary to maintain a high
concentration of detergent during the early stages of the
purification of OPT, an integral membrane protein, to
prevent the reaggregation of membrane fragments which could

46

47

interfere with the separation of the individual proteins.
It was decided that Cibacron Blue Sepharose Chromatography
should be used as a later step in the purification
procedure. At this time, it was also decided to omit the
hydroxyapatite column chromatography used in earlier
purifications because of a low enzyme recovery after this
step.

A high yield of CPT of high specific activity was
obtained following the chromatographic steps outlined in
Table III. Sephacryl S-300 was selected as the initial step
since its fractionation range is well suited to separate CPT
associated with detergent micelles (530,000 daltons) from
soluble CAT (60,000 daltons). As shown in Figure 4, CPT and
CAT activity elute as two well separated peaks. This step
also separates CPT from a bulk of the solubilized
mitochondrial protein giving an overall 20-fold purification
and a 75% recovery (see Table III). A further 2-fold
purification with a 60% overall recovery was obtained with
QAE—Sephadex Chromatography at pH 9.7. Cibacron Blue
Sepharose at a low concentration of Triton x-100 (0.1%)
yielded a final enzyme preparation with specific activity of
42 units/mg of protein which was apparently homogenous as
determined by SDS—gel electrophoresis (see Figure 5). The

overall recovery was 44%.

.382 a 5 gauging...
ou. moot—”>00 v.0 no.5 « 9.5280 09 in: 255.8 mo 959.0 93 ea Egauum no «...—5 g

 

48

82 3 0.9. 9 88958 33 5833

 

ooh 8 a." afifimmnmeo

0.8 fl. 8. oomum agate.

a . a no moo. ufiumfimflrn. £933

a." «m 3. 3.88505: 833338

fl 8“ no. 3.8.5805: 60.85.

Bow 3982 658.2: ovum 53.82385
0.588 3338 02868

 

gggaggaongHgghog

HHH g

49

.nconuoz cw connnomoc ma nn\na on no cums 30H“ m um nlfinoo
:0 0.0 M EU as a On monoumn me on Ca cuwanmm one: moHQEMm
.oomlm H>nomnaom :0 940 can amo confidanflnoa no ceaunumaum

(..ILU x 6m) ugade

.3632 5:02...
0?

on

.v unmask

ON

 

 

 

2T

 

 

 

522.. d
...40 O
.50 O

 

.1
O
[s

('Jw x I_u!|.u x smomri) Mglsgpv

 

 

50

Figure 5. SDS-polyacrylamide gl electrophoresis of purified beef
heart mitochorxirial CPT stained with Coanassie Blue.

Ieft lame: 10 pg pare CPT

Midile lane: (fran top to bottom): 10 (g each of myosin,
#Dspl'nrylase. BSA. ovalbimin.

Rigit lane: 5 1y pureCPT

 

Figure 5.

52

Physical Characterization of

Purified Carnitine Palmitoyltransferase

SDS—polyacrylamide gel electrophoresis of the purified
enzyme confirmed the previously reported subunit molecular
weight of 67,000 (24). Since CPT is a tightly membrane
bound protein, experiments were performed to determine if
the purified enzyme contained bound phospholipids. It
contained approximately 19 mol of phosphorus/mol of enzyme
as determined by protein ashing and phosphate quantitation.
All of the phosphorus prior to ashing was soluble in 20
volumes of chloroform:methanol (2:1), indicating that this
phosphate was not covalently bound. Two-dimensional thin
layer chromatography showed three major phospholipid
components identified by comparison to standards as
cardiolipin, phosphatidylcholine, and.
phosphatidylethanolamine. Three other unidentified minor
spots were also present. Though direct quantitation was not
done, the size and intensity of the spot indicated that
cardiolipin was the major component. A tracing of the TLC
plate is shown in Figure 6.

Using molecular sieving columns, it had been previously
shown (24) that in the presence of Triton x-1oo micelles the
detergent—protein complex has an apparent molecular weight

of 530,000. In this study, Triton x-1oo was substituted for

53

Figure 6. ‘Mo dimensicxal thin layer chnanatogramy of extracted
finspholipids frcm purified CPT. The solvent system used was:
(1) Chlorofomzmetl'anolmaterzaquemm mania (30:70:8:0.5)
(2) betraml :Aoetic acid:water : chloroform:mtcne

(20:20:10:100:40)

The plates were developed for 35-40 min, dried for 10 min, ani
mrayed with iblyhismm Blue Spray. A tracing arr! a paotograph
of the developed plate are shown.

gs

 

 

 

CL

Q PE
PC

(5
Q 0 (j!)

(1) If

 

 

 

 

Figure 6.

 

55

octylglucoside and the molecular weight of OPT in the
absence of detergent micelles was determined by gel
filtration in Fractogel TSK-55 at 12 mM octylglucoside.
Octylglucoside was chosen since the purified enzyme remains
stable at concentrations below the cmc of the detergent.
CPT activity migrates as a single peak of constant specific

activity with a molecular weight of 660,000 daltons (see i
Figure 7). The absence of detergent micelles in the eluant

was confirmed with the pinacyanol chloride method described

 

in Methods. Figure 8 shows that CPT activity with ‘
palmitoyl-CoA is optimal at 12 mM octylglucoside. Though

there are no detergent micelles at this concentration (see

below), the enzyme remains in suspension after

centrifugation at 100 x 9. However, the activity of OPT

with octanoyl-CoA decreased with increasing concentrations

of octylglucoside and was not optimal at 12 mM.

Kinetic Characterization of

Purified Carnitine Palmitoyltransferase

Preliminary Tests. For accurate determination of the
kinetic parameters of CPT using the TANKIN program, it is
necessary to obtain initial velocity measurements that are
linear for the duration of the assay, i.e., there should be
no hysteretic behavior and no significant product

inhibition. Figure 9 shows a time course for the DTNB assay

 

56

(ml/617’) ugaimd

 

.38. 5 constants 8 noun? 8!

Eu no BEE 8N "c3290 .3 can .48 3. 69:8? .8

.5588 .8 Swagger E "my? can: «288% £68:
81629. ..Eo unwashed «0 iguana 84: 22. none? .~. £99m .

.ooEsz cozoofi

OWF 0:.
in!)

 

 

loo...

. X
loos m

 

 

loom.
E E 8. a. 55.25 can

2.. . . . . . . Le.

 

 

 

(um/in/salowfi) humor: idO

57

H 100 ”M palmitoyl-CoA
H 100 m octanoyl- CoA

100-—

50—

% of CPT activity in TX— 100

 

 

J l l l J
6 12 18 24 30

octylglucoside (mM)

Figure 8. Effect of octylglucoside on CPT activity. Activity was
measuredwith the DTNB assay. th has 1.1 mMard
the acyl-00A was 100 m. Velocities are relative to the

Vmax in 0.1% Triton X—100 for each subtrate.

 

58

 

.dexu: “.0 mo Swvmuuguoo flag 5

5?. fine o.m n00 000m and o.m an an g :6 «In 40:102.”. 8.. 0: .20 a 00m

.48.?338 21 8a £5558...” :5 a; .8an copra. 56 so «9888 ufi 5
02988 on! can.“ and: 0E. .mmmum 893 5“: 83003 .50 98 00.500 9.3. .0 0.59m

 

 

 

 

AEEV 05;
00.n vmfi mad Nnd 0_.N 00.. ¢¢._ 00.. NHO 0nd 00.0
0.0.- u w m a 4. a m w w 0000
0.0:...
0.0.4.
of.
N 0.?
.N
ml 0.0..
m
D-
/
m. 0.~..
M.
of. I! 1
9:03.50 “mm...—
0.w... ..00N.0
o3. .83
0.0. 0 x u . q . . . 8N0

 

 

59

used in our studies. The initial rate (first derivative) is
constant throughout the duration of the assay, when the
substrates are saturating. End point deletions (123) of the
velocity-substrate data obtained from a substrate addition
experiment will detect enzyme inactivation. The deletions
in the data produce variation in the calculated Hill n and
the correlation coefficient if there is significant enzyme
inactivation. No effect of end point deletion on the Hill
coefficient can be detected in the substrate—velocity data
obtained for CPT, see Figure 10. In addition, Figure 11

shows that the vma is not affected by different amounts of

x
product accumulated in the reaction mixture. Therefore, it
can be concluded that there are no apparent hysteretic
effects or significant product inhibition under the

conditions of the DTNB assay.

Determination of Nonmicellar Assay Conditions. Suitable

 

conditions in which to investigate the kinetics of CPT in
the absence of detergent micelles, substrate micelles, or
mixed substrate detergent micelles were selected. The dye
pinacyanol chloride was used to detect the formation of
micelles in the assay mixture. The cmc of octylglucoside in
the DTNB assay mixture is 30 mM. In the absence of
detergent, palmitoyl-CoA forms a micelle at 5.4 uM. A
logarithmic relationship between the cmc of the acyl-CoAs

and their carbon chain length is observed (see Figure 12),

 

 

Figure 10.

60

 

2.0 -

0 GDP

I.8 ¢—.———'——’-——v——m

Hill

l.6 —

 

I l 1 l
O 5 IO IS 20 25

Number of end points deleted—9

 

 

Effect of and point deletion on apparent Hill n. Raw data
points fran a typical substrate-aidition experiment with
GT in which palmitoyl-CoA is the varied substrate were
progressively deleted from the std of tin data set, i.e.,
fran the highest suastrate ocucentration. The "new"

data set was analyzed with the TANCIN program

 

Figure 11.

61

 

4.0»

3.0"

VMOX/ E (muni15)

 

 

L 1
T

0625 01550 0100 (1:125 050
pmol Product

 

Effect of product accunulatim durirg a substrate adiiticn
experiment. The V of CPT reaction with L—camitine

as the varied subsgte aid octanyl-CoA as the cosubstrate
as determined for a series of substrate aidition experiments
inwhich the amount of enzyme acfled was varied to obtain

different anounts of prodict accumlation at the an! of
each rm.

 

62

2.5

LOG CMC (umolar)

 

 

:-

8 IO l2 I4 I6 I8

1 l

 

CHAIN LENGTH

Figure 12. The relationship between the carbon chain-lagth of the
acyl-CoAs and their cmc. The cmc's of the acyl-OoAs were

determined using the dye pinacyanol chloride as described
in Methods.

 

63

thus the cmc of medium-chain acyl-CoAs is much higher than
that of long chain acyl-CoAs. As shown in Figure 13, the
cmc of palmitoyl-CoA is reduced from 5.4 uM to 1 uM in the
assay mixture containing 12 mM octylglucoside. Octanoyl-CoA
has a cmc greater than 200 uM in the assay mixture in the
absence of detergent. Addition of octanoyl-CoA to the assay
mixture containing 12 mM octylglucoside shows no micelle
formation up to 150 uM octanoyl-CoA (see Figure 13). Since
octanoyl—CoA did not form micelles under these assay
conditions, it is used as substrate for kinetic studies in a

non-micellar environment.

Nonllnear Klnetlcs of Purified CPT. The kinetics of
purified beef heart mitochondrial CPT were found to be
sigmoidal with its acyl-CoA substrates as well as with
L-carnitine under all the different assay conditions used.
Double reciprocal plots showing non-hyperbolic
substrate-enzyme interaction are shown in Figures 14 and 15
for a variety of conditions. When octanoyl-CoA is the
varied substrate at pH=8.0 in 0.1% Triton K—100, a double
reciprocal plot as in Figure 14A, shows non-linearity. A
Hill plot of the same data shown in Figure 14B, gives a
linear fit when Hill n=2.3, indicating strong positive
cooperativity.

Hill plots also show non-hyperbolic enzyme-substrate

 

 

64

  
  

0
Li

 

1 2 mM octylglucoside + palmitoyl- CoA

Absorbance at 810 nm

1 2 mM octylglucoside + octanoyl- CoA

  

micelle
formation

 

l l l
5 1 O palmitoyl - Colt
(60) (1 20) octanoyl - CoA

0.0

[Acyl- CoA] pM

Figure 13. Determinatim of acyl-00A micelle fornation in octylglucoside.
Piracymol chloride (4uM) vasadiedtothem'NBassaymedia
omtainitg 12 w octylglucoside. The absorbance change at
610 rm was recorded tpon ocntinuous addition of plmitoyl-CoA
or octaioyl-CoA with a punp driven syringe.

 

65

Km- 63' pH

1'2”] v"...- u.49 munits
’ mK LOO
1.30»
A 3.92--

 

 

B. 581*

40V
0.2 *I
D 20
B 48*

 

  
       

50.58 4.65 pM
Vm a 8.56 m units
n= 230

1.20

        

-a.401

Figure 14. Bumble reciprocal plots of CPT reaction velocity versus
octanoyl-OoA emoentratim. L-carnitine is fixed at 3.0 uM.
Otherassayconditiasareasascribedforthemassay
inibthods. (a) 1/V versus 1/[S]. (b) 1/V versus
ms)“, 3111 n=2.3. The data were obtaimd
with the kinetic analyzer described in Authods.

66

     

Km '2.67 IBM
Vma‘. '7093 MUN".
Hi1! n- I.69

L l 1 __ __I

20.3” t'ofs'fiorsrt'fio
Io'5(—' )

(5)“

 
  
   
 

3.0). “0.5 - 8.55 all

I‘m - 27.75 munits

I-l.4

  

2.

/ 1u in 6.0
1041;
(5) a

 

~11!L

Figure 15. Double reciprocal plot of CPT reactim velocity versus

L—carnitine mtratim. (a) 1/V versus 1/[S]n.

Hill n-1.7. Palmitoyl-CoA is 100 m, Tgitm x-100 is

0.1 %, pH is 6.0. (b) l/Vversus 1/[S] , Hill m1.4.
Octmyl-OoA is 100 uM. octylglucoside is 12 mM. :8 is 8.
Other assay cmditions are as described in mthods.

 

67

interaction when L-carnitine is the varied substrate. The
data from a typical experiment in which palmitoyl-CoA is at
saturating levels at pH 6.0 in 0.1% Triton x-100 is shown in
Figure 15A. A Hill n=1.7 gives the best linear fit of the
experimental data. In 12 mM octylglucoside and 100 uM
octanoyl-CoA (non-micellar assay conditions) a Hill plot of
the L-carnitine data gives n=1.4 (see Figure 15 B).

In contrast, mouse liver peroxisomal COT exhibits linear
kinetics. A double reciprocal plot of data obtained using
the same assay, at pH 8.0 with stearoyl-CoA as substrate is
shown in Figure 16. A Hill n= 1.0 gives the best linear

fit.

Substrate Specificity. CPT substrate specificity profiles
in the forward direction in the presence of 0.1% Triton
X-100 and 12mM octylglucoside for two fixed concentrations
of L-carnitine, are shown in Figure 17. It is clear from
the profiles in Triton X-100 that the substrate specificity
will depend on the fixed concentration of L-carnitine. The
K0.5 and Vhax for L-carnitine were determined in the
presence of a fixed saturating concentration of various
even-chain acyl-CoAs in both detergents. These data are
given in Table IV. In Triton X-100, CPT has higher affinity
for L-carnitine in the presence of long-chain acyl—00A

derivatives, but it has the highest absolute catalytic rate

with hexanoyl-CoA. A logarithmic relationship exists at

68

 

 

.ggfiggguou 3300008939833

>83 H.230 .06 n... =0 05 on! 007* Guyana. 5.0 83:00 >080 can. .2: 06 0a

8382888 9:35.65 .fiounspaaooo «a ...... 6358.5 «5 AH: Smog >>
.Saumhufiwnxaoo 301300 90.5» 33ng 83002 .30 no #03 300.338 2900 .3 039.5

 

on... . .oum.

4‘

0.0..

 

No.7: 5:
2.55 8.3.3.5,
2.. No.5.

    

 

 

69

 

 

 

 

 

 

 

 

 

 

 

 

 

1 In 0.195 Triton mm
450—
‘9
0
T:
1!;
£
0
2
O
>
O
5
.2
a
C
O 8 10 12 14 16
AcyI - CoA chain length
200'-
In 1 2m“ octylglucoside
‘w'. I—— II—
“2
o —
-
0
if;
a 100—
B
2
0
>
0
3.
a
2
.2 so

 

 

 

 

 

6 13101214 1618
Acyl - 00A chain length

Figure 17. Specificity of OPT for acyl-CoAs of varyirg chain-lergth.
masursnents were male in 0.1% Triton x-100 and in 12 nM
octylglucoside. In staded areas, the activity was nussured
with 1.1 mIr-csmitine ani 100uMacyl-00A usim the DTNB
assay. Inmshaiedareasvmvmsdeteminedmimthe

kinetic analyzer, with Imitine as the varied swan—ate

and 200 m acyl-00A. The specific activity of OPT with
talmitoyl-CoA assured with the staniard D‘l‘tB assay in

Tritm x-100 was 40.

 

 

7O

 

.a to a! 87x 85.0. not ms «« n! «3803238 .2: 8« on! 28:32

 

92 o. 3
«.o« oi «.o «.o 3
«.o« 24 ed and 3
o.«.« «.4 «A and «H
«.8 «.3 o; 36 on
«..«« «.8 o.« a.” o
«.3 T: o.« o.« o
8.23 0.5 Bus... :3
x95 29. a! n.9—
ooamogeoflfioo 87x course 598753". «8.38

 

 

 

géggggho
aHfigggdzgfigggaggnin.g

Rug

71

saturating acyl-CoA levels between the K0.5 for L-carnitine
and the acyl- carbon chain-length (see Figure 18). In
contrast, mouse liver peroxisomal CAT has higher affinity
for L-carnitine in the presence of short chain acyl-CoAs and
the Km for L-carnitine of mouse liver peroxisomal COT, does
not show this type of relationship with the acyl-group
chain-length, see Table V.

It is apparent from these data that the specificity of
CPT for any given acyl-CoA depends both on the binding
capacity and on the catalytic rate for each individual
acyl-CoA substrate at a specific concentration of
L-carnitine. Therefore, a better determination of the
enzyme's capacity to use a particular acyl-00A could be made
if one determined the apparent kinetic constants within a
limited range of the cosubstrate L-carnitine. With this in
mind, the L-carnitine concentration was fixed at 1.0 mM
(nonsaturating for medium chain acyl-CoAs) or 5.0 mM
(overall saturating) and the kinetic parameters for several

acyl—CoAs were determined. The K with a

0.5 and vmax'

calculated vm /SO 5 ratio, are shown in Table VI, The data

ax
show that the beef heart enzyme is selective for long-chain
acylcarnitine formation at low concentrations of
L-carnitine. Low concentrations of L-carnitine decrease the
enzymes catalytic efficiency with hexanoyl—CoA and

octanoyl—CoA but do not affect the catalytic efficiency with

the long-chain acyl-CoAs.

 

72

91
O
I

log K0.5 (pM l-carniiine)

1

20

l I l

6 8 lb :2 IA I IS
Acyl-CoA chain-length

 

Figure 18. Relatim betvaen the chain la'gth of the acyl-00A and the
Kosofa’rforLr-centitine. Kosvaluesare
determined wing the kinetic analyzer described in
mthods with the DTIB assay. Acyl-CoAs are 200 (M.

 

LI

73

 

.o.o an an all 9.8 «5 swam:
E30 0322 05 53005300 0.53 @032, 0v. 62. «A 0! SEE

 

 

 

3 3
«« 3
o.o« :
no «a
o« «.« S
on « . o o
o« o . « o
S «.o 4
fl 2. «
e8 .80
m. on 50875.th 30.32

 

glgzgsgagho g n.
SHH§§§§<B¢MZHBH§LH§B¢US§B§ >024 0M

>935...

74

 

.8058. 5.. flag
8 >§m 9.8 0:» 058 fig 0395— 0:» 5%! 0858 and gum-Bung Dag

 

 

mm.o o.o n.fi ¢¢.o o.o m.” «H
o." a.“ a.” mm.” a." m.“ o"
NH.” v.fl a.” ou.H m.o ~.o «fl
m~.m «.« ~.o m~.¢ «.« ~.o «a
o~.~H «.ofl o.o o~.v ~.¢ o." ofl
««.o «.«H «.« om.o «.c m.¢ o
um.o o.oH o.m¢ mo.o a.» m.mo o

. .muuqna A . . .muficaa . . svucu~.:augu

m. oxxn§> 59> an? a oExua> xL.> an? fioég

 

SE 0.0 i 0.."

 

guELu no gauging

 

thfighdgHggg
Ehgh¢§ul§§§8§§§ho§mng

gag

75

When octylglucoside is substituted for Triton X-100,
changes in the substrate specificity pattern of CPT assayed
in the forward direction with 1.1 mM L-carnitine and 100 uM
acyl-CoA are apparent (compare shaded areas in Figure 17).
Although maximum activitity under these conditions is
obtained with decanoyl-CoA with both detergents, the
relative activity for octanoyl—CoA is much less in

octylglucoside and the relative activity with long-chain

 

acyl-CoAs is greater. Also shown in Figure 17 is the
acyl-CoA specificity of OPT in 12 mM octylglucoside at
saturating levels of L-carnitine (unshaded area).
Saturation of the enzyme with L-carnitine increased the Vm

ax

with medium-chain acyl-CoAs relative to its V

max with long-

chain acyl-CoA but to a significantly lesser extent than in
Triton x-100. Octylglucoside alters the kinetic constants
of the enzyme significantly. As shown in Table IV, at
pHsB.O and 200 uM palmitoyl—CoA the KO.5 for L-carnitine is
4.9 mM in 12 mM octylglucoside. This is in contrast to
assays in Triton x—ioo where the Ko.5 for L-carnitine is
only 0.2 mM. A 10-fold difference observed in Triton x—1oo
between the Ko.5 for L-carnitine with hexanoyl-CoA as
cosubstrate and the K0.5 for L-carnitine with palmitoyl-CoA
as cosubstrate is reduced to a 2-fold difference in
octylglucoside.

The effect of increasing concentrations of

octylglucoside on the kinetics with acyl-CoAs is shown in

76

Table VII and Figure 19. Octylglucoside lowers the K0 5 for

palmitoyl-CoA. At pH 6.25 the K for palmitoyl-CoA is

0.5
24.2 uM in 0.1x Triton X-100 in contrast to 3.1 pH in 12 mM
octylglucoside. However, octylglucoside inhibits CPT

activity with octanoyl-CoA raising its K0 5 for octanoyl-CoA

and acting like a competitive inhibitor with an apparent Ki

of 15 mM for octylglucoside (see Figure 19).

The Effect ofng on the Kinetics of OPT Since the response

 

of membrane bound CPT to malonyl-CoA is very pH dependent
(130), the effect of pH on the kinetics of OPT was
determined. Substrate-velocity plots at pH 6.0 and pH 8.0
are shown in Figure 20. In Figure 20A, palmitoyl-CoA is the
variable substrate and, in Figure 208, L-carnitine is the
variable substrate. The sigmoidal behavior of CPT with
palmitoyl-CoA is more apparent when low substrate
concentrations are used at pH 6.0 (see Figure 20A). The

insets show the differences in Vma at pH 6.0 and 8.0. The

X

K0 5 for both L-carnitine and palmitoyl-CoA is larger at pH

6.0 than at pH 8.0, the Vmax was greater at the lower pH

(see Table VIII). Likewise, in octylglucoside the K0 5 for
L-carnitine increases as the pH decreases from 2.2 mM at pH

8.0 to 9.8 mM at pH 6.25 and the highest Vma is at pH 7.0

x
(see Table IX).

Effect of Malonyl-CoA and TDGA—CoA on CPT Preliminary

 

1.:

 

'17

 

TABLE VII
EFFECTOFOCTYLGUJOOSIDEONMKINETICPAWOFCPTFOR
PAIMITOYL—CoA
Octylglucoside KO V Hill n
(W) “I“? nfi‘i‘t‘s
0 10.8 5.8 1.4
6 2.6 5.6 1.8
12 3.1 7.7 1.8
18 3.1 6.1 1.9

 

L—carnitine m 40 mM ani the palmitoyl-OoA concentration was varied.
Assays were due at [ii 6.25, with the DTBP assay using the kimtic

analyzer described mder Metlnds.

78

dd ufiiflflofiioefl gang
moment: 5 weapons fig 03% 05. gas Base 0.3 3.3.9
u 05. 48:16:88 .80 .88 no u «5 8 «Egan? no vacuum .2 9568

:25. 8.33.933

 

     

Swamp"?

l
O
N

I
l
O
Q

 

:—
O
CO
(mi) 9°): v03 - Mouezoo

79

 

 

IO'V. wholes/min
o
&

 

 

 

 

 

 

2.0 4.0 ' 6:0 ' 0:0 No.0 IZIO
(palmitoyl-CoA).pM

 

.
cl!-
III-

 

 

10‘ v,,.molomnin

 

 

 

 

 

 

 

A

 

I A

V v

6 l2
1 A l
U I

0.6 0.8 Ito
(L-carnitine).mM

l l A l
l I I 1

0.2 8.4

Figure 20. Effect of w m the velocity versus substrate concentration
curves for CPT. In A, palmitoyl—OoA oamtratim is varied.
The canoentratim of 12.-carnitine is 6.0 m. In B, the
mtratim of Ir-carnitine is varied. The canoentratim
of mlmitoyl-OoA is 200 pH. The insets show the velocity
versus sdastrate profile over a greater range of substrate
concentrations. The mits for the insets are the sauna as the
units of the figure. bperimtal details are described in
mthods.

80

.3052 0:0 5.. gugnam

3 3160 no H9: 0:0 95280 8. E00008: 855.0 no 0.30 may 0« rwbwuom no page 0.5

.0005»: 5 00920000 00 F000 09b 05 05.00 nonzda 03050— 0.? 5? 005.300 0003 300 05. .5 con
0n ¢8|H>0u§ .5. 0.0 0w menu? 00.3500... 000:» 08 g 0.! .008 any £800.50.» 38 one

 

.o.fiuo.~.~.fi Am.mno.~.o.o .m.~u~.~.m.« Ace.flnon.«.nm.fi in.muo.m.n~.m An~.oumfi.oso«.o o.m
An~.Hsno.H.o~.fi Ao.vfiuo.mfi.a.efi .o.~fiuo.ms¢.on loo.~nn¢.fison.fl Ao.nflnv.nasn.mfl .m.~u¢m.osoa.a o.~
Ao.a-~.dvn~.fl .«.on-~.ma.m.nfi .m.n~uc.o~.~.¢~ .n~.~uno.nvmo.fi Ao.m~um.sfl.o.ofl an.mun.~.mo.~ 0.0
_= Has: .ouscsg; xua> mans 0.6x _: as“: masseuse nuns. .22. 0.8x :0

 

 

eooufisoufiaamm acfiuficumoLq

 

SnlnggagmomfigfimOHEU—EZOEBE

HHH> mama

.00oaumzfca 009w00000
00 >008 men 05 9309 fig 030:9 05 5“: 0055—000 0003 83%
.30 003883 0380. an. .2... «H 3 03.22958 .5 8H 3 30.580230

 

81

o.

N
N

a.0u

m.«
04 oéa o6 as
a.” 0.2 a.» has
0." 0.2 0... o.~.
0.” «.9 oh and
m.” 53 vs 0.0
a; 0.3 06 8.0
0 £5
c :E mM.-she m on :0

 

mQHmOODQQQMHOO 2H NZHBHZMGULQ MOE 8&0 Q0 UHBNZHM M29 20 an MC BOBMMN

xH mam¢fi

82

initial rate assays at pH 8.0 showed no effect of

malonyl-CoA on the V . The purified enzyme has a low K

max 0.5

(1.9 uM) for palmitoyl-CoA at pH 8.0, and since malonyl-CoA
inhibition had been reported to be of a competitive type, it
was uncertain whether the lack of malonyl-CoA inhibition
resulted from oversaturation with the acyl-CoA and
malonyl-CoA concentrations used in the initial rate assay,
37.5 0M and 1-5 uM, respectively. Though membrane-bound CPT
had been reported to show significant inhibition with
similar assay conditions (34) the use of BSA or the presence
of a membrane could have reduced the effective concentration
of palmitoyl-CoA. Alternatively, the K0.5 for palmitoyl-CoA
could be higher in the membrane bound state. The kinetic
analyzer is very well suited for covering a range of
palmitoyl-CoA concentrations below and above the enzyme's
K0.5 for the substrate, conditions which allow the effects
of a "competitive" inhibitor to be best observed.
Experiments were performed with palmitoyl-CoA concentrations
above and below the [(0.5 using high and low malonyl-CoA
concentrations. No effect of malonyl-CoA on CPT activity
was obtained in any of the experiments, even when the
malonyl-CoA concentration was 600 uM. The malonyl-CoA
effects on the kinetics of CPT at pH 6.0 with saturating
levels of L-carnitine are summarized in Table X. When
subsaturating levels of L-carnitine are used, again no

effect of malonyl-CoA is obtained.

1.!

 

83

TABLE)!

EFFECIOFWDNYIrCOAONTHEHNETICSOFCPIWImPALMITOYIrCOA

 

 

Malonyl—CoA
nme 3000M 6000M
IS“ 24.2 22.1 25.4
Vmax 15.5 14.98 15.0
Hill n 1.73 1.8 1.9

 

Palmitoyl-OoA is the variable substrate and 12.—carnitine is 6.0 mM.
The kinetic determirations were done in 0.196 in Triton x-ioo at w
6.0asdescribedfortheDTBPassayinmthods.

84

In contrast, 2-tetradecylglycidyl-CoA does have an
effect on the kinetic parameters of CPT for palmitoyl-CoA.
As shown in Figure 21, TDGA-CoA altered the shape of the V
versus [S] curve for palmitoyl-CoA, changing the kinetics
from sigmoid to hyperbolic. The effect of TDGA—CoA on the
kinetic parameters of OPT with palmitoyl—CoA are summarized
in Table XI. At pH 6.0, the Hill coefficient for
palmitoyl-CoA is reduced to 1.0 in the presence of 40 0M

TDGA-00A.

Kinetic Characterization of Membrane Bound

Rat Heart Mitochondrial Carnitine Palmitoyltransferase

‘ffect of acyl-GOA on mitochondrial swelling; As shown in
Figure 22, continuous addition of low concentrations of
palmitoyl-CoA to the assay mixture containing intact heart
mitochondria causes a large drop in absorbance which is
apparently due to the swelling and rupturing of the
mitochondrial membrane caused by the detergent properties of
the palmitoyl-CoA. Addition of 1.4x BSA eliminated the
swelling but the enzymatic activity can only be measured at
much higher concentrations of palmitoyl-CoA (100 uM). With
such high concentrations of palmitoyl-CoA, BSA did not
completely prevent swelling. In contrast, decanoyl-CoA,
which has a higher cmc, does not induce swelling of the

mitochondria during the course of the reaction even with no

85

 

.ooHIX :09th &«.o MCQUCOU

00.303: >009... 05. .25 0.0 0w. 0032.301; .06 :0 #0 >080 00.5 93 5?. 8353.8
030:0. 05 0:0: 020538 0.8 832880 0395. ..Eo 8 400.3 «a 0.00000 .3 086E

.21 .48.. 322.60

 

0.0. 0w. $0 0.0 00.0
as}...
.5. u c ___z ...
£630.06: 0.5 u on> ......-

 

    
      

218 "of ..
48-52 218/. ...........
eon-ea 060}
O O. 0.
P800 It
7 m._ u c a:
C_E\m0_OEC 0.0_ u x05)
.21 mm u 8x
$6.35 9.

 

 

DON

U!w/ SBIOLUT” ‘AZOI

86

.8200:
E >88 008 05 .80 830.880 8 od :0 um 87x 80000. 5.0 3 88 as...
mama... 8.0 .2... od 00 80020810 Ba 38083 03209 £0 «0 3010000330

 

 

 

 

 

o.” 0.0 a; a
8 3 2 $0035 flm
8 3 «a :05 n
5 9 z: 8 08:
80:480.

 

EEQEEEUEHEMOWOHEEZOEOIgBEE

Hung

87

 

0.5-

 

 

 

 

 

 

 

Decency! - CoA
m
g "j
<
3%
o 0.25
(n
CD
<I
palmitoyl - CoA
Q _
1 1 1 1
O 25 50 75 IOO

Acyl " COA , PM

Figure 22. Effect of decamyl—Goh an: palmitoyl-OoA m mitocl'ndrial
swelling. Manor-bane dings were monitored at 412 nm in
a Gilford 2600. The acyl-00A cornentration was cmtinnusly
imreasedwithapnp drivan mime intoa 2 m1 cuvette
caxtainixg 20—50 pg of mitochcnirial protein in the D'INB
assay biffer with no detergents. 150 ‘4 K01, am w 8.

88

BSA added; see top curve of Figure 22. These data show that
decanoyl-CoA can be used in the kinetic study of membrane
bound CPT-I similar to the substrate addition experiments

with purified CPT.

Kinetics of CPT-I and Effect of Malonyl-CoA. Figure 23

shows that the kinetics of CPT-I with decanoyl-CoA at
saturating levels of L-carnitine are sigmoid rather than
hyperbolic. The Ko.5 for decanoyl-CoA is 3 pM and the Hill n
is 2.0 in both fed and fasted states. Figure 24A and 248
show the effect of varying concentrations of malonyl-CoA on
the V versus [S] curve. The saturation curve is displaced

to the right and the K0.5 for decanoyl-CoA increases without
affecting the Vhax or the Hill coefficient. At 20 uM
malonyl—CoA almost 100% of the activity is inhibited at low
decanoyl-CoA concentrations, but as shown in Figure 248,

this inhibition can be reversed by increasing the
decanoyl-CoA concentration. Table XII shows the effect of
malonyl-CoA on the Ko.5 for decanoyl-CoA of mitochondrial
CPT-I from fed and fasted rats determined under optimized
conditions. The K0.5 for decanoyl-CoA increases to 185 pH in

the presence of 20 pH malonyl-CoA in fed rats.

1 (app) —-f°r

Malonyl-CoA. Figure 25 shows the effect of feeding and

Effect of Feeding and Fasting on the K

 

fasting on malonyl-CoA inhibition of heart CPT-I. Fasting

89

 

 
 

 

 

4.0--
_|>
V
'9 ..
2.0.. K015 "' 38PM
Vmox‘7-5 munits
Hill n = 2.0
/ -o.:2 012 (5.4
IO"2( In)
" (S)

 

Figure 23. Dmble reciprocal plot of mitochondrial CPT—I reaction
velocity versus decanoyl-CoA conceltratim. 12.-carnitine
is 6.0 uM. Assay ocuiitions are as described in Figure 22.
Mitochordria were isolated from fed rats as moribed in

mthods.

 

90

 

 

 

 

 

0.8-
.5
g 0.6-
n
'3
E
1
50.4-
N
E
0.2-
...x e 21M
00 A I l

 

' 0 5 110 1‘5 20 25
Decanoyl - 60A. 11M

 

 

 

 

2.0 I I
a
N0 molonyl-COA
LS" "
.c
E
\
E»
g I 0 __ ZOpM ..
1
N).
9
0.5 - "
l l
0 50 IOO
Deconoyl-COA, FM
Figure 24. Eff of malonyl-CoA on manbraxe bouni CPT-I. L—carnitine

ect

6.0 uM. Activity is measured usixg the DTNB assay at
8.0. All other corditions are as described in Figire

2. 'me assay is started immdiately after the adiitim of
nlonyl-OoA. Heart mitochondria were isolated from fed rats
as described in mthods.

”2'5”

91

TABLEXII

mammmnmsrmoummn—cmmmmonor
W KID!) CPT-I

 

K0_5 for deoanoyl-CoA

 

 

mlonyl-OOA (114) Fed Fasted

none 3 3

0.5 5 -

1.0 9 -

5.0 34 7

10.0 - 14

15.0 — 18

20.0 185 25

The values are determined from Hill plots similar to the

cne '5n in Figure 23. The asay coalitions are as described in
Figire 22.

92

.mu 833 5 03288 8 2a 8.03850 .850 a2 .5 8 8: consign—oi:

.HIHLU Egon gang «0 Savanna.“ gflgg 8 8.38M Ea Dragon «0 Hoouum .mm 055?“

Si .48 - 388mm

 

om. om. om ov
_ _ _ O
1 no
mm
2
.A
1 0.. in
w
p
to“. w
/
m.
U
8.8... .. ma

 

 

 

93

significantly reduces the response of CPT-I to a fixed
concentration of malonyl-CoA. The secondary plots given in
Figure 26 show that there is a ten fold increase in the
Ki(app) for the inhibitor due to a 48 hr fast. CPT-I of
mitochondria from fed rats has a K for malonyl-CoA of

1(aPP)
0.3 uM while those from fasted rats is 2.5 uM. The replots of
Ko.5(app) against malonyl-CoA concentration are linear in
the fasted state, but appear to curve upward at high
inhibitor concentrations in the fed state. This suggests
that there are two inhibitor binding sites in the fed state.
However, from the shape of these curves, it cannot be

determined if there is any cooperativity between these sites

(131).

Kinetics of CPT-I with L-carnitine. Figure 27 shows a
substrate-velocity plot with L-carnitine as the varied
substrate. The K0.5 and Hill coefficient for L—carnitine
varied considerably from preparation to preparation of
mitochondria. The K0.5 varied between 0.2-0.7mM and the
Hill coefficient varied between 1.2-1.8. The cause for this
variability is unknown. Addition of malonyl-CoA did not
alter the K0.5 for L-carnitine. For example, for a given
preparation of mitochondria which had a K0.5 for L-carnitine
of 0.6—0.7 mM. the K005 for L-carnitine in the presence of

20 pH malonyl-CoA was an average of 0.65 mM.

94

 

 

 

1
l
/
150 -- /
/
=2; /
:53 /
3;. IOO-- /
g Fed /
é /
53 1
/
50-- /
/
//‘ . :osted
” 10 2'0

Molonyl 'CoA, pM

Figure 26. Replot of K0.5 versus malonyl-CoA concentration. The
K0 are determined from the best linear fit to the
H11? equation of the substrate—velocity data obtained as
described in Figure 23. Data collection is started
inmediately after addition of malonyl-CoA.

95

.mm enema 5 0.3308“. mm was 9833500 woman 05. .2: 03 we 30115508
.258 sofiumuucoguoo gag new“? 5.3ng 83089 H.450 853199552 .5 enema

.26 . 05:58 .. 4
QN 0..
_

 

_ A — O
Q. n =
£52 25... 65>
25 .v. u Ex

I
O
uyw/sauowfi 'AZOI

 

 

 

ON

 

DISCUSSION

The kinetic data show that purified beef heart
mitochondrial CPT exhibits non-Michaelis-Menten kinetics
with both substrates, the acyl-CoAs and L-carnitine. These
data suggest that CPT is allosterically regulated. This
suggestion is consistent with the proposed role for CPT in
the regulation of fatty acid metabolism. In contrast, when
soluble monomeric COT purified from mouse liver peroxisomes
(14) is used as a control in kinetic studies, with identical
assay conditions, linear kinetics are obtained with Hill
coefficients near 1.0.

The sigmoidal kinetics of CPT could result from subunit
interaction of varying degrees in purified CPT aggregates.
Alternatively, sigmoidal kinetics could arise from
allosteric interactions between more than one substrate
binding site on the enzyme. The high molecular weight of
purified beef heart mitochondrial CPT and of other forms of
CPT purified in detergents (see Table 1), suggests that CPT
is an aggregated enzyme. Taking into account the
phospholipid bound to the enzyme and the contribution of a
detergent micelle, one can estimate that an associated form
of CPT in Triton x-100 might contain at least 4-6 monomers.
If several monomers were anchored through a hydrophobic
region into a detergent micelle, an apparently associated

96

97

form of CPT could result, in which subunit interaction was
minimal. However, these studies show that in the absence of
detergent micelles, CPT has a molecular weight of 660,000 on
Fractogel TSK. These data suggest that CPT is an enzyme
aggregate that interacts with detergents. The higher
molecular weight of OPT in octylglucoside could be caused by
increased aggregation of CPT monomers in the absence of
detergent micelles, or perhaps more likely, from a large
amount of octylglucoside binding to the aggregates (membrane
proteins can bind up to their own weight in detergents
(132)). A simple explanation for the oligomerized form of
purified CPT is that CPT is an oligomer in the membrane,
though. the subunit structure of the native membrane bound
CPT might be different from the subunit structure of the
protein-detergent complex.

There are several other possible causes for
non-hyperbolic dependence of initial velocity upon substrate
concentration. Non-linear kinetics can be seen when two
enzymes catalyzing the same reaction have significantly
different affinities for the substrates. This situation
could arise in our system in the unlikely event that the two
forms of mitochondrial CPT were indeed separate enzymes with
very similar physical properties which had co-purified.
However, in this case double reciprocal plots would resemble
those of a system exhibiting negative cooperativity with

Hill n less than 1.0 (131). The data with purified CPT

98

shows strong positive cooperativity for CPT with both
acyl-CoA and L-carnitine and therefore does not indicate the
presence of isozymes.

The continuous addition of substrate and rapid mixing
enabled the calculation of initial rates with the TANKIN
program at low substrate concentrations (0.1-12 uM
palmitoyl-CoA) facilitating the study of the kinetic
properties of CPT. This is in contrast to previous studies
where the precision and assay sensitivity were limited due
to manual addition and mixing of the substrate, the rapid
conversion of the substrate to product, and a lower
sensitivity of the spectrophotometer used. In order to
determine whether the kinetic parameters obtained with the
TANKIN program were in any way artifactual several tests for
product inhibition and enzyme inactivation were conducted as
described by LeBlond et al. (123). In all instances, no
evidence was obtained that indicated a bias in the
parameters. Evidence has been presented by some
investigators that DTNB is inhibitory when preincubated with
pigeon breast muscle carnitine acetyltransferase (133) and
that DTNB can abolish the apparent cooperativity with
respect to palmitoyl-CoA of membrane bound CPT (36). Under
the assay conditions used in this study, no evidence of
enzyme inactivation was obtained. The rates and progress
curves obtained using the DTNB assay were compared to rates

and progress curves obtained by following the disappeareance

99

of the acyl—CoA substrates at 232 nm in the absence of
sulfhydryl reagents. Within experimental error, these were
identical. Therefore we can conclude that under the
conditions of our assay, DTNB does not cause enzyme
inactivation nor does it abolish substrate cooperativity.
These data also show that CPT exhibits very different
kinetic constants depending on the experimental conditions
in which its activity is assayed. As a consequence, the
choice of experimental conditions for the determination of
substrate specificity profiles can greatly influence the
results. This, undoubtedly, is a major factor in the large
differences in the data from different laboratories
reporting on the same enzyme, see Table II. When the
kinetics of purified beef heart mitochondrial CPT are
studied in micellar concentration (0.1%) of Triton x-1oo,
the data show that the affinity of the enzyme for
L—carnitine depends on the acyl-CoA chain-length, namely a
decrease in the chain-length of the acyl-CoA increases the
“0.5 for L-carnitine. Low concentrations of L-carnitine
decrease the catalytic efficiency with hexanoyl-CoA and
octanoyl-CoA shifting the specificity of the enzyme towards
long-chain acyl-CoAs. As shown in Table VI, at low
concentrations of L-carnitine, vmax/K0.5 is maximized, while
that of medium chain acyl-CoAs's is not. This would
indicate that the enzyme has evolved to use long-chain

acyl-CoAs at low, physiological L-carnitine levels. This

100

also suggests that fluctuations in the concentration of
L-carnitine can affect the "physiological substrate
specificity" in the forward direction. The data do not
support the suggestions that the enzyme is inhibited by
increasing substrate concentrations (30), nor that
increasing substrate inhibition (as the acyl- chain
lengthens) at the L-carnitine site is a valid explanation
(5) for the higher activity of OPT with medium-chain
acyl-CoAs.

Major differences in the kinetic parameters of CPT
occur when the detergent is changed from Triton X-100 to
octylglucoside and when the concentration of detergent is
altered thus affecting the substrate specificity pattern of
OPT. In 12 mM octylglucoside the K0.5 for L-carnitine with
medium-chain acyl-CoA's is near 10.5 mM. This large
increase in the K0.5 for L-carnitine results in very low
activity with the assay conditions in Figure 17 where the
L-carnitine is 1.1 mM. However saturation with L-carnitine
does not shift the specificity of the enzyme towards the
medium-chain acyl-CoAs, as it does in Triton X-100. The
shift in the substrate specificity of CPT towards higher
activity with long-chain acyl—CoAs in octylglucoside appears
to result from a combination of effects. Octylglucoside
lowers the K0.5 of OPT for long-chain acyl-CoAs and this may
be related to the formation of mixed micelles with these
substrates. The lower activity of CPT with medium-chain

acyl-CoAs apparently is due to a moderate competition by

 

101

octylglucoside monomers with octanoyl-CoA for the enzyme's
active site. In addition, octylglucoside may increase
catalysis with the long-chain acyl-CoAs by interacting with
the enzyme to give increased rates. In a submicellar
concentration of octylglucoside (12 mM) the enzyme remains
in solution after centrifugation at 100,000 g for 15 min,
while in the absence of detergent significant activity is
removed from the supernatant. This suggests that the enzyme
can interact with detergent monomers perhaps through
hydrophobic regions of the protein. Though direct
measurements of detergent binding to CPT were not made, the
effects of octylglucoside on the enzyme's kinetic
parameters, in particular its apparent competitive
inhibition of CPT reaction with octanoyl-CoA and its

capacity to increase the K for L-carnitine 20-fold

0-5(&PP)
from approximately 0.2 mM to 4.5 mM. indicate a strong
interaction at specific sites. One can speculate that
hydrophobic site(s) in the enzyme, may have a functional
role in the control of enzyme activity through allosteric
effects, by binding a membrane component, or acyl-CoA
substrate. Such an allosteric site could be involved in the
effect of varying hydrophobicity of the acyl-CoA substrates
on the enzyme's K0.5 for L-carnitine. It is interesting
that octylglucoside abolishes in part this effect, see Table
IV.

Studies with crude homogenates (134) and with isolated

102

mitochondria (82) have shown that the Km for L-carnitine
spans a 20 fold range depending on the tissue and species
examined. This variation in the Km for L-carnitine may be a
function of the membrane environment. It would be
interesting to see if, once solubilized from the membrane,
CPT from different tissues still shows significantly
different affinities for L-carnitine.

Palmitoyl-CoA and L-carnitine were used to investigate
the pH effects on the non-Michaelis-Menten kinetics of OPT.
DTBP was used rather than DTNB because it has a higher
extinction coefficient that is constant in the range of pHs
studied (28), and because it is not inhibitory to carnitine
acetyltransferase (133). This makes the assay more
sensitive, which allows the use of lower enzyme
concentrations with reduced product accumulation. At pH 6,
Hill coefficients for both substrates were in the vicinity
of 1.8. CPT activity is strongly pH dependent in the range
where respiring mitochondria develop a pH gradient during
oxidative phosphorylation. Since the matrix pH normally is
considerably higher than the inter—membrane pH in
mitochondria, this pH difference could affect the catalytic
properties of membrane bound carnitine palmitoyltransferase
depending on which surface of the inner membrane the enzyme
is located. As shown in Figure 20A, a low pH causes low
amounts of L-carnitine to greatly limit the enzyme's

catalytic capacity. However, at saturating amounts of

103

L-carnitine, the enzyme attains a vmax that is nearly twice
that attained at high pH (Table I). Thus low amounts of
L-carnitine might affect the activity of the outer form of
OPT more, since this form should be exposed to a lower pH.
This could be important in certain systemic carnitine
deficiencies where low tissue carnitine causes metabolic
abnormalities (135). The inner form of carnitine
palmitoyltransferase, being exposed to a higher pH, should
have a lower maximum capacity to catalyze the formation of
acylcarnitines conforming to its ascribed function in the
matrix of mitochondria, but it would not be as sensitive to
regulation by L-carnitine levels. Besides pH differences,
the membrane environment on both faces of the inner membrane
contain different protein and phospholipid composition (92).
These differences would also be expected to contribute to
the catalytic behavior of CPT ;g 3129.

A rise in pH has been reported to result in a marked
decrease in the sensitivity of membrane bound CPT (130) to
malonyl-CoA inhibition. Although lowering the pH increased
the “0.5 for palmitoyl-CoA of purified CPT, (this in
contrast to the data reported by Mills et al.(83) where no
effect of pH on the Km for palmitoyl-CoA of membrane-bound
CPT was observed) no effect of malonyl—CoA on the kinetics
of purified CPT was observed. The lack of malonyl-CoA
inhibition of the pure enzyme is not considered as evidence

that malonyl-CoA does not affect the membrane bound enzyme.

104

Rather, it seems likely that the purified enzyme has been
removed from some component which can influence its
catalytic activity, i.e., a regulator subunit or a special
membrane environment. Alternatively, the purified enzyme in
the high detergent concentrations may have lost a separate
malonyl-CoA binding site. We tested concentrations of
malonyl-CoA ranging from 20 uM to 600 pm at saturating and

non-saturating levels of L-carnitine over a wide range of

 

palmitoyl-CoA concentrations and neither K V or Hill 2

0.5'
n for palmitoyl-CoA were affected. This indicates that

max '

malonyl-CoA is not a competitive inhibitor in contrast to
the tentative conclusions by others (29). However, the fact
that the enzyme has a high affinity (K0.5'°'2—2‘ pH) for its
acyl-00A substrate and completely lacks a response to
malonyl-CoA at concentrations 100-500 times greater than
concentrations that are inhibitory with intact mitochondria
indicate malonyl-CoA normally must bind at some other site
on the native enzyme. Thus, it seems likely that
malonyl-CoA is exhibiting some allosteric interaction which
is lost upon solubilization of the enzyme. Recent data in
the literature strongly suggest a separate binding site for
malonyl-CoA. These reports have been discussed previously,
see Introduction.

Unlike malonyl-CoA, TDGA-00A, a substrate analog, did
have an effect on the kinetics of CPT. Figure 21 shows that

TDGA-00A altered the shape of the velocity versus

105

palmitoyl-CoA concentration curve, changing it from
sigmoidal to hyperbolic and increasing the “0.5 for
palmitoyl-CoA. These effects of TDGA-CoA on the kinetics of
OPT are those expected of a system with cooperative
substrate binding, in which an inhibitor mimics the
substrate (127) producing competitive inhibition at two
sites. As [I] increases. K0.5 increases, and the Hill n
approaches 1. Therefore the kinetic data indicate that
TDGA-CoA is inhibiting the enzyme by mimicking the acyl-CoA
substrate and destroying the substrate cooperativity, but it
is not inhibiting CPT irreversibly, since increasing

concentrations of palmitoyl-CoA restore V This

max'
observation is in contrast to the effect of TDGA-CoA on

membrane bound CPT, which appears to be irreversible (91).
Thus, it appears that TDGA-00A may have more than a single

site of action i vivo. The effect of TDGA-00A on the

 

purified enzyme, occurs at much higher concentrations of
inhibitor than are necessary to produce an irreversible
inactivation of membrane bound CPT. The effect of TDGA—CoA
on hepatic mitochondria and inverted submitochondrial
particles has been recently investigated by Brady et al.
(95). Their studies demonstrate that it is possible to
almost totally eliminate malonyl—CoA sensitivity in inverted
vesicles while retaining a relatively high degree of
sensitivity to TDGA—00A. This result also suggests multiple

sites of action for this inhibitor.

 

106

Reports in the literature about the kinetics of CPT-I
are discrepant. Substrate sigmoidicity has been reported in
(21.81) while other studies (29,34) report linear
hyperbolic kinetics, which become sigmoidal after addition
of malonyl-CoA to the assay medium. Therefore, it has been
suggested (34) that malonyl-CoA acts by a mechanism
involving cooperative inhibition. CPT of other nonhepatic
tissues such as heart and skeletal muscle have greater
sensitivity to inhibition by malonyl-CoA and contain
significant quantities of malonyl—CoA in the fed state
(113,83). Though these tissues do not have coordinated
control of fatty acid synthesis and oxidation, their energy
demands can vary greatly.

There is increasing interest in the study of the
mechanism of malonyl-CoA inhibition of CPT in both hepatic
and nonhepatic tissues. Our data for heart mitochondria
show that CPT exhibits the same degree of sigmoidicity with
decanoyl-CoA in the absence or in the presence of
malonyl-CoA. In addition, the kinetic parameters determined
for membrane bound CPT—I are similar to those of the
purified enzyme from beef heart mitochondria.

One possible explanation for the failure of some other
groups to detect sigmoidicity in the absence of malonyl-CoA
is that the relative sigmoidicity of two curves is not
always apparent when the displacement along the x- axis

differs due to different values of K Sigmoidicity

0-5(8PP)'

 

107

can remain undetected if the substrate range examined is

high compared to the K0.5(app)‘

progress curves for CPT-I with no malonyl-CoA because of the

This is common in the

low K0.5 for the acyl-CoAs. Addition of malonyl-CoA
increases the K0.5 sufficiently to make the substrate range
examined optimal for visual detection of sigmoidicity.
Analyses of the steepness of an apparently hyperbolic
progress curve of data reported by others shows considerable
cooperativity. This can be done by determining the
[S](.9vhax)/[S](.1Viax) ratio (131). This ratio is 81 for
a hyperbolic curve and 9 for a curve with a Hill n value of
2. Any value in between 9 and 81 indicates varying degrees
of sigmoidicity. For example, if this criteria is applied
to the data in Figure 1 of Reference (81), a [S].9/[S].1
ratio of about 9 is obtained for the palmitoyl-CoA
saturation curve with no added malonyl-CoA. However,
hyperbolic kinetics are described in the paper. Analysis of
the data by obtaining the best fit to the Hill equation
seems warranted in such situations.

The change in K for malonyl-CoA as a result of

1(app)
fasting has been shown for liver CPT-I. Our data suggest
that a similar mechanism for sensitization-desensitization
of mitochondrial CPT to inhibition by malonyl-CoA found in
liver exists in heart. Our studies were conducted with

freshly prepared heart mitochondria isolated by a new

procedure (128) using very young rats. Although the

 

108

conditions we used for studying CPT-I interaction with
malonyl-CoA, (DTNB assay and pH 8.0) have been avoided by
some because of the potential reduced sensitivity of OPT to
inhibtion by malonyl-CoA, our data clearly show strong
interaction of the same magnitude reported in (34) where a
different assay was used. These observations suggest that
the mitochondrial preparation procedure and/or the time
elapsed from the isolation of the mitochondria to the time
analyses are done can affect the interaction of malonyl—CoA
with CPT-I. Time dependent changes in the capacity of
malonyl—CoA to inhibit CPT-I and to bind to rat liver
mitochondria in giggg have been described (136,137). The
secondary plots presented in (34) for the rat liver enzyme
are very similar to the ones presented here for the heart
enzyme. In the fed state there appear to be two inhibitor
binding sites. Whether these sites can interact in a
cooperative manner cannot be determined from our data. In
the starved state, the data shows one inhibitor site of
lower affinity for malonyl-CoA.

Higher values of the Ho's for L-carnitine have been
measured in tissues with a lower sensitivity to malonyl-CoA
inhibition (82). In an attempt to correlate a change in the
K005 for L-carnitine of rat heart CPT-I to changes in the
sensitivity to malonyl-CoA inhibition, we encountered
considerable variability in the K0.5 for L—carnitine in the

control (fed state). This variability suggests a mechanism

109

to regulate the catalytic efficiency of CPT which remains to
be elucidated.

The data presented here can be explained by several
kinetic models which involve the binding of the inhibitor to
a regulatory subunit. For example in the "ligand exclusion"
model an inhibitor can bind to a separate regulatory site
which may have overlapping functional groups with two or
more substrate binding sites. For an enzyme with a high
degree of interaction among catalytic subunits, an infinite
amount of inhibitor would increase K0.5' but would not
significantly alter the sigmoidicity of the progress curve.
If there are multiple substrate binding sites on the enzyme,
not all catalytically active, then dissociation of the
enzyme would yield a catalytic subunit insensitive to
inhibitor and a regulatory subunit that binds both substrate
and inhibitor. An alternate model, would involve a partial
competitive inhibition by malonyl-CoA with different
substrate and regulatory binding sites. When there is
cooperative substrate binding in the absence of inhibitor,
addition of the inhibitory regulator makes the enzyme
behaves like a new enzyme with the same interaction between
subunits but a new K0.5' Both these models agree with our
data which show strong positive cooperativity for the
acyl-CoAs. The increasing amount of indirect evidence in
the literature for the interaction of malonyl-CoA with a

membrane component, possibly a regulatory subunit of CPT-I,

 

110

other than the catalytic subunit itself, strongly support

these conclusions.

 

SUMMARY AND CONCLUSIONS

These data strongly indicate that CPT is a highly
regulated, allosteric enzyme which exhibits cooperativity
towards its multiple substrates. The molecular weight of
CPT by Fractogel TSK chromatography in the absence of
detergent micelles indicates that CPT is an aggregate of 4—6
monomers. The similarity in the substrate kinetics of the
purified and membrane bound forms of CPT suggests similar
subunit aggregation in the membrane bound state.

The data show that the affinity of the enzyme for
L-carnitine depends on the acyl-CoA chain-length, namely a
decrease in the chain-length of the acyl-CoA increases the
K0.5 for L-carnitine. This suggests that the concentration
of L-carnitine can affect the "physiological substrate
specificity". In addition, the data do not support
suggestions that the enzyme is inhibited by increasing
substrate concentrations. These data indicate that some of
the differences observed in the specificity pattern of CPT
from various preparations or laboratories reflect the level
of saturation of the enzyme with its substrates which result
from the variable “0.58 with different assay conditions.

It is concluded that malonyl-CoA is not a competitive

inhibitor of CPT but acts like a negative allosteric

111

112

modifier of membrane bound CPT-I by binding at a site other
than the catalytic site. These data indicate that during
purification, CPT has been separated from the malonyl-CoA
binding component of the mitochondrial membrane since it has
lost all sensitivity to the inhibitor. The increased
sensitivity to malonyl-CoA inhibition with changes in the
physiological state of the enzyme result from changes in the
apparent K1 for the inhibitor of the enzyme.

The complex response to changes in substrate
concentration, pH, and hydrophobic environment, provides an
explanation for the observation that CPT from beef heart
mitochondria appears to be a single protein, yet in_gigg
exhibits different kinetic properties on the cytosolic face
as compared to the matrix face of the inner membrane of
mitochondria. The data support the proposal that CPT is the
key regulated step in the control of mitochondrial fatty

acid oxidation.

GNU)

CD

10.

11.

12.

13.

14.

15.

LIST OF REFERENCES

Gulewitsch,W., and Krimberg,R.(1905) §,Physigl.0hem.
15.326.

Fraenkel,G., Blewett,M., and Coles,M. (1948)

Nature (London) 161,981.

Tanphaichitr,V., and Broquist,H.P.(1973) J.Biol.Chem.
318.2176.

Bremer,J. (1961) Biochim.8iophys.Acta g1,622.
Bremer,J. (1983) Physiological Reviews §§,1420.
Fritz,I.B. (1955) Acta Physiol.Scand. 31,367.
Bremer,J. (1962) Nature 126,933.

Bremer,J. (1963) gggiol.0hem. g§§,2774.

Fritz,I.B., and Yue,K.T.N.,(1963) J.Lipid Res. 1,279
Bieber,L.L.,Emaus,R.K.,Valkner,K., and Farrell.S.(1982)
Fed.Proc. 11,2858.
Choi,Y.R.,Fogle,P.J.,Clarke,P.R.H.,and Bieber,L.L.
(1977) J.Biol.Chem. g_g,7930.

Markwell.M.A.K., Tolbert,N.E., and Bieber,L.L. (1976)
Arch.Biochem.Bigphys. 176,479.

Farrell,S., Vogel,J., and Bieber,L.L. (1986) (in press)
Farrell,S.0., and Bieber,L.L. (1983) Arch. giochem.
Biophys. 333,123.

Hiyasawa,S., Ozasa,H., Osumi,T., and Hashimoto,T.

(1983) J.Eiochem. 25,529.

113

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

114

Markwell,M.A.K., and Bieber,L.L.(1976) Arch.Biochem.

 

Biophys; 72,502.
Bieber,L.L. and Farrell,S.0. (1983) The Enzymes 19,627.
Hoppel,C.L. (1976) Enzymes Biol. Membr. 3,119.

Murphy,M.S.R., and Pande,S.V. (1984) J. Biol.Chem. 251,

 

6683.
Declercq,P.E., Venincasa,M.D., Mills,S.E., Foster,D.W.
and McGarry 3.0. (1985) J.Biol.Chem. 260,12516.

Saggerson,E.D. and Carpenter,C.A. (1982) BiochegLJ.

 

398,673.

McGarry,J.D., Stark,M.J. and Foster D.W. (1978)
J.Biol.Chem. g§§,8291.

McGarry,J.D., Leatherman,G.F., and Foster,D.W. (1978)
11§1ol.0hem. 253,4128.

Clarke,P.R.H., and Bieber,L.L. (1981) J.Biol.Chem.
g§§.9361.

Norum,K.R. (1964) Biochem.Biophys.Acta 89,95
Srere,P.A., Seubert,W. and Lynen,F. (1959)
Biochem.81ophys.Acta 13,313.

Bieber,L.L., Abraham,T., and Helmrath,T.(1972)
Anal.Biochem. 5Q,509.

Grassetti,D.R., and Murray,J.F.,Jr. (1967) 112h1
giochem.§iophys. 119,41.

Bremer,J. (1981) Biochem.Biophys.Actg 665,628.
Bremer,J. and Norum,K.R. (1967) J.Biol.Chem.

242,1744.

Crabtree,B. and Newsholme,E.A. (1972) giochem.J.

 

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

115

139,697.

Bergstrom,J.D., and Reitz,R.C. (1980)
Arch.Biochem.Biophys. 331,71.

Brosnan,J.T., and Fritz,I.B. (1971) Can.J.Biochem.
_3,1296.

Cook,G.A. (1984) J.Biol.Chem. 333,12030.

Bremer,J. and Norum,K.R. (1967) J.Biol.Chem.
3_3,1749.

Saggerson,E.D. and Carpenter,C.A. (1982) FEBS Lett.
131,124.

Norum,K.R. (1965) §iochem.Bigphys.Act§ 105,506.
McCormick,K. and Notar-Francesco,V.J. (1983) 31ochem.J.
_1§,495.

Zammit,V.A. (1985) Biochem.J. 229,273.

 

Zierz,S. and Engel,A.G. (1985) §_r.J.§1ochem.

113, 207.

Norum,K.R. (1965) Acta Phys1gl.Scand. 31, 172.

Van Tol,A. and Hulsman,W.C. (1969) Biochem.Biophys.Acta

1

CD

,342.

Norum,K.R. and Bremer,J. (1967) J.Biol.Chem. 313,407.
Yates,D.W. and Garland,P.B. (1970) Biochem.J. _13,547.
Hoppel,C.L. and Tomec,R.J. (1972) J.Biol.Chem.

311,832.

Brosnan,J.T.,Kopec,B. and Fritz,I.B.(1973) J.Bio1.Chem.
g_§,4075.

Brosnan,J.T.,Kopec,B. and Fritz,I.B.(1973) J.Biol.Chem.

2 8,4069.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

116

VanTol,A. (1974) Biochem.Biophys.Acta 357.14.

Fogle,P.J. and Bieber,L.L. (1978) Int.J.Biochem

19.547.

Harkwell.M.A.K., McGroarty,E.J., Bieber.L.L., and

Tolbert.N.E. (1973) J.Biol.Chem. 248.3426.

 

Farrell,S.0., Fiol,C.J., Reddy,J.K.,and Bieber,L.L.

(1984) J.Bio1.Cheg1 259.13089.

Bieber.L.L., Markwell,M.A.K., Blair.M. and

Helmrath,T.A.(1973) §iochem.Biophys.Acta 326,145.

Wood,J.M. and Chang,C.H. (1975) Fed.Proc. 31.673.

 

Choi,Y.R., Clarke.P.R., and Bieber.L.L.

J.Biol.Chem. 254.5580.

Harano,Y., Kowal.B. and Hiller,M. (1972) Fed.Proc

1,863.

(1979)

Harano,Y., Kowal.J., Yamazahi,R., Levine,L. and

Miller,M. (1972) Arch.§iochem.8iophys.

153.426.

Brady.L.J., Siverstein,S.J., Hoppel,C.L. and

Brady.P.S. (1985) Biochem. J. 232.445.

tt.

Saggerson E.D.. and Carpenter,C.A. (1981) FEBS Le

129.229.

Tomec,R.J. and Hoppel,C.L. (1975) Arch.§iochem.8iophys.

170,716.

DiMauro.S., and DiMauro,P.M. (1973) Science 182,929

 

Wood,J.M. (1973) 31ochegistry 13,5268

Patten,B.M.. Wood,J.M., Harati.Y., Heferan,P. and

HOW911,R.R. (1979) Am.J.Med. 33,610.

Trevisan,C.P., Angelini.c., Freddo,L..

Isaya,G..

and

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

117

Martinuzzi,A. (1984) Neurology 31,353.
Fiol,C.J.. and Bieber.L.L. (1984) J.Biol.0hem. 59,

13084-13088.

Clarke,P.R.H., and Bieber,L.L. (1981) J.Biol.Chem.

 

West,D.W., Chase,J.F.A., and Tubbs.P.K. (1971) Biochem.

Biophys.Res.Comm. 13,912.
Kopec,8.. and Fritz,J.B. (1973) J.Biol.Chem. 248.4069.

Ozasa,H., Miyazawa,s., and 0sumi,T. (1983) J.Biochem.

 

21.543.

Norum.K.R. (1965) giochem.8iophys.Acta 99 511.

4- -—'

Tubbs,P.K. and Chase.J.F.A. (1970) giochem. J. 11 .34.

 

Fritz,I.B.. Kaplan,E., and Yue.K.T.N. (1962)
11.3.Physio11 202.117.
Jenkins.D.L., and Griffith.0.W. (1986)

Proc.Natl.Acad.Sc11 3,290.

 

Yates.D.W. and Garland,P.B. (1966)
Biochem.Biophys.Res.Comg1 33,460.

Solberg,H.E. (1974) 31ochem.§1ophys.Act§ 360,101.
West,D.W., Chase,J.F.A.. and Tubbs,P.K. (1971)

Biochem.8iophys.Res.Cogg1 13912.

 

Woldegiorgis,G., Bremer.J., and Shrago,E.(1985)
Biochem.Bigphys.Act§ 837,135.
Saggerson,E.D. (1982) Biochem.J. 02,397.

Cook.G.A. (1984) §iocheg.J. 224.1015

 

Bremer.J., Woldegiorgis,G., Schanlinske,K., and

Shrago,E. (1985) Biochem.Bigphys.Acta 833,9.

 

-41n.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93

94.

118

Saggerson,E.D., and Carpenter,C.A. (1981) FEBS Lett.

32.166.
Paulson.D.J.. Ward.K.M. and Shug,A.L. (1984)

FEBS Lett. 176.381.

HcGarry.J.D., Mills.S.E.. Long,C.S., and Foster,D.W.

(1983) Biochem.J. 214.21.
Mills,S.E., Foster,D.W.. and McGarry.J.D.(1984)

Biochem.J. 219,601.

 

McGarry,J.D., and Foster,D.W. (1979) J.Biol.Chem.
54.8163.

NcGarry,J.D.. and Foster,D.W. (1981) 31ochem.J.

200,217

Brady.L.J.. Knoeber,C.M.. Hoppel,C.L.. and Brady,P.S.

(1986) (in press).

Mills.S.E., Foster,D.W., and McGarry.J.D. (1983)
Biochem.J. _11.83.

Chase.J.F.A.. and Tubbs.P.K. (1972) 310chem.J.
111.55.

Ciardelli.T.. Stewart.C.J., Seeliger,A., and
Wieland.T. (1981) Justus Liebigs 15n.0hem. p.828.

Karpati.8., et al. (1980) Neurology 33.16.

Kiorpes.T.C.. Hoerr.D., Ho.W.. Weaner.L.E..Inman,M.G..

and Tutwiler.G.F. (1984) J.Bio1.Chem. 259.9750.
Nilsson.0., and Dallner.0.(1977)
Biochem.8iophys.Act§ 464.453.

Daum.G. (1985) Biochem.Biophys.Acta 822,42.

Sire.0.. Mangenar,M.. Montagne.J., Nordmann.R.. and

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

119

Nordmann.J. (1983) Eur.J.Biochem. 136.371.

 

McGarry.J.D.. and Foster,D.W. (1980) Apnu.Rev.Biochem.

 

12.395.
Cook,G.A.. 0tto.D.A.. and Cornell.N.W. (1980)

31ochem.J. 192,955.

 

0ntko.J.A.. and Johns.M.L.(1980) 3iochem.J. 1 2.959.

 

Bremer.J. and Norum.K.R. (1967) gur.J.§iochem. 1.427.

 

Van Tol.A.. and Hulsmann.M. (1969) 31ochem.8iophys.Acta

1 9'3‘2e

Pande.S.V. (1971) J.Biol.Chem. 2 6.5384.

Fritz,I.B. (1961) Egysiol.Rev. 11,52.

Hanson.P.G.. Johnson.R.E.. and Zaharko,D.S.(1965)
Metabolism 11.1037.

0ntko.J.A..and Zilversmith.D.B.(1966) Proc.Soc1§§p1
Biol.Med. 131.319.

McGarry.J.D., and Foster,D.W. (1971) 11§iol.0hem.

 

31§.1149.

McGarry,J.D.. and Foster,D.W. (1971) J.Biol.Chem.
313.6247.

Robinson.I.N.. and Zammit,V.A. (1982) Biochem.J.
393.177.

Cook.G.A.. Stephens.T.W. and Harris.R.A.(1984)
Biochem.J. 19.337.

Cook.G.A., King.M.T. and Veech.R.L. (197a) J.Bio1.Chem.

 

253.2529.
McGarry,J.D.. Stark,J. and Foster,D.W. (1978)

J.Biol.Chem. 2 8.8291.

 

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

120

Saggerson,E.D. and Carpenter,C.A. (1981)FEBS Lett.
133,225.

Gamble.S.M., and Cook.G.A.(1985) J.Biol.Chem. 333.9516
Scow.R.0.. Chernick,S.S. (1960) Recent Prgg31 Hormone

Res. 6,497.

 

McGarry,J.D.. Takabayashi.Y. and Foster,D.W. (1978)
g4§l£l4§h£§z 222:8294-

Harano.Y.. Kashiwagi.A.. Kojima.H.. Suzuki.M..
Hashimoto,T. and Shigeta.Y. (1985) 3388 LETT. 1_3.267.
Brindle.N.P.J., Zammit,V.A.. and Pogson.C.I. (1985)
Biocheg1J. _33,177.

Tutwiler,G.F.. Kirsch.T., Mohrbacher,R.J., and Ho.W.
(1978) Metabolism 31.1539.

McGarry,J.D.. and Foster,D.W. (1973) J.C11g,1nvest.
33.377.

JenkinsD.L. and 0wen.G.W. (1986) Proc.Natl.Acad.Sci.
13.290.

Bohme.H.J.. Kopperschager.G., Schulz,J. and Hofman.E.
(1972) J. Chromatography 33.209.

Laemmli.U.K. (1970) Nature ngndon) 331.680.

Ames.B.N. (1966) "Methods of Enzymology" 3.115.

Dittmer,J.C.. and Lester.R.L. (1964) J. Lip1d Res.

 

 

3.696.

LeBlond.D.J.. Ashendel.C.L., and Wood,W.A. (1980)
Anal.§iochem. 104.355.

LeBlond.D.J.. Ashendel.C.L., and Wood.W.A.(1980)

Anal.Biochem. 104.370.

125.

126.

127.

128.

129.

130.

131.

132.

133.

134.

135.

136.

137.

121

Tippet.P.S..and Neet,K.E. (1982) J.Biol.Chem. 257.
12839.

Bohlen.P., Stein.S.. Durman.W.,and Udenfriend,S. (1973)
Arch.Biochem.§1gphys. 155.213.

Sedmak.J.J. and Grossberg.S.E. (1977) Anal.Biochem.

 

13,544.

Toth,P.P., Ferguson-Miller.S.M., and Suelter,C.H.
(1986) Methods Eggymol. 125.16

Robinson.J.B.Jr.. Strottman,J.M., Wick.D.G.. and
Stellwagen.E. (1980) Proc.Natl.Acag.Sc11 11.5847.
Stephens.T.W., Cook.G.A. and Harris.R.A. (1983)
Biochem.J. 212.521.

Segel.I.H. (1975) in "Enzyme Kinetics" pp.360-362.
John Wiley and Sons. Inc. New York.

Clarke.S. (1975) 113191.0hem. 333.5459.
Venkatraghavan. v.. and Smith, D.J. (1983)
Arch.Biochem.Biophys. 333,123.

Long,C.S., Haller.R.G.. Foster,D.W. and McGarry.J.D.
(1982) Neurology 33.663.

Ware,A.J., Burton.W.C.. McGarry.J.D.. Marks,J.F., and
Weinberg.A.G. (1978) J.Pediatrics 33.959.

Zammit,V.A. (1983) Biochem.J. 214.1027.

 

Zammit.V.A. (1984) 3iochem.J. 218.379.