Am J Physiol Endocrinol Metab 288: E133–E142, 2005.

First published September 21, 2004; doi:10.1152/ajpendo.00379.2004.

Malonyl-CoA and carnitine in regulation of fat oxidation

in human skeletal muscle during exercise
Carsten Roepstorff,1 Nils Halberg,1 Thore Hillig,1 Asish K. Saha,2 Neil B. Ruderman,2
Jørgen F. P. Wojtaszewski,1 Erik A. Richter,1 and Bente Kiens1
1
The Copenhagen Muscle Research Centre, Department of Human Physiology, Institute of Exercise and Sport Sciences,
University of Copenhagen, Copenhagen, Denmark; and 2Diabetes Research Unit, Departments of Medicine, Physiology,
and Biophysics, Section of Endocrinology, Boston University School of Medicine, Boston, Massachusetts
Submitted 16 August 2004; accepted in final form 17 September 2004

Roepstorff, Carsten, Nils Halberg, Thore Hillig, Asish K. Saha,
Neil B. Ruderman, Jørgen F. P. Wojtaszewski, Erik A. Richter,
and Bente Kiens. Malonyl-CoA and carnitine in regulation of fat
oxidation in human skeletal muscle during exercise. Am J Physiol
Endocrinol Metab 288: E133–E142, 2005. First published September
21, 2004; doi:10.1152/ajpendo.00379.2004.—Intracellular mecha-
nisms regulating fat oxidation were investigated in human skeletal
muscle during exercise. Eight young, healthy, moderately trained men
performed bicycle exercise (60 min, 65% peak O2 consumption) on
two occasions, where they ingested either 1) a high-carbohydrate diet
(H-CHO) or 2) a low-carbohydrate diet (L-CHO) before exercise to
alter muscle glycogen content as well as to induce, respectively, low
and high rates of fat oxidation. Leg fat oxidation was 122% higher
during exercise in L-CHO than in H-CHO (P ⬍ 0.001). In keeping
with this, the activity of ␣2-AMP-activated protein kinase (␣2-AMPK)
was increased twice as much in L-CHO as in H-CHO (P ⬍ 0.01) at
60 min of exercise. However, acetyl-CoA carboxylase (ACC)␤ Ser221
phosphorylation was increased to the same extent (6-fold) under the
two conditions. The concentration of malonyl-CoA was reduced 13%
by exercise in both conditions (P ⬍ 0.05). Pyruvate dehydrogenase
activity was higher during exercise in H-CHO than in L-CHO (P ⬍
0.01). In H-CHO only, the concentrations of acetyl-CoA and acetyl-
carnitine were increased (P ⬍ 0.001), and the concentration of free
carnitine was decreased (P ⬍ 0.01), by exercise. The data suggest that
a decrease in the concentration of malonyl-CoA, secondary to ␣2-
AMPK activation and ACC inhibition (by phosphorylation), contrib-
utes to the increase in fat oxidation observed at the onset of exercise
regardless of muscle glycogen levels. They also suggest that, with
high muscle glycogen, the availability of free carnitine may limit fat
oxidation during exercise, due to its increased use for acetylcarnitine
formation.
acetylcarnitine; acetyl coenzyme A; pyruvate dehydrogenase activity;
adenosine 5⬘-monosphosphate-activated protein kinase; acetyl coen-
zyme A carboxylase
FAT AND CARBOHYDRATE are the main sources of fuel for ATP
synthesis in human skeletal muscle. The ratio between fat and
carbohydrate oxidation during exercise depends on preexercise
substrate levels and exercise duration and intensity (12, 40, 42,
44). However, the intracellular determinants of the use of these
fuels are unclear.
It has been suggested that the fat oxidation rate in human
skeletal muscle during exercise is regulated intracellularly at
the entry of long-chain fatty acyl (LCFA)-CoA into the mito-
chondria (31, 36, 37). The transport of LCFA-CoA across the
inner mitochondrial membrane is preceded by the transfer of
the acyl moiety to carnitine, a process catalyzed by carnitine
palmitoyltransferase 1 (CPT-1), an enzyme located in the outer
mitochondrial membrane (23). Two mechanisms for the intra-
cellular regulation of CPT-1 have been proposed; one involves
malonyl-CoA, and the other involves the cytosolic concentra-
tion of carnitine.
Malonyl-CoA is an intermediate in the de novo synthesis of
fatty acids (FA) and an allosteric inhibitor of CPT-1 (14).
Studies in rat muscle suggest that changes in the glucose
supply and energy expenditure of the muscle cell regulate the
concentration of malonyl-CoA, in keeping with its need to
generate ATP from fat oxidation (28, 32, 33). Two types of
regulation of malonyl-CoA have been described: one involving
the cytosolic concentration of citrate and the other involving
5⬘-AMP-activated protein kinase (AMPK) (28, 32). Cytosolic
citrate is both an allosteric activator of acetyl-CoA carboxylase
(ACC), the key enzyme governing malonyl-CoA synthesis, and
a substrate for the malonyl-CoA precursor, cytosolic acetyl-
CoA. In rodents and humans, high glucose availability at rest
has been shown to elevate the cytosolic citrate concentration
and consequently the muscle malonyl-CoA concentration (3,
27, 32, 33). This is likely the mechanism whereby fat oxidation
is inhibited by high glucose availability in resting humans (3,
27). AMPK in turn regulates the concentration of malonyl-
CoA by phosphorylating and inhibiting ACC (28) and possibly
by activating malonyl-CoA decarboxylase (MCD), the major
enzyme regulating malonyl-CoA turnover in muscle (34). Dur-
ing muscle contractions in rodents, AMPK is activated and
ACC is phosphorylated and consequently inhibited, leading to
a decrease in muscle malonyl-CoA concentration, which may
induce the increase in fat oxidation at the onset of contraction
(28). In contrast, several human studies have failed to show a
reduction in the muscle malonyl-CoA concentration during
moderate-intensity exercise with normal muscle glycogen
stores when fat oxidation is increased from resting levels (7,
17, 18). Still, with low muscle glycogen content, there is a
higher AMPK activation during exercise (44), which may
decrease the malonyl-CoA level and thereby cause the higher
fat oxidation seen during exercise with low vs. high muscle
glycogen (44). So far, the muscle malonyl-CoA concentration
has never been measured in humans during submaximal exer-
cise with low muscle glycogen content.

Address for reprint requests and other correspondence: C. Roepstorff, The

Copenhagen Muscle Research Centre, Institute of Exercise and Sport Sciences, The costs of publication of this article were defrayed in part by the payment
Dept. of Human Physiology, Universitetsparken 13, DK-2100 Copenhagen Ø, of page charges. The article must therefore be hereby marked “advertisement”
Denmark (E-mail: croepstorff@ifi.ku.dk). in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

http://www.ajpendo.org 0193-1849/05 $8.00 Copyright © 2005 the American Physiological Society E133
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (122.171.080.167) on July 20, 2018.
Copyright © 2005 American Physiological Society. All rights reserved.
E134 MUSCLE MALONYL-COA, CARNITINE, AND FAT OXIDATION
Carnitine is a substrate for CPT-1. Consequently, the avail-
ability of cytosolic carnitine may limit fat oxidation. Another
function of carnitine is to buffer accumulated acetyl-CoA in a
reaction catalyzed by carnitine acetyltransferase (CAT) (26). It
has been proposed that, under exercise conditions with high
glycolytic flux and therefore excess formation of acetyl-CoA
compared with its utilization by the tricarboxylic acid cycle,
carnitine serves to buffer the excess acetyl-CoA, which leaves
less carnitine available to CPT-1 (40, 43). By this mechanism,
carnitine may play a role in adjusting the rate of fat oxidation
inversely to the rate of carbohydrate oxidation. Accordingly, in
humans it was shown that fat oxidation and muscle carnitine
concentration were tightly coupled during incremental exer-
cise: with increasing exercise intensity, they both decreased,
while, on the other hand, carbohydrate oxidation increased (40).
The present study investigated intracellular mechanisms to
regulate fat oxidation in response to altered carbohydrate
availability in human skeletal muscle during submaximal ex-
ercise. Healthy male subjects performed 60 min of submaximal
bicycle exercise with either high (H-CHO) or low (L-CHO)
preexercise muscle glycogen. During the two exercise proto-
cols, with expected similar energy output but marked differ-
ences in the relative fat and carbohydrate oxidation, we measured
in the same setup the malonyl-CoA, carnitine, acetylcarnitine, and
acetyl-CoA concentrations, the pyruvate dehydrogenase (PDH)
and AMPK activity, and the ACC␤ Ser221 phosphorylation in
skeletal muscle to examine the intramuscular mechanisms that
regulate the ratio between fat and carbohydrate oxidation
during exercise. To rule out marked differences between H-
CHO and L-CHO conditions in blood-borne substrate and
hormone levels and their possible confounding effects on the
interpretations of the results, we allocated a light meal 4.5 h
before exercise and infused glucose intravenously during ex-
ercise to keep the arterial blood glucose concentrations nearly
constant and similar between the two conditions. By this
approach, the levels of several other circulating metabolites
and hormones also were kept nearly similar between H-CHO
and L-CHO.
Fig. 1. Schematic diagram showing the experimental
protocol. Each protocol consisted of 2 experimental
days. Day 1: glycogen depletion and dietary manipu-
lation. Day 2: exercise experiment. All subjects com-
pleted 2 protocols [high (H-CHO) and low (L-CHO)
preexercise muscle glycogen].
AJP-Endocrinol Metab • VOL
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (122.171.080.167) on July 20, 2018.
Copyright © 2005 American Physiological Society. All rights reserved.
MATERIALS AND METHODS
Subjects. Eight young, healthy, moderately trained men [age, 26 ⫾
1 yr; height, 1.80 ⫾ 0.02 m; body mass, 76 ⫾ 3 kg; body mass index,
23.5 ⫾ 0.7 kg/m2; body fat, 14.6 ⫾ 1.9%; peak O2 consumption
(V̇O2 peak), 4.1 ⫾ 0.2 l/min; V̇O2 peak/body mass, 53.7 ⫾ 1.6
ml 䡠 kg⫺1 䡠 min⫺1] were recruited to participate in the study. All sub-
jects were nonsmokers and were engaged in 4 – 6 h/wk of regular
exercise training such as running, bicycling, and resistance training.
Before inclusion into the study, subjects were fully informed about the
nature and possible risks of the study and gave their written consent.
The study was approved by the Copenhagen Ethics Committee (KF-
01-078/01) and carried out in accordance with the code of ethics of the
World Medical Association (Declaration of Helsinki II). The results
presented in this study are part of a larger study on the regulation of
lipid metabolism during exercise in human skeletal muscle that has, in
minor part, been published previously (30, 31).
Preexperimental testing. All subjects initially performed an incre-
mental exercise test on a bicycle ergometer (Monark 839 Electronic
Ergometer; Monark Exercise Sweden) to determine peak O2 uptake
(V̇O2 peak). Respiratory measurements were carried out with the Doug-
las bag technique. Percent body fat was measured by dual-energy
X-ray absorptiometry (DEXA; Lunar, Madison, WI; DPX-IQ v. 4.6.6).
Experimental design. Subjects underwent two experimental proto-
cols separated by 2–3 wk. In both protocols, subjects completed a bout
of glycogen-depleting bicycling followed by a controlled diet for the
rest of the day (Fig. 1, day 1). In one protocol the controlled diet
consisted primarily of fat (L-CHO), and in the other protocol primar-
ily of carbohydrate (H-CHO) (see Glycogen depletion bout). The next
day, subjects performed a bicycle exercise test for 60 min at ⬃65%
V̇O2 peak (Fig. 1, day 2). The order of the L-CHO and H-CHO
protocols was randomized and stratified.
Dietary control. On the 3 days preceding the first protocol, subjects
determined their habitual energy and nutrient intake by a self-reported
dietary record. All food and beverage intakes were weighed to the
accuracy of 1 g and recorded. Subsequently, the energy intake and
composition of the habitual diet were calculated by means of a
computer database (Dankost 2000; Danish Catering Center, Copen-
hagen, Denmark). On the 3 days preceding the second protocol,
subjects followed a diet identical to the one ingested before the first
protocol.
Glycogen depletion bout. On the day of the glycogen depletion
bout, subjects arrived at the laboratory at 7:30 AM. All subjects
288 • JANUARY 2005 • www.ajpendo.org
 MUSCLE MALONYL-COA, CARNITINE, AND FAT OXIDATION
abstained from any heavy physical activity on the day before the
glycogen depletion bout. A breakfast meal was served that contained
25% of each subject’s daily energy intake and consisted of 30%
energy (30 E%) as fat, 55 E% as carbohydrate, and 15 E% as protein.
At 10 AM, subjects initiated the glycogen depletion bicycle bout,
which consisted of alternating periods of continuous and intermittent
exercise with a few short bouts of arm cranking in between, as
previously described (31). The glycogen depletion bout lasted for
3– 4.5 h and was well tolerated by all subjects. To determine the
muscle glycogen concentration after the glycogen depletion protocol,
a muscle biopsy was obtained under local anesthesia from the vastus
lateralis muscle by the needle biopsy technique. For the following 6 h,
subjects were allowed to move freely around at the laboratory. Meals
were served immediately after the biopsy procedure and 1, 2, 3.5, and
5 h after completion of the glycogen depletion bout. Then subjects
were given a light snack to ingest at 11 PM and left the laboratory.
The controlled diet during the recovery period consisted of 85 E% fat,
2 E% carbohydrate, and 13 E% protein in L-CHO and 8 E% fat, 80
E% carbohydrate, and 12 E% protein in H-CHO. The total energy
intake during this period was calculated as the individual daily energy
intake (DEI), as reported from the dietary record, minus the breakfast
(25% DEI) plus the estimated extra energy consumption during the
bicycle exercise to secure that subjects were in energy balance. During
the glycogen depletion bout and for the rest of the day, subjects were
allowed to drink water ad libitum.
Exercise experimental protocol. Subjects arrived at the laboratory
at 7:30 AM having abstained from any physical activity during the
morning. A light breakfast was served that contained 10% of the
individual DEI and consisted of 4 E% fat, 78 E% carbohydrate, and 18
E% protein. After 30 min of rest in the supine position, Teflon
catheters were inserted under local anesthesia into the femoral artery
and the contralateral femoral vein with a thermistor probe inserted
into the vein as described previously (29). During the second protocol,
femoral catheters were inserted in opposite legs. A venous catheter
was inserted into an antecubital vein for infusion of glucose.
Subjects then rested in the supine position. At 12 PM, femoral
venous blood flow was measured by the thermodilution technique,
using bolus injections of 5 ml of ice-cold sterile saline (1). Then,
expired air was collected in a Douglas bag, and blood samples were
obtained in duplicate for determination of resting blood concentra-
tions. A muscle biopsy was obtained from the vastus lateralis muscle.
Then subjects initiated a 60-min bicycle exercise test at ⬃65%
V̇O2 peak. Subjects exercised at the same absolute workload in L-CHO
and H-CHO groups. A variable continuous infusion of glucose into an
antecubital vein was initiated after 1 min of exercise and continued
throughout exercise to keep the blood glucose concentration constant
and similar to the basal resting level. The infusion rate of glucose was
evaluated by frequent arterial blood sampling (every 5 min) and
subsequent rapid analysis of the glucose concentration. Every 10 min
during exercise, expired air was collected in a Douglas bag, blood was
drawn from the femoral artery and vein, and femoral venous blood
flow was measured by the thermodilution technique with continuous
infusion of ice-cold sterile saline (1). After 30 min of bicycling,
subjects shortly terminated exercise, and a biopsy was obtained from
the vastus lateralis muscle (⬍15 s of rest). Exercise was continued
within 120 s. After completion of the 60-min bicycle test, another
biopsy was obtained from the vastus lateralis muscle (⬍15 s of rest).
All biopsies were obtained through separate incisions spaced at least
5 cm apart. Heart rate was monitored throughout the experiment with
a Polar Vantage XL heart rate monitor (Polar Electro Finland), and
blood pressure was measured through the femoral arterial catheter
connected to a pressure transducer (Spectramed P23XL, Statham).
During the first protocol, subjects were offered water ad libitum, and
their water intake was recorded. During the second protocol, subjects
repeated their water intake from the first protocol.
AJP-Endocrinol Metab • VOL 288 • JANUARY 2005 •
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (122.171.080.167) on July 20, 2018.
Copyright © 2005 American Physiological Society. All rights reserved.
E135
Breath samples. Expired air in the Douglas bags was processed,
and respiratory exchange ratio (RER) was calculated as described
previously (29).
Blood samples. Blood O2 and CO2 concentrations were deter-
mined, and leg respiratory quotient (RQ) was calculated as described
previously (29). Leg fat and carbohydrate oxidation rates were calcu-
lated by the following equations using the nonprotein RQ and then
expressed (in kJ/min) using standard caloric equivalents for fat and
carbohydrate, respectively (20)
fat oxidation rate (g/min) ⫽ 1.695 䡠 V̇O2 ⫺ 1.701 䡠 V̇CO2
carbohydrate oxidation rate (g/min) ⫽ 4.585 䡠 V̇CO2 ⫺ 3.226 䡠 V̇O2
where V̇O2 is O2 consumption, and V̇CO2 is CO2 production. Blood
glucose and lactate concentrations were measured automatically
(ABL510; Radiometer Medical, Copenhagen, Denmark). Plasma FA
concentrations were measured by a colorimetric commercial assay kit
(Wako Chemicals) performed by a COBAS FARA autoanalyzer
(COBAS FARA 2, Roche Diagnostic Switzerland). Concentrations of
insulin (Pharmacia Insulin Radioimmunoassay 100; Pharmacia &
Upjohn Diagnostics, Uppsala, Sweden) as well as epinephrine and
norepinephrine (KatCombi Radioimmunoassay; Immuno-Biological
Laboratories, Hamburg, Germany) in plasma were determined by
radioimmunoassay.
Muscle biopsies. The biopsies were quick-frozen (⬍10 s after being
obtained) in liquid nitrogen while still in the biopsy needle and stored
at ⫺80°C for subsequent biochemical analysis. Seventy-five to eighty
milligrams wet weight were freeze dried and dissected free of all
visible adipose tissue, connective tissue, and blood under a micro-
scope. The dissected muscle fibers were pooled and then divided into
subpools for the respective analyses. Malonyl-CoA content and PDH
activity in its active form (PDHa) were measured on 25 and 10 mg of
wet muscle tissue, respectively, which were not freeze dried.
Muscle glycogen. The glycogen concentration was determined on 2
mg dry wt by a fluorometric method (13).
Muscle lysates. Muscle lysates were prepared from 6 mg dry wt of
freeze-dried and dissected muscle tissue as described previously (10).
Due to the limited amount of tissue, muscle lysates were only
prepared from biopsies obtained at rest and at 60 min of exercise.
Western blotting. Phosphorylation of ␣-AMPK Thr172 and ACC␤
Ser221 was detected by Western blotting on muscle lysates. The
lysates were boiled in Laemmli buffer before being subjected to
SDS-PAGE and immunoblotting. Primary phosphospecific antibodies
were rabbit anti-␣-AMPK Thr172-phos (Upstate Biotechnology, Lake
Placid, NY) and rabbit anti-ACC␣ Ser79-phos. Secondary antibody
was horseradish peroxidase-conjugated anti-rabbit IgG (DAKO,
Glostrup, Denmark). Antigen-antibody complexes were visualized
using enhanced chemiluminescence (ECL⫹; Amersham Biosciences)
and quantified by a Kodak Image Station E440CF (Kodak, Glostrup,
Denmark). The anti-␣-AMPK Thr172-phos antibody detected a single
band at ⬃63 kDa as expected. With the phosphospecific ACC␣ Ser79
antibody, which probably recognizes the equivalent Ser221 in human
ACC␤ in the phosphorylated state, a single band was detected at
⬃260 kDa as expected.
AMPK activity. ␣-Isoform-specific AMPK activity was determined
in immunoprecipitates from muscle lysates as described previously
(44). Briefly, immunoprecipitates were prepared from 200 ␮g of
muscle lysate protein using anti-␣1- or anti-␣2-antibodies as described
previously (46) (kindly donated by Dr. D. G. Hardie, University of
Dundee, Dundee, Scotland). AMPK activity was measured in the
immunoprecipitates using SAMS-peptide (HMRSAMSGLHLVKRR,
200 ␮M) as previously described (45).
Malonyl-CoA. The muscle malonyl-CoA concentration was mea-
sured as described previously (15), modified to human muscle tissue
(7). In short, 25 mg of muscle tissue (wet wt) were homogenized in
250 ␮l of perchloric acid (10%) and neutralized with NaOH. Malonyl-
CoA was determined in the neutralized muscle homogenates by
www.ajpendo.org
E136 MUSCLE MALONYL-COA, CARNITINE, AND FAT OXIDATION
measuring malonyl-CoA-dependent incorporation of [3H]acetyl-CoA
into palmitate catalyzed by FA synthase (kindly donated by Dr. J.
Knudsen, University of Southern Denmark). Each sample was deter-
mined in duplicate. The coefficient of variation (CV) between three
single determinations on the same muscle extract was 18%, and the
CV between two different muscle extracts, determined in duplicate,
from the same biopsy was 11%.
PDHa. Due to limited amount of muscle tissue, PDHa activity was
only determined on the biopsies obtained at rest and at 60 min of
exercise. The frozen muscle tissue (10 mg) was homogenized in 300
␮l of buffer (pH 7.8) containing 200 mM sucrose, 50 mM KCl, 5 mM
MgCl2, 5 mM EGTA, 50 mM Tris 䡠 HCl, 50 mM NaF, 5 mM
dichloroacetate, and 0.1% Triton X-100. Subsequently, PDHa activity
was determined by a radioisotopic assay first described by Constantin-
Teodosiu et al. (6) and modified by Putman et al. (25). The CV
between five single determinations on the same muscle extract was
17%, and the CV between four different muscle extracts from the
same biopsy determined in duplicate was 17%.
Acetyl-CoA, acetylcarnitine, and free carnitine. Six milligrams dry
weight of freeze-dried and dissected muscle tissue were extracted with
0.6 M perchloric acid and neutralized with 2 M KHCO3. Acetyl-CoA,
acetylcarnitine, and free carnitine concentrations were determined by
radioisotopic assays (4). The CV between five single determinations
on the same muscle extract was 3.9, 3.1, and 5.8% for acetyl-CoA,
acetylcarnitine, and free carnitine, respectively. The CV between four
different muscle extracts, determined in duplicate, from the same
biopsy was 11.9, 8.1, and 6.2% for acetyl-CoA, acetylcarnitine, and
free carnitine, respectively. Recovery of acetyl-CoA, acetylcarnitine,
and carnitine added to muscle extracts was 90, 96, and 99%, respec-
tively.
Statistics. Data are presented as means ⫾ SE. For variables inde-
pendent of time, a paired t-test was performed to test for differences
between L-CHO and H-CHO. For variables measured before and after
exercise as well as variables measured before and during exercise, a
two-way analysis of variance (ANOVA), with repeated measures for
the time and protocol factors, was performed to test for differences
between protocols or changes due to time. When a significant main
effect of time was found, significant pairwise differences were per-
formed using Tukey’s post hoc test. Unless otherwise stated, a
probability of 0.05 was used as the level of significance.
RESULTS
Workload. During the 60-min bicycle exercise test, subjects
exercised at 180 ⫾ 7 W in both conditions. In L-CHO the
systemic V̇O2 averaged 2.8 ⫾ 0.1 l/min during exercise,
whereas in H-CHO the systemic V̇O2 averaged 2.5 ⫾ 0.1 l/min
during exercise (P ⬍ 0.001). The %V̇O2 peak during the bicycle
exercise test averaged 68 ⫾ 1 and 62 ⫾ 1% in L-CHO and
H-CHO, respectively (P ⬍ 0.001).
Intravenous glucose infusion. The glucose infusion rate
during exercise averaged 39 ⫾ 5 and 15 ⫾ 6 ␮mol䡠kg body
mass⫺1 䡠min⫺1 in L-CHO and H-CHO, respectively (P ⬍ 0.01).
Cardiovascular parameters. Heart rate was 62 ⫾ 2 and 56 ⫾
2 beats/min (bpm) at rest in L-CHO and H-CHO, respectively
(P ⬍ 0.05), and increased (P ⬍ 0.001) at initiation of exercise
to a plateau level averaging 163 ⫾ 5 and 152 ⫾ 6 bpm during
exercise in L-CHO and H-CHO (P ⬍ 0.001). At rest, femoral
venous blood flow was 0.51 ⫾ 0.09 and 0.41 ⫾ 0.02 l/min in
L-CHO and H-CHO, respectively [not significant (NS)]. At
onset of exercise, femoral venous blood flow increased (P ⬍
0.001) to 6.5 ⫾ 0.4 and 6.1 ⫾ 0.4 l/min in L-CHO and H-CHO,
respectively (NS), and remained at this level throughout exer-
cise.
AJP-Endocrinol Metab • VOL 288 • JANUARY 2005 •
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (122.171.080.167) on July 20, 2018.
Copyright © 2005 American Physiological Society. All rights reserved.
Fat and carbohydrate oxidation. RER and leg RQ did not
differ significantly from each other in either condition. At rest,
RER was 0.76 ⫾ 0.02 and 0.83 ⫾ 0.03 and leg RQ was 0.75 ⫾
0.01 and 0.85 ⫾ 0.03 in L-CHO and H-CHO, respectively.
Within the first 10 min of exercise, RER increased (P ⬍ 0.01)
to 0.82 ⫾ 0.02 and 0.91 ⫾ 0.01 in L-CHO and H-CHO,
respectively, and leg RQ increased (P ⬍ 0.01) to 0.85 ⫾ 0.02
and 0.93 ⫾ 0.02 in L-CHO and H-CHO, respectively. No
further significant changes were seen in RER or leg RQ during
exercise. There was a main effect of condition on RER and on
leg RQ, both being lower in L-CHO than in H-CHO (P ⬍ 0.001).
The fat oxidation rate across the leg did not differ signifi-
cantly between L-CHO and H-CHO at rest (Fig. 2). In H-CHO
the leg fat oxidation rate increased (P ⬍ 0.05) from rest to
exercise, and an even larger increase (P ⬍ 0.001) was seen in
L-CHO. The leg fat oxidation rate differed significantly be-
tween L-CHO and H-CHO at all exercise time points (P ⬍
0.001). The leg carbohydrate oxidation rate increased (P ⬍
0.001) from rest to exercise in both conditions. A main effect
of condition was observed in the leg carbohydrate oxidation
rate (P ⬍ 0.01).
Blood metabolite concentrations. At rest, the arterial con-
centration of plasma FA was 674 ⫾ 110 and 596 ⫾ 112 ␮M in
L-CHO and H-CHO, respectively. It decreased (P ⬍ 0.05) at
initiation of exercise to 468 ⫾ 61 and 334 ⫾ 34 ␮M at 10 min
in L-CHO and H-CHO, respectively, after which it increased
(P ⬍ 0.05) continuously during exercise to 703 ⫾ 58 ␮M in
L-CHO and 618 ⫾ 58 ␮M in H-CHO at 90 min. A main effect
of condition was observed on arterial plasma FA concentration,
it being slightly higher in L-CHO than in H-CHO (P ⬍ 0.01).
The arterial blood glucose concentration at rest was 4.9 ⫾
0.2 and 5.1 ⫾ 0.1 mM in L-CHO and H-CHO, respectively
(NS). During exercise, it was kept nearly constant because of
the intravenous glucose infusion.
Circulating hormones. At rest, the arterial plasma insulin
concentration was 6.6 ⫾ 1.1 and 6.5 ⫾ 0.6 ␮U/ml in L-CHO
and H-CHO, respectively (NS). In both conditions, the arterial
plasma insulin concentration decreased (P ⬍ 0.001) continu-
ously during exercise to 3.4 ⫾ 0.2 and 4.2 ⫾ 0.3 ␮U/ml at 60
min of exercise in L-CHO and H-CHO, respectively. The
Fig. 2. Leg fat and carbohydrate (CHO) oxidation rates at rest and during 60
min of bicycle exercise at 65% peak O2 consumption (V̇O2 peak) with L-CHO
or H-CHO. *Main effect of condition on CHO oxidation, P ⬍ 0.01. #Fat
oxidation different from L-CHO, P ⬍ 0.001. †CHO oxidation different from
exercise in both conditions, P ⬍ 0.001. ‡Fat oxidation different from exercise
in L-CHO (P ⬍ 0.001) and H-CHO (P ⬍ 0.05).
www.ajpendo.org
 MUSCLE MALONYL-COA, CARNITINE, AND FAT OXIDATION
arterial plasma insulin concentration was lower in L-CHO than
in H-CHO at 20, 30, 50, and 60 min of exercise (P ⬍ 0.05).
At rest, the arterial plasma epinephrine concentration was
0.60 ⫾ 0.10 and 0.54 ⫾ 0.08 nM in L-CHO and H-CHO,
respectively (NS). At initiation of exercise, it increased (P ⬍
0.001) to 1.79 ⫾ 0.19 and 1.54 ⫾ 0.19 nM at 10 min in L-CHO
and H-CHO, respectively, whereafter a further continuous
increase (P ⬍ 0.001) occurred to 2.92 ⫾ 0.49 and 2.05 ⫾ 0.31
nM at 60 min of exercise in L-CHO and H-CHO, respectively.
During exercise, the arterial plasma epinephrine concentration
did not differ significantly between L-CHO and H-CHO.
The arterial plasma norepinephrine concentration at rest was
1.22 ⫾ 0.29 and 1.49 ⫾ 0.29 nM in L-CHO and H-CHO,
respectively (NS). At onset of exercise, it increased (P ⬍
0.001) and averaged 14.88 ⫾ 1.39 and 11.04 ⫾ 0.87 nM in
L-CHO and H-CHO, respectively, during exercise. The arterial
plasma norepinephrine concentration was higher in L-CHO
than in H-CHO at 10 min (P ⬍ 0.05) and at 20, 40, 50, and 60
min (P ⬍ 0.01).
Muscle glycogen. The glycogen concentration in the vastus
lateralis muscle immediately after the glycogen depletion bout
was 35 ⫾ 10 and 84 ⫾ 42 mmol/kg dry wt in L-CHO and
H-CHO, respectively (NS) (Fig. 3). After the following dietary
manipulation and immediately before the 60-min bicycle ex-
ercise test, the glycogen concentration had increased (P ⬍
0.001) to 197 ⫾ 21 and 504 ⫾ 25 mmol/kg dry wt in L-CHO
and H-CHO, respectively. During the first 30 min of exercise,
a significant decrease (P ⬍ 0.01) was observed in the muscle
glycogen concentration in both conditions, but the rate of
glycogen breakdown was 46% higher in H-CHO than in
L-CHO (P ⬍ 0.05). From 30 to 60 min of exercise, the
glycogen concentration decreased further (P ⬍ 0.001) in H-
CHO but not in L-CHO (NS). At rest and at 30 and 60 min of
exercise, the glycogen concentration in the vastus lateralis
muscle was lower in L-CHO than in H-CHO (P ⬍ 0.001).
AMPK and ACC. ␣-AMPK Thr172 phosphorylation in the
vastus lateralis muscle at rest did not differ significantly be-
tween H-CHO and L-CHO (Table 1). From rest to 60 min of
exercise, a 306% increase (P ⬍ 0.01) occurred in ␣-AMPK
Thr172 phosphorylation in L-CHO, whereas no significant
change was observed in H-CHO. At 60 min of exercise,
␣-AMPK Thr172 phosphorylation was 126% higher (P ⬍ 0.01)
in L-CHO than in H-CHO.
Fig. 3. Glycogen concentration in the vastus lateralis muscle after glycogen
depletion and before and at 30 and 60 min of bicycle exercise at 65% V̇O2 peak.
d.w., Dry weight. *Different from L-CHO, P ⬍ 0.001. #Different from rest and
30 and 60 min, P ⬍ 0.001. Different from rest, †P ⬍ 0.01 and ††P ⬍ 0.001.
‡Different from 30 min, P ⬍ 0.001.
AJP-Endocrinol Metab • VOL 288 • JANUARY 2005 •
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (122.171.080.167) on July 20, 2018.
Copyright © 2005 American Physiological Society. All rights reserved.
E137
Table 1. ␣-AMPK Thr172 phosphorylation and activity
Rest 60 min of Exercise
L-CHO H-CHO L-CHO H-CHO
␣-AMPK Thr 172
phosphorylation, AU 8⫾1 6⫾1 33⫾11*† 15⫾4
␣1-AMPK activity, pmol䡠mg
prot⫺1䡠min⫺1‡ 2.7⫾0.4 1.7⫾0.1 3.4⫾1.0 2.6⫾0.4
␣2-AMPK activity, pmol䡠mg
prot⫺1䡠min⫺1 1.5⫾0.1 1.4⫾0.1 5.1⫾1.1*† 3.1⫾0.6§
Values are means ⫾ SE. ␣-AMPK Thr phosphorylation and activity in
172
the vastus lateralis muscle before and at 60 min of bicycle exercise at 65%
peak O2 consumption with low (L-CHO) or high (H-CHO) preexercise muscle
glycogen. AMPK, AMP-activated protein kinase; CHO, carbohydrate; AU,
arbitrary units; prot, protein. *Different from H-CHO, P ⬍ 0.01. †Different
from rest, P ⬍ 0.01. ‡Borderline-significant main effect of exercise, P ⫽ 0.06.
§Borderline significantly different from rest, P ⫽ 0.06.
␣1-AMPK activity in the vastus lateralis muscle did not
differ significantly between conditions at rest or at 60 min of
exercise (Table 1). A borderline-significant main effect of
exercise was observed in ␣1-AMPK activity, which increased
from rest to 60 min of exercise (P ⫽ 0.06).
␣2-AMPK activity was similar between L-CHO and H-CHO
at rest (NS) (Table 1). In L-CHO a 238% increase (P ⬍ 0.01)
occurred from rest to 60 min of exercise, whereas the exercise-
induced 119% increase in ␣2-AMPK activity in H-CHO only
reached borderline statistical significance (P ⫽ 0.06). At 60
min of exercise, ␣2-AMPK activity was 62% higher in L-CHO
than in H-CHO (P ⬍ 0.01).
ACC␤ Ser221 phosphorylation did not differ significantly
between L-CHO and H-CHO at rest or at 60 min of exercise,
but it increased (P ⬍ 0.001) approximately sixfold from rest to
exercise irrespective of condition (Fig. 4A).
Malonyl-CoA. At rest the malonyl-CoA concentration in the
vastus lateralis muscle did not differ significantly between
L-CHO (0.32 ⫾ 0.03 ␮mol/kg wet wt) and H-CHO (0.30 ⫾
0.03 ␮mol/kg wet wt) (Fig. 4B). From rest to 30 min of
exercise, there was a 13% decrease (P ⬍ 0.05) in the muscle
malonyl-CoA concentration irrespective of condition. The
lower malonyl-CoA concentration persisted throughout exer-
cise (0 vs. 60 min, P ⬍ 0.05). During exercise there was no
significant effect of condition on the muscle malonyl-CoA
concentration.
PDH. At rest PDHa activity was lower in L-CHO (0.17 ⫾
0.06 mmol䡠min⫺1 䡠kg wet wt⫺1) than in H-CHO (0.52 ⫾ 0.10
mmol䡠min⫺1 䡠kg wet wt⫺1) (P ⬍ 0.01) (Fig. 5). In both
conditions, PDHa activity was higher after 60 min of exercise
than at rest (P ⬍ 0.001). After 60 min of exercise, PDHa
activity was still lower in L-CHO (0.86 ⫾ 0.12 mmol䡠min⫺1 䡠
kg wet wt⫺1) than in H-CHO (1.48 ⫾ 0.21 mmol䡠min⫺1 䡠kg
wet wt⫺1) (P ⬍ 0.01).
Acetyl-CoA. The resting acetyl-CoA concentration in the
vastus lateralis muscle did not differ significantly between
L-CHO (18.8 ⫾ 2.6 ␮mol/kg dry wt) and H-CHO (13.5 ⫾ 3.0
␮mol/kg dry wt) (Fig. 6A). In L-CHO it was ⬃35% lower (P ⬍
0.05) during exercise than at rest, whereas in H-CHO it was
⬃70% higher at 30 min (P ⬍ 0.001) and at 60 min (P ⬍ 0.01)
of exercise than at rest. The muscle acetyl-CoA concentration
was lower in L-CHO than in H-CHO at 30 min (49%, P ⬍
0.001) and 60 min (45%, P ⬍ 0.01) of exercise.
www.ajpendo.org
E138 MUSCLE MALONYL-COA, CARNITINE, AND FAT OXIDATION
Fig. 4. Acetyl-CoA carboxylase (ACC) phosphorylation and malonyl-CoA
concentration in the vastus lateralis muscle at rest and during 60 min of bicycle
exercise at 65% V̇O2 peak with L-CHO or H-CHO. A: ACC␤ Ser221 phosphor-
ylation. Arb units, arbitrary units. #Main effect of exercise, P ⬍ 0.001. B:
malonyl-CoA concentration. w.w., Wet weight. †Different from rest, P ⬍ 0.05.
Carnitine and acetylcarnitine. At rest, the acetylcarnitine
concentration in the vastus lateralis muscle was higher in
L-CHO (8.1 ⫾ 0.9 mmol/kg dry wt) than in H-CHO (4.4 ⫾ 1.0
mmol/kg dry wt) (P ⬍ 0.01) (Fig. 6B). In L-CHO it did not
change significantly from rest to exercise, whereas it increased
(P ⬍ 0.001) ⬃138% in H-CHO. At 30 and 60 min of exercise,
the muscle acetylcarnitine concentration was ⬃37% lower in
L-CHO than in H-CHO (P ⬍ 0.01).
The free carnitine concentration in the vastus lateralis mus-
cle at rest was lower in L-CHO (16.1 ⫾ 1.4 mmol/kg dry wt)
than in H-CHO (20.2 ⫾ 1.3 mmol/kg dry wt) (P ⬍ 0.01) (Fig.
6C). In L-CHO it did not change significantly from rest to
exercise, whereas it decreased (P ⬍ 0.001) ⬃46% in H-CHO.
The muscle free carnitine concentration was higher in L-CHO
than in H-CHO at 30 min (55%, P ⬍ 0.001) and 60 min (43%,
P ⬍ 0.01) of exercise.
DISCUSSION
The intracellular mechanisms regulating the ratio between
fat and carbohydrate oxidation in human skeletal muscle dur-
ing exercise have not yet been fully clarified. A likely site of
regulation of fat oxidation is CPT-1 at the entry of LCFA-CoA
into mitochondria. The present study is the first to simulta-
neously measure several factors related to the possible role of
malonyl-CoA and carnitine in regulation of fat oxidation at
CPT-1 in humans during exercise. Exercise was performed
under two conditions at comparable work loads but with
marked differences in the relative fat and carbohydrate oxida-
AJP-Endocrinol Metab • VOL
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (122.171.080.167) on July 20, 2018.
Copyright © 2005 American Physiological Society. All rights reserved.
tion rates. The latter was accomplished by manipulating the
muscle glycogen content to high (H-CHO) or low (L-CHO)
levels before a 60-min bicycle exercise bout at 65% V̇O2 peak.
It was previously suggested that low preexercise muscle
glycogen content might be associated with high AMPK activity
in rodents during muscle contraction in vitro (8) and in humans
during exercise (44). However, in the latter study, circulating
epinephrine was also dependent on muscle glycogen content
(44), wherefore the possible effect of muscle glycogen on
␣2-AMPK activity may have been confounded by adrenergic
stimulation of ␣2-AMPK. Adrenergic stimulation of AMPK
has been demonstrated in 3T3-L1 adipocytes (47) and has also
been reported for rat skeletal muscle (19). In the present study,
circulating epinephrine did not differ significantly between
conditions, and, still, ␣2-AMPK activity during exercise was
markedly higher in L-CHO than in H-CHO (Table 1). This
suggests that altered muscle glycogen alone can influence
␣2-AMPK activity in human skeletal muscle during exercise
independently of adrenergic stimulation, in agreement with
studies in rodent muscle (8). On this basis, AMPK could be a
likely candidate mediating the effect of muscle glycogen on fat
oxidation through regulation of ACC activity and malonyl-
CoA formation, as suggested earlier (34, 41). However, the
similar ACC␤ Ser221 phosphorylation and malonyl-CoA con-
centration between L-CHO and H-CHO in the present study
(Fig. 4) suggest that the effect of muscle glycogen on fat
oxidation occurs primarily by other mechanisms in human
skeletal muscle during exercise.
AMPK Thr172 phosphorylation closely paralleled ␣2-AMPK
activity in the present study (Table 1), indicating that the
markedly higher ␣2-AMPK activity during exercise in L-CHO
than in H-CHO was due to a higher stimulatory effect on
AMPK by an upstream AMPK kinase that phosphorylated
␣2-AMPK on Thr172. However, altered muscle glycogen did
not influence the large increase in ACC␤ Ser221 phosphoryla-
tion seen with exercise (Fig. 4A), suggesting that allosteric
regulation of AMPK, which is not detected in the AMPK
activity assay, may have overridden the covalent regulation of
AMPK by an upstream AMPK kinase. This is in contrast to a
previous study in which ␣2-AMPK activity and ACC␤ Ser221
phosphorylation were closely associated during exercise with
low vs. high muscle glycogen (44). The major difference
between that study (44) and the present study was that, in the
present study, subjects ingested a preexercise meal and had
Fig. 5. Pyruvate dehydrogenase activity in the active form (PDHa) in the
vastus lateralis muscle at rest and at 60 min of bicycle exercise at 65% V̇O2 peak
with L-CHO or H-CHO. *Main effect of condition (L-CHO vs. H-CHO), P ⬍
0.01. #Main effect of exercise, P ⬍ 0.001.
288 • JANUARY 2005 • www.ajpendo.org
 MUSCLE MALONYL-COA, CARNITINE, AND FAT OXIDATION
Fig. 6. Acetyl-CoA, acetylcarnitine, and carnitine concentrations in the vastus
lateralis muscle at rest and at 30 and 60 min of bicycle exercise at 65% V̇O2 peak
with L-CHO or H-CHO. A: acetyl-CoA. B: acetylcarnitine. C: carnitine.
Different from L-CHO, *P ⬍ 0.01 and **P ⬍ 0.001. Different from rest, #P ⬍
0.05, †P ⬍ 0.01, and ‡P ⬍ 0.001.
glucose infused intravenously during exercise, which resulted
in similar blood glucose and epinephrine levels between con-
ditions. In the previous study, blood glucose and plasma
epinephrine concentrations differed markedly between condi-
tions (44). It follows that, in the present study, the different
rates of intravenous glucose infusion in H-CHO and L-CHO
leading to similar levels of circulating glucose and epinephrine
between H-CHO and L-CHO may have contributed to the
dissociation between measured ␣2-AMPK activity and ACC␤
Ser221 phosphorylation. If so, the in vivo ␣2-AMPK activity in
the present study may have been better reflected by the ACC␤
Ser221 phosphorylation state than by the in vitro ␣2-AMPK
activity. Alternatively, ACC␤ Ser221 phosphorylation may
have been affected by factors other than AMPK activity, such
AJP-Endocrinol Metab • VOL
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (122.171.080.167) on July 20, 2018.
Copyright © 2005 American Physiological Society. All rights reserved.
E139
as spatial dissociation of AMPK and ACC or phosphatase
activity acting on ACC␤ Ser221. Those factors may in some
cases render ACC␤ Ser221 phosphorylation a not-so-optimal
predictor of in vivo ␣2-AMPK activity. Altogether, the present
study suggests that during exercise, covalent regulation of
␣2-AMPK on Thr172 is glycogen dependent, but that ACC␤
Ser221 phosphorylation does not parallel ␣2-AMPK activity
when circulating glucose and epinephrine levels are standard-
ized. This is supported by findings in isolated contracting
rodent skeletal muscle with high and low muscle glycogen
levels, where differences in ␣2-AMPK activity did not translate
into differences in ACC activity (8).
Despite markedly different ␣2-AMPK activity during exer-
cise between H-CHO and L-CHO (Table 1), ACC␤ Ser221
phosphorylation and malonyl-CoA concentration did not differ
significantly between L-CHO and H-CHO (Fig. 4). Still, the fat
oxidation rate was higher during exercise with low muscle
glycogen even though 160% more glucose was infused intra-
venously under this condition. This suggests that the AMPK-
ACC-malonyl-CoA pathway is not likely to mediate the regu-
lation of fat oxidation by muscle glycogen availability in
human skeletal muscle during prolonged exercise. This con-
clusion is also supported by other human exercise models
where no association between malonyl-CoA content and fat
oxidation rate was obtained in skeletal muscle during pro-
longed moderate-intensity exercise (17) or during graded-
intensity exercise (7, 18). However, compartmentalization of
malonyl-CoA may occur in skeletal muscle, i.e., concentrations
of malonyl-CoA in the vicinity of CPT-1 may vary without any
detectable changes in the measurement of total muscle malo-
nyl-CoA concentration. In resting human skeletal muscle,
changes in malonyl-CoA concentration have occurred with
opposite changes in fat oxidation. Thus a significant negative
correlation was obtained between muscle malonyl-CoA con-
tent and fat oxidation rate in healthy middle-aged men during
a two-step euglycemic-hyperinsulinemic clamp (3). Further-
more, in healthy, young individuals, the inhibition of fat
oxidation induced by hyperglycemia with hyperinsulinemia
occurred together with a threefold increase in muscle malonyl-
CoA concentration (27). These results suggest that, in resting
human skeletal muscle, the downregulation of fat oxidation by
insulin may be mediated by malonyl-CoA.
The absolute increase in fat oxidation rate that occurred from
rest to exercise in both conditions in the present study (Fig. 2)
was associated with increased ACC␤ Ser221 phosphorylation
and decreased malonyl-CoA concentration (Fig. 4). This sug-
gests that inactivation of ACC␤ by AMPK and the consequent
decrease in malonyl-CoA concentration from rest to exercise
may have relieved malonyl-CoA inhibition of fat oxidation in
both conditions (Fig. 7). There was a tendency (P ⫽ 0.10) for
malonyl-CoA to decrease more from rest to exercise in L-CHO
than in H-CHO. This could have been due to lower availability
of cytosolic acetyl-CoA, the precursor of malonyl-CoA, in
L-CHO than in H-CHO. Alternatively, the activity of MCD,
the enzyme catalyzing the turnover of malonyl-CoA, may have
differed between the two conditions. It has been suggested that
AMPK phosphorylates and activates MCD during muscle con-
traction (19, 34). Consequently, the higher ␣2-AMPK activity
in L-CHO than in H-CHO might have induced higher malonyl-
CoA turnover in L-CHO via an effect on MCD. The tendency
toward a more pronounced decrease in muscle malonyl-CoA
288 • JANUARY 2005 • www.ajpendo.org
E140 MUSCLE MALONYL-COA, CARNITINE, AND FAT OXIDATION

Fig. 7. Metabolic events that may determine the fat oxidation rate in response to muscle glycogen availability in human skeletal muscle during exercise. A:
individuals with high muscle glycogen. Exercise activates glycogenolysis, glycolysis, and PDH, leading to markedly increased mitochondrial generation of
acetyl-CoA and, secondarily, acetylcarnitine. The latter results in a decrease in mitochondrial free carnitine and, consequently, cytosolic carnitine. Whole tissue
malonyl-CoA levels are reduced by exercise due to increased phosphorylation (inactivation) of ACC by AMP-activated protein kinase (AMPK), which is
activated moderately by exercise. The combination of reduced free carnitine content and reduced malonyl-CoA content leads to a moderate increase in fat
oxidation rate from rest to exercise. B: individuals with low muscle glycogen. The increases in glycogenolytic and glycolytic rates and PDH activity by exercise
are less than in individuals with high glycogen. Consequently, mitochondrial acetyl-CoA and acetylcarnitine generation appears to be lower, and no decrease in
whole tissue free carnitine content occurs from rest to exercise. Whole tissue malonyl-CoA levels are reduced by exercise due to increased phosphorylation
(inactivation) of ACC by AMPK, which is activated markedly by exercise. The reduction in malonyl-CoA may be slightly more pronounced compared with
high-glycogen individuals because of the lower concentration of acetyl-CoA, the source from which malonyl-CoA is formed by ACC. The combination of
unchanged free carnitine content and reduced malonyl-CoA content leads to a marked increase in fat oxidation rate from rest to exercise. Muscle concentration
of pyruvate was not measured in the present study but was previously shown to increase more from rest to exercise with high than with low muscle glycogen
(25). The muscle concentration of citrate, the source of the acetyl-CoA from which malonyl-CoA is generated (35), was not measured in the present study but
was previously shown not to differ during exercise with high and low muscle glycogen, although an increase with exercise occurred (11). AcCoA, acetyl-CoA;
MaCoA, malonyl-CoA; FACoA, fatty acyl-CoA; CAT, carnitine acetyltransferase; CPT, carnitine palmitoyltransferase-1/2 system; AMPK, 5⬘-AMP-activated
protein kinase; CL, citrate lyase; ⫹, activating effect; ⫼, inhibiting effect.

by exercise in L-CHO than in H-CHO may have contributed to
the difference between conditions in fat oxidation rate during
exercise (Fig. 7). However, the increase in absolute fat oxida-
tion rate by exercise was much more marked in L-CHO than in
H-CHO. Therefore, factors other than malonyl-CoA seem to be
more important in “fine tuning” the fat oxidation rate during
prolonged exercise with different muscle glycogen stores.
Availability of carnitine, a necessary substrate for CPT-1,
could be one such factor. The importance of carnitine in fat
oxidation is evidenced by the fact that, in the pathology of lipid
storage myopathy, a marked reduction in skeletal muscle
carnitine content by 85% was associated with a 75% reduction
in palmitate and oleate oxidation in rectus femoris muscle
homogenates (9). On the other hand, the usual fluctuations in
free carnitine content in skeletal muscle of healthy humans
between 1 and 4 mM are not expected to influence CPT-1
activity, since it has been reported that the Km of CPT-1 for
AJP-Endocrinol Metab • VOL
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (122.171.080.167) on July 20, 2018.
Copyright © 2005 American Physiological Society. All rights reserved.
carnitine is ⬃0.5 mM at pH 7.4 in human skeletal muscle (14).
Nevertheless, partitioning of free carnitine between the cytosol
and the mitochondrial matrix makes it very difficult to estimate
the absolute carnitine concentration in the vicinity of CPT-1.
Furthermore, a drop in muscle pH during high-intensity exer-
cise (2, 39) leads to an increase in the Km of CPT-1 for
carnitine (16). Thus carnitine may still be a potent regulator of
CPT-1 during exercise. In support, the skeletal muscle free
carnitine concentration as well as the fat oxidation rate was
lower during prolonged moderate-intensity exercise in H-CHO
than in L-CHO in the present study (Figs. 2 and 6C). Also, in
several previous human studies, similar alterations in muscle
free carnitine content and fat oxidation rate occurred in parallel
during exercise. Thus muscle carnitine content and fat oxida-
tion rate increased to a similar extent during graded-intensity
exercise (5, 40). During bicycle exercise at 75% V̇O2 peak to
exhaustion with either high or low muscle glycogen, both
288 • JANUARY 2005 • www.ajpendo.org
 MUSCLE MALONYL-COA, CARNITINE, AND FAT OXIDATION
muscle carnitine content and fat oxidation rate were markedly
higher with low muscle glycogen (25). Taken together, the
present and previous studies suggest a relation between muscle
free carnitine availability and fat oxidation rate in human
skeletal muscle during exercise.
Protein oxidation was not measured in the present study,
and, consequently, absolute rates of fat and carbohydrate oxi-
dation across the leg were calculated using the nonprotein
respiratory quotient (20). Protein oxidation may have differed
between L-CHO and H-CHO and, in that case, would have
slightly confounded the fat and carbohydrate oxidation rates
reported in Fig. 2. However, during moderate-intensity exer-
cise, protein usually covers ⬍5% of oxidative metabolism (22).
Therefore, the fact that protein oxidation was ignored in the
present study has probably not influenced the results and
interpretations to any major extent.
Several potentially confounding blood metabolites and hor-
mones were kept almost similar between L-CHO and H-CHO
in the present study due to a light carbohydrate-rich preexercise
meal and a variable intravenous glucose infusion during exer-
cise. Thus the higher preexercise muscle glycogen content
seems to have been the primary cause of the higher PDH
activity seen in H-CHO vs. L-CHO (Fig. 5). In accordance, it
has previously been suggested that PDH activity is sensitive to
the muscle glycogen level (21, 25). PDH activity seems to be
important in limiting glycolysis during exercise (24, 38, 42). In
the present study, the high PDH activity during exercise in
H-CHO may have, consequently, enhanced glycolytic flux and
production of acetyl-CoA from carbohydrate compared with
L-CHO (Fig. 7). The accumulation of acetyl-CoA in H-CHO
compared with L-CHO most likely decreased the muscle free
carnitine content via the carnitine acetyltransferase equilibrium
(26). If muscle free carnitine is an important regulator of the fat
oxidation rate during exercise, then PDH, acetyl-CoA, and
acetylcarnitine could have linked the regulation of fat oxidation
to carbohydrate availability and glycolytic flux (Fig. 7). The
present findings therefore support the hypothesis that regula-
tion of fat oxidation in response to carbohydrate availability in
human skeletal muscle during exercise is at least partly medi-
ated via the availability of free carnitine to CPT-1.
In conclusion, the present study is the first to show a decline
in muscle malonyl-CoA concentrations from rest to moderate-
intensity exercise in humans, which may contribute to the
increase in absolute fat oxidation at the onset of exercise.
However, malonyl-CoA does not seem to be a major factor in
fine tuning fat oxidation in human skeletal muscle during
prolonged exercise with different preexercise muscle glycogen
content, since muscle malonyl-CoA concentrations did not
depend on muscle glycogen levels despite marked differences
in fat oxidation rates between low and high muscle glycogen
conditions. On the other hand, the present findings support the
assertion that the availability of free carnitine to CPT-1 may
participate in regulation of fat oxidation in human skeletal
muscle during prolonged moderate-intensity exercise, since
muscle carnitine and fat oxidation rate were both lower during
exercise with high compared with low glycogen.
ACKNOWLEDGMENTS
We thank Dr. Jens Knudsen, University of Southern Denmark, for kindly
donating the purified fatty acid synthase and Dr. D. Grahame Hardie, Univer-
sity of Dundee, Scotland, for kindly donating the ␣-isoform-specific AMPK
AJP-Endocrinol Metab • VOL 288 • JANUARY 2005 •
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (122.171.080.167) on July 20, 2018.
Copyright © 2005 American Physiological Society. All rights reserved.
E141
antibodies. We are grateful to Dr. Henriette Pilegaard, University of Copen-
hagen, Denmark, for help on the PDHa assay. We acknowledge the skilled
technical assistance of Irene Bech Nielsen, Betina Bolmgren, and Winnie
Taagerup.
GRANTS
This study was supported by The Danish National Research Foundation
(grant no. 504-14), The Copenhagen Muscle Research Centre, The Novo
Nordisk Research Foundation, The Danish Diabetes Association, a Research
and Technological Development Project (QLG1-CT-2001-01488) funded by
the European Commission, The Danish Sports Research Council, The Danish
Medical Research Council, and a Hallas Møller Fellowship from The Novo
Nordisk Foundation (J. F. P. Wojtaszewski).
REFERENCES
1. Andersen P and Saltin B. Maximal perfusion of skeletal muscle in man.
J Physiol 366: 233–249, 1985.
2. Bangsbo J, Madsen K, Kiens B, and Richter EA. Effect of muscle
acidity on muscle metabolism and fatigue during intense exercise in man.
J Physiol 495: 587–596, 1996.
3. Bavenholm PN, Pigon J, Saha AK, Ruderman NB, and Efendic S.
Fatty acid oxidation and the regulation of malonyl-CoA in human muscle.
Diabetes 49: 1078 –1083, 2000.
4. Cederblad G, Carlin JI, Constantin-Teodosiu D, Harper P, and
Hultman E. Radioisotopic assays of CoASH and carnitine and their
acetylated forms in human skeletal muscle. Anal Biochem 185: 274 –278,
1990.
5. Constantin-Teodosiu D, Carlin JI, Cederblad G, Harris RC, and
Hultman E. Acetyl group accumulation and pyruvate dehydrogenase
activity in human muscle during incremental exercise. Acta Physiol Scand
143: 367–372, 1991.
6. Constantin-Teodosiu D, Cederblad G, and Hultman E. A sensitive
radioisotopic assay of pyruvate dehydrogenase complex in human muscle
tissue. Anal Biochem 198: 347–351, 1991.
7. Dean D, Daugaard JR, Young ME, Saha A, Vavvas D, Asp S, Kiens B,
Kim KH, Witters L, Richter EA, and Ruderman N. Exercise dimin-
ishes the activity of acetyl-CoA carboxylase in human muscle. Diabetes
49: 1295–1300, 2000.
8. Derave W, Ai H, Ihlemann J, Witters LA, Kristiansen S, Richter EA,
and Ploug T. Dissociation of AMP-activated protein kinase activation and
glucose transport in contracting slow-twitch muscle. Diabetes 49: 1281–
1287, 2000.
9. Engel AG and Angelini C. Carnitine deficiency of human skeletal muscle
with associated lipid storage myopathy: a new syndrome. Science 179:
899 –902, 1973.
10. Frosig C, Jorgensen SB, Hardie DG, Richter EA, and Wojtaszewski
JFP. 5⬘-AMP-activated protein kinase activity and protein expression are
regulated by endurance training in human skeletal muscle. Am J Physiol
Endocrinol Metab 286: E411–E417, 2004.
11. Gibala MJ, Peirce N, Constantin-Teodosiu D, and Greenhaff PL.
Exercise with low muscle glycogen augments TCA cycle anaplerosis but
impairs oxidative energy provision in humans. J Physiol 540: 1079 –1086,
2002.
12. Helge JW, Watt PW, Richter EA, Rennie MJ, and Kiens B. Fat
utilization during exercise: adaptation to a fat-rich diet increases utilization
of plasma fatty acids and very low density lipoprotein-triacylglycerol in
humans. J Physiol 537: 1009 –1020, 2001.
13. Lowry OH and Passonneau JV. A Flexible System of Enzymatic Anal-
ysis. New York: Academic, 1972.
14. McGarry JD, Mills SE, Long CS, and Foster DW. Observations on the
affinity for carnitine, and malonyl-CoA sensitivity, of carnitine palmitoyl-
transferase I in animal and human tissues. Demonstration of the presence
of malonyl-CoA in non-hepatic tissues of the rat. Biochem J 214: 21–28,
1983.
15. McGarry JD, Stark MJ, and Foster DW. Hepatic malonyl-CoA levels
of fed, fasted and diabetic rats as measured using a simple radioisotopic
assay. J Biol Chem 253: 8291– 8293, 1978.
16. Mills SE, Foster DW, and McGarry JD. Effects of pH on the interaction
of substrates and malonyl-CoA with mitochondrial carnitine palmitoyl-
transferase I. Biochem J 219: 601– 608, 1984.
17. Odland LM, Heigenhauser GJ, Lopaschuk GD, and Spriet LL. Human
skeletal muscle malonyl-CoA at rest and during prolonged submaximal
exercise. Am J Physiol Endocrinol Metab 270: E541–E544, 1996.
www.ajpendo.org
E142 MUSCLE MALONYL-COA, CARNITINE, AND FAT OXIDATION
18. Odland LM, Howlett RA, Heigenhauser GJ, Hultman E, and Spriet
LL. Skeletal muscle malonyl-CoA content at the onset of exercise at
varying power outputs in humans. Am J Physiol Endocrinol Metab 274:
E1080 –E1085, 1998.
19. Park H, Kaushik VK, Constant S, Prentki M, Przybytkowski E,
Ruderman NB, and Saha AK. Coordinate regulation of malonyl-CoA
decarboxylase, sn-glycerol-3-phosphate acyltransferase, and acetyl-CoA
carboxylase by AMP-activated protein kinase in rat tissues in response to
exercise. J Biol Chem 277: 32571–32577, 2002.
20. Peronnet F and Massicotte D. Table of nonprotein respiratory quo-
tient—an update. Can J Sport Sci 16: 23–29, 1991.
21. Peters SJ, St. Amand TA, Howlett RA, Heigenhauser GJ, and Spriet
LL. Human skeletal muscle pyruvate dehydrogenase kinase activity in-
creases after a low-carbohydrate diet. Am J Physiol Endocrinol Metab
275: E980 –E986, 1998.
22. Phillips SM, Atkinson SA, Tarnopolsky MA, and MacDougall JD.
Gender differences in leucine kinetics and nitrogen balance in endurance
athletes. J Appl Physiol 75: 2134 –2141, 1993.
23. Price NT, Jackson VN, van der Leij FR, Cameron JM, Travers MT,
Bartelds B, Huijkman NC, and Zammit VA. Cloning and expression of
the liver and muscle isoforms of ovine carnitine palmitoyltransferase 1:
residues within the N-terminus of the muscle isoform influence the kinetic
properties of the enzyme. Biochem J 372: 871– 879, 2003.
24. Putman CT, Jones NL, Lands LC, Bragg TM, Hollidgehorvat MG,
and Heigenhauser GJF. Skeletal muscle pyruvate dehydrogenase activity
during maximal exercise in humans. Am J Physiol Endocrinol Metab 269:
E458 –E468, 1995.
25. Putman CT, Spriet LL, Hultman E, Lindinger MI, Lands LC, McK-
elvie RS, Cederblad G, Jones NL, and Heigenhauser GJ. Pyruvate
dehydrogenase activity and acetyl group accumulation during exercise
after different diets. Am J Physiol Endocrinol Metab 265: E752–E760, 1993.
26. Ramsay RR and Naismith JH. A snapshot of carnitine acetyltransferase.
Trends Biochem Sci 28: 343–346, 2003.
27. Rasmussen BB, Holmback UC, Volpi E, Morio-Liondore B, Paddon-
Jones D, and Wolfe RR. Malonyl coenzyme A and the regulation of
functional carnitine palmitoyltransferase-1 activity and fat oxidation in
human skeletal muscle. J Clin Invest 110: 1687–1693, 2002.
28. Rasmussen BB and Winder WW. Effect of exercise intensity on skeletal
muscle malonyl-CoA and acetyl-CoA carboxylase. J Appl Physiol 83:
1104 –1109, 1997.
29. Roepstorff C, Steffensen CH, Madsen M, Stallknecht B, Kanstrup IL,
Richter EA, and Kiens B. Gender differences in substrate utilization
during submaximal exercise in endurance-trained subjects. Am J Physiol
Endocrinol Metab 282: E435–E447, 2002.
30. Roepstorff C, Vistisen B, Donsmark M, Nielsen JN, Galbo H, Green
KA, Hardie DG, Wojtaszewski JFP, Richter EA, and Kiens B. Regu-
lation of hormone-sensitive lipase activity and Ser563 and Ser565 phos-
phorylation in human skeletal muscle during exercise. J Physiol 560:
551–562, 2004.
31. Roepstorff C, Vistisen B, Roepstorff K, and Kiens B. Regulation of
plasma long-chain fatty acid oxidation in relation to uptake in human
skeletal muscle during exercise. Am J Physiol Endocrinol Metab 287:
E696 –E705, 2004.
AJP-Endocrinol Metab • VOL 288 • JANUARY 2005 •
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (122.171.080.167) on July 20, 2018.
Copyright © 2005 American Physiological Society. All rights reserved.
32. Saha AK, Kurowski TG, and Ruderman NB. A malonyl-CoA fuel-
sensing mechanism in muscle: effects of insulin, glucose, and denervation.
Am J Physiol Endocrinol Metab 269: E283–E289, 1995.
33. Saha AK, Laybutt DR, Dean D, Vavvas D, Sebokova E, Ellis B, Klimes
I, Kraegen EW, Shafrir E, and Ruderman NB. Cytosolic citrate and
malonyl-CoA regulation in rat muscle in vivo. Am J Physiol Endocrinol
Metab 276: E1030 –E1037, 1999.
34. Saha AK, Schwarsin AJ, Roduit R, Masse F, Kaushik V, Tornheim K,
Prentki M, and Ruderman NB. Activation of malonyl-CoA decarbox-
ylase in rat skeletal muscle by contraction and the AMP-activated protein
kinase activator 5-aminoimidazole-4-carboxamide-1-␤-D-ribofuranoside.
J Biol Chem 275: 24279 –24283, 2000.
35. Saha AK, Vavvas D, Kurowski TG, Apazidis A, Witters LA, Shafrir
E, and Ruderman NB. Malonyl-CoA regulation in skeletal muscle: its
link to cell citrate and the glucose-fatty acid cycle. Am J Physiol Endo-
crinol Metab 272: E641–E648, 1997.
36. Sidossis LS, Gastaldelli A, Klein S, and Wolfe RR. Regulation of
plasma fatty acid oxidation during low- and high-intensity exercise. Am J
Physiol Endocrinol Metab 272: E1065–E1070, 1997.
37. Sidossis LS, Wolfe RR, and Coggan AR. Regulation of fatty acid
oxidation in untrained vs. trained men during exercise. Am J Physiol
Endocrinol Metab 274: E510 –E515, 1998.
38. Stellingwerff T, Watt MJ, Heigenhauser GJ, and Spriet LL. Effects of
reduced free fatty acid availability on skeletal muscle PDH activation
during aerobic exercise. Pyruvate dehydrogenase. Am J Physiol Endocri-
nol Metab 284: E589 –E596, 2003.
39. Street D, Bangsbo J, and Juel C. Interstitial pH in human skeletal muscle
during and after dynamic graded exercise. J Physiol 537: 993–998, 2001.
40. van Loon LJ, Greenhaff PL, Constantin-Teodosiu D, Saris WH, and
Wagenmakers AJ. The effects of increasing exercise intensity on muscle
fuel utilisation in humans. J Physiol 536: 295–304, 2001.
41. Vavvas D, Apazidis A, Saha AK, Gamble J, Patel A, Kemp BE,
Witters LA, and Ruderman NB. Contraction-induced changes in acetyl-
CoA carboxylase and 5⬘-AMP-activated kinase in skeletal muscle. J Biol
Chem 272: 13255–13261, 1997.
42. Watt MJ, Heigenhauser GJ, Dyck DJ, and Spriet LL. Intramuscular
triacylglycerol, glycogen and acetyl group metabolism during 4 h of
moderate exercise in man. J Physiol 541: 969 –978, 2002.
43. Winder WW. Intramuscular mechanisms regulating fatty acid oxidation
during exercise. Adv Exp Med Biol 441: 239 –248, 1998.
44. Wojtaszewski JF, MacDonald C, Nielsen JN, Hellsten Y, Hardie DG,
Kemp BE, Kiens B, and Richter EA. Regulation of 5⬘AMP-activated
protein kinase activity and substrate utilization in exercising human
skeletal muscle. Am J Physiol Endocrinol Metab 284: E813–E822, 2003.
45. Wojtaszewski JF, Nielsen P, Hansen BF, Richter EA, and Kiens B.
Isoform-specific and exercise intensity-dependent activation of 5⬘-AMP-
activated protein kinase in human skeletal muscle. J Physiol 528: 221–
226, 2000.
46. Woods A, Salt I, Scott J, Hardie DG, and Carling D. The alpha 1 and
alpha 2 isoforms of the AMP-activated protein kinase have similar
activities in rat liver but exhibit differences in substrate specificity in vitro.
FEBS Lett 397: 347–351, 1996.
47. Yin W, Mu J, and Birnbaum MJ. Role of AMP-activated protein kinase
in cyclic AMP-dependent lipolysis in 3T3-L1 adipocytes. J Biol Chem
278: 43074 – 43080, 2003.
www.ajpendo.org