Localization and solubilization of a rat liver microsomal carnitine acetyltransferase.

Carnitine acetyltransferase activity had been previously shown to occur in peroxisomes, mitochondria, and a membranous fraction of rat and pig hepatocytes. When components of this third subcellular fraction (plasma membranes, components of the Golgi apparatus, and microsomes) were further separated, carnitine acetyltransferase fractionated with the microsomes. Microsomes isolated by three different methods (isopycnic sucrose density zonal centrifugation, high-speed differential centrifugation, and aggregation with Ca²⁺ followed by low-speed differential centrifugation) all contained carnitine acetyltransferase activity. The lability of carnitine acetyltransferase in microsomes isolated by different methods and in different isolation media is reported.When total microsomes were subfractionated into rough and smooth components, carnitine acetyltransferase activity was found to the same extent in both and was tightly associated with the microsomal membrane. The microsomal enzyme was rapidly inactivated in 0.25 m sucrose or 0.1 m phosphate, but was stable for at least 2 weeks in 0.4 m KCl. Extensive treatment with high ionic strength salt solutions, 1% Triton X-100, or a combination of the two was used to solubilize microsomal carnitine acetyltransferase activity.Carnitine octanoyltransferase activity was also found in the microsomal fractions isolated by three different methods, but no carnitine palmitoyltransferase was detected in the microsomal fractions. It is proposed that microsomal carnitine acetyl- and octanoyltransferases could be involved in the transfer of acyl groups across the microsomal membrane, thereby providing a source of acetyl and other acyl CoA's at sites of acetylation reactions and synthesis.     

In vivo and in vitro intervention with L-carnitine prevents abnormal energy metabolism in isolated diabetic rat heart: chemical and phosphorus-31 NMR evidence

The beneficial effects of in vivo injections (200 mg/kg, twice daily) or in vitro perfusion (5.0 mM) of L-carnitine on an intrinsic abnormality in energy metabolism was investigated in isolated, perfused diabetic rat heart. Hearts were aerobically perfused for 60 min with elevated fatty acid substrate to simulate diabetic conditions. Phosphorus-31 nuclear magnetic resonance spectroscopy revealed a temporal decline in myocardial ATP levels (to approx 82%) during perfusion of diabetic hearts, but not in control hearts. This reduction was prevented by prior treatment in vivo with L-carnitine or by providing L-carnitine acutely in the perfusion medium. Chemical analysis of tissue extracts indicated that L-carnitine injections were effective in replenishing the decrease in total myocardial carnitine content which was present in diabetic hearts and in preventing the accumulation of long chain fatty acyl CoA. Perfusion with L-carnitine also attenuated the elevation of long chain fatty acyl CoA in diabetic hearts. This study gives additional support to the hypothesis that decreases in ATP which occur in the isolated, perfused diabetic heart are correlated with a concomitant elevation in long chain fatty acyl CoA, a known inhibitor of adenine nucleotide translocase. In the presence of elevated exogenous fatty acids, a primary deficiency in the total myocardial carnitine pool would result in elevations in tissue concentrations of long chain fatty acyl CoA since carnitine is a required carrier for transport of fatty acids into mitochondria. Replenishment of the carnitine in vivo was shown to be sufficient to prevent subsequent alteration in long chain fatty acyl CoA and ATP in isolated perfused diabetic hearts despite the burden of elevated fatty acid substrates.     

Purification and properties of carnitine octanoyltransferase and carnitine palmitoyltransferase from rat liver

The activities of carnitine octanoyltransferase (COT) and carnitine palmitoyltransferase (CPT) in rat liver were markedly increased by administration of di(2-ethyl-hexyl)phthalate. COT and CPT were purified from the enzyme-induced rat liver. COT was a 66,000-dalton polypeptide. The molecular weight of native CPT was 280,000--320,000 daltons, and the enzyme consisted of 69,200-dalton polypeptides. CAT, COT, and CPT were immunologically different. COT exhibited activity with all of the substrates tested (acyl-CoA's and acylcarnitines of saturated fatty acids having carbon chain lengths of C2--C20), though maximum activity was observed with hexanoyl derivatives. CPT exhibited catalytic activity with medium- and long-chain acyl derivatives. 2-Bromo-palmitoyl-CoA inactivated COT but not CPT. Malonyl-CoA inhibited CPT but not COT. CPT was confined to mitochondria, whereas COT was found in peroxisomes and the soluble compartment but not in mitochondria.     

Changes in tissue levels of carnitine and other metabolites during myocardial ischemia and anoxia

The induction of ischemia in the open chest dog, or anoxia in the perfused rat heart, causes dramatic changes in the tissue levels of free acyl carnitine and related metabolites. During the early phase of ischemia or anoxia the tissue levels of free carnitine decline, while acetyl carnitine rapidly increases. These changes are accompanied by elevation in long-chain acyl carnitine, long-chain acyl CoA, and lactate and by decreases in acetyl CoA, CoA, ATP, and creatine phosphate. As the degree of ischemia becomes more severe, carnitine appears to be lost from the myocardium. A scheme is presented which relates carnitine-linked mitochondrial metabolism to the activity of carnitine acyl transferase, ANT, carnitine/acyl carnitine translocase, creatine phosphokinase, and pyruvate dehydrogenase. It is suggested that the conversion of carnitine to acyl carnitine during the onset of ischemia may play an important role, by virtue of its effect on these enzymes, in the regulation of metabolism during the early or reversible phase of ischemia.     

Surplus acylcarnitines in the plasma of starved rats derive from the liver

The method used here to assess the contribution of liver to plasma acylcarnitine is based on the idea that in rat, shortly after administration of [3H]butyrobetaine the [3H]carnitine appearing in the plasma derives from the liver and so does the acyl moiety of [acyl-3H] carnitine. In the perchloric acid extracts of plasma and liver, the ester fraction of total carnitine was determined by enzymatic analysis and that of [3H]carnitines was determined by high performance liquid chromatography. The ester fraction of total carnitine in the plasma of fed rats was 32.6% while that of [3H]carnitines was 67.9%, 1 h following injection of [3H]butyrobetaine. For 48 h starved rats the equivalent values were 54.2 and 84.0%, respectively. 24 h after the administration of [3H]butyrobetaine, the ester content became the same in the total and [3H]carnitines. That the newly synthesized carnitine was more acylated (67.9 versus 32.6%, fed) indicates that liver exports acyl groups with carnitine as carrier. The observation that the ester fraction in the newly synthesized plasma carnitine increased with fasting (84.0 versus 67.9%) indicates that the surplus plasma acylcarnitine in fasting ketosis derives from the liver. Perfused livers, however, released carnitine with the same ester content (60-61%) whether they were from fed or fasted animals. Probably, the increased plasma [acylcarnitine] in fasting develops not by an increased ester output from the liver but by an altered handling in extrahepatic tissues.     

The synthesis of palmitoylcarnitine by etio-chloroplasts of greening barley leaves

CoASH, Mg²⁺, ATP and (-)-carnitine were found to be essential for the production of palmitoylcarnitine from palmitate by purified barley etio-chloroplasts. It was concluded that long-chain acyl CoA synthetase (palmitoyl CoA synthetase, EC 6.2.1.3) and carnitine long-chain acyl-transferase (carnitine palmitoyltransferase, EC 2.3.1.21) activity were present in the etio-chloroplasts. It is suggested that the long-chain acylcarnitine formed may move more easily through membrane barriers than the long-chain acyl CoA compound. Also or alternatively this enzyme may spare CoA by transferring long-chain acyl groups from long-chain acyl CoA to carnitine.     

Monoacylglycerol lipase activity in cardiac myocytes

Monoacylglycerol lipase activity in homogenates of isolated myocardial cells (myocytes) from rat hearts was recovered in both particulate and soluble subcellular fractions. The activity present in the microsomal (100,000 X g pellet) fraction was solubilized by treatment with Triton X-100 and combined with the 100,000 X g supernatant fraction; the properties of monoacylglycerol lipase were investigated with this soluble enzyme preparation. The Km for the hydrolysis of a 2-monoolein substrate was 16 microM. The rates of hydrolysis of 1-monoolein and 2-monoolein were identical, and 1-monoolein was a competitive inhibitor (Ki = 20 microM) of the hydrolysis of 2-monoolein. Monoacylglycerol lipase activity was regulated by product inhibition according to the following order of potency: fatty acyl CoA greater than free fatty acids greater than fatty acyl carnitine.     

Effect of vitamin B-12 deficiency on the hepatic tissue concentration of acyl carnitines.

Concentrations of carnitine, acetyl carnitine, propionyl carnitine, and long chain acyl carnitines have been measured in hepatic tissue of normal and vitamin B-12 deficient rats using radiolabelled butyrobetaine to label carnitine pools. Increased levels of propionyl carnitine were seen in the livers of vitamin B-12 deprived animals when compared to those from normal animals. Methylmalonyl carnitine was not detected in the B-12 deprived animals. Free carnitine levels were no different in the livers from the B-12 deprived animals than from the normal control animals.     

Inhibitory effect of long chain fatty acyl CoAs on RNA polymerase from Escherichia coli

RNA polymerase from Escherichia coli was inhibited by long chain fatty acyl CoAs, such as myristoyl CoA (Ki = 17.2 microM), palmitoyl CoA (Ki = 8.9 microM), oleoyl CoA (Ki = 5.5 microM), and stearoyl CoA (Ki = 0.94 microM). The inhibition by these CoA thioesters was non-competitive against nucleoside triphosphates. Short chain fatty acyl CoAs, such as acetyl CoA, propionyl CoA, acetoacetyl CoA, butyryl CoA, and decanoyl CoA, failed to inhibit RNA polymerase. CoA, Na-myristate, Na-palmitate, Na-oleate, Na-stearate, palmitoyl carnitine, and carnitine did not inhibit the enzyme. The inhibition of RNA polymerase by long chain fatty acyl CoAs was competitive against template DNA.     

Carnitine acyltransferases in rat liver peroxisomes

Carnitine acyltransferase activities, as well as acetyl-CoA, octanyl-CoA, and palmityl-CoA hydrolase activities, were assayed in mitochondrial, peroxisomal, and endoplasmic reticulum fractions after isopycnic density sucrose gradient fractionation of rat liver homogenates. Both the forward and reverse assays show that carnitine acetyltransferase and carnitine octanyltransferase are associated with peroxisomes, mitochondria, and endoplasmic reticulum, while carnitine palmityltransferase was detected in mitochondria. Palmityl-CoA and octanyl-CoA hydrolase activities were found in all but the leading edge of the peroxisome peak of the gradient. The palmityl-CoA hydrolase in peroxisomal fractions was due to lysosomal contamination since the activity coincided with the lysosomal marker, acid phosphatase. The substrate specificity for carnitine octanyltransferase activity was maximum with medium-chain-length derivatives (about 20 nmol/ min/mg protein) and decreased as the acyl length increased until very low activity (<1 nmol/min/mg protein) was obtained with palmityl-CoA. When acyltransferases in peroxisomes were assayed by measuring acylcarnitine formation, nearly theoretical amounts of acetylcarnitine and octanylcarnitine were formed, but lesser quantities of 12 and 14 carbon acylcarnitines and very low amounts of palmitylcarnitine were detected. The presence of a broad spectrum of medium-chain and short-chain carnitine acyltransferases in peroxisomes is consistent with a role for carnitine for shuttling short-chain and medium-chain acyl residues out of peroxisomes. Carnitine acyltransferase activity was not detected in peroxisomes from spinach leaves.     

The synthesis of short- and long-chain acylarnitine by etio-chloroplasts of greening barley leaves

Etio-chloroplasts of barley, purified on sucrose density gradients were shown to possess carnitine long-chain acyltransferase (carnitine palmitoyltransferase, EC 2.3.1.21) activity and carnitine short-chain acyltransferase (carnitine acetyltransferase EC 2.3.1.7) activity. These enzymes may play a role in the transport of acyl groups as acylcarnitines through the membrane barrier of barley etio-chloroplasts and also ‘or alternatively’ may spare CoA by transferring short- and long-chain acyl groups from short-and long-chain acyl CoA to carnitine.     

Purification and properties of peroxisomal carnitine palmitoyltransferase in chick embryo liver

Peroxisomal carnitine palmitoyltransferase was purified by solubilization using Tween 20 and KCl from the large granule fraction of the liver of clofibrate-treated chick embryo, DEAE-Sephacel and blue Sepharose CL-6B column chromatography. The peroxisomal carnitine palmitoyltransferase was an Mr 64,000 polypeptide; the mitochondrial carnitine palmitoyltransferase had a subunit molecular weight of 69,000 on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The carnitine acetyltransferase was an Mr 64,000 polypeptide. Antibody against purified peroxisomal carnitine palmitoyltransferase reacted only with peroxisomal carnitine palmitoyltransferase, but not with mitochondrial carnitine palmitoyltransferase or carnitine acetyltransferase. In addition, anti-peroxisomal carnitine palmitoyltransferase reacted only with the protein in peroxisomes purified from chick embryo liver by sucrose density gradient centrifugation. Thus, it was confirmed that purified peroxisomal carnitine palmitoyltransferase was a peroxisomal protein. Compared with mitochondrial carnitine palmitoyltransferase, peroxisomal carnitine palmitoyltransferase was extremely resistant to inactivation by trypsin. The pH optimum of peroxisomal carnitine palmitoyltransferase was 8.5, differing from that of mitochondrial carnitine palmitoyltransferase. The Km value of peroxisomal carnitine palmitoyltransferase for palmitoyl-CoA (32 microM) was similar to that of the mitochondrial one, whereas those values for L-carnitine (140 microM), palmitoyl-L-carnitine (43 microM) and CoA (9 microM) were lower than those of mitochondrial carnitine palmitoyltransferase. Peroxisomal carnitine palmitoyltransferase exhibited similar substrate specificities in both the forward and reverse reactions, with the highest activity toward lauroyl derivatives. Furthermore, this enzyme showed relatively high affinities for long-chain acyl derivatives (C10-C16) and similar Km values (30-50 microM) for acyl-CoAs, acylcarnitine and CoA, and a constant Km value (approximately 150 microM) for carnitine. These results indicate that peroxisomal carnitine palmitoyltransferase played a role in the modulation of the intracellular CoA/long-chain acyl-CoA ratio at the hatching stage of chicken when long-chain fatty acids are actively oxidized in peroxisomes.     

Structural and biochemical studies of the substrate selectivity of carnitine acetyltransferase

Carnitine acyltransferases catalyze the exchange of acyl groups between coenzyme A (CoA) and carnitine. They have important roles in many cellular processes, especially the oxidation of long-chain fatty acids, and are attractive targets for drug discovery against diabetes and obesity. These enzymes are classified based on their substrate selectivity for short-chain, medium-chain, or long-chain fatty acids. Structural information on carnitine acetyltransferase suggests that residues Met-564 and Phe-565 may be important determinants of substrate selectivity with the side chain of Met-564 located in the putative binding pocket for acyl groups. Both residues are replaced by glycine in carnitine palmitoyltransferases. To assess the functional relevance of this structural observation, we have replaced these two residues with small amino acids by mutagenesis, characterized the substrate preference of the mutants, and determined the crystal structures of two of these mutants. Kinetic studies confirm that the M564G or M564A mutation is sufficient to increase the activity of the enzyme toward medium-chain substrates with hexanoyl-CoA being the preferred substrate for the M564G mutant. The crystal structures of the M564G mutant, both alone and in complex with carnitine, reveal a deep binding pocket that can accommodate the larger acyl group. We have determined the crystal structure of the F565A mutant in a ternary complex with both the carnitine and CoA substrates at a 1.8-A resolution. The F565A mutation has minor effects on the structure or the substrate preference of the enzyme.     

The substrate specificity of carnitine acetyltransferase

1. A study of the acyl group specificity of the carnitine acetyltransferase reaction [acyl-(-)carnitine+CoASH right harpoon over left harpoon (-)-carnitine+acyl-CoA] has been made with the enzyme from pigeon breast muscle. Acyl groups containing up to 10 carbon atoms are transferred and detailed kinetic investigations with a range of acyl-CoA and acylcarnitine substrates are reported. 2. Acyl-CoA derivatives with 12 or more carbon atoms in the acyl group are potent reversible inhibitors of carnitine acetyltransferase, competing with acetyl-CoA. Lauroyl- and myristoyl-CoA show a mixed inhibition with respect to (-)-carnitine, but palmitoyl-CoA competes strictly with this substrate also. Palmitoyl-dl-carnitine shows none of these effects. 3. Ammonium palmitate inhibits the enzyme competitively with respect to (-)-carnitine and non-competitively with respect to acetyl-CoA. 4. It is suggested that a hydrophobic site exists on the carnitine acetyltransferase molecule. The hydrocarbon chain of an acyl-CoA derivative containing eight or more carbon atoms in the acyl group may interact with this, which results in enhanced acyl-CoA binding. Competition occurs between ligands bound to this hydrophobic site and the carnitine binding site. 5. The possible physiological significance of long-chain acyl-CoA inhibition of this enzyme is discussed.     

Carnitines slow down tumor development of colon cancer in the DMH‐chemical carcinogenesis mouse model

Dietary agents are receiving much attention for the chemoprevention of cancer. While curcumin is known to influence several pathways and affect tumor growth in vivo, carnitin and its congeners play a variety of important metabolic functions: are involved in the oxydation of long‐chain fatty acids, regulate acyl‐CoA levels and influence protein activity and stability by modifying the extent of protein acetylation. In this study we evaluated the efficacy of carnitines in the prevention of cancer development using the 1,2,‐dimethylhydrazine (DMH)‐induced colon carcinogenesis model. We also assessed whether their combination was able to give rise to increased protection from cancer development. Mice treated with DMH were dosed orally with curcumin and/or carnitine and acylcarnitines for 20 weeks. At the end of the treatment colon samples were collected, and scored for multiple ACF and adenomas. We observed that carnitine and acyl‐carnitines had same, if not higher, efficacy than curcumin alone in inhibiting the formation of neoplastic lesions induced by DMH treatment. Interestingly, the combination of curcumin and acetyl‐L ‐carnitine was able to fully inhibit the development of advanced adenoma lesions. Our data unveil the antitumor effects of carnitines and warrant additional studies to further support the adoption of carnitines as cancer chemopreventative agents. J. Cell. Biochem. 114: 1665–1673, 2013. © 2013 Wiley Periodicals, Inc.     

Effect of ischemia on fatty acid metabolism in fetal lung

The effects of ischemia on in vivo fatty acid metabolism in fetal lung were studied using rabbit fetuses of 25 to 28 gestational age. Ischemia was produced by inflating the aortic balloon thereby reducing the uterine blood flow. Ischemic insult resulted significant increase in lactate/pyruvate and NADH/NAD ratios and decrease in ATP/ADP ratio in fetal lung. Levels of CoA, acetyl CoA, carnitine and acetyl carnitine decreased while those of long chain acyl CoA and long chain acyl carnitine enhanced. Tissue content of these metabolites returned to normal after 2 hr stabilization following 20 min of ischemic insult. Ischemia also caused small increase in lipogenesis and neutral lipid content of fetal lungs. Our results thus suggest that β-oxidation in fetal lung is inhibited and becomes rate-limiting for fatty acid oxidation during ischemia.Sudden occurrence of hypoxia or ischemia in the fetus is a typical challenge for the obstetricians. The patients occasionally suffer from neurological injury following cerebral hypoxemia. The hypoxic insult may also affect the respiratory activity significantly. For example, acute alveolar hypoxia causes pulmonary vasoconstriction by damaging pulmonary vascular smooth muscle (1) and results in reduction of fatty acid oxidation by limiting the ATP supply required for metabolic processes (2). Hypoxia has also been shown to decrease the rate of palmitate incorporation into phospholipids (3), inhibit rate of fatty acid synthesis (3) and depress rate of incorporation of fatty acid and phosphatidic acid into lipids (4). Despite the fact that fatty acids represent a major substrate for energy metabolism in lung, no work has been done on the fatty acid metabolism in fetal lung. The present study was designed to determine the fate of fatty acid oxidation in fetal lung during ischemic challenge. The levels of acyl CoA and acylcarnitine intermediates were also measured in order to determine the rate-controlling steps of fatty acid metabolism in the fetal lung.     

Effect of clofibrate treatment on carnitine acyltransferases in different subcellular fractions of rat liver

The subcellular distribution of carnitine acetyl-, octanoyl-, and palmitoyltransferase in the livers of normal and clofibrate-treated male rats was studied with isopycnic sucrose density gradient fraction.In normal liver 48% of total carnitine acetyltransferase activity was peroxisomal, 36% of the activity located in mitochondria and 16% in a membranous fraction containing microsomes. Carnitine octanoyltransferase and carnitine palmitoyltransferase were confined almost totally (77–81%) to mitochondria in normal liver.Clofibrate treatment increased the total activity of carnitine acetyltransferase over 30 times, whereas the total activities of the other two transferases were increased only 5-fold.From the three different subcellular carnitine acetyltransferases the mitochondrial one was not responsive to clofibrate treatment, i.e. the rise in mitochondrial activity was over 70-fold as contrasted to the 6- and 14-fold rises in peroxisomal and microsomal activities, respectively. After treatment mitochondria contained 79% of total activity.It is concluded that the clofibrate-induced increase of carnitine acetyltransferase activity is not due to the peroxisomal proliferation that occurs during clofibrate treatment. The rise in peroxisomal activity contributed only 8% to the total increase.After clofibrate treatment the greatest part of carnitine octanoyl- and palmitoyltrnasferase activities were located in mitochondria but a considerable amount of both activities was found also in the soluble fraction of liver.     

Carnitine acyltransferase and acyl-coenzyme A hydrolase activities in human liver. Quantitative analysis of their subcellular localization.

The subcellular localizations of carnitine acyltransferase and acyl-CoA hydrolase activities with different chain-length substrates were quantitatively evaluated in human liver by fractionation of total homogenates in metrizamide density gradients and by differential centrifugation. Peroxisomes were found to contain 8-37% of the liver acyltransferase activity, the relative amount depending on the chain length of the substrate. The remaining activity was ascribed to mitochondria, except for carnitine octanoyltransferase, for which 25% of the activity was present in microsomal fractions. In contrast with rat liver, where the activity in peroxisomes is very low or absent, human liver peroxisomes contain about 20% of the carnitine palmitoyltransferase. Short-chain acyl-CoA hydrolase activity was found to be localized mainly in the mitochondrial and soluble compartments, whereas the long-chain activity was present in both microsomal fractions and the soluble compartment. Particle-bound acyl-CoA hydrolase activity for medium-chain substrates exhibited an intermediate distribution, in mitochondria and microsomal fractions, with 30-40% of the activity in the soluble fraction. No acyl-CoA hydrolase activity appears to be present in human liver peroxisomes.     

Identification by site-directed mutagenesis of a hydrophobic binding site of the mitochondrial carnitine/acylcarnitine carrier involved in the interaction with acyl groups

The role of hydrophobic residues of the mitochondrial carnitine/acylcarnitine carrier (CAC) in the inhibition by acylcarnitines has been investigated by site-directed mutagenesis. According to the homology model of CAC in cytosolic opened conformation (c-state), L14, G17, G21, V25, P78, V82, M85, C89, F93, A276, A279, C283, F287 are located in the 1st (H1), 2nd (H2) and 6th (H6) transmembrane α-helices and exposed in the central cavity, forming a hydrophobic half shell. These residues have been substituted with A (or G) and in some cases with M. Mutants have been assayed for transport activity measured as [(3)H]carnitine/carnitine antiport in proteoliposomes. With the exception of G17A and G21M, mutants exhibited activity from 20% to 100% of WT. Among the active mutants only G21A, V25M, P78A and P78M showed Vmax lower than half and/or Km more than two fold respect to WT. Acylcarnitines competitively inhibited carnitine antiport. The extent of inhibition of the mutants by acylcarnitines with acyl chain length of 2, 4, 8, 12, 14 and 16 has been compared with the WT. V25A, P78A, P78M and A279G showed reduced extent of inhibition by all the acylcarnitines; V25M showed reduced inhibition by shorter acylcarnitines; V82A, V82M, M85A, C89A and A276G showed reduced inhibition by longer acylcarnitines, respect to WT. C283A showed increased extent of inhibition by acylcarnitines. Variations of Ki of mutants for acylcarnitines reflected variations of the inhibition profiles. The data demonstrated that V25, P78, V82, M85 and C89 are involved in the acyl chain binding to the CAC in c-state.     

Effects of carnitine on respiratory chain and metabolism of oxygen radical in mitochondria of liver after exhaustive running in training rat

The aim of the present study was to investigate the effect of carnitine on cellular respiratory chain and metabolism of oxygen radical in mitochondria of liver after exhaustive running in training rats. Forty male Wistar rats were randomly divided into 4 groups (n = 10): control, carnitine, training and training+carnitine groups. The training and training+carnitine groups received 6-week treadmill training, whereas carnitine and training+carnitine groups were administrated intragastrically with carnitine (300 mg/kg per d) for 6 weeks. After exhaustive running, all the groups were sacrificed to obtain liver samples, and liver mitochondria were extracted by differential centrifugation. Spectrophotometric analysis was used to evaluate activities of respiratory chain complexes (RCC) I-IV, superoxide dismutase (SOD), glutathione peroxidase (GSH-Px) and the content of malondialdehyde (MDA) in the liver mitochondria. The results showed that, compared with the control group, the carnitine group exhibited increased RCCIV activity (P < 0.05), the training group exhibited increased RCCI, RCCIII and RCCIV activities, and the training+carnitine group showed increased RCCI-IV activities (P < 0.05 or P < 0.01). Moreover, RCCI and RCCIV activities in the training+carnitine group were higher than those in the carnitine and training groups (P < 0.05 or P < 0.01). Compared with the control group, the training+carnitine group showed increased SOD activity (P < 0.01), the carnitine, training and training+carnitine groups showed increased GSH-Px activities (P < 0.05, P < 0.01), and the training and training+carnitine groups showed decreased MDA contents (P < 0.05, P < 0.01). The training+carnitine group showed increased GSH-Px activity compared to that in the carnitine group (P < 0.01).The SOD activity in the training+carnitine group was higher than those in the carnitine and training groups (P < 0.05 or P < 0.01), and the MDA level in the the training+carnitine group was lower than those in the carnitine and training groups (P < 0.01). These results suggest that training and carnitine can increase the function of respiratory chain and antioxidant capacity in liver mitochondria, and the improving effects of training and carnitine can be synergistic.