”—— L1 8833'! "E. in: 5': 3 '22. .22 3 State 2.33.13 a; sit)! THESIS This is to certify that the dissertation entitled Carnitine Octanoyltransferase and carnitine Acetyltransferase of Mouse Liver Peroxisomes presented by Shawn O. Farrell has been accepted towards fulfillment of the requirements for PhoDo degreein BiOChemiStry / / iMSU is an Affirmatt'vr Action/Equal Opportunity Institution 0-12771 MSU LlBRARlES “ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. CARNITINE OCTANOYLTRANSFERASE AND CARNITINB ACETYLTRANSPERASE OF HOUSE LIVER PEROXISOMES by Shawn O. Farrell A DISSERTATION Submitted to flichigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry ABSTRACT CARNITINE OCTANOYLTRANSPERASE AND CARNITINE ACETYLTRANSFBRASE OF HOUSE LIVER PEROXISOMBS by Shawn O. Farrell The purpose of this study was to establish the existence of a separate carnitine octanoyltransferase in mouse liver peroxisomes. purify and characterize it. and compare it to the carnitine acetyltransferase purified from the same source. Carnitine octanoyltransferase (COT) and carnitine acetyltransferase (CAT) were solubilized from livers of mice treated with the hypolipidemic drug Wy-l4.643 by homogenization and freezing in 8.5% sucrose. 10 an sodium perphosphate. pH 7.5. COT and CAT were separated using Cibacron Blue Sepharose and purified to homogeneity. Both have a molecular weight of 60.000 by Sephadex G-ioo chromatography and SOS-polyacrylamide gel electrOphoresis. Both have similar pH Optima. 8.0 to 8.5. but the pIs are different. 5.2 for COT and 6.8 for CAT. COT and CAT have maximum activities in the forward direction with hexanoyl-CoA and butyryl-CoA. respectively: and in the reverse direction with hexanoylcarnitine and prepionylcarnitine. respectively. The Kms for acyl-CoA are low and suggest that formation of acylcarnitine is favored i vivo. With COT. using acyl-CoA with chain-lengths of 4-12. the Kms for acyl-CoA are between 2 and 4 pH. and are C -I and C higher for C2-. 16 lB—COA' With CAT. the acyl-CoA Kms are between 15—29 uh for C -CoA through C -CoA. For both 2 lo enzymes. the Km for L-carnitine varies with the acyl-CoA used. With CAT the Km for carnitine increases from 86 pH with C2-CoA to 519 pH with Clo-CoA. With COT the Km for L—carnitine is lower with long-chain acyl-CoAs as cosubstrate. COT retained its maximum activity when preincubated with DTNB at pH 7.0 or 8.5. In contrast. CAT was inactivated at both pH values but could be protected by the substrates. CAT was unaffected by preincubation with trypsin while COT was activated at low trypsin concentrations and inactivated at higher concentrations. Neither enzyme was inhibited by malonyl-CoA. Antibodies raised against the purified COT did not precipitate with purified CAT. purified beef heart mitochondrial CPT. or solubilized mouse liver mitochondria. The anti-COT serum did react with 10.000g supernatant fluids from homogenates of mouse kidney. mouse intestine. rat liver. dog liver. and beef liver. It is concluded that carnitine octanoyltransferase is a separate enzyme in mouse liver peroxisomes. with kinetic prOperties that favor formation of medium-chain acylcarnitines. Peroxisomal COT and CAT probably function in the transport of peroxisomal B-oxidation products out of the peroxisome. DEDICATION To my father. George Farrell. who instilled in me a great joy of learning and the will to persevere when the going gets tough. 11 ACKNOWLEDGEMENTS The support of the Department of Biochemistry. its faculty and students. is gratefully acknowledged. The time and interest of the members of my thesis guidance committee-—Drs. Richard Anderson. William Wells. Jack Holland. and Al Pearson-- in following this research are greatly appreciated. Special thanks go to Professor Loran Bieber. my Doctoral Thesis Advisor and abogado del diablo. who gave me the sopport and input necessary to develOp as a scientist and also the freedom necessary to develOp as an individual. I wish to thank past and present lab members Kim Valkner. Pat Sabourin. Sara Morrison-Rowe. and Janos Kerner for their help and support. I would also like to thank my favorite secretaries. Theresa Fillwock and Julie Doll. for making my stay more enjoyable: Dr. Pam Fraker. who got me through some tough sledding; and Betty Brazier. who made sure that I always got paid. On a more personal note. I would like to thank my parents. George and Aurel Farrell. who financially and genetically supported this venture: Rob Page. my best friend of years past. who kept me in school many times when I wanted to chuck it all to become a bike racer: y lo mas 111 importante. al P.R.C.. las originales y los adOptados. Carolita Josefina Fiol Lay Bigas de Buendia. Pensativa. Rebelde. Bobito Buendia. Marco. Eric. Aileen. Carmen. Mildred. Tanya. y Ouisqui Haravillosa Perrita Calcetines Descosidos Buendia: los amigos de mi presents. Quiero agradecerles de manera especial a las tres mas bellas mujeres del mundo-- Arlyn. quien siempre estaba cuando necesitaba un abrazo: Carol. mi mejor amiga y compafiera. quien me ayudo a mantener la sanidad en esta porqueria de Michigan: and my wife. Lynn. who put Up with me during the first years. kept me going during the middle years. and who really makes it all worthwhile. iv TABLE OF CONTENTS Page LISTOP TABLES .OOOOOOOOOCOOOOOCOO-OOOOOOOOOO ...... .0. Vii LIST OF FIGURES 0......OOOOOOOOOOOOOOOOO0.0.0.0000... Viii LISTOF ABBREVIATIONS 000......OOOOOOOOOOOOOOOOOOOOO. x INTRODUCTION 0.0...O...COOOOIOOOOCOOOOOOOOOOO00....0.0.01 Background on Carnitine and Carnitine Acyltransferases ................................l Localization and Functions ........................2 Existence of Carnitine Octanoyltransferase ........4 Peroxisomal 8-Oxidation ...........................5 THESIS STATEMENT .IOCCOOCCOC000.......OOOCOOCOCOOOOOOOOO7 EXPERIHENTAL pROCEDURES .0.90...0....0000.0.0.00000000008 Materials .........................................8 Methods ...........................................8 Induction of Carnitine Acyltransferases ......8 Sucrose Gradients ............................9 Purification of COT ..........................9 Purification of CAT ..........................ll Assays .......................................12 Preparation of Antibodies and Immunodiffusion 13 Trypsin Inactivation .........................13 DTNB Inactivation ............................l4 Isoelectric Focusing .........................14 Other Methods ................................14 Amino Acid Analysis ..........................14 RESULTS 0.000....0000..OOOOOOOOOOOOOOOOOOO ..... 0.0.0....15 Preliminary Investigations ........................lS Effect of Hypolipidemic Drugs on Total Liver Carnitine Acyltransferases ...........15 Subcellular Distribution of Acyltransferases .15 Effect of Hypolipidemic Drugs on Subcellular Distribution of Carnitine Acyltransferases .19 Purification of Carnitine Acyltransferases of Mouse Liver Peroxisomes .........................22 Solubilization and Purification of Peroxisomal Carnitine Octanoyltransferase ..22 Purification of Peroxisomal Carnitine Acetyltransferase ..........................29 Characterization of Carnitine Octanoyltransferase and Carnitine Acetyltransferase .................29 Determination of Molecular Weight ............29 Stability of COT and CAT .....................29 Specificity of COT and CAT for Various Substrates .................................34 Isoelectric Focusing of COT and CAT ..........38 PH Optima of COT and CAT .....................44 Effect of DTNB. Malonyl-CoA. and Divalent Cations ....................................44 Trypsin Inactivation .........................49 Inhibition by D-Carnitine ....................49 Amino Acid Analysis ..........................49 Immunology of Carnitine Octanoyltransferase .......58 DISCUSSION 00000000000000.00000000000000000000000000000070 LIST OF REFERENCES 0...00000000000000.000000000 ..... 000087 APPENDIX--COLLABORATIVE REVIEW OF CARNITINE ACYLTRANSFERASES eeeeeeeeeeeeoeeeeeeeeeeeeeeeegs LIST OF TABLES TABLE Page I. Effect of Clofibrate and NafenOpin on Mouse Liver Carnitine Acyltransferase Activities ........16 II. Specific Activities of Carnitine Acyltransferases in Peak Sucrose Gradient Fractions ................23 III. Summary of the Purification of Carniitine Octanoyl- transferase of Mouse Liver Peroxisomes ............25 Iv. Ammonium Sulfate Precipitation of COT from 10.0009 sup.rnatant FIUids 000000.0000000000000000000000.0.26 V. Km Values of Acyl-CoA Substrates for COT and CAT ..37 VI. K Values of Acylcarnitine Substrates for COT and CAT 0000000.0..00.000I0.0000.000000000000000.0039 v11. Inhibition of cow and car by zn2+ .................so VIII.Amino Acid Analyses of COT and CAT ................57 Ix. ImmunOprecipitation of COT with Rabbit Anti-COT s‘rum 0000000000000000000000000000000000000.000000059 vii LIST OF FIGURES FIGURE Page 1. Sucrose gradient separation of mouse liver organell‘s 00000000000000...000.000.000.0000000000.18 2. Sucrose gradient separation of organelles from livers of mice treated with clofibrate. nafenopin. and wy-14'643 00000000000.0000000000.0000000000000021 3. Separation of solubilized COT and CAT on Cibacron Blue Sepharose CL-6B ..............................28 4. SDS-polyacrylamide gel electrophoresis of purified carnitine octanoyltransferase and carnitine ac.ty1tran8f.r68‘ 0.00000000000000000.000000000000031 5. Temperature stability of COT and CAT ...... ..... ...33 6. Specificity of COT and CAT for acyl-CoAs and acylcarnitines of varying chain—lengths .. ......... 36 7. Isoelectric focusing of COT ....... ............... .41 8. Isoelectric focusing of CAT ..................... ..43 9. PH Optima for COT and CAT ... ...................... 46 10. Effect of DTNB on COT and CAT ....... . .......... ...48 11. Effect of trypsin on COT and CAT ................. .52 12. Effect of D-carnitine on COT and CAT ..............54 13. Replot of slope versus D—carnitine concentration for inhibition of COT and CAT .....................56 14. Immunoprecipitation of purified COT with rabbit anti-COT serum .0000000.000.000000000000000000000.061 lS. Immunoprecipitation of purified COT and mouse tissues with rabbit anti-COT serum ................64 16. ImmunOprecipitation of purified mouse liver COT 10.0009 sUpernatant fluids from rat liver .........66 17. Immunoprecipitation of purified mouse liver COT and 10.000g sUpernatant fluids from rat. beef. viii 18. 19. and dog liver .. ..... .................... ..... .....68 Acyl-CoA specificity of COT with saturating and subsaturating L-carnitine .....................79 PrOposed scheme of relation between COT. CAT. and perOXisomal B-DXidation ......OOOOOOCOOOOOOOIOOO0.085 ix BSA CAT CoA CoASH COT CPT DEAE- DTNB EDTA HEPES p1 OAE- SDS TCA Tris Wy—l4.643 LIST OF ABBREVIATIONS Bovine Serum Albumin Carnitine acetyltransferase Coenzyme A Reduced coenzyme A Carnitine octanoyltransferase Carnitine palmitoyltransferase Diethylaminoethyl- 5.5'—dithiobis-(2-nitrobenzoic acid) (Ethylenedinitrilol—tetraacetic acid N-Z-hydroxyethylpiperazine-N'-2-ethanesulfonic acid Inhibition constant Michaelis constant Reduced nicOtine adenine dinucleotide phosphate Isoelectric pH Diethyl-(2-hydroxypr0pyl)aminoethyl- Sodium dodecyl sulfate Trichloroacetic acid tris-(hydroxymethyl)aminomethane [4-chloro—6-(2.3—xylidino)-2-pyrimidinylthio]- acetic acid I NTRODUCTION Background on Carnitine and Carnitine Acyltransferases Carnitine (gamma-trimeth1yamino-B-hydroxybutyrate) was first isolated in 1905 (1) from mammalian muscle. but its function remained unknown for forty years. Fraenkel et a1. (2) discovered that carnitine was an essential nutrient for larvae of the beetle Tenebrio molitor. and it was subsequently given the trivial name vitamin BT' In 1955 Friedman and Fraenkel (3) discovered a possible enzymatic role for carnitine when it was found to be acetylated by acetyl-CoA in pigeon and sheep liver. Fritz (4) reported a carnitine dependent stimulation of long-chain fatty acid oxidation in liver preparations. but found little effect of carnitine on medium-chain fatty acid oxidation. Almost simultaneously the laboratories of Fritz and of Bremer provided evidence that carnitine plays a role in mitochondrial S-oxidation of long-chain fatty acids (5-7). Carnitine acyltransferases are a class of enzymes that catalyze the reversible reaction: L-(-)-carnitine + acyl-CoAHacyl-L-(-)-carnitine + CoASH. where the acyl grOUp has a carbon chain-length of 2 to more than 20 and can be straight chained. branched chained. or unsaturated. The reactions are defined as: Forward reaction: acyl-CoA + L-carnitine...—Aacyl—L-carnitine + CoASH. Reverse reaction: acyl-L-carnitine + CoASH.__)acyl-CoA + L-carnitine. The initial investigations dealt with enzymes utilizing predominantly long-chain acyl moieties and acetyl moieties. These enzymes were then given the names carnitine palmitoyltransferases (CPT)(6-8). and carnitine acetyltransferases (CAT)(3.9-ll). respectively. Although mammalian systems have a broad acyl specificity of carnitine acyltransferases. in some yeasts. plants. and insects the acyl specificity may be more restricted and depends on the fuel source and metabolism involved (12-18). Localization and Functions Studies on the intracellular distribution of CPT have shown that it is a mitochondrial enzyme (19-22). CPT is a membrane associated enzyme that requires the use of detergents for solubilization and stability (21.23.24). but its distribution in the inner mitochondrial membrane is still the subject of debate. It has been well established that CPT functions in the mitochondrial B-oxidation of fatty acids. The inner mitochondrial membrane is permeable to free fatty acids. but only short- and medium-chain fatty acids are activated to acyl-CoAs inside the mitochondria (25). as the long-chain acyl-CoA synthetases are associated with microsomes (26) and the outer mitochondrial membrane (27). CPT I (outer form) converts cytosolic acyl-CoAs to acylcarnitines which pass into the matrix of the mitochondria where they are reconverted to acyl-CoAs by CPT II (inner form) and then undergo s-oxidation (6.7.19.28-30)(see appendix for further review). CAT is the predominant acyltransferase in most tissues (31). Early studies indicated a mitochondrial location (32-36). but CAT has also been found in peroxisomes and microsomes from many tissues (20.37.38). The microsomal CAT is tightly membrane associated and labile (37). In contrast. the peroxisomal enzyme is a stable. soluble. matrix enzyme that is easily released by treatments that disrupt the fragile peroxisomal membrane (37.38). Although the functions of CAT are still being elucidated. several possibilities have been suggested. It is thought that mitochondrial CAT functions to buffer the CoASH/acetyl-CoA ratio by shuttling acetyl groups out of the mitochondria as acetylcarnitine (18.33.39-41). This would relieve the ”acetyl pressure” in the mitochondria that inhibits pyruvate dehydrogenase and the citric acid cycle. A probable function of peroxisomal CAT is the shuttling of acetyl moieties out of the peroxisome after peroxisomal a-oxidation (20.42-50). The hypolipidemic drugs that increase peroxisomal B-oxidation also increase CAT in both peroxisomes and mitochondria (46.51.52). It is also thought that CAT plays a role in the development of spermatozoa. where acetylcarnitine has been found in large quantities (53.54). The Existence of Carnitine Octanoyltransferase While working with a commercial preparation of CAT. Solberg found a contaminating activity specific for medium-chain (Cs-Clo) acyl-CoA. which was subsequently referred to as carnitine octanoyltransferase (COT)(SS). This activity was also found in mitochondria of rat liver. heart. and testis. In subsequent studies. COT activity was found in calf heart mitochondria (21) and in an acyltransferase preparation from calf liver (56). In the latter study. four acyltransferase fractions were partially purified that had chain-length optima of 3.6.12. and 16. Three ranges of Km were found for short—chain. medium-chain. and long-chain acyl residues. but complete separation of the activities could not be achieved. Early work in our lab indicated that COT activity exists in mitochondria. peroxisomes. and microsomes of mammalian liver and kidney (20). A microsomal preparation containing COT and CAT activity was purified free of CPT. and when treatment that solublilized the CAT activity completely destroyed the COT activity. a separate COT enzyme was postulated (38). Similarly. with a peroxisomal acyltransferase system free of CPT. purification of CAT with 0.4 M KCl and DEAE-cellulose caused the loss of all COT activity (37). In contrast. COT activity has been purified from beef heart mitochondria. where at least 90% of the activity was shown to be due to a combination of CAT and CPT (23). Clofibrate and other hypolipidemic drugs were found to increase the levels of carnitine acyltransferases (46.51.52.57.58). however all three activities increased to different extents. We interpreted these results as further evidence that a separate COT enzyme must be present. Peroxisomal s-Oxidation Peroxisomes have 6-oxidation capability (44-48). and it has been estimated that as much as 50% of the total s-oxidation activity of mouse liver may occur in peroxisomes (49). The peroxisomal s-oxidation process apparently terminates at medium-chain acyl-CoAs indicating that peroxisomes contain a chain-shortening capacity. This capacity seems particularly important in the metabolism of fatty acids with carbon chain-lengths greater than 20 (59-62). Following peroxisomal f-oxidation. the products. medium-chain acyl moieties and acetyl moieties. would need to be exported out of the peroxisome (50). Medium-chain and short-chain acylcarnitines formed via medium-chain and short-chain carnitine acyltransferases could serve this purpose (45.59-64). THES I S STATEMENT Solberg originally suggested the existence of a separate carnitine octanoyltransferase enzyme after he found an activity specific for medium-chain acyl—coenzyme As in a commerical CAT preparation (55). Other studies then demonstrated this activity in calf heart mitochondria and calf liver (21.56). Harkwell found that rat liver peroxisomes and microsomes also contained a medium-chain activity (37.38). However. in none of the above studies were these activities purified. Ironically. the only previous attempt to purify COT showed that beef heart mitochondria did not have a separate COT enzyme and that the observed activity was due to the combination of the broad substrate specificity of CPT and CAT (23). As enzyme purification and characterization is the only definitive way of demonstrating the existence of an enzyme. the purpose of this thesis was to isolate the carnitine octanoyltransferase enzyme in mouse liver peroxisomes. purify and characterize it. and compare it to its ”nearest neighbor”. peroxisomal carnitine acetyltransferase. EXPERIMENTAL PROCEDURES Materials Acyl-CoAs were purchased from PL Biochemicals. L-carnitine was a generous gift from the Otsuka Pharmaceutical Company. Naruto. Tokushima. Japan. Nafenopin (2-methyl-2-[p—(l.2.3.4.-tetrahydro-l-napthyl)-phenoxy1l-pro pionic acid) was a gift from N.E. Tolbert (Michigan State University. East Lansing. MI) and Wy—l4.643 ([4-chloro-6-(2.3-xylidino-)-2-pyrimidinylthiol-acetic acid) was a generous gift from J.K. Raddy (Northwestern University Medical School. Chicago. IL.). Ultrathin Serva Precoats pH 3-10. accessories and protein standards were purchased from Serva Feinbiochimica GMBH and Co. QAE-Sephadex was purchased from Pharmacia Fine Chemicals and Sigma Chemical Company. All other reagents were analytical grade. £22222 Induction of Carnitine Acyltransferases. Five to six week old male Swiss mice or male Sprague Dawley rats were fed diets of ground Purina Chow containing 0.5% w/w clofibrate (p—chlorOphenoxy-isobutyrate). 0.125% nafenOpin. or 0.1% Wy-l4.643 for two weeks. The animals were sacrificed and the livers rapidly removed and minced. washed. and homogenized in ice cold lOmM sodium pyrophosphate. pH 7.5. Homogenates were frozen at -80° C until used. Sucrose Gradients. Livers from animals that had been treated with hypolipidemic drugs as above. or from control mice that had been starved for 24 hours were removed and rinsed with ice cold 0.25 M sucrose. lmM sodium phosphate. pH 7.5. The livers were minced and suspended in 10 volumes w/v of the same buffer. and homogenized by one pass of a loose-fitting Teflon pestle with a glass Potter-Elvejhem homogenizer. After centrifugation at 4° C~for.20 minutes at 5009. 6 ml of the supernatant fluid were loaded onto the sucrose step gradients prepared as in (48). These were spun for 3 hours at 25.000 rpm in a Beckman SW 25.2 swinging bucket rotor and 60 drOp (2 ml) fractions were collected from the bottom of the gradient. Purification of COT. Liver homogenates from mice treated with Wy-l4.643 as above were thawed and centrifuged at 500g for 15 minutes and the supernatant fluid collected and centrifuged at 10.000g for 15 minutes. The supernatant fluid was made 40% in ammonium sulfate. centrifuged. and the pellet discarded. This supernatant was then made 60% in 10 ammonium sulfate. centrifuged. and the pellet suspended in lOmM sodium pyrophospahte. 0.25 mM EDTA. 0.02% sodium azide. pH 7.5 (blue buffer). The dissolved pellet was dialyzed overnight in 20 volumes of blue buffer. and applied at a flow rate of 1 ml/min to a column (40 x 2.5 cm) of Cibacron Blue Sepharose CL-GB equilibrated with blue buffer. After washing with 500 ml blue buffer. a 500 ml linear gradient of O-lM KCl in blue buffer was used to elute the enzyme. COT fractions (58-68. Fig. 3) were pooled. dialyzed overnight in 20 volumes of blue buffer. and applied at a flow rate of 1 ml/min to a column (35 x 1.5 cm) of GAS Sephadex A-25 (Pharmacia) equilibrated with blue buffer. COT activity washed through without binding. It was pooled. dialyzed overnight in 20 volumes 20 mM sodium phosphate. 0.25 mM BDTA. 0.02% sodium azide. pH 7.5 (HAP buffer). and applied at a flow rate of 1 ml/min to a column (20 x 2.5 cm) of hydroxylapatite equilibrated with 20 mM sodium phosphate. 0.02% sodium azide. pH 7.5. After 400 ml of MAP buffer were passed through. a 500 ml gradient of 20mM sodium phosphate. 60 mM KCl. pH 7.5 to 900 mM sodium phosphate. 60 mM KCl. pH 7.5 was applied. Fractions containing COT were pooled and dialyzed overnight in 20 volumes 20 mM sodium pyrophosphate. 0.02% sodium azide. pH 7.5 (Seph buffer). concentrated to 3 ml using an Amicon PM-lo filter. and applied at a flow rate of 1 ml/min to a column (95 x 2.5 cm) of Sephadex 6-100 equilibrated with Seph buffer. COT activity was eluted with 11 Seph buffer. the fractions pooled. and dialyzed overnight in 20 volumes 2 mM sodium pyrophosphate. 0.002% sodium azide. pH 7.5. and applied at a flow rate of 1 ml/min to a column (2.5 x 8 cm) of OAS-Sephadex A-25 (Sigma) equilibrated with the same buffer. The column was washed with 200 ml of the starting buffer before a 200 ml linear gradient of 2 mM-lOO mM sodium pyrophosphate. pH 7.5 was applied to elute the enzyme. Fractions containing COT were pooled. Purification of CAT. The first carnitine acyltransferase peak to elute from the Cibacron Blue Sepharose column above was used for further purification of CAT (see fig. 3). The fractions were pooled. dialyzed overnight in OAS buffer (25 mM sodium pyrophosphate. 0.25 mM EDTA. 0.02% sodium azide. pH 7.5). and applied at a flow rate of 1 ml/min to a column (35 x 1.5 cm) of OAS Sepahdex A-25 (Pharmacia) that had been equilibrated with QAE buffer. CAT washed through without binding. The effluent was pooled. dialyzed overnight in 20 volumes of CM buffer (5 mM MEPES. 60 mM KCl. 0.25 mM EDTA. 0.02% sodium azide. pH 7.3) and applied at a flow rate of 1 ml/min to a column (35 x 1.5 cm) of CM Sephadex equilibrated in CM buffer. The column was washed with 300 ml CM buffer and the enzyme eluted with a 400 m1 linear gradient of 60-560 mM KCl in CM buffer. Fractions containing CAT were pooled and dialyzed overnight in 20 volumes Seph buffer. concentrated to 3 m1 using an Amicon PM-lo filter. and applied at a flow rate of 1 ml/min 12 to a column (95 x 2.5 cm) of Sephadex G-100 equilibrated with Seph buffer. CAT activity was eluted with Seph buffer and the fractions pooled. Assays. Assays were performed as previously described for catalase (65). fumarase (66). NADPH cytochrome c reductase (67). and glutamate dehydrogenase (68). Carnitine acyltransferases were measured in the forward direction by the DTNB method of (69). The 0.2 ml cuvette volume contained 0.1 mM Acyl-CoA. 1.25 mM L-carnitine. 0.1 mM DTNB. and 115 mM Tris buffer. pH 8.0. All assays were corrected for the carnitine independent release of CoASH (hydrolase). The reverse reaction was assayed as described in (70). Acylcarnitine. 500 uM. and CoASM. 120 uM. were used. For the Km profiles with various substrates. COT and CAT were assayed in the forward direction by continuously monitoring the release of Coenzyme A with DTNB at 412 nm. The 2.0 ml reaction volume contained 115 mM Tris buffer. 1.1 mM EDTA. 0.1% Triton x-ioo. 0.1 mM DTNB with varying amounts of L-carnitine and acyl-CoAs at 25° C. pH 8.0. Kinetic data were obtained with a semiautomated system described in (71.72). Raw absorbance-time data were obtained by continuously increasing the substrate concentration of a stirred enzyme assay mixture. with the use of a precision syringe drive during a reaction time of 3.6 minutes. and collected with a Gilford model 2600 spectrOphotometer. The 13 raw data were transformed into velocity-substrate data by a tangent slope procedure and then analyzed as linear plots using the TANKIN program with a Hewlett—Packard 9815 calculator. When Clz-CoA was used with the semiautomated system and the TANKIN program. the Km could not be defined due to strong inhibition by C -moieties. so a substrate 12 depletion method was used (73). Preparation of Antibodies ang_;mmunodiffusion. Rabbit antiserum was raised against purified mouse liver COT by Dr. J.K. Reddy. After collecting pre-immune serum. two New Zealand white male rabbits weighting 2-3 kg were immunized with purified COT. Approximately 500-600 pg of COT were emulsified in Freund's adjuvant (Difco. Detroit. MI) and injected subcutaneously at multiple sites at weekly intervals for 4 weeks (74). One week after the last injection a booster dose of COT was administered intravenously and the rabbits bled 5 days later. Double diffusion plates were prepared using 1% agarose in 0.1 M sodium phosphate. 0.02% sodium azide. 0.15% NaCl. pH 7.4. and develOped overnight at room temperature. Trypsin Inactivation. Inactivation by trypsin was accomplished by incubating the enzymes with varying amounts of trypsin (0-2 mg/ug enzyme) for 15 min at 37° C in 10 mM sodium pyrophosphate. pH 7.5. The reaction was stOpped by 14 addition of a 3-fold excess of trypsin inhibitor. DTNB Inactivation. COT and CAT were preincubated for 15 min at room temperature with 0-1 mM DTNB in 10 mM sodium phosphate buffer pH 7.0 or 8.5. The transferase reaction was then started by addition of a premix containing all of the substrates as described previously (69). Isoelectric Focusing. Concentrated COT and CAT were focused for 3 hours on ultrathin gels pH 3-10 at 1 1/2 watts per gel on an LKB Multiphor 2117. Bands were fixed with 20% TCA and then stained with 0.1% Coomassie Brilliant Blue. Other Methods. SDS-polyacrylamide gel electrophoresis was performed as described (75) employing bovine serum albumin. phosphorylase b. ovalbumin. and trypsinogen as molecular weight standards. Native molecular weight was estimated with chromatography on Sepagdex G-100 as in (76). Protein was determined by the fluorescamine method (77) with the exception that 0.2 M borate. pH 9.25 was used. Bovine serum albumin was used as the protein standard. Amino Acid Analysis. Purified. lyophilized COT and CAT from mouse liver peroxisomes were subjected to acid hydrolysis for 24 hours and the amino acids measured using a Beckman model 121 by Doris Bauer. Michigan State University Biochemistry Dept. RESULTS Preliminary Investigations Effect of Mypglipidemic Drugs on Total Liver Carnitine Acyltransferase Activity. In order to determine if hypolipidemic drugs could be used to increase the level of carnitine octanoyltransferase in mouse liver. mice were fed control diets. or diets containing clofibrate or nafenopin. Table I shows the increase in the specific activities of carnitine acyltransferases in liver 500g sUpernatant fluids due to these drugs. All three activities increased with treatment of hypolipidemic drugs. COT. which is present at higher levels than CAT and CPT. increased 3.8- and 11.1-fold with clofibrate and nafenOpin. respectively. CAT increased 3.1- and 9.6-fold. while CPT increased to a lesser extent. Subcellular Distribution of Acyltransferases. Previous studies have shown that rat liver has a multiorganelle distribution of short-chain and medium-chain carnitine acyltransferases (20.42.52). To determine the location or locations of carnitine octanoyltransferase in mouse liver. sucrose gradients were prepared to separate the mouse liver organelles. Figure 1 shows the distribution of carnitine 15 16 .Amnv eccefiusaeou eanwuase saw use» 9 e.uueccso an DeCAEueuev so no.0 v m uo Ae>ea 0 us unsus> Acuucou EOuu uceueuuuv xaucsu«uficmfiu eue3 eue>e~ Deueeuu an» ass» ueusomvca .eeHnEse xue uOu .zmm. asses sue ce>am sesas> s .weusonOLm deuceeuueaxm ca Donahueev as onus Deuces» was «Ouucou um eue>a~ eesoe EOuu echo-u unsuscueasu woom cu Dec-Ensuem one: useeueuecsuua>us sauuacusu s H.v A.N «AA.AV O.NA Ao.av H.o An.ov 0.N FLU 0.6 A.n sav.nv h.hN saa.ov 9.0 Ao.ov 0.N 840 H.HH o.n s-o.nv N.No sam.nv N.QN Ab.ov v.5 FOO Caucasusz eusunwquU auuoceusz eusunquAU Acuucou eeseuuc« Odom: waflmmmmlmewcfiE\aoecvm>u«>«uut Meeuufi>uuo< ensueuecsuua 04 asauucusu ue>uq eeso: :0 cu oceusz was unsunfiquU no uueuuu H HAG‘B 17 Figure l. Sucrose gradient separation of mouse liver organelles. (A) Activity of marker enzymes in mmol/min/ml and protein in mg/ml. (B) Activity of carnitine acyltransferases in nmol/min/ml. (C) Specific Activity of carnitine acyltransferases in nmol/min/mg. Sucrose gradients were prepared and enzymes were assayed as described in Experimental Procedures. (mmol/min/ml 1 (nmol/min/ml 1 (moi/whim) H Catalose e—e Fumarase x I e—e NADPH cytochro c reductase x ~§~ IO ' ' I5 ' fraction number l W is o--o protein cone. (mg/ml) 19 acyltransferase activities from liver homogenates of control mice. The catalase. fumarase. and NADPH cytochrome c reductase distributions shown in Figure 1A represent the distribution of peroxisomes. mitochondria. and microsomes. respectively. A large peak of COT activity coincides with the distribution of catalase. as can be seen in Figure 1B. The COT specific activity in peroxisomes is 10-fold greater than the other two carnitine acyltransferases. Three different control gradients were assayed and the fractions of the carnitine acyltransferases associated with the peroxisomal. mitochondrial. and soluble fractions of the gradient were determined. Seventy percent of the particulate COT was associated with peroxisomal fraction. and 30% with the mitochondria and microsomes (data not shown). The amount of transferase in the soluble fraction varied with the grinding technique. but the fraction of COT in the soluble fraction coincided with that of catalase. Effect of Hypglipidemic Drugs on the Subcellular Distribution of Carnitine Acyltransferases. In order to determine in which organelles the enzymes are affected by the hypolipidemic drugs. organelles from livers of mice treated with hypolipidemic drugs were separated on sucrose gradients. Figure 2 shows the distribution of carnitine acyltransferases from mice fed clofibrate. nafenopin. or Wy-l4.643. COT and CAT increased in the peroxisomes and 20 .eeusveUOum deuceeuuenxn ca Denuuuaev we measure euez seahwce was Teuoneun one: eucewvsum eeouusm .~E\cwE\AOEc cu wensueuucsuua>us ecuuucuou no aeauu>uuu< “unease E09u0m .AE\mE aw cueu0un use HE\C«E\HOEE cg eeeauce nexuse no sewn->«uot ”unease n08 .nvo.val>3 Dcs .caaoceusc.eusunaquo cu“: Teaseuu sows uo eue>w~ eouu sen-ecsmuo no cowususnee uceuvsum secuosm .N ousmwm "mud CW (nu/bun) #- -- - _°-—_——————— n¢w.v. u>~> .363... 5:8: c0. 31 088.69 a 95.50;“. 1842 I 5.2a 0:6 no. x 308.5... I 3928 I 20520.0 8 § 9 Q N l (luv “11W 101““) “WWW 10W) 22 mitochondria with all 3 hypolipidemic drugs. The largest increase. however. was mitochondrial CAT. For all of the drug treatments the peroxisomal breakage was greater. and this was reflected by the catalase and COT activities found in the soluble fractions. Table II gives the specific activities of the carnitine acyltransferases in the peak peroxisomal and mitochondrial fractions. Clofibrate. nafenopin. and Wy—l4.643 increased peroxisomal COT 1.1-. 2.9-. and 3.5-fold and mitochondrial COT 4.7-. 8.0-. and ll-fold. respectively. Peroxisomal CAT increased 6-. 12-. and ll-fold. while mitochondrial CAT increased 19-. 54-. and 44-fold. respectively. Since carnitine octanoyltransferase activity was found in higher levels in mouse liver than could be accounted for by the combination of CAT and CPT. and since hypolipidemic drugs were shown to raise the levels of the 3 carnitine acyltransferases to different extents in different organelles; it was decided that COT must be a separate enzyme. and its purification was undertaken. Purification of Carnitine Acyltransferases of Mouse Liver Peroxisomes Solubilization and Purification of Peroxisomal Carnitine Octanoyltransferase. Mice were fed a diet of 0.1% Wy-l4.643 for 2 weeks to increase the absolute level of liver 23 .ueusveUOLm Asuceewuenxn cu Denauuuefl as veasems eue3 sensueuecsuua>us enauwcusu .>Ae>wuuemeeu .eesus59u was eusasuso 5n vecAEueueD we .~ vcs A eehsmwu ca c3Oce euceuvsum secuuse ecu uo acoauusuu .zo Asuuflc0:009ue was as. AsEOeAxouea Jean ecu cg eeusueuecsuua>us sawuacusu 0:» van DeCAEueueD sue: eeuuu>auus uauuuenm om mm on on 0v 0 0.5 h.v FLU end ova OON 00H Nb on b." m.nH 840 No 000 mm mmv 0v mod m.o and 800 2 h m 2 m 2 m t In nvo.v~IN3 cumoceusz eusunuuouu acuucou AcweuOHQ mEVcaE\AOEchuu>uuu< ICOMUUQUE UCCMVGHU .IOth-m XQCG ca seesueuecsuuaxo< ecauqcusu uo eeuuw>uuu< uuuuuenm HH ”4&48 24 carnitine acyltransferases. Preliminary studies indicated that peroxisomal but not mitochondrial carnitine acyltransferases are solubilized by the combination of homogenization and freezing in 8.5% sucrose. 10 mM sodium pyrophosphate. Also. previous investigations had indicated that rat liver microsomal COT and CAT were tightly bound and could not be easily released (37.38). Freeze-thawing in the sucrose. sodium pyrophosphate buffer was therefore employed to solubilize peroxisomal COT and CAT. which were then separated from mitochondria by centrifugation at 15.0009 (Table III). Eighty percent of the COT was recovered after this step. but only 30% of the CAT and CPT was recovered. Preliminary results showed that most of the COT in 10.0009 supernatant fluids from mouse liver homogenates could be precipitated by ammonium sulfate at concentrations of 40-60% (Table IV). This fractionation range was used to precipitate COT from the 15.0009 supernatant fluid above. Cibacron Blue Sepharose was then used to separate the solubilized COT and CAT. As can be seen in Figure 3. one peak (CAT) contained carnitine acyltransferase with high activity for octanoyl-CoA but a higher activity for acetyl-CoA. while the second peak (COT) contained carnitine acyltransferase with a high activity for octanoyl-CoA and a much lower activity for acetyl-CoA. This second peak was then purified to homogeneity with the series of chromatographic steps outlined in Table III. 25 .euocfle a :a sawuacuequue Ou (GUI-mus nose A uue>coo Ou >ueuueoec sexuce no unease ecu «- >uu>wuue uo vac: eco .eeusveUOHm deuceeuuenxm ca Denmuueev es Deuuausn e03 eeeueueceuuaaoceuuo ecauucueu ms oom.p o.o ne-.H onm mmo.~n om onn.-n Ha xenosnomumOUOK 8&0 >h0>000¢ 8‘0 %U«>muu< UwquOflm Nh0>000¢ POU OOEOQMXOMCQ h.>wd 0.30: EOHN OIQHCUOCQHUH%OCGUUO OCMUNCHUU H0 C0uUOUuUwth 0:9 U0 NHOEEDM HHH fldfl cuu3 cue ma pan 0 ohm us fieuecoucA eue: 8(0 use POO .840 use 800 :0 caenauu uo uueuuu .AH epomum OOON . _ CON ._.<0 .5 ...00 9358?... a: co. on ...40 e ...00 e D N O in 96 I) [C 0N. o 9 av IOJlUOO Km .oeusveUOum deuceewueaxn Cu oecAuUeeu memes DueIuOu ecu cuuz ecmuucueUIo :E m.~ no .EE mN.~ .zE no.o .o no euceeeua ecu an ecoaueuuceucou ecuuucueUIq mc«>ue> cuuz Deacons one: 840 was 900 3 5 ado Am. 900 54v .840 use POO c0 ecwuucuoUID no uueuum .Na eusmum name. x Hangs—tool: \_ e n 5.. l m \\n .\ 0‘6 OO’V/ O .r/ . e .999 . 04/ ‘7 u 90 e are e e [1.4. e/iw/ . .1. l... a? e Own» .. ., ...... .1. ... .... .. u we. . .W m o. (IOWU/“imiA/l 55 Figure 13. Replot of lepe versus inhibitor concentration for D-carnitine inhibition of COT and CAT (see figure 12) (A) COT (B) CAT K; -.84 mM d-carnitine o l 2 3 (D-carnitine) mM K; - l mM d-cornitine l O I 2 3 (D- carnitine) mM 57 TABLE VIII Amino Acid Compositions of COT and CAT Amino Acid COT CAT :2; """"""""""""" §T§;""""""""ISTE3 """" Thr 4.88 4.55 Ser 6.45 6.83 Glx 15.16 14.03 Pro 5.16 5.34 Gly 7.39 5.33 Ala 7.00 7.97 Val 4.37 3.98 Met 2.54 2.58 Ile 3.86 4.70 Leu 10.68 9.98 Tyr 3.50 2.57 Phe 4.87 4.10 His 3.79 3.98 Lys 6.28 7.87 Ar9 5.66 4.80 Values for each amino acid are mole percent of detected amino acids. Cysteine and Tryptophan were not determined. Glx and Asx are the combined glutamine/glutamate and asparagine/aspartate values. respectively. 58 largest differences seen with lysine. glycine. glutamate and glutamine. and aspartate and aSparagine: which all varied by more than 1 mole percent of the determined amino acids. Immunology of Carnitine Octanoyltransferase Antibodies were prepared against purified mouse liver COT. and their specificity for COT was investigated. Table IX shows the selective immunOprecipitation of COT from 10.0009 supernatant fluids of mouse liver homogenates. The original supernatant was high in COT and CAT. but the antibody-antigen complex was low in CAT activity. Figure 14 shows a characteristic immunOprecipitation reaction between rabbit antiserum raised against pure COT and various acyltransferase preparations. Anti-COT serum reacted only with mouse liver COT. There was no reaction with purified CAT from mouse liver nor with mitochondria that had been sonicated. treated with Triton x-100. or both: and no reaction was obtained with untreated mitochondria from mouse liver. Triton x-lOO (0.1%) did not interfere with immunoprecipitation of COT by anti-COT serum. No reaction was seen between anti-COT serum and beef heart mitochondrial CPT purified as in (23) (data not shown). In order to determine if other mouse tissues have proteins immunologically similar to COT. the anti-COT antiserum was diffused towards homogenates of various mouse 59 TABLE IX Immunoprecipitation of COT with Rabbit Anti-COT Serum Fraction CAT/COT ratio 10.0009 supernatant 16.4 Immunoprecipitate 4.4 Rabbit anti-COT serum and 10.0009 supernatant fluids were mixed to give maximum precipitation. incubated at 37 C. for 15 min.. and stored overnight at 4 C. The precipitate was spun down. washed twice with agar buffer. and redissolved in 5mM sodium perphosphate. pH 7.5. Activity for octanoyl-CoA and acetyl-CoA were performed as described in Experimental Procedures. 60 Figure 14. ImmunOprecipitation of purified COT with rabbit anti-COT serum. Center well: 20 p1 Wells 1.3.5: 10 pg Well 2: 10 pg Well 4: 20 pl Tissues were taken the agarose plates Procedures. Well 6 rabbit antiserum. pure mouse liver COT mouse liver CAT. solubilized mitochondria from mouse liver. from a mouse treated with Wy-l4.643 and developed as described in Experimental was unused. 62 tissues. Figure 15 shows the reaction between anti-COT serum and 10.0009 supernatant fluids from kidney. heart. intestine. and skeletal muscle. Kidney and intestine gave precipitin bands of identity with the purified COT. No reaction was detected with heart or muscle. Livers of other species were then used to see if they contained proteins immunologically similar to mouse liver COT. Figure 16 shows a reaction of partial identity with mouse liver COT and 10.0009 supernatant fluids from livers of rats that had been treated with Wy-l4.643. The direction of the precipitin spurs indicates that the rat liver preparation is the weaker cross reactant. as expected. Figure 17A compares the reaction between anti-COT serum and mouse liver COT to that with 10.0009 supernatant fluids from beef and dog liver. Again the non-mouse tissues gave reactions of partial identity with the pure mouse liver COT. Figure 17B shows the relationship between the preparations from non-mouse livers. The beef and dog liver give reactions of identity with each other. and both give reactions of partial identity with the rat liver supernatant. Although the rat liver titer was much lower than the other two. it also appears to be immunologically more similar to mouse liver COT because the preparations from dog liver and beef liver are weaker cross reactants than the rat liver preparation. For all of the tissues and species tested. mitochondrial pellets were collected. 63 Figure 15. Immunoprecipitation of purified COT and mouse tissues with rabbit anti-COT serum. Center well: 20 pl anti-COT serum. Wells 1.4: 10 pg pure mouse liver COT. Wells 2.3.5.6: 10.0009 supernatant fluids from mouse kidney. intestine. heart. and skeletal muscle. respectively. Tissues were taken from a mouse treated with Wy-14.643 and the agarose plates developed as described in Experimental Procedures. 65 Figure 16. Immunoprecipitation of purified mouse liver COT and 10.0009 supernatant fluids from rat Liver. Center well: 20 pl anti-COT serum. Wells 2.4.6: 10 p9 mouse liver COT. Well 1: 20 p1 10.0009 supernatant fluid from rat liver. Rat liver was taken from a rat treated with Wy-l4.643 and the agarose plates developed as described in Experimental Procedures. 67 Figure 17. Immunoprecipitation of purified mouse liver COT and 10.0009 supernatant fluids from rat. beef. and dog liver. (A) Center well: 20 p1 anti-COT serum. Wells 1.3.5: 10 pg mouse liver COT. Wells 2.4.6: 20 pl 10.0009 supernatant fluid from rat. beef. and dog liver. respectively. (B) Center well: 20 pl anti-COT serum. Wells 1.4: 20 p1 beef liver 10.0009 supernatant fluid. Well 2: 20 pl dog liver 10.0009 supernatant fluid. Well 3: 20 pl rat liver 10.0009 supernatant fluid. Well 5: 10 pg mouse liver COT. Rat liver was taken from a rat treated with Wy-l4.643 and the agarose plates developed as described in Experimental Procedures. 69 Samples of each were solublized by sonication. addition of Triton X-100. or both. None of the whole mitochondria or solublized mitochondria reacted with anti-COT serum. DISCUSSION pocaiization and Purification of COT and CAT Previous studies demonstrated that carnitine octanoyltransferase activity is widely distributed in rat tissues. with the exception of the brain (31). and other investigations with liver indicated a separate COT enzyme in peroxisomes and microsomes (20.37.38). Extensive investigation of beef heart mitochondria. however. attributed this activity to a combination of long-chain and short-chain carnitine acyltransferase instead of a true medium-chain enzyme (23.24). This work shows that with density gradient separation of mouse liver organelles. most of the COT parallels the distribution of catalase. with 60% in the peroxisomes and 10% in the soluble fractions. The 20-fold greater specific activity of COT in peroxisomes compared to mitochondria in mouse liver is in contrast to the distribution in rat liver where the specific activities of COT and CAT are about equal in peroxisomes (52) and about 60% of the COT activity is associated with mitochondria with a specific activity about 2.5-fold greater than peroxisomal COT. Clofibrate and nafenOpin. two drugs which induce 70 71 peroxisome proliferation (86.87). increase B-oxidation (44.46.88.89). and increase carnitine acyltransferases (52.57.58.89.90) increased the specific activity of COT in mouse liver 5009 supernatant fluids 4- and ll-fold. respectively. When mice were fed clofibrate. nafenOpin. or Wy-l4.643. and the liver organelles separated on sucrose gradients. COT levels were increased in both the peroxisomes and mitochondria with all treatments. The relationship between the increase in carnitine acyltransferases and the lipid-lowering effects of the drugs is not known. Sucrose gradients from mice fed hypolipidemic drugs contained much more catalase. COT and CAT in the soluble fraction than the controls. If all of the soluble catalase came from peroxisomes as has been suggested (91). then 50—60% of the peroxisomes were broken from drug treated animals. This is in contrast to the controls where a maximum of 20% were broken. Thus. it appears that hypolipidemic drugs enhance peroxisomal membrane fragility. This fragility aided in the separation of the peroxisomal enzymes COT and CAT from the mitochondrial enzymes. Momogenization and freezing released 80% of the COT present in the crude homogenate (Table III). whereas only 30% of the initial CPT and CAT remained in the supernatant fraction after the mitochondria were removed by centrifugation. This remaining activity with palmitoyl-CoA was due to the chain-length specificity of COT rather than 72 the CPT enzyme on the inner membrane of the mitochondria. as harsher methods are required to liberate and stabilize CPT (21.23.24.92-94). Mitochondria do not appear to rupture during the isolation procedure. as the mitochondrial matrix marker. glutamate dehydrogenase. is not detected in the supernatant after centrifugation at 15.0009. The CAT activity remaining is therefore also due to peroxisomal CAT. and the low percentage recovery after centrifugation is a reflection of the high levels of CAT found in the mitochondria of drug treated mice. The solubilized peroxisomal COT and CAT were easily separated with the Cibacron Blue Sepharose column. and then both purified to apparent homogeneity by the column chromatographic steps described. The final purified COT preparation contained 20% of the activity initially present in the crude homogenate. This represents a greater than 20% recovery of the peroxisomal enzyme because the mitochondria have COT activity. The final purification of 530-fold. although low for a liver enzyme. is reasonable considering the lO-fold increase in absolute COT levels in the livers of mice treated with WY-l‘: 643s 73 Physical Characterization 9; COT and CAT Mouse liver peroxisomal COT and CAT both have molecular weights of 60.000. No aggregation is apparent as Sephadex G-100 and SDS-polyacrylamide gel electrOphoresis give the same molecular weight. Recently. carnitine octanoyltransferase was purified from rat liver peroxisomes. and its molecular weight was found to be 66.000 (95). The properties of several carnitine acetyltransferases have been reported. The molecular weight of CAT is 58.000 from pigeon breast muscle (96). 67-69.000 from rat liver mitochondria (97). 62.600 from beef heart mitochondria (23). and 59.000 from rat liver peroxisomes and microsomes (37). Thus. most of the COT and CAT enzymes studied to date are of a similar size. An exception is alkane grown yeast. where both mitochondria and peroxisomes contain a CAT enzyme that has a molecular weight of 420.000 by gel filtration and ultracentrifugation (98). SDS-gel electrophoresis gives subunit weights of 64.000 and 57.000 for the peroxisomal CAT and 64.000 and 52.000 for the mitochondrial CAT. however (15.98). Peroxisomal COT and CAT have pIs of 5.2 and 6.8. respectively. Other studies have reported pIs for carnitine acyltransferases near 8.3. such as CAT from beef heart mitochondria (23).and from rat liver microsomes and 74 peroxisomes (37). In (37) a second peak at 5.3 was observed for the microsomal enzyme. and with alkane grown yeast the pIs were 5.11 for peroxisomal CAT and 5.22 for mitochondrial CAT. Proteolytic degradation of carnitine acyltransferases from liver can occur during purification. Rat liver CAT was reported to consist of two non-identical subunits of 34.000 and 25.000 daltons (99). This observation was confirmed in (97). but further analysis showed that CAT is a single polypeptide that can be degraded during purification. When we used mouse liver homogenates stored frozen in the absence of protease inhibitors. COT had 3 major isoelectric peaks at pI 5.9.6.0. and 6.1 (data not shown). It also eluted from Sephadex G-100 as a wide peak spanning a range of 3000 daltons. When fresh liver preparations were used in the presence of 1 mM EDTA. 0.1 mM PMSF. and 5 mg/l pepstatin: COT eluted as a much narrower band from Sephadex G-100 and only 1 band appeared after isoelectric focusing. Therefore. some of the differences between molecular weights and pIs reported in different studies could be due in part to proteolytic degradations during purification. but they could also be due to organelle and species differences. and differences between membrane associated and matrix enzymes. Although apparent proteolytic degradation led to COT eluting as a wide band on Sephadex G-100. fractions taken from the extremes of the peak had the same substrate specificity for acyl-CoAs and similar Km' for octanoyl-CoA and L-carnitine 75 (data not shown). Using a semiautomated kinetic analyzer we determined the Kms for the even chain-length acyl-CoA substrates and the corresponding Kms for L-carnitine. For both COT and CAT the Kms for L-carnitine varied with the acyl-CoA cosubstrate used. With CAT the Km for L-carnitine increased with increasing acyl-CoA carbon chain-length. This is in contrast to the results with pigeon breast-muscle CAT (100) where no such pattern existed. With COT the Kms for L-carnitine also varied with the acyl-CoA used but were lower for long-chain acyl-CoAs. as has been reported for rat liver COT (95). With COT. acyl-CoAs from C to C had very 4 12 2 ls-CoA. and Cla-CoA were large. This is different from the pattern reported for rat low Kms. but the Kms for C -CoA. C liver COT (95) where the Kms decreased with increasing acyl-CoA carbon chain-length.‘ Mouse liver CAT had maximum activities with butyryl-CoA and propionylcarnitine. All of the mammalian carnitine acetyltransferases characterized to date have specificities that are similar for the forward and reverse reactions. Maximum activities occur with C3 or C‘ and the activity drops as the carbon chain-length increases until almost no activity exists at C or C (23.37.95). 8 10 Exceptions are with non-mammalian sources. such as alkane grown yeast (15) and T.bovina (17). where virtually no carnitine acyltransferase activity exists with acyl-CoAs of carbon chain-lengths greater than 3. 76 COT had a maximum activity with hexanoyl-CoA and hexanoylcarnitine similar to the rat liver enzyme (95). We found a biphasic substrate specificity curve. with a local minimum for C12 moieties. When this pattern occurred during the reverse reaction. it was noticed that the abnormality occurred at the substrate concentration where acylcarnitines should form micelles. as had been shown in (24). However this transition to micelles should not be present in the mixed micelle environment use for the forward assay containing 0.1% Triton x-lOO. yet the biphasic specificity pattern existed in the presence and absence of detergents. Also. when Clz-CoA was used with the semiautomated system and the TANKIN program the Km could not be defined due to strong inhibition by C12 moieties. In fact. a substrate depletion method (73) was used to determine the Km for Clz-CoA in order to avoid high concentrations of the substrate. The Kms for the acyl-CoA substrates of COT and CAT are all very low compared to the Kms for the corresponding acylcarnitines. which are all over 100 pM. This indicates that acylcarnitine formation would be favored i vivo. The combination of the Km profiles and the specificity for substrates of varying chain-length under saturating conditions indicates that in the forward direction both the KIn for acyl-CoA and the V affect the acyl-CoA max specificity. With both enzymes there are large differences 77 in velocities in the range where the Km for acyl-CoA does not vary with the carbon chain-length. It is also apparent that the concentration of L-carnitine could be important for the i9 yiyg substrate specificity. With CAT. for example. at concentrations of L-carnitine below 50 pM. acetyl-CoA would become a more favored substrate. This effect can be seen with COT. When subsaturating levels of L-carnitine are used. the acyl-CoA substrate specificity changes so that the relative specificity for long-chain acyl-CoAs increases (see Figure 18). Inhibitors and Inactivators It has been suggested that rat liver mitochondria contain a separate carnitine octanoyltransferase because of differences in enzymatic activities with C8-CoA or Cls-CoA in the presence of malonyl-CoA. divalent cations. and DTNB (80.82-85). With purified peroxisomal COT from mouse liver. no effects of malonyl-CoA are seen. Malonyl-CoA was not an inhibitor at concentrations Up to 100 pM. which agrees with data reported for the rat liver enzyme (95). This is in contrast to some reported effects of malonyl-CoA on ”mitochondrial COT” activity (80.82.83.85) where malonyl-CoA inhibited COT activity and CPT activity to different extents. However. these studies used whole mitochondria that contain at least two carnitine acyltransferases with 78 Figure 18. Acyl-CoA specificity for COT at saturating and unsaturating levels of L-carnitine. Carnitine octanoyltransferase was measured in the forward direction with 100 pM acyl-CoA and L-carnitine at 1.25mM or 50 pM. % Activity 0 I25 mM L-carnitine '00 ' A 50 gm L-carnitins 50 " l0 * I I I I I I 2 4 6 8 l0l2|4l6 Acyl-CoA carbon chain-length 80 overlapping specificities for acyl-CoAs of different chain-lengths. along with the possibility of contamination by peroxisomal COT and CAT. It is therefore not suprising that such differences exist between activities with one acyl-CoA versus another. but little information can be attained regarding the actual enzymes involved. Rat liver mitochondrial CAT shows latent and overt activity. and the overt activity can be inhibited by malonyl-CoA (101). We did not find any malonyl-CoA 2+ or M92+ for mouse liver CAT. but both COT and CAT were inhibited by 2n2+ at inhibition. nor inhibition by Ca concentrations greater than 50 pM in the reverse direction. In (102) trypsin was shown to inactivate COT and CAT from rat liver microsomes. although the membrane-bound CAT was activated at low trypsin levels. We found that COT from mouse liver peroxisomes is activated by incubation with 3-5 ug trypsin/ug COT. but is inactivated at higher concentrations. CAT was not affected by trypsin under our conditions. Some sulfhydryl reagents. including DTNB (81). iodoacetimide (99). and p-chloromercuribenzoic acid (15). can inactivate CAT. Mouse liver CAT is inactivated by preincubation with DTNB in the absence of the substrates. but COT was not affected. In the presence of the substrates. no inactivation occurred with either enzyme. Thus DTNB does not affect the initial rates under the 81 conditions of our standard forward assay. D-carnitine is a competitive inhibitor of both COT and CAT with respect to L-carnitine. Our data agree with the original investigations of Fritz and coworkers (103). but are in contrast to the data in (104) where parallel Lineweaver-Burk plots were obtained. The Kis were high but might be physiologically important. D-carnitine is virtually absent in biological systems except where it has been introduced. Recently DL-carnitine has become available as a dietary supplement. raising questions about the metabolic consequences of oral ingestion of the D-isomer (105-110). These data show that D-carnitine is inhibitory to both short-chain and medium-chain carnitine acyltransferases. and thus might have adverse effects on systems which involve medium-chain acyl-CoAs and acylcarnitines. Immunology and COT pocalization Antibodies raised against purified COT from mouse liver peroxisomes do not react with peroxisomal CAT nor with various mitochondrial fractions from mouse liver. In (111) we discussed the possibliity that mitochondria might have a separate COT enzyme. This work. as well as those of (95.97). make it doubtful that a separate COT exists in mouse or rat liver mitochondria. The effects of DTNB. 82 malonyl-CoA. and divalent cations reported in (80.82-85) are probably not due to a separate enzyme. rather the overlapping specificities of mitochondrial carnitine palmitoyltransferase and acetyltransferase. and possible contamination of the mitochondrial preparations with carnitine acyltransferases from non-mitochondrial sources. All of the livers tested have proteins that share antigenic determinants with COT from mouse liver. Rat liver has a protein that is more similar to mouse liver COT than the protein from dog or beef liver. However. no mitochondrial fractions reacted with our anti-COT serum. If there is a mitochondrial COT. it would have to be antigenically very different from the COT from liver peroxisomes. It is interesting to note that beef liver has a high titer of a protein antigenically similar to mouse liver COT. but beef heart. which has been studied extensively in the attempt to locate COT (23.24). does not. Peroxisomal B-Oxidation Peroxisomes have B-oxidation capability (44-48). Recent studies indicate that the contribution of peroxisomes to the total fatty acid oxidation in mouse liver could be as much as 50% (47-49). If large quantities of fatty acids are undergoing s-oxidation in the peroxisomes. and the long-chain fatty acyl-CoAs are not oxidized completely to short-chain 83 acyl-CoAs. as has been reported (44.46.59-63). then medium-chain acyl-CoAs and acetyl-CoA would be formed in the peroxisome and would need to be disposed of (50). Our data are consistent with the prOposed pathway of peroxisomal a-oxidation diagrammed in Figure 19. Long-chain acyl coenzyme As are directed to a-oxidation rather than acylcarnitine formation due to the Km and vmax effects discussed. Acetyl moieties produced would then be transferred to carnitine by peroxisomal CAT and presumably shuttled out of the peroxisome. Medium-chain acyl coenzyme As are directed to acylcarnitine formation rather than further B-oxidation. as the enzymes of peroxisomal B-oxidation are specific for longer chain-length substrates (59.112-115). Medium-chain acylcarnitines should then be shuttled out of the peroxisome and presumably into the mitochondria for further s-oxidation (30.50.62). COT activity in rat liver peroxisomes ranges from 6 to 15 nmol/min/mg (37.58). In comparison. COT in mouse liver peroxisomes has a very high specific activity. 250 nmol/min/mg. The finding that mouse liver has more COT than rat liver is consistent with the higher levels of peroxisomal B-oxidation in mouse liver (49) than in rat liver (46). It is not clear. however. why there is an apparent excess of COT activity compared to CAT activity since much more acetyl-CoA should be produced and shuttled out of peroxisomes than medium-chain acyl-CoA. We therefore cannot 84 .evaue xuueu uo couuewuxOIa aeEOnwaMea was 840 was 800 AeEOeAxOHOQ ceezuec couueaeu uo eEecue meeoaoue .oa eunuch .338 <81? £28-55 28.55:: 2.822: (\ 22.985 25:2: 8253.? 525.5282 159:2 5.35-2282 3.2.6» ..... 832.85.21.18- 3.2 m. -—-----—-—-> 8:85... 23898:: 2.822: 5:58.? 5.5-9.3 .F . €0-32 5.5-9.3 86 discount the possibility that COT has an additional unknown function beyond that proposed for eXport of fatty acid oxidation intermediates out of peroxisomes. 11. 12. 13. 14. LIST OF REFERENCES Gulewitsch.W.and Krimberg.R.(l905) z.Physiol.Chem. 1§.326. Fraenkel.G..Blewett.M..and Coles.M.(l948) Nature(pondon) 161.981. Friedman.S.and Fraenkel.G.(l955) Arch. Biochem. BIOphys. 59.491. Fritz.I.B.(1955) Acta Physiol.Scand. 11.367. Bremer.J.(l962) Nature(London) 196.993. Bremer.J.(l963) J.Biol.Chem. 238.2774. Fritz.I.B.and Yue.K.T.N.(1963) J.Lipid Res. 1.279. Norum.K.R.(1965) Biochim.Biophys.Acta 23.511. Bremer.J.(l962) J.Biol.Chem. 237.2228. Fritz.I.B..Schultz.S.and Srere.P.A.(l963) J.Bioi.Chem. 238.2509. Fritz.I.B.and Schultz.s.(l963) Fed. Proc. 11.637. Beenakkers.A.M.T.and Klingenberg.M.(l964) Biochim. Biophys.Acta 11.205. 1 Bieber.L.L..Sabourin.P..Fogle.P.J..Valkner.K.and Lutnick.R.(l980) in Carnitine Biosynthesis. Metabolism. and Functions (Frenkel.R.A.and McGarry. J.D..eds.) p.159.Academic Press. New York. Worm.R.A.A..Luytjes.W..and Beenakkers.A.M.T.(l980) 87 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 88 Insect.Biochem. 19.403. Ueda.M..Tanaka.A..and Fukui.S.(l982) Eur.J.Biochem. i11.2os. Claus.R..Kappeli.O..and Fiechter.A.(l982) Anal.Biochem. 127.376. Emaus.R.K.and Bieber.L.L.(l983) J.Biol.Chem. 111. 13160. Childress.C.C..Sacktor.B.. and Traynor.D.R.(l966) J. Biol. Chem. 242.754. MOppel.C.L.(l976)Enzymes giol. Membr. 1.119. Markwell.M.A.K..McGroarty.E.J..Bieber.L.L.. and Tolbert.N.E.(l973) J. Biol.Chem. 248.3426. Kopec.B. and Fritz.I.B.(l97l) Can. J. Biochem _1.941. Brosnan.J.T..KOpec.B..and Fritz.I.B.(l973) J.Biol.Chem. 111.4075. Clarke.P.R.M. and Bieber.L.L.(l981) J. giol. Chem. 111.9861. Clarke.P.R.M.and Bieber.L.L.(l981) J.Biol.Chem. 111.9869. Aas.M.(l97l) Biochim.Biophys.Acta 231.32. Martin.P.A.,Temple.N.J..and Connock.M.J.(l979) 1ur.J.Ceil Biol. 12.3. Norum.K.R..Farstad.M..and Bremer.J.(l966) Biochem.BiOphys.Res.Comm. 4.797. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 89 Arch.Biochem.BiOphys 170.716. Bieber.L.L.and Farrell.S.(l983) in The Enzymes 11.627. Bremer.J.(l983) Physiol.Reviews 11.1420. Choi.Y.R..Fogle.P.J..Clarke.P.R.H.. and Bieber.L.L.(l977) J.Biol.Chem. 1_1.793o. Barker.P.J..Fincham.N.J.. and Hardwick.D.C.(1968) Biochem. J. 111.739. Snoswell.A.M. and Koundakjian.P.P.(1972)Biochem J. 111.133. Solberg.H.E.(1974)Biochim. gigphys. Acta 360.101. Brdiczka.D..Gerbitz.K.. and Pette.D.(1969) Eur. J. Biochem. 11.234. Edwards.Y.M.,Chase.J.F.A..Edwards.M.R.. and Tubbs.P.K.(l974) Eur. J.Biochem. 11.209. Markwell.M.A.K..Tolbert.N.E.. and Bieber.L.L. (1976)Arch. Biochem.Biophys 111.479. Markwell.M.A.K.. and Bieber.L.L.(l976) Arch. Biochem. Biophys 111.502. Alkonyi.I..Kerner.J.. and Sandor.A. (1975) FEBS Lett. 11.265. Pearson.D.J. and Tubbs.P.K.(l967) 1iochem J. 191.953. Choi.Y.R..Clarke.P.R.M.. and Bieber.L.L.(l979) J. Biol. Chem. 254.5580. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 90 Bieber.L.L..Krahling.J.B..Clarke.P.R.H..Valkner.K.J.. and Tolbert.N.E.(l981) Arch. Biochem. BiOphys. 2111599. Kahonen.M.T.(l976) Biochim. BiOphys.Acta 428. 690. Lazarow.P.B. and DeDuve.C.(1976) Proc. Nat. Acad. Sci. 7 .2043. Lazarow.P.B.(l978) J. Biol. Chem. 111.1522. Mannaerts.G.P..Debeer.L.J..Thomas.J. and DeSchepper.P.J.(l979) J. Biol. Chem. 254.4585. Krahling.J.B..Gee.R..Gauger.J.A.. and Tolbert.N.E. (1979) J. Cell Physiol. 101.375. Krahling.J.B..Gee.R..Murphy.P.A..Kirk.J.R.. and Tolbert.N.E.(l978) Biochem. Biophys. Res. Comm. 1;. 136. Murphy.P.A..Krahlin9.J.B.Gee.R..Kirk.J.R.. and Tolbert.N.E.(l979) Arch. Biochem. piophys. 111.179. Tolbert.N.E.(l981) Annu. Rev. piochem. 11.133. Kahonen.M.T. and Ylikahri.R.M.(l979) Atheroscierosis 11.47. Markwell.M.A.K..Bieber.L.L.. and Tolbert.N.E.(l977) Biochem. Pharm. 11.1697. Casillas.E.R.and Erickson.B.J. (1975) 1111.Reprod. _1.275. Mutson.S.M..Van Dop.C..and Lardy.H.A.(1977) 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 91 Arch.Biochem.Biophys. 181.345. Solberg.H.E.(l97l) FEBS Lett. 11.134. Solber9.M.E. (1972) Biochim Biophys Acta. 280.422. Solberg.H.E..Aas.M.. and Daae.L.N.W.(l972) Biochim; BiOphys. Acta. 280.434. Moody.D.E. and Reddy.J.K.(l974) Res. Comm. Chem. Pathoi. and Pharm. 1.501. Bremer.J. and Norum.K.R.(l982) J. pipid Res. 11.243. Osmundsen.H..Neat.C.E..and Borrebaek.B.(l980) Int.J.§iochem. 11.625. Thomas.J..Debeer.L.J..De Schepper.P.J..and Mannaerts. G.P.(l980) 1iochem.J. 1_1. 485. Christiansen.R.z.Christiansen.E.N..and Bremer.J. (1979) Biochim.Biophys.Acta 573.417. Osmundsen.M..Christiansen.R.z.. and Bremer.J.(l980) in Carnitine giosynthesig. Metabolism. ang Functions (Frenkel.R.A. and McGarry.J.D..eds.) p.127. Academic Press. New York. Bieber.L.L.,Emaus.R..Valkner.K. and Farrell.s. (1982) Fed. Proc. 11.2858. Tolbert.N.E.(l974) in Methods in Enzymology (F1eischer.S.and Packer.L..eds.). 11.p.734. Academic Press. New York. Racker.E.(1950) 1iochim.Biophys.Acta 1.211. Pederson.T.C.(l973). J.Bioi.Chem. 248.7134. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 92 Leighton.F.(1968) J.Cell Biol. 11.482. Bieber.L.L..Abraham.T..and Melmrath.T.(l972) Anal.Biochem. 11.509. Srere.P.A..Seubert.W..and Lynen.F.(l959) Biochim. BiOphys.Acta 11.313. LeBlond.D.J..Ashendel.C.L..and Wood.W.A. (1980) Anal.Biochem. 104.355. LeBlond.D.J..Ashendel.C.L..and Wood.W.A. (1980) Anal.Biochem. 104.370. Yun.S.L.and Suelter.C.H.(l977) Biochim.Biophys. 5.25.9. 5.92.1. Reddy.M.K..Mollenberg.B.F.. and Reddy.J.K.(l980) Biochem.J. 188.731. Laemmli.U.K.(l970) Nature(London) 227.680. Andrews.P.(l965) Biochem.J. 11.595. Bohlen.P..Stein.S..Dairman.W..and Udenfriend.s. (1973) Arch.Biochem.Biophys. 155.213. Dunnett.C.W.(l955) J.Amer.Stat.Assoc. 11. 1096. Fiol.C.J.and Bieber.L.L.(l984) J.Biol.Chem. (submitted for publication) Saggerson.E.D.and Carpenter.C.A.(l981) FEBS Lett. 111.124. Venkatraghavan.V.and Smith.D.J.(l983) 15111 Biochem.Biophys. 220.193. Saggerson.E.D.and Carpenter.C.A.(l981) FEBS 83. 84. 85. 86. 87. 88. 89. 90. 91. 92.. 93. 94. 95. 96. 93 Lett. 129.225. Saggerson.E.D.and Carpenter.C.A.(l981) £111 11111 111.166. Saggerson.E.D.(l982) Biochem.J. 202.397. Saggerson.E.D.(l981) 1118 1ett. 129.229. Reddy.J.K..Moody.D.E..Azarnoff.D.L.and Rao.M.s. (1977) Proc.Soc.§1p.Biol.Med. 154.483. Reddy.J.K.and Krishnakantha.T.P.(l975) Science 1_1. 787. Lazarow.P.B.(l977) Science 111. 580. Mannaerts.G.P..Thomas.J..Debeer.L.J..McGarry.J.D.. and Foster.D.W.(l978) Biochim.Biophys.Acta 111. 201. Reddy.J.K.,Warren.J.R..Reddy.M.K..and Lalwani.N.D. (1982) Ann. N.Y. Acad.Sci. 386.81. Roels.F. (1976) J.Mistochem.Cytochem. 11.713. McCarry.J.D..Leatherman.G.F..and Foster.D.W.(l978) J.Biol.Chem. 111.4128. Moppel.C.L.and Tomec.R.J.(l972) J.Biol.Chem. 111. 832. Bergstrom.J.D.and Reitz.R.C.(l980) Arch.Biochem. Bithys. 1_1.71. Miyazawa.S..Ozasa.H..Osumi.T..and Mashimoto.T.(l983) J.Biochem. 11.529. Chase.J.F.A.and Tubbs.P.K.(l969) 1iochem.J. 111. 225. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 94 Hiyazawa.S..Ozasa.fl..Furuta.S..Osumi.T..and Hashimoto. T.(1983) J.Biochem. 11.439. Ueda.M..Tanaka.A..and Fukui.S.(l984) gur.J.Biochem. 111.445. Hitta1.B.and KurUp.C.K.R.(1980) Biochim.BiOths.Acta 111.90. 3 Chase.J.F.A.and Tubbs.P.K.(1967) 1iochem.J. 104. 510. Lund.fl.and Bremer.J.(l983) 11ochim.8ioghys.Acta 111.164. Valkner.K.J.and Bieber.L.L.(l982) Biochim.BioPhys. Acta 89.73. Fritz.I.B.and Schultz.S.K.(1965) J.Biol.Chem. 111. 2188. Chase.J.F.A.and Tubbs.P.K.(1966) Biochem.J. 11.32. Rebouche.C.J.(1983) J.Nutr. 111.1906. Haccari.P. et a1 (1982) 11st Congresso De11a Societa Italiana di Farmacolog11 (abstract) Paulson.D.J.and Shug.A.L.(1981) Life Sci. 11.2931. Garzya.6.and Amico.R.H.(1980) Int.J.Tissue React. 1.17s. ' De Grandis.D. et a1 (1980) J.Neur01.Sci. 11.365. Bazzato.G. et a1 (1979) Lancet 1.1041. Parrell.S.O.and Bieber.L.L.(l983) Arch.§iochem. Bioehxa. 222.123. Neat.C.E..Thomassen.n.s..and Osmundsen.u.(1981) 95 Biochem.J. 196.149. 113. Osumi.T.and Hashimoto.T.(1978) Biochem.BiOphys.Res. 12111 11.479. 114. Osumi.T.and Hashimoto.T.(1979) J.Biochem. 11.131. 115. Osumi.T..Hashinoto.T..end Ui.N.(1980) J.Biochem. 11.1735. APPENDIX COLLABORATIVE REVIEW OF CARNITINE ACYLTRANSFERASES 96 PLEASE NOTE: Copyrighted materials in this document have not been filmed at the request of the author. They are available for consultation, however, in the author's university library. These consist of pages: Appendix (Reprint) Un’ Micr. Ilms Int 3'30 N. ZEEB 80.. ANN ARBOR, Ml 48106 (313) 761-4700 II Carnitine Acyltransferases L. L. BIEBER ‘ SHAWN FARRELL I. Introduction ........................... 627 ll. Reactions Catalyzed ....................... 628 111. Number of Carnitine Acyltransferases ............... 629 A. General Comments ...................... 629 B. Carnitine Palmityltransferase (CPT) ............... 630 C. Carnitine Acetyltransferase (CAT) ............... 632 D. Carnitine Octanyltranst‘erase (COT) ............... 635 IV. CPT: Purification. Properties. and Regulation ............ 636 A. Pr0perties of Purified CPT ................... 637 8. Properties of Membrane-Bound CP’I' .............. 540 C. CPT Assays ......................... 642 V. Pathophysiology and Clinical Aspects ............... 642 I. Introduction Carnitine (y-trimethylamino-fl-hydroxybutyrate), (CH3)3N*-CH;- CHOH-Cl-Iz-COOH, was first isolated in 1905 (I) from muscle. Forty years later Fraenkel et al. (2) established that carnitine is an essential nutrient for larvae of the beetle Tenebrio molitor. and it was given the trivial name, vitamin B.. In 1955 Friedman and Fraenkel (3) presented evidence for a possible enzymatic role for carnitine, and Fritz (4) reported I. Gulewitsch, W.. and Kn'mberg. R. (I905). Hoppe-Seyler’s Z. Physiol. C Item. 45, 326. 2. Fraenkel. (3.. Blewett, M.. and Coles. M. (I948). Nature (London) 161, 981. 3. Friedman. S., and Fraenkel. G. (1955). A83 59, 49]. 4. Fritz. I. B. (I955). Acta Physiol. Scand. 3‘, 367. 627 THE ENZYMES. VOL. XVI Copyright c I”! by Academic Puss. Inc. All rights of reproduction in any form reserved. ISBN 0|2-I227I62 628 L. L. BIEBER AND SHAWN FARRELL a carnitine-dependent stimulation of palmitate oxidation by liver prepara- tions. Almost simultaneously the laboratories of Bremer and of Fritz provided convincing evidence for a role in mitochondrial [3 oxidation (5—7) of long-chain fatty acids. II. Reactions Catalyzed Carnitine is a cosubstrate for a family of enzymes that catalyze the reversible reaction acylcarnitine + CoASH —.-9 acyl-CoA + carnitine (I) The enzyme. which has a high transfer capacity for palmityl residues. is named carnitine palmityltransferase (CPT). and the one with a large acyl transfer capacity for acetyl residues is called carnitine acetyltransferase (CAT). The reactions catalyzed by the carnitine acyltransferases are de- fined: Forward reaction acyl-CoA + carnitine ———+ acylcarnitine + CoASH (2) Reverse reaction acylcarnitine + CoASH ——. acyl-CoA + carnitine (3) The acyl moieties are aliphatic hydrocarbons that range from 2 carbons to more than 20 carbons in length. They can be straight-chained. branch- chained. or unsaturated. There have been some reports that hydroxyacyl- camitines or CoA are substrates or products (8—10), but compounds such as succinyl- and malonyl-CoA are not substrates. The broad acyl specific- ity of the carnitine acyltransferases in mammalian systems does not occur in all living systems. In some yeasts, plants. and even insects, the acyl specificity may be more restricted and appears to depend on the fuel 5. Bremer. 1. (I962). Nature (London) 196, 993. 6. Bremer. 1. (I963). JBC 238, 2774. 7. Fritz. I. 8.. and Yue. K. T. N. (I963). J. Lipid Res. 4, 279. 8. Horak. H.. and Pritchard. E. T. (I971). 88.1 248. SIS. 9. Portenhauser. R.. Schafer. (3.. and Lamprecht. W. (I969). Hoppe~$e_vler's Z. Physio]. Chem. 350, 64]. IO. Bressler. R.. and Katz. R. J. (I965). JCI 44, 840. 18. CARN ITIN E AC YLTRAN SF ERASES 629 source and metabolism involved (II-I4). The functions of short-chain carnitine acyltransferases require elucidation, especially in non-fatty acidoxidizing systems such as trypanosomes (l5) and certain yeasts (l3. l6), and also in mammals. III. Number of Carnitine Acyltransferases A. GENERAL COMMENTS It would be incorrect to state that there are a specific number of cami- tine acyltransferases. Rather. the number and nature of the transferases depend on the tissue and animal source. For example. trypanosomes (l5) and yeast (13, 16) contain a short-chain carnitine acyltransferase of very narrow acyl specificity. In contrast, mammalian systems contain more than one enzyme (17—19) and the short-chain acyl-specific activity often greatly exceeds that of the long-chain. An exception appears to be honey- bee flight muscle mitochondria where short-chain carnitine acyltrans- ferase (12) was not detected. Even for mammals the number and nature of transferases vary and are tissue specific. For example, heart may contain as little as two carnitine acyltransferases (20-22) and liver clearly has a more complex distribution (23-27), whereas sperm has a predominance of carnitine acetyltransferase (28-31). 11. Childress. C. C.. Sacktor. B.. and Traynor. D. R. (1967). JBC 242, 754. 12. Beenakkers. A. M. T.. and Klingenberg. M. (1964). BBA 84, 205. 13. Bieber. L. L.. Sabourin. P.. Fogle. P. 1.. Valkner. K.. and Lutnick. R. (1980). In “Carnitine Biosynthesis. Metabolism. and Function“ (R. A. Frenkel and J. D. McGan’y. eds.). p. 159. Academic Press. New York. 14. Worm. R. A. A.. Luytjes. W.. and Beenakkers. A. M. T. (1980). Insect Biochem. 10. 15. Gilbert. R.. and Klein. R. A. (1982). FEBS Len. 141, 271. I6. Ueda. M.. Tanaka, A.. and Fukui. S. (I982). J. Biochem. (Tokyo) 124, 205. 17. Choi. Y. R.. Fogle, P. 1.. Clarke. P. R. 11.. and Bieber. L. L. (1977). 186‘ 252, 7930. 18. Solberg. H. E. (1972). 88.4 280, 422. 19. Fritz. I. B. (1963). Adv. Lipid Res. 1, 285. 20. Kopec. B.. and Fritz, I. B. (1971). Can. J. Biochem. 49, 941. 2]. Clarke. P. R. H.. and Bieber. L. L. (1981). JBC 256, 9861. 22. Clarke. P. R. H.. and Bieber. L. L. (1981). JBC 256, 9869. 23. Markwell. M. A. K.. McGroarty, E. 1.. Bieber. L. L.. and Tolbert. N. E. (1973). JBC 248, 3426. 24. Brosnan. .1. T.. Kopec, B., and Fritz. I. B. (1973). JBC 248, 4075. 25. Hahn. P.. and Seccombe. D. (1980). In "Carnitine Biosynthesis. Metabolism and Functions“ (K. A. Frenkel and D. J. McGarry. eds.). p. 177. Academic Press. New York. 26. Solben. H. E. (1971). FEBS Len. 12, I34. 27. Solberg. H. E.. and Bremer. J. (1970). BBA 222, 372. 28. Milkowski. A. L.. Babcock, D. F.. and Lardy. H. A. (1976). A88 176, 250. 630 L. L. BIEBER AND SHAWN FARRELL B. CARNITINE PALMITYLTRANSFERASE (CPT) I. Background Although the intramitochondrial membrane is permeable to long-chain fatty acids, only short- and medium-chain fatty acids (32) are activated to CoA esters within the mitochondrial matrix. The long-chain acyl-CoA synthetases are associated with microsomes (33), the outer mitochondrial membrane (34), and peroxisomes. Some variability exists for the activa- tion of medium-chain fatty acids since rat skeletal muscle mitochondria. in contrast to rat liver mitochondria, oxidize only octanoic acid in the presence of carnitine (35). Thus carnitine is required for translocation of the acyl residues across the acyl-CoA barrier of the inner membrane of mitochondria via CPT (6, 7, 36, 37), which converts cytosolic long-chain acyl-CoA to long-chain acylcarnitines; they subsequently enter the mito- chondrial matrix from the cytosol compartment and are then reconverted to acyl-CoA that can undergo )3 oxidation: Cytosol Matrix (catalyzed by the outer form of CPT) (catalyzed by the inner form of CPT) long-chain acyl-CoA + camitinej long-chain acylcarnitine + CoASlD long-chain acylcarnitine + CoASH long-chain acyl-CoA + carnitine Careful intracellular distribution studies (20, 23, 24, 36) have shown that CPT is a mitochondrial enzyme. Reports of CPT in microsomes were amended after more careful studies excluded microsomal location (38. 39). An exception is the report that some CPT is associated with rat heart microsomes (40). Small amounts of extramitochondrial CPT. which in- crease with changes in physiological states such as diabetes. fasting. or high-fat diet, have been reported (4]). However, it seems likely that ex- 29. Casillas. E. R. (1973). JBC 248, 8227. 30. Brooks. D. E. (1980). In “Carnitine Biosynthesis. Metabolism and Functions“ (R. A. Frenkel and J. D. McGarry. eds.). p. 219. Academic Press. New York. 31. Marquis. N. R.. and Fritz. l. B. (1965). JBC 240, 2197. 32. Aas, M. (1971). 88.4 231, 32. 33. Martin, P. A., Temple, N. 1.. and Connock, M. J. (1979). Eur. J. Cell Biol. 19, 3. 34. Norum. K. R.. Farstad, M.. and Bremer. .l. (1966). BBRC 24, 797. 35. Great. P. H. E., and Hulsmann. W. C. (1973). BBA 316, 124. 36. Hoppel. C. L. (1976). Enzyme: Biol. Membr. 2, 119. 37. Tomec. R. .l.. and Hoppel. C. L. (I975). .488 I70, 716. 38. Norum, K. R. (1966). Acta Physiol. Scand. 66, 172. 39. Norum. K. R.. and Bremer. J. (1967). JBC 242, 407. 40. Fogle. P. 1.. and Bieber. L. L. (1978). Int. J. Biochem. 9, 761. 41. Bremer, J. (1981). BBA 665, 628. rs. CARNITINE ACYLTRANSFERASES ‘ 631 tra-mitochondrial CPT activity, especially in liver. is due to the broad- Specificity carnitine octanyltransferase associated with peroxisomes. Recent data show that purified. homogenous. medium-chain carnitine acyltransferase from mouse liver peroxisomes has some activity with palmityl-CoA as substrate, but the K"l for this substrate is exceedingly large. 2. Membrane Distribution of CPT Although CPT is associated with mitochondria, how it is distributed with respect to the intramitochondrial membrane has not been unequivo- cally established. Several approaches have been used to measure the relative proportion of CPT on the matrix and cytosolic face of the inner membrane. Studies using digitonin or low amounts of detergent to remove the easily extractable CPT (presumably the transferase associated with the cytosolic face of the inner membrane) and other more direct assays have yielded distributions between 10 and 35% of the CPT associated with the cytosolic face of the inner membrane (36, 41-46); other investigations using malonyl-CoA inhibition (47) and DTNB under nonswelling condi- tions [(48, 49); see also Ref. (50)] give values of approximately 1 : l for distribution of the two activities. Thus, the fraction of CPT exposed to the cytosolic face of the inner membrane of mitochondria is between 15 and 50% of the total activity in normal liver, heart. and skeletal muscle mito- chondria. This ratio may depend on the type of tissue, the stage of animal development, the concentration of carnitine. and possibly the dietary or hormonal State of the animal. Fasting (4], 51-53), diabetes (54 ), and diet (55-57) can all affect CPT levels. 42. Bergstrbm. J. D.. and Reitz. R. C. (1980). ABB 204, 71. 43. Yates. D. W.. and Garland. P. B. (1966). BBRC 23, 460. 44. Harano. Y.. Kowal. J.. and Miller. M. (1972). FR 31, 863. 45. Layzer. R. B.. Havel. R. J.. and Mcllroy. M. B. (1980). Neurology 30, 627. 46. Wood. J. M.. Wallick. E. T.. Schwartz. A., and Chang. C. H. (1977). 88.4 486. 331. 47. McGarry. J. D.. Leatherrnan. G. F.. and Foster. D. W. (1978). JBC 253, 4128. 48. Bieber. L. L., Markwell. M. A. K.. Blair. M.. and Helmrath. T. A. (I973). BBA 326, 145. 49. Choi. Y. R.. Clarke. P. R. H.. and Bieber. L. L. (1979). JBC 254, 5580. 50. Patten. B. M.. Wood. J. M.. Harati. Y.. Hefferan. P.. and Howell. R. R. (1979). Am. J. Med. 67, I67. 51. Aas. M.. and Daae. L. N. W. (1971). BBA 239, 208. 52. Norum. K. R. (1965). BBA 98, 652. 53. Hahn. P.. and Skala. J. P. (1981). Can. J. Physio]. Pharmaeo]. 59, 355. 54. Harano. Y.. Kowal. J.. Yamazaki. R.. Lavine. L.. and Miller. M. (1972). ABB 153. 426. 55. Ishii. H.. Fukumori. N.. Horie. S.. and Suga. T. (1980). BBA 617, l. 56. Wolfe. R. C.. Maxwell. C. V.. and Nelson. E. C. (1978). J. Mm. 108, I621. 57. Bremer. J.. and Norum. K. R. (1982). J. Lipid Res. 23. 243. 632 L. L. BIEBER AND SHAWN FARRELL As discussed in Section III. the amount of catalytically active outer CPT becomes quite critical when one considers its potential effect on the flux of fatty acids through mitochondrial 3 oxidation and its regulation by malonyl-CoA. A major factor contributing to the different values obtained for the ratios of the outer and inner forms of CPT is undoubtedly the variable, and sometimes inadequate, methodology. An extreme example of assay variability is a report in which it was estimated that using digito- nin approximately 20—25% of rat liver mitochondrial CPT is the outer form, yet using a flavoprotein reduction method the ratio changed from approximately 1 :5 to l :450 [see Ref. (42), Table 3]. C. CARNmNE ACETYLTRANSFERASE (CAT) 1. Location Carnitine acetyltransferase is the predominant acyltransferase in most tissues (17). It was initially described by Friedman and Fraenkel (3) and was subsequently partially purified from pig heart (58). Early studies indicated a mitochondrial location in tissues such as mammary gland, liver, heart, skeletal muscle, and kidney (59-62). At least two forms of the enzyme were reported, one outer and the other inner, presumably similar to CPT (59. 6], 62). However, Tubbs and co-workers (63) questioned the concept of more than one CAT after partially purifying the activity from liver, heart, and muscle. Their data suggested the existence of a single type of CAT. The finding that liver from several species contains extrami- tochondrial CAT associated with peroxisomes and to a lesser degree with endoplasmic reticulum (23) provides a possible explanation for the pre- vious findings. Whereas the microsomal enzyme is tightly membrane as- sociated and very labile (64), the peroxisomal enzyme is a stable. soluble enzyme located in the matrix and is readily released by treatments that disrupt the fragile peroxisomal membrane (64, 65). Thus in tissues such as liver, CAT is associated with at least three different subcellular structures. while in tissues such as heart and skeletal muscle the distribution of CAT appears to be more limited. 58. Fritz. I. B.. Schultz. S. K.. and Srere, P. A. (1963). JBC 238, 2509. 59. Barker, P. J.. Fincham. N. J.. and Hardwick. D. C. (1968). BJ 110, 739. 60. Snoswell, A. M.. and Koundakjian. P. P. (1972). BJ 127, 133. 61. Solberg. H. E. (1974). BBA 360, 101. 62. Brdiczka. D.. Gerbitz. K.. and Pette, D. (1969). EJB II, 234. 63. Edwards, Y. 11., Chase, J. F. A., Edwards, M. R.. and Tubbs. P. K. (1974). EJB 46, 209. 64. Markwell. M. A. K.. Talbert. N. E.. and Bieber. L_. L. (1976). ABB 176. 479. 65. Markwell, M. A. K.. and Bieber. L. L. (1976). A88 172, 502. rs. CARNITINE ACYLTRANSFERASES 633 2. Substrate Specificity, Properties. and Mechanism Partially purified and purified preparations of CAT from mammalian sources Show a broad short-chain acyl-CoA Specificity with maximum activities for acetyl- and propionyl-CoA and declining Vmax values for the longer-chain acyl-CoA derivatives up to about ten carbons in length (58. 65, 66). The molecular weights of most of the preparations, regardless of the source, are 58,000 to 61 .000 (5 , 66, 67). Some exceptions are a com- mercial pigeon breast enzyme. which Showed a molecular weight of 51,000 with the identical procedures used for the microsomal and peroxi— somal enzyme (65 ). and the report by Mital and Kurup Where partially purified rat liver mitochondrial CAT was found to be a dimer of unequal subunits with molecular weights of 25,000 and 34,000 (68). However, neither of these studies eliminated proteolytic degradation. Dr. Furuta (Shinshu U.) and his colleagues have submitted for publication extensive studies that Show that the enzyme from rat liver mitochondria reported by Mital and Kurup is probably a proteolytic artifact obtained from a 67,500- dalton CAT (recent personal communication). Some kinetic constants have been determined for purified CAT. The equilibrium constant is 0.6 (reverse direction) for this reversible reaction (58) and the apparent K,,I values for the substrates, where reported. ap- pear to be in the physiological range (see Table 1). CAT from different sources can be divided into two general categories relative to substrate specificity (see Table I). The mammalian trans- ferases, regardless of organelle source, have a broad acyl specificity from 2 to approximately 10 carbons, with maximum Vm, values for the shorter acyl chain lengths. In contrast, CAT from yeast has a very narrow acyl- CoA specificity, being optimal for acetyl-CoA and propionyl—CoA with limited butyryl and isobutyryl activity and no acyltransferase activity with acyl carbon lengths greater than 4. As expected for an enzyme that uses a spectrum of acyl-CoA as substrates, the various acyl-CoA deriva- tives act as competitive inhibitors toward each other (69. 70). Some mechanism studies with the pigeon breast muscle enzyme have been performed (67, 70, 71). The enzyme can exist in two or more ternary enzyme complexes in rapid equilibrium with the free substrates. The interconversion of ternary complexes appears to be the rate-limiting step. Evidence has been presented that the enzyme contains a reactive sulfhy- 66. Bremer. J.. and Norum. K. R. (1967). JBC 242, 1744. 67. Chase, 1. F. A., and Tubbs. P. K. (1969). BJ 111, 225. 68. Mittal. B.. and Kurup. C. K. R. (1980). BBA 619, 90. 69. Chase. J. F. A. (1967).BJ 104, 510. 70. Chase. J. F. A., and Tubbs, P. K. (1966). BJ 99, 32. 71. Chase. 1. F. A. (1967). BJ 104, 503. .PCU cc .eLu> £3.39: .9 .U c. ”U 55:38:... .3.ququ Eras—ta 603 Sun 3932 28mm.“ 28380.). 1| 1. .1 can 3 c Emu 2.33 H :6. 9:30 we 6.9.00.0 co m... R ice—Bot c. N .0...me 693 ea 0.85: .1 22m .38.: Ben 18.... 2352...: 95.9. :3. S. a. a S .32.. .58.... 2.2—Soc o. .585 255 l a: l e: ... Q ...n 52. a: 58:. 88.3 .932... .53 + 95.3... 2....an 3.9.2.8.? l .I .68...“ 23.5.52 Sadr. I .l 138.. 35 an we 5at. .3— :ESoo o. 3353 use 8...? ate—552.8 l 3.85 :85 18...... 8895...: :8. .... l l I l m. .a 52. Loom .3538 o. mac—8.8.3.. min 285 .585 one: 035-0338 68.3 .I. I 9. ac we . .82. .3— xm ..o 38.. 59. cc .0... 2.2.28 2 -833 253 35322:. .1. 88m .383 285 -2552: :25 89% l | an. e. .5 .3 L2... 3. 825:8 565099. 933 .3 95:58 19.60 25.: <00 3:80.3— uflaom 1.. 22.5.5 .37.»...— -_>.oo< -3011: -..:uo< .31. ...k amu—PKNEOKE magmum2<¢hfi>u< wZ-P_Z¢(J l8. CARNITINE ACYLTRANSFERASES 635 dryl group at the catalytic site (72). It is rapidly inactivated by bromoace- tylcarnitine and bromoacetyl-CoA via formation of an S-carboxymethyl- CoA (-)-carnitine ester which binds very strongly to the substrate binding sites (67). Numerous studies with isolated mitochondria using acetylcarnitine as substrate show that considerable CAT is located on the matrix side of the inner membrane of mitochondria. The data are ambiguous as to whether mitochondria contain some CAT on the cytosolic face that is loosely associated because when most of these studies were done precautions were not taken to eliminate peroxisomal contamination. However. inves- tigations with insect flight mucle mitochondria (11 ) and heart (73) strongly indicate that little, if any. mitochondrial CAT is exposed to the cytosolic surface of the inner membrane. Again. species and organ differences may exist. D. CARNITINE OCTANYLTRANSFERASE (COT) Although mammalian tissues from several sources contain high amounts of carnitine acetyltransferase activity that has some activity with octanyl-CoA as substrate, unequivocal data demonstrating the existence of a separate (unique) COT had not been published (I7, 18, 20, 23, 26, 74); that is. COT had not been isolated and purified. However, the liver data from various sources (18, 20, 23 , 26) strongly indicate the occurrence of a separate COT. The finding that rat liver microsomes and peroxisomes contain COT and CAT activity with little, if any, CPT activity was very suggestive, especially when one considers the acyl-CoA chain-length spe- cificity of peroxisomes. The activity is high for short-chain and medium- chain acyl-CoA derivatives and very low for long-chain derivatives. The acyl-CoA specificity of peroxisomes is much broader than the acyl-CoA specificity of partially purified CAT from peroxisomes (65), clearly indi- cating the existence of a second enzyme. The existence of a separate COT protein has recently been established by Mr. Shawn Farrell of this labora- tory, who has purified to homogeneity a medium-chain carnitine acyl- transferase from mouse liver; he finds it has high V..... and very low K.. values for medium-chain acyl-CoA derivatives, but high K.. and low V..,“ 72. Fritz. l. B.. and Schultz. S. K. (1965). JBC 240, 2188. 72a. Valkner. K. J.. and Bieber. L. L. (1982). 38.4 689, 73. 72b. Emaus. R.. and Bieber. L. L. (1983). Submitted for publication. 72c. Emaus. R. K. (I982). PhD. Thesis. Michigan State University. East Lansing. 72d. Tanalta. A., Osumi, M.. and Fukui. S. (1982). Ann. N. Y. Acad. Sci. 386, I38. 73. Wmhaw. J. B. (1970). BRA 223, 409. 74. Saggerson, E. D. (1982). B] 202, 397. 636 L. L. BIEBER AND SHAWN FARRELL values for short-chain and long-chain acyl residues (74a). The possible occurrence of a unique COT in beef heart mitochondria was extensively investigated (2]. 22) and the data indicated the presence of only two carnitine acyltransferases, one a broad-specificity CAT and the other a broad-specificity CPT. More than 90% of the COT activity of beef heart mitochondria was accounted for by the combined activities of CAT and CPT. Thus the data indicate the existence of a separate COT in liver peroxisomes and rat liver microsomes, but its occurrence in tissues such as heart and skeletal muscle remains to be established. The occurrence of a more complex carnitine acyltransferase pattern in liver (a very active catabolic and anabolic tissue) as compared to heart or skeletal muscle, which are more catabolic in nature, seems reasonable. The finding that liver peroxisomes have considerable B-oxidation ca- pacity (75, 76) that does not go to completion but terminates at medium- chain acyl-CoA derivatives provides a role for the peroxisomal medium- chain carnitine acyltransferase activity, namely, to form acylcarnitines that shuttle acyl residues out of peroxisomes (57, 77—79). W. CPT: Purification, Properties. and Regulation The kinetic and catalytic properties of both membrane-bound and puri- fied CPT have been investigated extensively. As indicated in Section 11.8, it is well established that CPT exists in at least two forms on the inner membrane of mitochondria. One, referred to here as the outer CPT, is located on the cytosolic face, and the other, referred to as inner CPT, is located on the matrix face of the inner membrane of mitochondria. The two forms of the enzyme show different kinetic and catalytic properties consistent with their different functions in the cytosol and matrix com- partments. The data for both the isolated enzyme and the mitochondrial enzyme vary from one laboratory to another. Some data are summarized in Table II in which specific properties of purified or partially purified CPT are 74a. Farrell, S. and Bieber. L. L. (1983). A88. 222. 123. 75. Lazarow. P. B.. and deDuve, C. (1976). PNAS 73, 2043. 76. Lazarow. P. B. (1978). JBC 253. 1522. . 77. Leighton, F.. Brandan. E., Lazo, 0.. and Branfman. M. (1982). Ann. N. Y. Acad. Sci. 306, 62. 78. Bieber. L. L.. Emaus, R. K.. Valkner. K.. and Farrell. S. (1982). FP 41. 2858. 79. Osmundsen, B.. Christiansen. R. 2.. and Bremer. J. (1980). In “Carnitine Biosynthe- sis. Metabolism and Functions" (R. A. Frenkel and J. D. McGary. eds.). p. 127. Academic Press. New York. 18. CARNITINE ACYLTRANSFERASES 637 given, with selected comments about specific properties of the enzyme preparations used. The bottom portion of Table II summarizes selected investigations with isolated mitochondria that have given specific insights into the nature of this membrane-bound enzyme. A. PROPERTIES or PURIFIED CPT The kinetic properties of the inner and outer forms of CPT (primarily from liver) have been investigated (20, 33, 36, 42. 80), and other studies ' have been done with purified CPT without designation of the form (2, 22, 41, 66, 81, 82). As shown in Table II. the K,,, values for palmityl-CoA and CoASH are small, while the K", values for L-carnitine and palmitylcarni- tine are larger. With all of the substrates. the K," values reported are greater than lO-fold from one laboratory to another. The large differences are at least in part attributable to different assay conditions. Investiga- tions by Bremer, Norum. and colleagues (66, 81) demonstrated that palmityl-CoA can be a competitive inhibitor of L-carnitine. and that deter- gents can alter the palmityl-CoA kinetic parameters. It was recognized by Fritz and co-workers (20) that partially purified CPT showed variable K, values, particularly for carnitine and acylcarnitine, depending on the con- centration of the cosubstrate. In a study with highly purified, detergent- bound CPT from beef heart mitochondria (21. 22) it was shown that the K, values for carnitine (particularly the acylcarnitines) can vary greatly depending on the experimental conditions. For example, the Km for myristylcamitine varies between 14 and 2000 uM depending on the sub- strate concentration and detergent concentration. Both the K", and Vma, values for specific substrates are dependent on whether the long-chain acyl-substrate, either the acyl-CoA or acylcarnitine, is above or below its critical micelle concentration (cmc), the amount of detergent used, and whether the detergent is above or below its cmc. Curiously. the least- affected kinetic constants were those for the acyl-CoA derivatives. Such data also provide an explanation for the apparent discrepancy between the significantly different specificity profiles reported for CPT. Alterna- tively, the different chain-length specificities of CPT in the two directions could be explained by an increasing substrate inhibition by acyl-CoA (J. Bremer, personal communication). Almost all data are in agreement that CPT catalyzes reversible reactions involving acyl-CoA and acrylcarnitines , as substrates using acyl residues from approximately 6 to 20 carbons in 80. West. D. W.. Chase. J. F. A., and Tubbs. P. K. (l97l). BBRC 42, 912. 8l. Bremer. J.. and Norum. K. R. (1967). JBC 242, 1749. 82. Norum. K. R. (1964). BBA 89, 9S. 638 CARNITINE PALMITYLTRANSFERASE (CPT) PROPERTIES L. L. BIEBER AND SHAWN FARRELL TABLE II K.. (uM ) Palmityl- Palmityl- Preparation Source M, carnitine CoASH CoA Carnitim Outer CPT Ox liver 59,000 12 — 0.59 140 Inner CPT Ox liver 65 .000 60 — 9 2600 CPT Calf liver — 40 50 10 250 mitochondria Outer CPT Calf liver 150,000 136 5.5 17.6 450 Outer CI’I' Calf liver — 170 45 31 210 Inner CPT Calf liver 150.000 ' — — — -— Outer CPT Rat liver 430.000 11 35 2.8 280 mitochondria Inner CPT Rat liver 430.000 11 34 3.5 300 mitochondria Inner and Beef heart 67.000 — — 2“ — outer CPT mitochondria (Detergent CPT complex = 510.000 Intact Bovine heart — — — — — mitochondria and rat liver mitochondria Heavy Rat liver —- — — 1.7 170 mitochondria Mitochondria Rat liver — — — — — and heart Mitochondria Rat liver — — -— — — Mitochondria Rat liver — — — — — from fed and fasted animals Mitochondria Rat heart — — — — — and liver ‘ Unpublished data from Carol Fiol of this laboratory. length. However, some evidence has been presented indicating the inner form of CPT is not reversible (83). Investigations by Kopec and Fritz (83), Bremer and Norum (8] ), Berg- strom and Reitz (42), and Clarke and Bieber (21. 22) have all resulted in data indicating that the outer and inner forms of CPT represent kinetically 83. Kopec. B.. and Fritz. J. B. (1973). JBC 248. 4069. 18. CARNITINE ACYLTRANSFERASES 639 TABLE II (Continued) References Acyl group specificity and comments 36. 80 Very broad with moderate C. and medium-chain activity. 80. 36 Broad C. -’ C.. with maximum activity with Cu. 66. 8]. 82 Broad with moderate medium-chain activity; inhibition by palmityl-CoA. 20. 83 Low medium-chain activity, high long-chain activity. 82 20. 83 CPT” -' CPT. with urea. K... values for carnitine. CoASH. and acylcamitine vary. 42 Kinetic data indicate CPT outer and CPT inner are the same enzymes and in siru factors alter properties of CPT. 42 21. 22 Broad acyl specificity with highest activity with C.., in forward direction and great- est with C .. in reverse direction. CPT outer and inner are the same protein. Great variability in the K.., values for carnitine and acylcarnitines depending on experimental conditions. 37, 83a Biphasic rates of palmitylcarnitine formation versus palmityl-CoA concentrations. Very high acyl-CoA concentrations required for I’m; lag before maximum palmityl-CoA oxidation. 84 Low K... for palmityl-CoA at low (0.25 mM) carnitine and no biphasic saturation curve. 46 Concentration-dependent lag in palmityl-CoA oxidation: possible substrate inhibition. 47. 85 . 86 Malonyl-CoA is a potent inhibitor of outer CPT; malonyl-CoA may be the key intermediate for coordinating fatty acid synthesis and degradation in liver. 4!. 86—89 The sensitivity of outer CPT to malonyl-CoA may vary depending on the physio- logical state of the animal. 74 CPT very sensitive to ionic composition of media: CPT sensitive to malonyl-CoA. different forms of the same protein. Such data provide a possible explana- tion for the apparent conversion of the inner form to the outer form by treatment with urea (83). If CPT is a single protein in mitochondria. then the kinetically different forms must be determined by the membrane and the membrane environment, reminiscent of the concept of allotopy used 83a. Hoppel. C. L.. and Tomec. R. J. (1972). JBC 247, 832. 640 L. I... BIEBER AND SHAWN FARRELL to explain the different catalytic and kinetic properties of other mem- brane-bound enzymes. B. PROPERTIES OF MEMBRANE-BOUND CPT Some of the extensive studies of mitochondrial CPT are summarized in the bottom portion of Table II. A biphasic saturation curve for palmi- tylcarnitine formation results when the palmityl-CoA concentration is varied (37). In some studies extremely high amounts of palmityl-CoA were required for attaining V0,“, yet in what appear to be essentially identical experimental conditions except the carnitine concentration was approximately 10—fold lower (84), no biphasic saturation curve was de- tected and a K," of less than 5 11.11! for palmityl-CoA was obtained. The contribution of lysis or swelling due to detergent effects of palmityl-CoA to the biphasic saturation curves has not been determined. Other studies with intact mitochondria have shown (37, 46) concentration-dependent lags in palmityl-CoA oxidation; such lags are not obtained when palmi- tylcarnitine is the substrate. 1. Efi'ects of MalonyI-COA The data showing variable kinetics of CPT indicate that other, as yet unidentified factors may affect catalysis by CPT. This has been reinforced by the results of McGarry and Foster (47, 85 . 86), which show that ma- lonyl-CoA can be a potent inhibitor of the outer form of CPT. The inhibi- tion by malonyl-CoA is lost when the enzyme is solubilized, indicating the importance of membrane factors. Although the physiological significance of the malonyl-CoA inhibition has been questioned by some, because of the amounts of malonyI-CoA in situ, others have confirmed and extended the investigations. Very recently, it has been shown that outer CPT sensi- tivity to malonyI-CoA may vary depending on the physiological state of the animal (41, 86-89), including different thyroid states (90), and the ionic composition of the assay media (74). In all of these studies signifi- cant inhibition of CPT by added malonyl-CoA was obtained at concentra- tions at or below those considered physiological. . VanTol, A. (1974). BBA 357, 14. McGarry, J. D.. and Foster. D. W. (1979). JBC 254, 8163. . McGarry. J. D., and Foster. D. W. (1981). BJ 2110, 217. . Robinson, 1. N., and Zarnmit, V. A. (1982). 8.] N6, 177. Cook, 6. A., Otto, D. A., and Cornell, N. W. (1980). B] 192, 955. Saggerson, E. D.. and Carpenter. C. A. (1981). FEBS Len. 129, 225. . Stakkestad. S. A., and Bremer, J. (1982). BBA 711, 90. annexe: 18. CARNITINE ACYLTRANSFERASES 641 a. Speculation. The studies that show that the sensitivity of CPT to malonyl-CoA can vary depending on the physiological state of the animal (86-90) have an exceedingly important ramification, namely, that an outer CPT must exist in at least two interconvertible forms on the cytosolic surface of the inner membrane of mitochondria. One is catalytically active both in the presence and absence of malonyl-CoA, and the other is cata- lytically active but sensitive to malonyl-CoA. Therefore it seems likely that membrane-bound CPT may be a regulated, possibly allosteric, en- zyme. If so, modulation of the activity through covalent modification or the existence of a membrane-bound regulator component(s) similar to classical regulator subunits seems plausible. This could provide an expla- nation for the lack of malonyl-CoA sensitivity when the enzyme is sol- ubilized with detergents, the apparent lack of sensitivity on the matrix side of the inner membrane, and the variability in sensitivity by different physiological states. These speculations are mentioned because most data with intact mitochondria in which latent and overt CPT have been investi- gated are usually interpreted in terms of two distinct, catalytically differ- ent proteins (91). Nevertheless, it is evident that mitochondrial CPT can be inhibited by low concentrations of malonyl-CoA, but whether this is the major regulator in coordinating fatty acid synthesis and fatty acid catabolism and ketogenesis remains to be unequivocally determined. 2. Substrate Analogs and Inhibitors Removal of the B-hydroxyl group (deoxycarnitine) from carnitine abol- ishes its activity (92) and produces a competitive inhibitor of the reaction (93). Substitution of the B-hydroxyl group with a thiol group does not cause loss of activity (94), but removal of a methyl group (conversion of the quaternary ammonium to a tertiary amine) produces an inhibitor norcarnitine (93). Palmityl-(+)-carnitine inhibits CPT (43, 95) with a larger K,- for solublized CPT than the membrane-bound enzyme. Fatty acyl-CoA esters of 2-tetradecylglycidic acid (96, 97) and l-pyrenebutyryl- CoA (98) are potent inhibitors of carnitine palmityltransferase; the former affects the outer form of CPT. The l-pyrenebutyrylcarnitine derivative is 91. Hoppel, C. L. (1982). F? 41, 2853. 92. Fritz. I. 8., Kaplan, B.. and Yue. K. T. N. (1962). Am. J. Physiol. 202, 117. 93. Norum. K. R. (1965), REA 9, 511. 94. Tubbs. P. K.. and Chase. J. F. A. (1970). B.) 116. 34p. 95. Fritz, l. B.. and Marquis. N. R. (1965). PNAS 54, 1226. 96. Tutwiler, G. F.. and Ryglak, M. T. (1980). Life Sci. 26. 393. 97. Tutwiler, G. F.. Ho, W.. and Mahrbrachr, R. J. (1981). “Methods in Enzymology." Vol. 72, p. 533. 642 L. L. BIEBER AND SHAWN FARRELL a potent inhibitor of the carnitine : acylcarnitine translocase (98). The 2- substituted oxiran-Z—carbonyl-CoA esters (99) and palmityl-CoA analogs in which the carbonyl group is replaced by methylene groups (100) are also potent inhibitors of outer CPT. C. CPT Assays The large differences in CPT data from one laboratory to another are undoubtedly partly due to the use of different assays and assay condi- tions. Assay problems include (a) nonlinear rates (i.e., lags in activity, which in our hands are not always reproducible; (b) variable K", values depending on experimental conditions; (c) latent and overt activity with conversion of one form to the other; and (d) effects of ionic strength, cations, inhibitors, etc. With such an enzyme, rigorously controlled assay conditions must be used that assure linearity during the reaction time. Since there is no ideal assay for CPT, several methods have been used, which are summarized in Table 111. Some investigators have used end- point assays and assumed the assay conditions adequate for controls are also adequate for experimental samples. Such assumptions are not always valid and the data obtained may be equivocal. V. Pathophysiology and Clinical Aspects It is now well established that low carnitine levels (also low carnitine acyltransferase levels) in human muscle are associated with certain types of myopathies (101, 102), and low carnitine levels in liver are associated with serious metabolic problems (103, 104). The fact that some muscle myopathies are apparently related to low levels of CPT (105, 106) or low levels of carnitine (102) is consistent with a lesion in the B oxidation of long-chain fatty acids. However. in systemic carnitine deficiency the de- creases in liver carnitine indicate other roles, with difierent etiologies and 98. Wolkowicz, P. B.. Pownall, H. J.. and McMillin-Wood. J. B. (1982). Biochemistry 21, 2990. 99. Bartlett. K.. and Meredith, P. (1981). Biochem. Soc. Trans. 9, 574. 100. Ciardelli, T.. Stewart, C. J.. Seeliger. A., and Wieland. T. (1981). Justus Liebigs Ann. Chem. p. 828. 101. Karpati. 6.. et al. (1975). Neurology 25. 16. 102. Engel, A. 6., and Angelini. C. (1973). Science 179, 899. 103. Ware, A. J.. et al. (1978). J. Pediatr. 93, 959. 104. Angelini. C.. Lucite. 8., and Cantarutti, F. (1976). Neurology 26, 633. 105. DiMauro. S., and DiMauro, P. M. M. (1973). Science 182, 929. 106. Bertoni, T.. et al. (1980). Neurology 30, 263. l8. CARNITINE ACYLTRANSFERASES TABLE III 643 CARNITINE PALMITYLTRANSFERASE ASSAYS Method Principle Comments Exchange 232 Forward 232 Reverse Radioactive product formation Hydroxamate formation CoASH release Exchange of radiolabeled cami- tine into a pool of acylcarni- tine. Rate assay following thioester formation spectrophotometri- cally at 232 nm. Same as above but measured in opposite direction by monitor- ing thioester disappearance. Determine the formation of radioactive product using either radiolabeled carnitine or radiolabeled acyl groups. Product acyl-CoA or the remain- ing substrate (acyl-CoA de- pends an assay direction) converted to hydroxymate that is quantitated. Initial rate assay in which the CoASH released is monitored speetrophotometrically with SH reagents such as DTNB and DTBP. Usually performed as an end- point assay. Many do not ensure that the experimental samples respond as controls. Acyl-CoA hydrolase (present in mitochondria. microsomes. and lysosomes) affects final result due to hydrolysis of the acyl-CoA [see Ref. (90)]. Particulate preparations can be used. Difficult to perform with non- purified systems because of high 232 backgrounds. Prepa- rations should be solubilized. Deviations from linearity and lags occur. See comments above. Requires separation and quanti- tation of radioactive product. Can be used as an initial rate assay. but is usually used as an end-point assay where the limitations of exchange assay can occur. Same limitations as stated above for end-point assays. Some inhibition by hydroxylamine has been reported. Can be used with nonpurified systems. Total CPT can be assayed in crude systems; requires deter- gent solubilization. Can be used to estimate outer C PT in absence of detergent. under nonswelling conditions. Inhibi- tion by DTNB at pH 7.4. but not a problem at pH 8.0. 644 p L. L. BIEBER AND SHAWN FARRELL possibly secondary effects of carnitine. Examples are neurological symp- toms due to an apparent inhibition of pyruvate dehydrogenase (107), Reyes-like syndrome (108), and the effects of carnitine deficiency on propionic acid acidemia ( 109). Apparently there are no examples of deleterious effects due to tissue carnitine increases. Even when the carnitine levels increase by an order of magnitude such as that in muscle and liver of streptozotocin-induced diabetic sheep (110). the carnitine pool appears to function normally. Rather, clinical problems arise due to severe reductions in either mito- chondrial carnitine palmityltransferase (the medium- and short-chain ac- tivities have not been thoroughly investigated) or tissue carnitine levels. Other possible roles for carnitine unrelated to mitochondrial )3 oxidation of long-chain fatty acids have been proposed [see Ref. (78) for dis- cussion]. 107. DiDonato, S. D., Rimoldi. A., Moise, D.. Bertagnoglio. B.. and Uziel. C. (1979). Neurology 29. 1578. 108. Chapoy, P. R., et al. (1980). N. Engl. J. Med. 303, I389. 109. Roe, C. R., and Bohan. T. P. (1982). Lancet. 1411. 110. Fishlock, R. C.. Snoswell, A. M.. Valkner. K.. and Bieber. L. L. (1982). Int. J. Biochem. 4, 451.