X 1.8 cm) of Sephadex G-200 equilibrated with isotonic sucrose buffer .... 8.95. 181. 30. 1.95. 19.6. 97. 16. 0.819. 8.23. 143. 23. 4.09. 41.2. 59. 10 24.7. 248. 115.
T H E JOURNAL OF
Vol. 256, No. 19, Issue of October 10,pp. 9861-9868,1981 Printed in U.S.A.
Isolation and Purificationof Mitochondrial Carnitine Octanoyltransferase Activitiesfrom Beef Heart* (Received for publication, January 22, 1981)
Peter R. H. Clarke and Loran L. Bieber From the Biochemistry Department, Michigan State University, East Lansing, Michigan 48824
The purpose of this study was to characterize the mitochondrial membrane to the site of mitochondrial /3 oximitochondrial carnitine octanoyltransferase of beef dation. This role involves a reversible reaction catalyzed by heart. Carnitine acyltransferase activities were solu- carnitine palmitoyl transferase located on both inner and bilized from isolated beef heart mitochondria using KC1 outer surfaces of the inner mitochondrial membrane. This and the nonionic detergent, Triton X-100, at final con- shuttling requires a translocase for carnitine and its esters (3, centrations of 1 M and 2%, respectively. Uponfraction- 4). Conflicting reports describe the “inner” and “outer” caration of the solubilized protein on Cibacron Blue Seph- nitine palmitoyltransferase activities as two distinct transferarose, t w o protein peaks with carnitine octanoyltrans- ase activities (5, 6) and as identical proteins with regard to ferase were obtained. These two fractions accounted kinetics and molecular weight (7). for all carnitine acyltransferase activity present in the The second role proposed for carnitine involves the mitooriginal mitochondrial suspension. The first eluting chondrial carnitine acetyltransferase which has been postupeak w a s purified 400-fold b y Sephadex G-100 gel fil- lated to allow the movement of activated acetate groups across tration, CM-Sepharose ion exchange, and hydroxylapatite chromatography to a single protein of greater the inner mitochondrial membrane (1).In this model, carnithan 95% purity. This carnitine acetyltransferase tine accepts activated short chain fatty acids and carries them shows highest activity with acetyl and butyryl carni- into the cytoplasm, to free intramitochondrial CoA for further tine and coenzyme A esters. It has a subunit molecular /3 oxidation. Thus, carnitine acetyltransferase is thought to weight of 62,600 on sodium dodecyl sulfate polyacryl- provide an “acetyl sink” (8,9). Recently, a more generalized amide gel electrophoresis, a native M, of 60,500 on role for carnitine has been proposed to explain the broad Sephadex G-200 gel filtration, and a PI of 8.20 on su- spectrum of carnitine acyltransferase activity that occurs in mammalian tissues (10, 11).This involves modulation of the crose density gradient isoelectric focusing. The second peak of carnitine acyltransferase activity intracellular CoASH/acylcoenzyme A ratio. In 1971, Solberg suggested the existence of a third carnitine from Cibacron Blue Sepharose was purified 1600-fold by fractionationon Sephadex G-100 gel filtration, QAE- acyltransferase, carnitine octanoyltransferase (12). The conSephadex ion exchange, and hydroxylapatite chroma- clusion wasbased on substrate specificity profiiesof commertography toa single protein of greater than 95% purity. cial pigeon breast muscle mitochondrial carnitine acetyltransThis enzyme was carnitine palmitoyltransferase. It is ferase and of disrupted mitochondria from rat liver, heart, and most active with decanoyl and lauryl ester substrates, testis. She subsequently attempted a partial purification and has subunit molecular weight of 67,000 on sodium do- separation of carnitine acetyltransferase, carnitine octadecyl sulfate-polyacrylamide gel electrophoresis, a PI noyltransferase, and carnitine palmitoyltransferase activities of 8.05 on sucrose densitygradient isoelectric focusing, from calf liver, again concluding the existence of a separate and migrates as part of a detergent micelle ofapparent carnitine medium chain transferase (13).Also in 1971, Kopec M, 510,000 on Sephadex G-200 gel filtration. and Fritz noted in theirreport of the purification of a carnitine It is concluded that there are only two carnitine palmitoyltransferase protein from calf liver, the finding of acyltransferase proteins present in beef heart mitocarnitine acyltransferase activity most specific for octanoyl chondria, one membrane-bound (carnitine palmitoylcarnitine in extractsof mitochondrial protein from beef heart; transferase) and one membrane-associated (carnitine acetyltransferase). Eachhas significant activity toward they also concluded the presence of an “intermediary chain hexanoyl, octanoyl, and decanoyl carnitine and coen- length carnitine acyltransferase distinct from carnitine palzyme A esters. The presence of a separate medium mitoyltransferase and carnitine acetyltransferase” (14). Results of Markwell from this laboratory investigating carchain length-specific carnitine acyltransferase in beef nitine acyltransferase activities in liver organelles have shown heart mitochondria is not confirmed by our results. in peroxisomes and in microsomes the absence of detectable carnitine palmitoyltransferase activity, while at the same time, More than one role has been proposed for 1( -)-carnitine in sigdlcant carnitine octanoyltransferase activity from which intermediary metabolism. One role,fiist suggested by Bremer carnitine acetyltransferase was successfullyseparated and par(1)and by Fritz (2) requires carnitine to shuttleactivated long tially purified (15).Solubilizationand stabilization of carnitine chain fatty acids across the coenzyme A-impermeable inner octanoyltransferase activity from these organelleswas not achieved. A survey of carnitine acetyltransferase, carnitine * This work was supported in part by grant AM 18427 from the octanoyltransferase, and carnitine palmitoyltransferase activNational Institutes of Health and in part by the Michigan Heart ities in homogenates from 10 rat tissues by this laboratory Association. Paper No. 9753 from the Michigan Agricultural Experi- (16) revealed significant carnitine octanoyltransferase activity mental Station. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore in all tissues assayed except brain. In theovary, intestine, and be hereby marked “advertisement” in accordance with 18 U.S.C. liver, carnitine octanoyltransferase was the highest of the activities seen. Section 1734 solely to indicate this fact.
The possible existence of a medium chain carnitine acyltransferase is intriguing in light of the proposed roles for carnitine to date. Medium chain length fatty acids are rare in the normal diet. Fatty acids such as octanoate are able to cross the inner mitochondrial membrane (17) and can be activated for3!, oxidation by thiokinases within the matrix. It is possible that mitochondrial p oxidation of long chain fatty acids to medium chain lengthsis followed bytheir transfer to carnitine and export from the organelle. Significant intracellular concentrationsof medium chain acyl esters of CoA have been reported in rat liver (18, 19).Reliable estimates of tissue concentrations of medium chain length carnitine estershave not been available, however. Octanoyl carnitine is observed to be roughly equally divided between the acid-soluble and insoluble fractions obtainedduring isolation and quantitationof short versus long chain carnitine esters (20). Many recent studies withperoxisomes indicate that peroxisomes, like glyoxysomes, are capable of fi oxidation of long chain fatty acids (21, 22). The work of Lazarow (23) and of Osmundsen et al. (24) suggests incomplete breakdown of palmitoyl-CoA to acetyl-coAby peroxisomes. Thus, theproducts of peroxisomal fi-oxidation in vivo may be medium chain acyl-CoA esters and, via peroxisomal carnitine octanoyltransferase, medium chain carnitine esters. These esters may be the substrate for an inner mitochondrial carnitine octanoyltransferase, providing a function for an otherwise enigmatic carnitine acyltransferase enzyme. In order to compare the kinetic and physical characteristics of such a mitochondrial carnitine octanoyltransferase protein with those of carnitine acetyltransferase and carnitine palmitoyltransferase, we chose to investigate beef heart mitochondria, fromwhich Kopec and Fritz (14) extracted fractionsrich in carnitine octanoyltransferase activity.This paper describes the isolation and purification of carnitine octanoyltransferase activity-containing proteins from beef heart mitochondria. A portion of these results has been reported elsewhere in an abbreviated form ( 2 5 ) .
sucrose buffer, assayed for enzyme activities and protein, and stored at -80 "C. Solubilization of Mitochondria-Mitochondrial suspensions from five beef hearts were thawed, pooled, and added to one-half volume of 3 M KC1in 6% Triton X-100. The suspension was inverted for mixing and then treatedbatchwise with six passes of the Teflon-glass homogenizer at 15 "C. After centrifugation for 90 min at 89,000 x g (29,000 rpm) in the type 30 rotor, a clear apricot-colored supernatant fluid containing solubilized mitochondrial protein was pipetted from between floating lipid and the flocculant surface of the mitochondrial pellet. Carnitine Acyltransferase Assays-The forward reaction was assayed at 412 nm by the DTNBmethod of Bieber et al.(26). The 200pl reaction volume contained 115 mM Tris-HC1, 1.1 mM EDTA, 0.1% Triton X-100,1.25 mM I(-)-carnitine, 250 p~ DTNB and 1 0 0 acyl~ ~ CoA at pH 8.0 and 25 "C. The reverse reaction was assayed by the method of Srere et al.(27) in which CoA thioester bond formation is monitored at 232 nm. The 200-4 reaction volume contained 0.2 M Tris-HC1, 1 mM dithlothreitol, 0.5 mM EDTA, 0.1% Triton X-100, 60 PM CoASH, and 500 PM acylcarnitine at pH7.45 and 35 "C. Sephadex G-200 Chromatography-Solubilized protein from beef heart mitochondria (10 ml) was chromatographed on a column (90 X 1.8 cm) of Sephadex G-200 equilibrated with isotonic sucrose buffer containing 1.0% Triton X-100 and 1 M KC1; 4.0-ml fractions were collected. Catalase and hemoglobin were similarly chromatographed as standards. Void and bed volumes were determined using blue dextran and KaFe(CN)G,respectively. Comparison of Beef Liver and Heart CarnitineAcyltransferase Activities-Heart mitochondria were isolated and solubilized as described above. Beef liver was treated in an identical manner, with two exceptions: 1) the ratio of liver tissue to sucrose isolation buffer used for homogenization was 140 g:420 ml, and 2) the Polytron homogenization/sonication step was omitted. The solubilized protein solutions from liver and heart mitochondria were chromatographed separately on a column (112 X 4.8 cm) of Sephadex G-100. Samples of 100 ml containing approximately 15 mg of protein/ml were applied to the columns and 24-1111 fractions were collected. Purification of Carnitine Octanoyltransferase Activity-containing Proteins from Beef Heart Mitochondria-Beef heart mitochondria were solubilized as described above. The mitochondrial protein solution was equilibratedwith blue buffer (2% Triton X-100,2.5 mM Hepes, 0.25 mM EDTA, 60 mM KCl, pH 7.5) by exhaustive dialysis and applied at a flow rate of 130 ml/h to a column (75 X 4.1 cm) of Cibacron Blue Sepharose 4B equilibrated with blue buffer. Four bed volumes of buffer were then passed through the column to wash off EXPERIMENTALPROCEDURES unbound proteins before a 1500-ml linear gradient of60 to 860 mM KC1 in blue buffer was used to elute carnitine octanoyltransferase Materials activity. Coenzyme A and coenzyme A esters were from P-L Biochemicals. Fractions eluting from the Blue Sepharose column containing both Carnitine was a generous gift from Otsuka Pharmaceutical Co. and carnitine octanoyltransferase and carnitine acetyltransferase activiSigma Tau farmaceutici. Ampholines and hydroxylapatite were from ties (fractions 31 to 34) were pooled and passed through a column Bio-Rad. Sephadex and Sepharose chromatography media were from (112 X 4.8 cm) of Sephadex G-100 equilibrated with CM buffer (1% Pharmacia. Fluorescamine, Triton X-100, and 5,5'-dithiobis(2-nitro- Triton X-100,5.0 mM HEPES, 0.25 mM EDTA, 60 mM KC1, pH 7.3). benzoic acid) were from Sigma. Octanoyltransferase-containing fractions from the gel filtration column were pooled and applied a t a flow rate of 80 ml/h to a column Methods (29 x 4.1 cm) of CM-Sepharose CL-GB equilibrated with CM buffer. Mitochondrial Isotation-Beef hearts were obtained from a local The column was washed with 4 bed volumes of CM buffer and then abbatoir. Immediately upon removal from the carcass, the heart was eluted with a 1000-ml linear gradient of CM buffer containing 60 to sliced and packed in ice for transport to the laboratory. All subsequent 560 mM KCl. Fractions containing carnitine octanoyltransferase acprocedures were performed at 4 "C. The ventricle was cleaned of all tivity were pooled (117 ml), diluted with 117 ml of20mM KPOI, pH fat and connective tissue and cut into thin slices. Batches of 60 g of 6.6, and applied at a flow rate of24 ml/h to a column (13 X 2.2 cm) ventricle were homogenized in 500 ml of sucrose isolation buffer (0.25 of hydroxylapatite previously equilibrated wtih with HAP(-) buffer (0.1% Triton X-100, 10 mM KPO,, 60 mM KCl, pH 6.8). Four bed M sucrose, 5.0 mM HEPES,' 0.25 mM EDTA, pH 7.7) for 45 s at high speed in a Waring Blendor. T o insure more complete release of volumes of HAP(+) buffer (identical with HAP(-) buffer but with mitochondria, this was followed by a 30-s Polytron homogenization the addition of 0.25mM EDTA) were used to wash the column before using the large probe in batches of 220 ml.Mitochondria were isolated elution with a 200-ml linear gradient of HAP(+) buffer containing 10 from the homogenate by differential centrifugation employing 15-min to 510 mM KPO,, pH 6.8. Fractions having carnitine octanoyltransspins at 500,15,000,500, 11,OOO, and 7,000 X g in a Sorvall RC-2 ferase of a constant specific activity were pooled. Fractions eluting from the BlueSepharose column with both centrifuge using the 55-34 and the GSA rotors. High speed pellets were resuspended in isotonic sucrose buffer (0.25 M reagent grade carnitine octanoyltransferase and carnitine palmitoyltransferase acsucrose, 2.5 mM HEPES, 0.25 mM EDTA, pH 7.5) using two strokes tivities (fractions 31 to 48) were pooled and passed through a column of a loose fitting Potter-Elvehjem Telfon-glass homogenizer. The (112 X 4.8 cm) of Sephadex G-100 equilibrated with QAE buffer (1.0% 7,000 x g pellet was resuspended in a minimal volume of isotonic Triton X-100, 5.0 mM Bis-Tris propane buffer, 0.25 mM EDTA, 20 mM KC1, pH 9.7). Carnitine octanoyltransferase-containing fractions ' The abbreviations used are: HEPES, 4-(2-hydroxyethyl)-l-piper-were pooled and applied at a flow rate of 30 ml/h to a column (48 X azineethanesulfonic acid DTNB, 5,5'-dithiobis(2-nitrobenzoicacid; 4.1 cm) of QAE-Sephadex 6-25-120 equilibrated with QAE buffer. HAP, hydroxylapatite; SUS, sodium dodecyl sulfate; Bis-Tris, bis(2- The carnitine octanoyltransferase activity which washed through the column was pooled, dialyzed against HAP(+) buffer, and loaded at a hydroxyethyl)aminotris(hydroxymethyl)methane.
Carnitine Octanoyltransferase flow rate of 24 ml/h onto a column (11.5 X 2.2 cm) of hydroxylapatite which was equilibrated with HAP(-) buffer. The column was then washed with 4 bed volumes of HAP(+) buffer and eluted with400 ml of HAP(+) buffer containing a 10 to 510 mM linear gradient of KPOn, pH 6.8. Fractions containing carnitine octanoyltransferase of a constant specific activity were pooled. Other Methods-Blue Sepharose 4B was synthesized by the method of Bohme et al. (28). Carnitine esters were synthesized as described (29, 30). Isoelectric focusing were performed in a sucrose density gradient after the method of Vesterberg (31). SDS-polyacrylamide gel electrophoresis of purified proteins was performed by the method of Lemmli (32) employing bovine serum albumin, catalase, fumarase, and chicken egg albumin as molecular weight standards. Catalase was assayed by the method of Boudhuin et al. (33). Protein was estimated with bovine serum albumin as standard according to the method of Lowry et al. (34) and, for solutions containing detergent, by the method of Bohlen et al. (35). All chromatographic columns were treated with dimethyldichlorosilane.
LENGTH CARBON CHAIN
CARBON CHAIN LENGTH
Comparison of Carnitine Actyltransferase Actiuities in Beef Liver and Heart Mitochondria-Substrate specificity profies of carnitine acyltransferase activities in liver and heart mitochondria are presented in Fig. 1. Fig. 1,A and E , shows the profiles of intact liver and heart mitochondria. To ascertain whether or not these profiles reflect the activity of one enzyme or more than one enzyme, solubilized liver and heart mitochondria were chromatographed separately on Sephadex G-100. Fractions were assayed for the distribution of carnitine octanoyltransferase activity (B and F ) . Substrate specificity profies of the two carnitine octanoyltransferase-containing peaks separated by gel filtration are pictured in C and D and in G and H for liver and heart, respectively. For each tissue, gel filtration of solubilized mitochondria gave two protein peaks containing transferase activities. Virtually all carnitine acyltransferase activity was released from mitochondria during solubilization (data not shown, but see results from purification of heart mitochondrial transferase activities in Table I). There was no appreciable loss of any of the transferase activities as a resultof gel filtration. The fiist peak of carnitine octanoyltransferase activity to emerge from each Sephadex column contains enzyme or enzyme aggregates of apparent molecular weight greater than 300,000. Based on the analysis by other investigators (36) of the gel filtration behavior of detergent-soluble proteins in the presence of nonionic detergent, we conclude that this peak represents predominately proteins associated with micelles of Triton X-100. For solubilized mitochondrial protein from both tissues, the first peak contains all of the long chain and some of the medium length carnitine acyltransferase activities. All
CARBON CHAIN LENGTH ”
D. FREE PROTEIN FRETION
5 E1 5 0
CARBON CHAIN LENGTH
&E FRACTION & E 7
FIG. 1. Substrate specificity profile of intact and partially fractionated beef heart and liver mitochondrial carnitine acetyltransferases. Carnitine acyltransferase activities were determined by the DTNBassay described under “Methods.” Activities are given as nanomoles min” ml” a t 25°C. The substrate specificity profies of intact liver mitochondria are given in A and intact heart mitochondria in E . Band Fshow the transferase activities determined with octanoyl-CoA of the fractions obtained when liver and heart mitochondria were solubilized in 2% Triton X-100 containing 1M KC1 and chromatographed on a SephadexG-100 column (112 X 4.8 cm) as described under “Methods.” The micellar peak (the first peak) from B and F were analyzed for substrate specificity; these profiles are shown in C and G, respectively. The substrate specificity of the free protein fraction (the second peak) are given in D and H.
TABLEI Summary of the purificationof carnitine acyltransferase proteins from beef heart mitochondria Acetyltransferase
Purification step Recovery units %
Thawed mitochondria 100 1113 Solubilized mitochondria 833 Dialysis supernatant 1.24 0.225 49 547 Dialysis pellet0.114 24 270 Sepharose Blue (No. 31-34)“ 22.3 4.0644496 Blue Sepharose (No. 40-48)* Sephadex G-100” Sephadex G-lWb CM-Sepharose” QAE-Sephadexb Hydroxylapatite” 166 Hydroxylapatiteb
Specific activity -fold
Recovery units %
1 100608 0.95 78 474 55 333 0.63 130 109 181 16 97 23 143 59 19 115 44 431 14 87
Recovery units %
100 179 1 0.0995 0.99 0.0987 161 79 157 1.380.1370 4 7 21 0.0549 0.55 18 8.95 0.891 30 47.0 1.38 19.6 64 128 1.95 8.23 0.819 41.2 4.09 248 10 24.7 373 37.1 2107 20.9 537 53.4 162047.5 39 77
0.0293 90 1.14 0.0335 2.21 0.0646 0.0030
Indicates purification procedures performed on blue Sepharose fractions (No. 31 to 34) containing both carnitine octanoyltransferase and carnitine acetyltransferase activities. Also corresponds to B, C, and D of Fig. 2. Indicates purification procedures performed on blue Sepharose fractions (No.40 to 48) containing both carnitine octanoyltransferase and carnitine palmitoyltransferase activities. Also corresponds to E , F, and G of Fig. 2.
of the short chain transferase activities and the remainder of the medium chain activities eluted with an apparentmolecular weight of60,000. The substrate specificity profilesof micelleassociated activities solubilizedfrom mitochondria ofbeef liver and heart are similar (see Fig. 1, C and G ) as are the profiles of the activities of the 60,000-Mrpeak eluted (Fig. 1, D and H). FRACTION NUMBER Purification of Carnitine Octanoyltransferase Activity from Beef Heart Mitochondria Solubilization-Preliminary studies (data not shown) indicated that beef heart mitochondrial carnitine acetyltransferase activity is mostly soluble. More than 90% is released when the mitochondria were disrupted by freeze-thaw or sonication in the presence of 0.5 to 2.0 M KCI. Once the mitochondria were disrupted, 60 m~ KC1 was required to keep proteins containing carnitine acyltransferase activity in solution. In contrast, carnitine palmitoyltransferase is poorly solubilized by treatment with KC1, requiring instead Triton X-100 (1.0to 2.0%) to release greater than 90% ofthe carnitine palmitoyltransferase into the90,000 X g supernatant fluid. A significant fraction of total carnitine octanoyltransferase activity is found in both supernatant fluid and pellet with either the salt or the detergent extraction procedure. Therefore, in order to maximize the release of carnitine octanoyltransferase activities by one extraction procedure, thawed mitochondria were solubilizedin 2%Triton XFRACTION NUMBER FRACTION NUMBER 100 in the presence of 1 M KCl. The results in Table I show that this combined detergent/salt extraction solubilized 79% of the mitochondrial protein and 75, 78, and 90% of carnitine acetyltransferase, carnitine octanoyltransferase, and carnitine palmitoyltransferase activities, respectively. 0.4 In Dialysis-Dialysis of the solubilized protein solution preparatory to Blue Sepharose chromatography lowers the effective KC1 concentration to 60 n i ~ At . this concentration, ap00 0 2 4 6 8 1 0 proximately half of all solubilized protein precipitated out of MIGRATION DISTANCE (cm) solution, including a fraction of the carnitine acyltransferase FIG. 2. Purification of carnitine octanoyltransferase from activities. Our preliminary experiments (data not shown) inbeef heart mitochondria. In A, transferase was solubilized as described for Fig. 1E and the solution was dialyzed exhaustively and dicated that in dilute ( t l mg of protein/ml) solution, carnitine applied to a column Cibacron Blue Sepharose 4B column (75 X 4.1 acyltransferase activity from solubilizedbeef heart mitochoncm) equilibrated with the blue buffer described under “Methods.” dria is not appreciably precipitated at a KC1 concentration of The sample was applied and the column was washed with 4 bed 60mM. There is measureable precipitation of carnitine acevolumes of blue buffer and theneluted with a 60 to 860 mM KC1 linear tyltransferase and to alesser extent of carnitine octanoyltransgradient in blue buffer. The numbers in parentheses represent car- ferase at 50 mM KCl, with near total precipitation of carnitine nitine acetyltransferase activity (0) and the other numbers represent acetyltransferase, a lesser fraction of total carnitine octanoylcarnitine octanoyltransferase (0) and carnitine palmitoyltransferase activity (0). The solid thin line represents the protein. The KC1 transferase and insignificant loss from the supernatant fluid gradient is indicated by the dots; the units mS represent millisemens of carnitine palmitoyltransferase at 20mM KCI. The precipitation of Significant carnitine acetyltransferase conductivity. In B , the first transferase peak, fractions 31 to 34 of A, were pooled and passed over a Sephadex G-100 column to equilibrate and carnitine octanoyltransferase activities during this prothe protein with the CM buffer described under “Methods.” The cedure may be due to areduction of the actual concentration protein in CM buffer was applied to a column (29 X 4.1 cm) of CM- of KC1 inside the dialysis bag because of charge contributions Sepharose CL-GB and washed with 4 bed volumes of CM buffer. The activity was then eluted by 60 to 560 mM linear gradient of KC1 in of the protein. The protein pelleted after dialysis contains a CM buffer. The symbols are identical with those described in A. The third of the total carnitine acetyltransferase activity, a smaller open circles represent carnitine octanoyltransferase activity. Frac- fraction of the carnitine octanoyltransferase and negligible tions 37 to 41 of B were pooled and diluted with an equal volume of carnitine palmitoyltransferase activity (see Table I). The ratio 20mM K2HP04,pH 6.6. The solution was applied to a hydroxylapatite of Carnitine octanoyltransferase to carnitine acetyltransferase column (13 X 2.2 cm) which wasequilibrated with the HAP(-) buffer activity assayed in the pellet is less than 0.5. described under “Methods.” The column was washed with 4 bed The dialysis supernatant fluid which was loaded directly volumes of HAP(+) buffer and then eluted with a 10 to 510 mM KHP04,pH 6.8, linear gradient. The symbols are identical with those onto the Blue Sepharose column contains 67,72, and 96%, described in B. Fractions 17 to 19 (the peak ones) of C were pooled respectively, of the initially solubilized carnitine acetyltransand analyzed for purity by SDS-polyacrylamide gel electrophoresis ferase, carnitine octanoyltransferase, and carnitine palmitoylas by the method of Laemmli, D. T represents the top of the gel and transferase activities. w
B represents the bottom of the gel. The DYE was bromphenol blue. In E , fractions 40 to 48 of A were pooled and passed through a Sephadex G-100column equilibrated with the QAE buffer described under “Methods.” The protein solution was then passed through a column (48 X 4.1 cm) of QAE-Sephadex 6-25-120 equilibrated with QAE Buffer. The open circles represent carnitine octanyltransferase activity. Fractions 23 to 29of E were dialyzed exhaustively against several changes of HAP(+) buffer and then applied to a column (11.5 X 2.2 cm) of hydroxylapatite and processed as described in C. In F,
the open circles represent carnitine octanyltransferase, but the solid line represents protein. For the peak fractions protein and transferase activity were superimposable. In G, the peak fractions (12 to 14 of F) were combined and subjectto SDS-polyacrylamide electrophoresis as described for D . The data in Fig. 2 are summarized in Table I where Footnote a represents purification described in B, C, and D and Footnote b represents the purification shown in E, F, and G.
Results of Column Chromatography-Greater than 95% of carnitine octanoyltransferase activity present in the soluble heart mitochondrial protein fraction after dialysis against 60 n u v ~KC1 is retained on Blue Sepharose. This activity elutes as two well separated peaks after administration of a linear KC1 gradient (see Fig. 2 A ) . Carnitine acetyltransferase and carnitine palmitoyltransferase activities are completely separated; each co-elutes with one of the carnitine octanoyltransferase peaks. Acyl-CoA hydrolase activities are partially removed during the dialysis and are nearly totally absent in the Blue Sepharose peaks of carnitine acyltransferase activity (data not shown). Both carnitine acetyltransferase and carnitine palmitoyltransferase are purified about 20-fold by this step. Purification of theCarnitineOctanoyltransferaseand Carnitine Acetyltransferase-Containing Blue Sepharose Peak (Fractions 31 to 34)-A further 20-fold purification of the carnitine acetyltransferase/carnitine octanoyltransferase activity is effected by CM-Sepharose ion exchange chromatography (Fig. 23). Chromatography of the peak from CMSepharose on hydroxylapatite yields nearly superimposable peaks of protein and carnitine octanoyltransferase activity (see Fig. 2 0 . Fractions from the hydroxylapatite column of constant specific activity were pooled and a sample was analyzed for purity by SDS-polyacrylamide gel electrophoresis. This precedure revealed the presence of one major protein band and two discernible contaminating proteins of much lesser abundance. The densitometer scan of the gel stained for protein (Fig. 2 0 ) indicates that the purified carnitine acetyltransferase/carnitine octanoyltransferase protein is approximately 95% pure. Purification of theCarnitineOctanoyltransferase and Carnitine Palmitoyltransferase-containing Blue Sepharose Peak (Fractions 40 to 48)-A 2-fold purification was achieved during Sephadex G-100 chromatography of the pooled carnitine palmitoyltransferase/carnitine octanoyltransferase fractions from the Blue Sepharose column. A further 9-fold purification was obtained by passing the protein solution through a column of QAE-Sephadex at pH 9.7 (see Fig. 2E). Hydroxylapatite chromatography of the protein effluent from the QAE-Sephadex column revealed a sharp, symmetrical peak of constant specific activity (Fig. 2 0 . The carnitine palmitoyltransferase/carnitine octanoyltransferase protein is estimated to be greater than 95% pure by SDS-polyacrylamide gel electrophoresis (see Fig. 2G). Molecular Weight Determinations-Comparison of the elution of carnitine acetyltransferase/carnitine octanoyltransferase and carnitine palmitoyltransferase/carnitine octanoyltransferase activities after Sephadex G-200 chromatography with those of catalase and hemoglobin indicate apparent
FIG. 3. Chromatography of solubilized carnitine octanoyltransferase from beef heart mitochondria. Thesolid circles representcarnitineoctanoyltransferase activity measuredby the DTNB assay. The thin solid line is protein and the arrows indicate the peak positions of molecular weight markers.
0.5 RELATIVE MOBILITY
FIG. 4. Weight determination of carnitine acyltransferases from beef heart mitochondria. The peak fractions from Fig. 2, C and F, were subjectedto the polyacrylamide gel(10%gel) electrophoresis as described by Laemmli (32). CPT represents the carnitine palmitoyltransferase/carnitine octanoyltransferase protein andCAT represents the carnitine acetyltransferase/carnitine octanoyltransferase protein describedunder “results.”BSA is bovine serine albumin.
FIG. 5. Isoelectric point determination of the carnitineacyltransferases from beef heart mitochondria. Aliquots of the pooled purified proteins described for Fig. 4 were subjected to isoelectric focusing (see “Methods”).
the two transferases of 60,500 and
510,000, respectively (see Fig. 3). Subunit molecular weight
determination by SDS-polyacrylamide gel electrophoresis using bovineserum albumin, catalase, fumarase, and ovalbumin as standards yields values of 62,600 and 67,000, respectively, for beef heart mitochondrial carnitine acetyltransferase and carnitine octanoyltransferase (Fig. 4). Isoelectricpoint-Aliquots of the two purifiedproteins were combined and subjected to isoelectric focusing in a sucrose density gradient. The results shown in Fig. 5 give PI values of 8.20 for the carnitine acetyltransferase protein and 8.05 for carnitine palmitoyltransferase of mitochondria from beef heart. Isoelectric Focusing: Approach to Equilibrium-In order to investigate the possibility that the micelle-associated carnitine octanoyltransferase and carnitine palmitoyltransferase activities are associated with two distinct proteins which may have co-purified during our purification procedure, the purified carnitine palmitoyltransferase/carnitine octanoyltransferase protein was subjected to isoelectric focusingon a pH 3
to 10 gradient and assayed before the migration of the protein($ had reached equilibrium. Similar but different proteins which have indistinguishable PI values may show different patterns in approaching isoelectric equilibrium. The results presented in Fig. 6 show that the two activities are superimposable, suggesting that both activities reside ina single enzyme. Substrate Specificity-The substrate specificities of the two carnitine acyltransferases purified from beef heart mitochondria are presented in Table 11. Included are results obtained
L ~ N G T H
FIG. 7. Comparison of carnitine acyltransferase (CAT) activity of beef heart mitochondria with the purifiedtransferase enzymes. See “Substrate Specificity” under “Results” for details of this experiment. A, whole mitochondria; B, purified proteins. COT, carnitine octanoyltransferase.
for commercial pigeon breast muscle mitochondrial carnitine acetyltransferase (Sigma) assayed under identical conditions. FRACTION NUMBER For purposes of comparison, literature datafor a preparation FIG. 6. Approach to isoelectric equilibrium of the carnitine of purified carnitine palmitoyltransferase are listed, showing palmitoyltransferase/carnitine octanoyltransferaseenzyme. relative rates of the reverse reaction. The carnitine palmiA-A represents carnitine octanoyltransferase and the ) ” . toyltransferase/carnitine octanoyltransferase enzyme has represents carnitine palmitoyltransferase activity. The pooled peak fractions of Fig. 2 F were used. The numbers inparentheses represent high medium chain transferase activity in the forward direction in contrast to the reverse reaction. the palmitoyltransferase. In Fig. 7, the substratespecificities of the purified carnitine acyltransferase proteins(lowerpanel) arecompared with that TABLEI1 obtained initially for the thawed suspension ofbeef heart Relative activities of beef heart mitochondrial carnitine mitochondria. The carnitine acetyltransferase and carnitine acyltransferases The carnitine acetyltransferase/carnitine octanoyltransferase palmitoyltransferase enzyme profiies have been normalized to (CAT/COT) activity represents the peak fractions from Fig. 2C and equalize the heights of C-2 and C-16 in upper and lower the carnitine palmitoyltransferase/carnitine octanoyltransferase are portions of the figure. The cross-hatching indicates overlapthe peak fractions from Fig. 2F. The activity of acetyltransferase ping of medium chain activities. 0.00
(CAT) was arbitrarily set at 100 for CAT/COT and the activity of palmitoyltransferase(CPT) was set at 100for the CPT/COT enzymes so relative activities could be expressed. Carnitine acetyltransferase Acyl residue
Acetyl Butyryl Hexanoyl Octanoyl Decanoyl Lauryl Myristoyl Palmitoyl
Forward m a y
100 109 51 27 5 1 0 0
100 98 41 20 3
0 0 0
100 88 64 54 26 2 0 0
100 86 63 50 20 2
0 1 43 113 226 166 100 100
0 0 3 8
38 31 79 100
0 0 0 2 26 31 99
Our initial studies on acyl coenzyme A - carnitine acyltransferase profiles with beef liver and heart mitochondria agree with the reports by Solberg (12, 13) who used mitochondria from calf liver and from three rattissues. Substrate specificity profiies for the forward reaction show a peak of carnitine acyltransferase activity which indicates the presence of a medium chain-specific transferase protein or carnitine octanoyltransferase inmitochondria from both tissues. Gel filtration of the detergent-solubilized mitochondria separates short and long chain acyltransferase activities, with each fraction contributing carnitine octanoyltransferase activity. In beef liver, the majority of carnitine octanoyltransferase activity chromatographs with carnitine palmitoyltransferase while in heart, in which carnitine acetyltransferaseis much more abun-
dant relative to long chain transferase activity, the majority of carnitine octanoyltransferase is found with the shortchain transferase. The substratespecificity profiles ofthe detergent micelle-associated protein from both mitochondrial sources reveal a peakof transferase activity with medium chain length CoA thioesters, with palmitoyl transferase activity less than half that for decanoyl-CoA (see Fig.1, C and G). A medium chain-specific carnitine acyltransferase in calf heart had previously been suggested by the findings of Kopec and Fritz ( 14).
9867 TA%LE 111 Amino acid compositions of beef heart mitochondrial carnitine acyltransferases Values for each m i n o acid are mole per cent of total amino acid detected after 24-h hydrolysis; nd = not determined. The carnitine acetyItransferase/carnitine octanoyltransferase (CAT/COT) is from the pooled peak fractions from Fig. 2C and the carnitine palmitoyltransferase/carnitine octanoyltransferase (CPT/COT) is from the pooled peak fractions from Fig. 2F. 7.72 4.49
8.99 4.70 6.97 11.36 6.23 5.04 8.65 6.71 3.13 4.40 10.77 4.83 5.09 2.74 5.67 4.74 nd nd
Because of the high carnitine octanoyltransferase activity Thr in heart mitochondria, carnitineoctanoyl transferase-containSer 6.33 ing proteins were purified to near homogeneity from beef Glx 9.46 Pro 5.95 heart mitochondria. The two proteins obtained can account 5.26 G~Y for nearly all carnitine octanoyltransferase activity initially 9.76 Ala detectedin the intact beef heart mitochondria. Carnitine 5.03 Val acyltransferase activities which were precipitated during diMet 2.53 alysis of solubilized mitochondrial protein preparatory tocol- 5.00 Ile umn chromatography were not further purified. These precipLeu 11.52 itated proteins contained 33% of the initially solubilized car4.89 Tyr 8.01 Phe nitine acetyltransferase, 28% of carnitine octanoyltransferase, 2.40 His and 4% of total carnitine palmitoyltransferase activities. It is LYS possible that this fraction contains a carnitine octanoyltrans- 5.70 5.94 -4% ferase protein different from those finally purified from the nd CYS dialysis supernatant fluid. However, based on the carnitine nd TrP octanoyltransferase activities present in the pure carnitine acetyltransferase/carnitine octanoyltransferase and carnitine palmitoyltransferase/carnitine octanoyltransferase proteins, protein assayed under identical conditions shows a strikingly more than half (17 of the 28%) of total carnitine octanoyltrans- similar pattern (see Table 11).For the forward reaction, the ferase found in this fraction canbe accounted by the carnitine purified carnitine palmitoyltransferase/carnitine octanoylacetyltransferase/carnitine octanoyltransferase and carnitine transferase protein gives a markedly different substrate specpalmitoyltransferase/carnitine octanoyltransferaseproteins ificity profiie witha preference for medium chainlength which are partially precipitated by this procedure. Although substrates. (An investigation of the apparent difference bethe remaining 11%of carnitine octanoyltransferasecould rep- tween substrate specificity in the forward and reverse direcresent a separate acyltransferase protein, it should be noted tions is described in an accompanying paper.) Based on the that the precipitate contained all of the acyl-CoA hydrolase similarity of the substrate specificity of the carnitine palmiactivities (data notshown) so that theblanks measuredin this toyltransferase/carnitine octanoyltransferase protein to that fraction are very high and the error for the transferase activ- of carnitine palmitoyltransferase purified by other investigaities reported in the precipitate is correspondingly large. tors, it is concluded that thebeef heart mitochondrial carnitine The two carnitine octanoyltransferase-containing proteins palmitoyltransferase/carnitine octanoyltransferase protein is are separated by column chromatography on Cibacron Blue a carnitinepalmitoyltransferase (EC 18.104.22.168), also not a novel Sepharose and are further purified to near homogeneity by carnitine octanoyltransferase enzyme. CM-Sepharose and QAE-Sephadex chromatography, respecHoppel (37) has reviewed the evidence for multiple forms tively, for the acetyl- and palmitoyltransferase-containing en- of mitochondrial palmitoyltransferase. Mitochondria, includzymes (see Fig. 2 4 ) . The two proteins have similar isoelectric ing beef heart mitochondria, contain two forms of carnitine points and amino acid compositions, although the carnitine palmitoyltransferase activity, carnitinepalmitoyltransferase I palmitoyltransferase/carnitine octanoyltransferase protein is associated with the cytosol side of the inner membrane and slightly richer in hydrophobic amino acid residues (see Table carnitine palmitoyltransferase I1 located on the matrix side of III), as expected from its affinity for the detergent Triton X- the inner membrane.Although carnitine palmitoyltransferase 100. The native molecular weight (60,500) and PI (8.2) of the I and carnitine palmitoyltransferase 11, exhibit different kicarnitine acetyltransferase/carnitine octanoyltransferase pro- netic properties in intact mitochondria (37-39), it hasnot been tein are similar to thosereported by Markwell (15) M, = determined whether carnitinepalmitoyltransferase I and car59,000, PI = 8.3) for carnitineacetyltransferasespartially nitine palmitoyltransferase I1 are different proteins or whether purified from rat liver peroxisomes and microsomes. A com- theyare identicalproteins which exhibit different kinetic parison of the relativeactivities of the purified carnitine properties due to the localized environment of the membrane. acetyltransferase/carnitine octanoyltransferase protein using In a study of the same question in regard to apparentmultiple acylcarnitine and acyl-CoA substrates with those of commer- forms of carnitine acetyltransferase, Edwards et al. (40) concial purified pigeon breast muscle carnitine acetyltransferase cluded that the differences observed could be attributed to (see Table 11) leads us to conclude that this beef heart mito- differentenvironments for the same enzyme, i.e. different chondrial protein is a carnitine acetyltransferase (EC 22.214.171.124) enzyme forms are observed when the carnitine acetyltransferand not a novel carnitine octanoyltransferase enzyme. ase proteinis associated with a membraneas opposed to being Carnitine palmitoyltransferase has been purified to homo- free in solution. It is known that membrane faces are different geneity from calf liver mitochondria by Kopec and Fritz (14) on the cytosol and matrix side of the inner membrane (41). and therelative rates of the reverse reaction using acylcarni- Bergstrom and Reitz have reported that carnitine palmitoyltine substrates are reported therein. A comparison of these transferase I and carnitine palmitoyltransferase I1 isolated data with those obtained for the beef heart mitochondria from rat liver mitochondria show no kinetic differences once carnitine palmitoyltransferase/carnitine octanoyltransferase theyare freed from membrane vesicles (7). Our dataare
consistent with a single protein exhibiting different properties depending on its membrane environment. Although no attempthas been made to determine the relative amounts of subsarcolemmal and interfibrillar mitochondria released by our mitochondrial isolation procedure, the finding of a single protein with carnitine palmitoyltransferase activity also implies that the carnitine palmitoyltransferase of the two types of mitochondria found in heart (42) are very similar if not identical. If heart mitochondria contain one carnitine palmitoyltransferase protein which has different properties, depending on its membrane environment, then reconstitution studies and studies comparing the membrane associated enzyme to the solubilized enzyme are required for a betterunderstanding of the role of carnitine palmitoyltransferase in regulating and facilitating of the /3 oxidation of long chain fatty acids. In the accompanying paper, some kinetic properties and substrate specificities of the purified beef heart mitochondrial carnitine palmitoyltransferase/carnitine octanoyltransferase protein are described. 1. 2. 3. 4. 5.
7. 8. 9. 10. 11.
12. 13. 14. 15.
16. Choi, Y. R., Fogle, P. J., Clarke, P. R. H., and Bieber, L. L. (1977) J. Biol. Chem. 252, 7930-7931 17. McGarry, J. D., and Foster, D.W. (1974) J. Biol. Chem. 249, 7984-7990 18. Williamson, J. R., Walajtys-Rode, E., and Coll, K. E. (1979) J. Biol. Chem. 254, 11511-11520 19. Walajtys-Rode, E., Coll, K. E., and Williamson, J . R. (1979) J . Biol. Chem. 254, 11521-11529 20. Choi, Y. R., h d Bieber, L. L. (1977) Anal. Biochem. 79,413-418 21. Lazarow, P. B., and de Duve, C. (1976) Proc. Natl. Acad. Sci. U. S. A. 73,2043-2048 22. Murphy, P. A., Krahling, J. B., Gee, R., Kerb, L. R., and Tolbert, N. E. (1979) Arch. Biochem. Biophys. 193,179-185 23. Lazarow, P. B. (1978) J. Biol. Chem. 253, 1522-1528 24. Osmundsen, H., Neat, C. E., and Norum, K. R. (1979) FEBS Lett. 99,292-296 25. Clarke, P. R. H., and Bieber, L. L. (1979) Fed. Proc. 38,352 26. Bieber, L. L., Abraham, T., and Helmrath, T. (1972) Anal. Biochem. 50,509-518 27. Srere, P. A., Seubert, W., and Lynen, F. (1959) Biochim. Biophys. Acta 33, 313 28. Bohme. H. J.. KoDDerschaaer. G.. Schulz.. J... and Hofman. E. (1972) J. Chromatogr. 69;2&-214 REFERENCES 29. Bohmer, T., and Bremer, J. (1968) Biochim. Biophys. Acta 152, Bremer, J. (1962) J. Biol. Chem. 237,2228-2231 559-567 Fritz, I. 3. (1963) Adu. Lipid Res. 1, 285-334 30. Bremer, J. (1968) Biochem. Prep. 12,69-73 31. Vesterberg, 0. (1971) Methods Enzymol. 22,389-412 Ramsay, R. R., and Tubbs, P. K. (1975) FEBS Lett. 54,21-25 32. Laemmli, U. K. (1970) Nature (Lond.)227,680-685 Pande, S. V. (1975) Proc. Natl. Acad. Sci.U. S. A. 72, 21-22 West, D. W., Chase, J. F. A., and Tubbs, P. K. (1971) Biochem. 33. Boudhuin, P., Beaufay, H., Rahman-Li, Y., Sellinger, 0. Z., WatBiophys. Res. Commun. 42,912-918 tiaux, R., Jacques, P., and de Duve, C. (1964) Biochem. J. 92, 179-184 Hoppel, C. L., and Tomec, R. J. (1972) J. Biol. Chem. 247, 83234. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. 841 (1951) J. Biol. Chem. 193,265-275 Bergstrom, J. D., and Reitz, R. C. (1980)Arch. Biochem. Biophys. 204,71-79 35. Bohlen, P., Stein, S., Dairman, W., and Undenfriend, S. (1973) Arch. Biochem. Biophys. 155,213-220 Pearson, D. J., and Tubbs, P. K. (1967) Biochem. J. 105.953-963 Snoswell, A. M., and Henderson, G. D. (1970) Biochem. J. 119, 36. Springer, T. A., Mann, D. L., DeFranco, A. L., and Strominger, J. L. (1977) J. Biol. Chem. 252,4682-4693 59-65 Choi, Y. R., Clarke, P. R. H., and Bieber, L.L. (1979) J. Biol. 37. Hoppel, C. L. (1979) Enzymes Biol. Membranes 2, 119-143 Chem. 254,5580-5583 38. Edwards, M. R. (1977) Prod. Aust. Biochem. SOC.10, 18 Van HensGergh, V. W., Veeskamp, S. H., Engelen, P. M. J., and 39. McGarry, J. D., Leatherman, G. F., and Foster, D. W. (1978) J. Ghysen, W. J. (1978) Biochem. Med. 20, 115-124 Biol. Chem. 253,4128-4136 Solberg, H. E. (1971) FEBS Lett. 12, 134-136 40. Edwards, Y. H., Chan, J. F. A., Edwards, M. R., and Tubbs, P. K. (1974) Eur. J. Biochem. 46,209-215 Solberg, H. E. (1972) Biochim. Biophys. Acta 280,422-433 41. Rocker, E. (1970) Essays Biochem. 6,1-22 Kopec, B., and Fritz, I. B. (1971) Can. J. Biochem. 49,941-948 Markwell, M. A. K., Tolbert, N. E., and Bieber, L. L. (1976)Arch. 42. Palmer, J. W., Tandler, B., and Hoppel, C. L. (1977) J. Biol. Chem. 252,8731-8739 Biochem. Biophys. 176,479-488