Fatty Acyl Coenzyme A-sensitive Adenine Nucleotide Transport in a Reconstituted Liposome System *

The adenine nucleotide translocase was purified from bovine heart mitochondria and incorporated into membranes of phospholipid liposomes. The rate of transport of the adenine nucleotides was competitively inhibited by oleoyl coenzyme A with an approximate Ki of 1.0 microM. Significant inhibition was limited to those fatty acyl coenzyme A esters which are carnitine dependent for their oxidation in isolated mitochondria. Octanoyl coenzyme A was almost completely inactive as was palmitic acid and palmitoyl carnitine. By comparing the inhibitory characteristics of carboxyatractylate and bongkrekic acid with those of oleoyl-CoA, it was determined that the fatty acyl-CoA esters could produce inhibition whether the carrier was inserted into the liposome in either the conventional (65%) or reverse (30%) orientation. The results demonstrate that the interaction of long chain fatty acyl-CoA esters with the ADP/ATP carrier in a purified reconstituted system mimics their effects with isolated mitochondria and inverted submitochondrial particles. In general, these findings are consistent with the role of acyl-CoA esters acting as natural ligands and biological effectors of the translocator.

The adenine nucleotide translocase was purified from bovine heart mitochondria and incorporated into membranes of phospholipid liposomes. The rate of transport of the adenine nucleotides was competitively inhibited by oleoyl coenzyme A with an approximate Ki of 1.0 p~.
Significant inhibition was limited to those fatty acyl coenzyme A esters which are carnitine dependent for their oxidation in isolated mitochondria. Octanoyl coenzyme A was almost completely inactive as was palmitic acid and palmitoyl carnitine. By comparing the inhibitory characteristics of carboxyatractylate and bongkrekic acid with those of oleoyl-CoA, it was determined that the fatty acyl-CoA esters could produce inhibition whether the carrier was inserted into the liposome in either the conventional (65%) or reverse (30%) orientation. The results demonstrate that the interaction of long chain fatty acyl-CoA esters with the ADP/ATP carrier in a purified reconstituted system mimics their effects with isolated mitochondria and inverted submitochondrial particles. In general, these findings are consistent with the role of acyl-CoA esters acting as natural ligands and biological effectors of the translocator.
Long chain fatty acyl coenzyme A esters are potent reversible inhibitors of adenine nucleotide translocation in mitochondria (l), submitochondrial particles with inverted sidedness (2), and mitoplasts prepared by lubrol extraction (3). Because of their strikingly similar effects to those of the classical inhibitors atractylate and bongkrekic acid, long chain fatty acyl-CoA esters have been proposed to function as natural ligands for the ADP/ATP carrier (4). However, not aLl investigators agree with this concept and some regard the acyl-CoA esters to be only nonspecific inhibitors of the translocase (5, 6). In view of the increasing amount of evidence which indicates that the adenine nucleotide translocase is the rate-limiting step for oxidative phosphorylation ( 6 4 , it is very important to document the biochemical specificity of the acyl-CoA effect if the ligands are to be seriously considered as biological effectors of the carrier. Analytical techniques for the isolation and purification of the ADP/ATP carrier from bovine heart mitochondria and its subsequent reconstitution in a liposome system have recently been developed (9)(10)(11). The use of these procedures offers a unique opportunity to establish rigorous criteria with * This work was supported by United States Public Health Service Grants GM-14033 and GM-26227. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
* To whom reprint requests should be addressed at the University of Wisconsin, Department of Medicine, Madison, WI 53706. which to test the specificity and sensitivity of the adenine nucleotide translocase to the fatty acyl-CoA esters. The present communication provides results which demonstrate that the reconstituted carrier is sensitive to the acyl-CoA in a specific manner. Furthermore, the interaction of the ligand with the receptor corresponds appropriately with the asymmetric orientation of the carrier in the liposome membrane.
Heavy beef heart mitochondria were prepared according to the procedure of Green et al. (12) and protein was determined by the method of Lowry et al. (13). The ADP/ATP carrier was purified according to published procedures (9-11) with slight modification. A mitochondrial pellet containing 50-60 mg of protein was dissolved in 2.5% Triton X-100,100 mM NaZS04,lO mM Tricine-KOH, pH 7.4, and 1.0 mM EDTA at a concentration of 7 mg/ml. The suspension was incubated at 0 "C for 20 min and then centrifuged at 20,000 X g for 10 min. A 1-ml aliquot of the supernatant was placed on a hydroxyapatite column (6 cm x 0.7 cm) that had been equilibrated with an elution buffer containing 0.5% Triton X-100, 100 mM Na2S04, 10 mM Tricine-KOH, pH 7.4, and 0.1 mM EDTA, and the carrier protein which was eluted in the pass through fraction within 15 min was monitored by protein determinations.
The purified ADP/ATP carrier protein was incorporated into liposomes composed of phosphatidylethanolamine/phosphatidylcholine/cardiolipin in a ratio of 7022:8 (9)(10)(11). A chloroform solution of the phospholipids was dried under nitrogen at room temperature and then Vortex suspended in a medium containing 20 mM ATP, 100 mM NaCI, 0.5 mM MgC12, and 10 mM Tricine-KOH, pH 7.5, at a concentration of 33 mg/ml. The phospholipid suspension was sonicated at 40-50 watts using a Branson Sonifier with a microtip for 15 min in an ice bath and centrifuged at low speed for 10 min. Approximately 200-300 pg of carrier protein in 0.2-0.4 ml of Triton X-100 solution was added to the sonicated liposome suspension and incubated at 0 "C for 20 min. The mixture was then subjected to a second sonication of 15-20 s in an ice bath. External ATP was removed by passage of the reconstituted carrier through an anion exchange column (25 cm x 0.7 cm) (AG 1-X8, 100-200 mesh, formate form) that had been equilibrated with 136 mM glycerol solution (14). The fractions containing the proteoliposomes which were eluted with the glycerol buffer were pooled and used for the transport studies.
Adenine nucleotide transport was measured by a standard radioactive forward exchange assay (1). In a typical assay, 0.2-0.5 ml of the reconstituted carrier was added to a 1.0-ml incubation media containing 250 mM sucrose, 0.5 mM EDTA, and 20 mM 4-morpholinepropanesulfonic acid, pH 7.0, Vortex dispersed, and incubated for 4 min at room temperature. The reaction mixture was then incubated for an additional 2 min with or without the inhibitors and the forward exchange initiated by addition of 3.3 pCi of [I4C]ADP at a final the Reconstituted ADP/ATP Carrier concentration of 10 PM. After incubation for 15 min a t room temperature, the exchange was terminated by addition of 10 PM carboxyatractylate and 10 p~ bongkrekic acid and the external ["CIADP removed by passage of the reaction mixture through an anion exchange column (6 cm x 0.7 cm) (AG I-XS, 100-200 mesh, formate form) which was equilibrated with 136 mM glycerol. The proteoliposomes were eluted with 136 mM glycerol and 1.5-ml aliquots were collected until the liposomes were quantitatively recovered. The radioactivity of a 0.2-ml aliquot dissolved in 10 ml of Bray's solution (15) was determined using a Tri-Carb liquid scintillation spectrometer. Sodium dodecyl sulfate gel electrophoresis was performed on 10% acrylamide and 0.27% bisacrylamide gels using the discontinuous buffer system of Neville (16) as previously described (17).

RESULTS
In order to prevent denaturation of the carrier protein following its extraction from the inner mitochondrial membrane, it is characteristically preloaded with the tight binding ligand, carboxyatractylate, which confers almost complete protection during the purification (18). This procedure is not applicable when the objective is to obtain a protein which carries out active transport. Conditions have, therefore, been modified and with the use of a very small hydroxyapatite column, a more rapid purification has been accomplished permitting a considerable amount of the unliganded carrier to be actively reconstituted into the liposome system (11). The carrier protein purified by hydroxyapatite chromatography as described under "Materials and Methods" is a dimer of 60,000 daltons (18). Fig. 1 is a representative example of the purified preparation used in these experiments which has been subjected to sodium dodecyl sulfate gel electrophoresis. A distinct protein band with the mobility equivalent to a molecular weight of 30,000 is the subunit of the carrier (18). Using the standard more lengthy purification procedure, an almost homogeneous protein (70-8056 pure) can be obtained by chromatography on hydroxyapatite (17,18).
The results in Table I demonstrate rates of adenine nucleotide transport in the reconstituted liposome system which are comparable to values reported (9)(10)(11)  series of experiments is the one showing that adenine nucleotide translocation is also extremely sensitive to oleoyl-CoA. The concentration curve for oleoyl-CoA inhibition of adenine nucleotide transport in the liposome system shown in Fig. 2 is almost identical with that obtained with either mitochondria or submitochondrial particles (2). Furthermore, the inhibition is competitive with ADP as it is in isolated   mitochondria (2, 3). The K , for ADP of 14 p~ obtained in these experiments is similar to the value recently reported for the reconstituted carrier system (11,19). A K; for oleoyl-CoA was calculated to be 1.0 p~. The value is similar to that reported for rat liver mitochondria (3) but lower than the figure given for palmitoyl-CoA inhibition of a reconstituted carrier protein prepared from cholate-extracted bovine heart mitochondria (20). Fatty acyl-CoA inhibition of adenine nucleotide translocation in isolated mitochondria is limited to those thioesters which are carnitine dependent for their subsequent oxidation (2,3). In fact, inhibition is released upon addition of carnitine to the incubation medium which permits the transacylation of the acyl-CoA to the acylcarnitine ester by the carnitine palmitoyltransferase enzyme (2,3). In the experiments shown in Table 11, all of the acyl-CoA esters tested except octanoyl-CoA produced a strong inhibition of adenine nucleotide transport, again illustrating the marked similarity of the responses in isolated mitochondria and the reconstituted liposome sys-

TABLE I Effect of inhibitors on the rate of ADP transport in the reconstituted ADP/ATP carrier
The assay for transport was carried out as described under "Materials and Methods." The amount of protein in the reconstituted carrier was 0.2 pg and the concentration of the inhibitors in the reaction medium 10 p~ each.  tem. Furthermore, palmitoyl carnitine and palmitic acid were completely inactive in this reconstituted system, thus providing additional evidence for the specificity of the action of the acyl-CoA esters. Results from other types of experiments such as these using spin-labeled acyl-CoA esters have led to similar conclusions (2 1).
The results comparing the effects of the various inhibitors (Table 111) indicate that the majority of the carrier molecules in the liposome maintain the same asymmetric arrangement as they do in isolated mitochondria. Atractylate and carboxyatractylate which are highly hydrophilic are not membrane permeable, whereas, bongkrekic acid is lipophilic and can permeate phospholipid membranes (2, 10). These facts partially explain the asymmetric binding and inhibition of the ADP/ATP carrier by the diverse ligands. In isolated intact mitochondria, atractylate and carboxyatractylate inhibit the translocase from the cytosolic side of the inner membrane, whereas bongkrekic acid inhibits from the matrix side. Shertzer and Racker (22) were the first to observe that unlike intact mitochondria the submitochondrial particles with inverted sidedness were more sensitive to bongkrekic acid and insensitive to atractylate. This finding indicated that the asymmetry of inhibition was not merely due to the degree of penetration of the ligands but intrinsic to the carrier itself. It was subsequently shown that whereas each of the two specific receptors or conformations of the carrier recognized either atractylate or bongkrekic acid, both c0ul.d interact with long chain fatty acyl- CoA esters (2, 23, 24). In the present experiments, a high concentration of carboxyatractylate inhibited only 65% of the transport activity, whereas bongkrekic acid at a much lower concentration was effective in producing almost total inhibition. These results can be interpreted to indicate that following insertion into the liposome, ?h of the carrier molecules are in the conventional orientation, whereas about Yi are in the reverse orientation. Carboxyatractylate which is nonpenetrant can only interact with the carrier in the conventional orientation. Bongkrekic acid which can penetrate the membrane is able to inhibit the reconstituted carrier in either orientation. Most important is the finding that in the liposome system oleoyl-CoA inhibits adenine nucleotide transport with the carrier inserted in either orientation. In this case, the acyl-CoA ligand appears to act like bongkrekic acid in producing almost complete inhibition. However, since the acyl-CoA does not penetrate the lipid bilayer (25), it interacts with the carrier both in the conventional orientation and in the reverse onentation from the outer side of the liposome membrane. The experiment showing more complete inhibition of the adenine nucleotide transport with various combinations of the inhibi-36

DISCUSSION
The present results provide evidence that the interaction between the ADP/ATP carrier and long chain fatty acyl-CoA esters in a purified reconstituted system is identical with that observed in isolated mitochondria and submitochondrial particles. An important finding brought out in previous studies (2, 23, 24) was that, unlike atractylate and bongkrekic acid which bind and inhibit the carrier asymmetrically, long chain fatty acyl-CoA esters recognize receptor sites on both the cytosolic and matrix side of the inner mitochondrial membrane. Moreover, on the cytosolic side of the membrane, the acyl-CoA esters resemble atractylate in their kinetic effects, while on the matrix side, they resemble bongkrekic acid (4). Whereas the present results might seem predictable, it was, nevertheless, reassuring to find that they confirm the fact that in a reconstituted system as in the natural environment, the acyl-CoA esters were still able to recognize either conformation of the receptor as dictated by the orientation of the carrier in the membrane.
The role of the ADP/ATP carrier in energy-linked respiration is an important factor in assessing the potential significance of the present study. Whereas there has been disagreement in the past, more recent evidence (6)(7)(8) favors the probability that the adenine nucleotide translocase is the ratelimiting step in oxidative phosphorylation. This condition implies that the carrier, like other rate-limiting enzymes, must be carefully regulated. It also seems reasonable to consider the possibility that the receptors on the carrier fortuitously recognize atractylate and bongkrekic acid, and that the fatty acyl-CoA esters may represent the natural ligands and biological effectors of the translocator. The present study supports this concept and future efforts w i l l be directed to providing evidence for the existence of physiological or pathophysiological conditions under which the ADP/ATP carrier is regulated and, in turn, affects energy-linked respiration.