In Vitro Import into Mitochondria of the Precursor of Mitochondrial Aspartate Aminotransferase*

The import of the precursor of mitochondrial aspar- tate aminotransferase was reconstituted in vitro with isolated mitochondria thus corroborating the earlier conclusion of a post-translational uptake. The higher M, precursor was synthesized in a reticulocyte lysate programmed with free polysomes from chicken liver. After incubation with intact mitochondria from chicken heart about 50% of the precursor was converted to the mature form in a time-dependent process, its rate being a function of the amount of mitochondria added. The same amount of precursor was processed to the mature form on addition of a mitochondrial extract. No conversion to the mature enzyme took place when the precursor was incubated with intact mitochondria in the presence of the uncoupling agent carbonyl cyanide m-chlorophenylhydrazone or of the chelator o-phenanthroline which penetrates the mito- chondrial inner membrane. In contrast, the chelator bathophenanthroline disulfonate which does not dif- fuse into the mitochondrial matrix did not inhibit the appearance of the mature form. The results indicate that that precursor must pass through an energized inner mitochondrial membrane before it is processed by a chelator-sensitive protease in the mitochondrial matrix. Excess mature mitochondrial aspartate ami- notransferase did not compete with the precursor for its uptake into mitochondria. Mature mitochondrial aspartate aminotransferase is an aa-dimer with M, = 2 X 45,000. Both the precursor synthesized in a rabbit reticulocyte lysate and the precursor accumulated in the cytosol

(pre-mAspAT, A M r -3,000). Several lines of evidence have indicated that the precursor is imported into the mitochondria post-translationally: 1) pre-mAspAT was found to be exclusively synthesized on free polysomes (5); 2) a small pool of pre-mAspAT was detected in the cytosol of cultured chicken embryo fibroblasts (6); 3) in the presence of the uncoupler carbonyl cyanide rn-chlorophenylhydrazone (CCCP) pre-mAspAT accumulated in the cytosol of these cells and on dilution of CCCP was quantitatively taken up into the mitochondria (6).
In the present paper we extend previous studies on the uptake of in vitro synthesized pre-mAspAT into isolated mitochondria (7,8) and show that the in vitro translocation of chicken pre-mAspAT follows the mechanism delineated for precursors of other matrix and inner membrane proteins. Extensive studies particularly with yeast and Neurospora (for reviews, see Refs. 9 and 10) have shown that the import of these proteins occurs post-translationally (11,12), and requires an energized inner mitochondrial membrane fdr transport (13-16) as well as a metal-dependent matrix protease for processing (17,18). In addition, the aggregation state of in vivo and in vitro synthesized pre-mAspAT was investigated.

EXPERIMENTAL PROCEDURES
Preparation of Polysomes-Estradiol treatment of roosters for the preparation of polysomes from liver and disruption of liver tissue was as described ( 5 ) except for the composition of the media. Medium A contained 20 mM Tris chloride, 75 mM potassium chloride, 100 mM potassium acetate, 5 mM magnesium acetate (pH 7.6). Medium B contained in addition to medium A, 4 mg ml" of yeast RNA, 4 mg ml" of heparin, 200 mM sucrose, and 5 mM dithiothreitol. The tissue suspension was homogenized with a Dounce homogenizer by six strokes with the loosely fitting pestle A and two strokes with the tightly fitting pestle B. Subsequently, the homogenate was centrifuged in a Sorvall SS-34 fixed angle rotor at 10,000 rpm (12,000 x g-) for 5 min at 0 "C. The supernatant was again centrifuged in the same rotor at 12,000 rpm (17,400 X g-) for 10 min at 0 "C. For isolation of free polysomes, the supernatant was overlaid onto a discontinuous gradient comprising 6 ml each of 1.38 and 2 M sucrose containing 40 mM Tris chloride, 25 mM potassium chloride, 5 mM magnesium acetate, 4 mg ml" of yeast RNA, 5 mM dithiothreitol (pH 7.6). The 1.38 M sucrose solution contained in addition 4 mg ml" of heparin. After centrifugation in a Sorvall T 865 rotor at 52,000 rpm (200,000 X gav) for 20 h at 2 "C the polysomal pellets were resuspended in polysome buffer (50 mM Tris chloride, 25 mM potassium chloride, 5 mM magnesium acetate, 2 mM dithiothreitol, pH 7.6) and stored at -70 "C. To assess polysome size and integrity, their sedimentation profiles were analyzed on 1533.2% sucrose isokinetic gradients as described (5), except that the sucrose solutions contained the same salt concentrations as the polysome buffer, and centrifugation was in a Kontron TST 41 rotor at 40,000 rpm (204,000 X gav) for 40 min at 4 "C.
Cell-free Protein Synthesis-Read-out translation of free polysomes in a reticulocyte lysate in the presence of [36S]methionine was performed as described (5) except that incubation was at 29 "C for 40 min. Protein synthesis was stopped by chilling or, when specified, by the addition of 0.2 mM cycloheximide. Postribosomal supernatants 257 Precursor of ~i t o c h o~r i u l Aspartate Aminotransferase were obtained by centrifuging the chilled translation mixture in a Kontron TST 60 rotor at 45,000 rprn (190,000 X gay) for 1 h at 2 "C. I n Vitro Uptake into Mitochondria and Processing of Pre-mAspAT-The incubation mixture for in oitro import contained 60 pl of the labeled reticulocyte lysate (-7 X lo6 cpm of [35S]methionine incorporated into protein) or 80 p1 of a labeled postribosomal supernatant which had been adjusted to 210 mM sucrose, 110 mM potassium acetate, and 1.5 mM magnesium acetate. M i t~h o n d r i a were isolated from chicken heart as described (19) except that EGTA was omitted from the isolation medium. The freshly prepared mitochondria (250 pg of mitochondrial protein in 25 pl of isolation medium) were added to the translation products and incubated at 29 "C for 1 h. The mitochondria were then reisolated by centrifugation in a SS-34 rotor at 12,500 rpm (15,000 X gmU) for 15 min at 4 "C. To both the mitochondrial pellet (suspended in 50 pl of water) and the supernatant the protease inhibitors chymostatin, antipain, leupeptin, pepstatin (final concentration 0.2 mg ml" each), and aprotinin (0.55 mg m1-I) were added. Subsequently, 10% SDS was added to a final concentration of 4%. Both samples were treated for 4 min at 95 "C and then diluted 40-fold with Triton buffer (1% Triton X-100, 150 mM sodium chloride, 5 mM EDTA, 50 mM Tris chloride, pH 7.5) to 0.1% SDS. The particulate material was removed by centrifugation at 45,000 rpm (140,000 X gay) for 1 b at 15 "C, and the supernatant was mixed with 50-100 pl of anti-mAspAT antiserum for 14-16 h at 4 "C. After incubation, 100-200 p1 of a 10% (w/v) suspension of Staphylococcus aureus (20) were added and the incubation continued for 1 h at room temperature. After two washings with Triton buffer containing 0.1% SDS and 1 mg ml" of methionine the S t~~y~o ccus-antibody-antigen complexes were dissociated in 3-5% SDS for 4 min at 95 "C and the immunoprecipitation was repeated. The final pellet was dissolved in sample buffer for SDS-polyacrylamide gel electrophoresis (21).
For processing assays, the labeled translation products were incubated with the supernatant (15,000 X gmax, 10 min, 4 "C) of a hypotonic m i t o c h o n~a l extract (17) as described above for the import assay. After incubation, the total mixture was dissociated in SDS and subjected to immunoprecipitation.
Preparation of Samples for Gel Permeation Chromatography-For analysis of pre-mAspAT synthesized in vivo five Petri dishes of cultured chicken embryo fibroblasts were pulsed with 0.3 mCi ml" of [39S]methionine for 15 min at 37 'C in the presence of 20 p M CCCP (6). The cells were harvested in the cold, centrifuged for 7 min at 2,400 rpm, and taken up in 300 pl of homogenization buffer containing 80 pg ml" each of chymostatin, antipain, leupeptin, pepstatin, 175 pg ml" of aprotinin, 5 mM o-phenantbroline, 2 mM EGTA, 2 mM EDTA, 2 mM phenylmethanesulfonyl fluoride, 150 mM potassium chloride, 1 mM dithiothrei~l, 10 mM Tris chloride (pH 7.1). Homogenization was performed by 20 strokes with a Teflon pestle in a Potter-Elvehjem homogenizer at 1750 rpm and 4 "C. A postmitochondrial supernatant was obtained by centrifugation in an Eppendorf centrifuge for 10 min at 12,000 rpm (10,000 X gmU) and 4 "C.
The sample for chromatography of pre-mAspAT synthesized in a cell-free system was obtained by diluting the postribosomal supernatant of a [%]methionine-labeled reticulocyte lysate containing 120 pg ml" each of chymostatin, antipain, leupeptin, pepstatin, and 260 pg ml" of aprotinin, 1 mM dithiothreitol with 1 volume of 150 mM potassium chloride, 1 mM dithiothreitol, 10 mM Tris chloride (pH 7.1). P r e~r a t i o n of Samples for Suerose Gradient C e n t r i~~w n -T h e sample of in vivo synthesized pre-mAspAT was prepared in the same way as the sample for gel permeation chromatography except that the homogenization buffer was adjusted to pH 7.4 and contained no dithiothreitol. The sample of in vitro synthesized pre-mAspAT was the same as for gel permeation chromatography except that the postribosomal supernatant was diluted with 1 volume of 150 mM potassium chloride, 10 mM Tris chloride (pH 7.4). For analysis of newly processed mAspAT, one Petri dish of chicken embryo fibroblasts was pulsed with 0.2 mCi ml" of ["Slmethionine for 90 min. The cells were harvested in 300 pl of sonication buffer containing 40 pg ml" each of chymostatin, antipain, leupeptin, pepstatin, 87.5 pg ml" of aprotinin, 150 mM potassium chloride, 10 mM Tris chloride (pH 7.4). After treatment for 10 min in an ultrasonic ice-water bath at 35 kHz the sample was cleared by centrifugation for 10 min at 12,000 rpm (10,000 X g-) in an Eppendorf centrifuge.
ographed on a TSK-G 3000 SW column from LKB (7.5 X 600 mm) Gel Permeation C~r o m u~o g r~~y -S a m p l e s (200 pl) were chromatand eluted with 150 mM potassium chloride, 1 mM dithiotbreitol, 10 mM Tris chloride (pH 7.1). The flow rate was 1 ml min". Fractions of 0.5 ml were collected and analyzed by immunoprecipitation.
Sucrose Gradient Centrifugation-Samples (200 pl) were layered on 5-35% (w/w) sucrose gradients (11 ml) and centrifuged for 20 h at 35,000 rpm and 2 ' C using a Kontron TST 41 rotor. The sucrose solution contained 150 mM potassium chloride, 10 mM Tris chloride (pH 7.4). In the case of the in oiuo synthesized precursor, the sucrose solution contained 40 pg mi" each of chymostatin, antipain, leupeptin, pepstatin, 87.5 pg ml" of aprotinin, 5 mM o-phenanthroline, 2 mM EGTA, 2 mM EDTA, and 2 mM phenylmethanesulfonyl fluoride. After centrifugation in the absence of the protease inhibitors, no precursor could be recovered because its half-life in the cell homogenate at 4 "C was only 5 h. Fractions were collected by puncturing the gradient tubes at the bottom. Sucrose concentrations were determined in a Bausch and Lomb refractometer. The fractions were analyzed by immunopr~ipitation.
Mi.scellhneous--mAspAT was isolated from chicken heart as described (22). Aspartate aminotransferase activity was measured by the coupled assay with malate dehydrogenase (23). Mitochondrial protein concentration was estimated assuming a specific mAspAT content of 4.6 pg/mg of mitochondrial protein (19) and a specific activity of mAspAT of 220 units mg" at 25 "C.
The reticulocyte lysate was prepared and pre-treated with nuclease from S. aureus as described by Pelham and Jackson (24).
Antisera against mAspAT were obtained as described (5) except that an equimolar mixture of native and denatured mAspAT was used as immunogen. SDS-polyacrylamide gel electrophoresis was performed in 10 or 12.5% slab gels (21). For fluorography of the dried gels in Enlightning the recommendations of the manufacturer (New England Nuclear) were followed. Radioactivity in protein bands on SDS-polyacrylamide gels was determined by counting the radioactivity eluted from cutout gel pieces (5). Peroxidase activity was measured according to Ref. 25. "C-Labeled M, marker proteins and ["SI methionine (-1000 Ci/mmol) were from New England Nuclear; ophenanthroline from Fluka; bathophenanthrolinedisulfonate disodium salt from Serva. Bovine trypsin (~-l-tosylamido-2-phenylethyl chloromethyl ketone treated, 3.5 units mg") was from Merck. The protease inhibitors chymostatin, antipain, leupeptin, and pepstatin were from the Peptide Institute, Osaka. Aprotinin was from Bayer.

In Vitro Uptake of the Precursor of Mitochondrial Aspartate
Aminotransferase by Chicken Heart Mitochondria-A rabbit reticulocyte lysate was programmed with free polysomes from chicken liver in the presence of [36S]methionine. After translation, a sample of the lysate was incubated with isolated mitochondria. The incubation mixture was centrifuged, and the supernatant and the mitochondrial pellet were analyzed for labeled pre-mAspAT and mature mAspAT (Fig. lA). The supernatant contained only precursor, whereas the mitochondrial pellet contained both precursor and mature mAspAT. Apparently, part of the precursor had been taken up by the organelles and processed to the mature protein while part of it had been merely bound to the mitochondria. Exogenous mature mAspAT did not compete with the precursor for uptake into mitochondria (Fig. 1B). In order to exclude cotranslational uptake of pre-mAspAT, protein synthesis in the translation mixture was stopped by cycloheximide and a postribosomal supernatant was prepared. Incubation of this postribosomal supernatant with mitochondria also yielded mature mAspAT (Fig. IC). The possibility to inhibit the import in chicken embryo fibroblasts by CCCP (see Introduction) offered a second source of precursor for the in vitro study of the translocation process. However, the precursor in the cell homogenate was neither imported into isolated mitochondria nor was it processed by a mitochondrial extract (not shown).
Rates of Import and Processing of Pre-mAspAT-On incubation of a labeled translation mixture with exogenous mitochondria the precursor in the supernatant decreased concomitantly with an increase of labeled mature mAspAT in the mitochondria during the first 90 min of incubation (Fig. 2 4 ) .
No mature labeled or unlabeled mAspAT was detected in the supernatant indicating that most mitochondria remained intact during the incubation. A quantitative evaluation of the results is given in Fig. 2B. After an incubation of 120 min about 50% of the total precursor had been converted to the mature form. The percentage of pre-mAspAT co-sedimenting with mitochondria increased from -10% after 15 min to -20% after 120 min. The co-sedimented precursor, like the precursor FIG. 1. In vitro import of pre-mAspAT into isolated chicken heart mitochondria. A, reticulocyte lysate programmed with chicken liver free polysomes and labeled with ["Slmethionine (1 mCi ml") was incubated with or without chicken heart mitochondria at 29 "C for 60 min as described under "Experimental Procedures." The incubation was stopped by chilling and the lysate containing the mitochondria was fractionated into mitochondrial pellet and supernatant. All samples were analyzed for mAspAT by immunoprecipitation, SDS-polyacrylamide gel electrophoresis, and fluorography: Total labeled reticulocyte lysate incubated without mitochondria The sample was incubated with mitochondria for 60 min as described for A. Subsequently, the mitochondrial pellet and the supernatant (not shown) were analyzed as described above. A remaining in the supernatant, was completely digested by trypsin in a very low concentration (10 pg ml-I, 20 min, 4 "C; not shown); apparently it had not entered the mitochondria but was merely bound to the outer mitochondriai membrane. The proteolytic susceptibility of pre-mAspAT (see also Ref. 6) is in marked contrast to the resistibility of the mature protein either obtained by in vitro processing or as isolated from chicken heart (see Ref. 26).
Incubation of a labeled reticulocyte lysate with increasing amounts of mitochondria or with corresponding increasing amounts of hypotonic mitochondrial extract, results in an accelerated generation of mature mAspAT (Fig. 3, A and B, respectively). However, a t 200-400 pg of mitochondrial protein added a maximum value of imported or processed precursor is reached. This maximum value corresponds to 50% of the total precursor and equals the maximum fraction of pre-mAspAT found to be imported on prolonged incubation (Fig. 2). Addition of unlabeled reticulocyte lysate (nucleasetreated or untreated; up to 6 times the amount of labeled reticulocyte lysate) as well as addition of chicken heart cytosol and chicken fibroblast cytosol to the import assay failed to increase the yield of import. Such additions were reported to stimulate the in vitro import of the precursors of ornithine transcarbamylase (27) and of cytochrome bp (28). Addition of Zn2+ or Co2+ to the mitochondrial extract (18) also did not increase the fraction of processed precursor. Apparently, under the present experimental conditions the extent of import or processing of pre-mAspAT is limited neither by time nor by the amount of mitochondria or mitochondrial extract added nor by auxiliary factors. Only part of the precursor synthesized in the reticulocyte lysate is present in a form that can be imported or processed. The same fraction of importable precursor was obtained with different preparations of polysomes. However, the fraction was even smaller when other batches of reticulocyte lysate were used.
Inhibitors of Import and Processing-In order to show that of hypotonically disrupted mitochondria (final volume 85 pk for details, see "Experimental Procedures"). After 90 min a t 29 "C the supernatant and mitochondrial fractions of the import assays A and the processing assays B were made 4 % (w/v) in SDS and analyzed by immunoprecipitation, SDS-polyacrylamide gel electrophoresis, and fluorography. The bands corresponding to pre-mAspAT and mAspAT were cutout from the gels and the radioactivity counted as described (5). 0, mAspAT in the mitochondria or processed mAspAT, respectively; 0, pre-mAspAT in the supernatant of import assay A or pre-mAspAT in processing assay B; A, pre-mAspAT associated with the mitochondria; m, sum of total pre-mAspAT plus mAspAT.
Prolonged incubation ( 2 h) of the samples containing 400 pg of mitochondrial proteins did not increase the amount of imported or processed mAspAT.  Fig. 2A. pre-mAspAT is processed to the mature form only after having entered the matrix of the mitochondrion, a protocol with inhibitors of import and processing was applied (see Refs. 13 and 29). Following read-out translation in a reticulocyte lysate the labeled translation mixture received either the uncoupler CCCP or one of the chelating agents, o-phenanthroline or bathophenanthroline disulfonate. The samples were incubated with either intact mitochondria or with a supernatant of hypotonically disrupted mitochondria. In the presence of CCCP no mature labeled enzyme could be detected in the mitochondria (Fig. 4A) in agreement with results obtained in vivo (6). The effect of CCCP had been shown previously to be due to depletion of the electrochemical potential across the inner mitochondrial membrane (13-16). As expected, CCCP did not prevent the generation of mature enzyme if pre-mAspAT was incubated with a mitochondrial extract (Fig. 4B). Of the two chelating agents tested only the more hydrophobic o-phenanthroline which penetrates the mitochondrial membranes, but not the charged bathophenanthroline disulfonate which cannot enter the mitochondria, inhibited the appearance of mature enzyme in intact mitochondria (Fig. 4A). When tested with the mitochondrial extract both agents inhibited the processing of pre-mAspAT (Fig. 4B).
Aggregation State of in Vitro and in Vivo Synthesized Pre-mAspAT-A postribosomal supernatant of a labeled reticulocyte lysate or a postmitochondrial supernatant of chicken embryo fibroblasts pulsed in the presence of CCCP (6) were applied to a high performance gel permeation chromatography column. In both cases, the bulk of pre-mAspAT was eluted with the void volume (Fig. 5, A and B ) corresponding to a MI > 300,000. Apparently, after synthesis pre-mAspAT becomes part of a high molecular weight complex. A minor fraction of pre-mAspAT from the reticulocyte lysate was eluted in a broad peak corresponding to an apparent M , of 75,000 (Fig.  5A). This elution volume corresponds with that of mAspAT ( Fig. 5C) which for unknown reasons does not elute corresponding to its M , of 90,000. In the case of the in vivo synthesized product, faint bands of pre-mAspAT were detected which corresponded to a M, of -40,000 (Fig. 5R) suggesting a small pool of monomer.
Results similar to those of gel permeation chromatography were also obtained by zone velocity centrifugation in sucrose gradients. About half of pre-mAspAT from the reticulocyte lysate sedimented in a M , range around 85,000 and the other half in the high MI range of the gradient (Mr > 240,000), suggesting that under the conditions used pre-mAspAT exists partly as a homodimer or hetero-oligomer and partly as a higher M , aggregate (Fig. 6, A and D). The distribution of pre-mAspAT between the two forms was not constant; in another experiment the ratio between the two peaks was 4 1 in favor of the putative dimeric form. On centrifugation of the postmitochondrial supernatant from CCCP-treated chicken embryo fibroblasts, the precursor was distributed over a MI range from -40,000 to >240,000 corresponding to 1 to 6 times its MI in SDS containing gels (Fig. 6, B and D). For determining the M , of newly processed mAspAT, a homogenate of pulsed chicken embryo fibroblasts was sonicated in order to release the labeled enzyme from the mitochondria. After centrifugation, mAspAT was detected as a distinct peak in the 85,000 region of the gradient (Fig. 6, C and D) indicating that the newly processed mAspAT exists exclusively as a homodimer. Mature endogenous unlabeled mAspAT was immunoprecipitated from the same fractions as newly processed mAspAT (not shown).

DISCUSSION
In this study the import of pre-mAspAT into mitochondria was reconstituted with precursor that had been synthesized in a rabbit reticulocyte lysate programmed with free polysomes from chicken liver and with mitochondria that had been isolated from chicken heart. The following lines of experimental evidence indicated that the observed processing of pre-mAspAT was executed by a chelator-sensitive protease in the mitochondrial matrix. 1) The uncoupling agent CCCP abolished the appearance of labeled mAspAT when pre-mAspAT was incubated with intact mitochondria. 2) Both chelators, bathophenanthroline disulfonate and o-phenanthroline (13, 29), inhibited the processing by a mitochondrial extract; however, only o-phenanthroline which in contrast to the charged bathophenanthroline disulfonate can diffuse indicated are from the calibration run of C. The denotation rn indicates the elution volume of marker mAspAT as visualized by protein staining. It corresponds to a M , of 75,000. The same elution volume of mAspAT was found in the calibration run. C, calibration of the column. The following M , markers were used: catalase (240,000); lactate dehydrogenase (140,000); cytosolic aspartate aminotransferase (92,000); mitochondrial aspartate aminotransferase (90,000; however, its elution volume corresponds to a M , of 75,000); carboanhydrase (30,000); myoglobin (17,000); cytochrome c (13,000). In some experiments including that of A , pre-mAspAT appeared as a double band. The additional band with a slightly lower M, was analyzed by chemical cleavage of the protein in the gel slice and reelectrophoresis of the fragments (41). The peptide pattern was very similar to those of cleaved pre-mAspAT and mAspAT. The appearance of the additional band was sporadic and independent of the presence of mitochondria. through the inner membrane inhibited the processing by intact mitochondria.
The reconstitution of the import of pre-mAspAT corroborates the previous conclusion, that the precursor is post-translationally translocated into the mitochondria (see Introduction). The notion of a post-translational uptake is confirmed by the finding that import occurred also with a postribosomal supernatant of a reticulocyte lysate where protein synthesis had been stopped by cycloheximide.
The estimated half-life of pre-mAspAT with respect to import in the i n vitro system was -30 min (Fig. 2), whereas in chicken embyro fibroblasts the half-life of the pre-mAspAT was found to be only 0.5 min (6). Similar slow i n vitro translocation and processing has been observed in the case of several other precursors, e.g. pre-ornithine transcarbamylase (30,31) pre-carbamoylphosphate synthetase (31,32), and premethylmalonyl-CoA mutase (33).
The i n vitro system is also less efficient than intact cells with respect to the yield of import. In CCCP-treated chicken fibroblasts pre-mAspAT which has been accumulated in the cytosol is quantitatively chased into mitochondria on release of the import block (6). In contrast, only about 50% of the total in uitro synthesized precursor was recovered as mature enzyme on incubation with intact mitochondria. The bulk of the nonimported precursor was about equally distributed between the mitochondrial fraction and the supernatant. Processing of in vitro synthesized pre-mAspAT to the mature form by a mitochondrial extract was also found to be limited to -50% of total precursor (Fig. 3). The coincidence of the yields of import and processing suggests that the precursor itself rather than auxiliary factors is limiting. In agreement with this conclusion, addition of more mitochondria, more reticulocyte lysate, or a cytosol from chicken heart or fibroblasts all failed to increase the yield of import. Conceivably, the nascent polypeptide chains in the polysome preparation used might be damaged. However, this explanation appears unlikely because the same polysome preparation was found to produce precursor with an even lower yield of import and processing in other reticulocyte lysates. The finding that precursor from chicken embryo fibroblasts after homogenization of the cells was neither imported into isolated mitochondria nor processed by a mitochondrial extract also might indicate that under cell-free conditions the precursor is subject to post-translational alterations prohibiting its import and processing. Such alterations could include covalent modifications, complexation with other cellular constituents, or denaturation. Denatured precursors have been reported not to be processed (18).
The studies on the aggregation state of the precursor showed that pre-mAspAT in both a reticulocyte lysate and a postmitochondrial supernatant from chicken embryo fibroblasts is not a homogeneous species. Centrifugation and gel filtration experiments indicate that pre-mAspAT may exist as high (M, > 300,000) and low M, aggregates. With both methods the apparent M, of the low M, form corresponds with the M , determined for mature mAspAT. Thus, the low M, material may represent a dimeric form of pre-mAspAT although an association of the precursor with other proteins cannot be excluded.
For several other mitochondrial precursor proteins, M, values ranging from that of a monomer up to 500,000 have been found. Each protein under a given set of conditions yielded a characteristic aggregation profile. On gel filtration analysis, the bulk of the precursors of the P-subunit of yeast F1-ATPase (34), of rat liver ornithine transcarbamylase (35), of Neurospora ATP/ADP carrier (12), and of rat liver adenine nucleotide carrier (36) behaved as high M, aggregates. On centrifugation the precursors of rat liver ornithine transcarbamylase and carbamoyl phosphate synthetase (35) as well as the precursor of subunit 9 of Neurospora ATPase (37) showed broad M , profiles ranging from monomer up to high M, aggregates. For the precursors of rat liver mAspAT (38) and malate dehydrogenase (39) M, values only slightly larger than those for the mature dimeric form of the corresponding enzyme have been determined by gel filtration. The marked variation in the aggregation profile of a given precursor under different conditions of analysis on the one hand and among different precursor proteins on the other suggests that the aggregation state depends critically on the experimental conditions, e.g. duration of experiment, temperature, availability of cytosolic factors, or ionic conditions. In the case of pre-mAspAT, gel filtration which was performed in the short time of 12 min (elution time of high M, aggregates) and at 20 "C favored the detection of high M, aggregates, while after centrifugation for 20 h at 2 "C, the ratio between high and low M , material was shifted towards the latter.
In conclusion, all experimental evidence including the present data indicate that mitochondrial precursor proteins are present in the cytosol as low M, forms (homo-or heterooligomers) and as high M, forms. The formation of aggregates may protect the precursors against proteolysis (34), or in analogy to the signal recognition particle for secretory proteins (41) to be involved in the specific translocation of the precursor from its site of synthesis to the mitochondrial outer membrane. In vitro import assays such as that established in this study will allow us to address this question experimentally.