Lipophilic Proteins of Mitochondria from Microaerobic and Aerobic Continuous Cultures of Saccharomyces cerevisiae

Transcription

1 JOURNAL OF BACrERIOLOGY, Sept. 1974, p Copyright American Society for Microbiology Vol. 119, No. 3 Printed in U.S.A. Lipophilic Proteins of Mitochondria from Microaerobic and Aerobic Continuous Cultures of Saccharomyces cerevisiae P. J. ROGERS AND P. R. STEWART Department of Developmental Biology, Research School of Biological Sciences, Australian National University, Canberra, A.C.T., 2601, Australia, and Department of Biochemistry, School of General Studies, Australian National University, Canberra, A.C.T., 2600, Australia Received for publication 21 March 1974 Products of the mitochondrial protein-synthesizing system have been labeled in vivo in the presence of cycloheximide in microaerobic cells and in cells from glucose-limited and glucose-repressed aerobic continuous cultures of Saccharomyces cerevisiae. Lipophilic proteins were extracted from labeled mitochondrial membranes with aqueous methanol and neutral and acidic chloroformmethanol solvents. In glucose-limited aerobic and microaerobic cells, about half of the total mitochondrial products were soluble in organic solvents; in contrast, almost all of the labeled products were extracted from glucose-repressed mitochondria. Only trace amounts of labeled product were formed in mitochondrial membranes of a petite mutant. Lipophilic proteins were examined by polyacrylamide gel electrophoresis under dissociating conditions. Most of the label was associated with components of apparent molecular weights 12,000, 14,000 and 16,000. The relative proportions of these species in mitochondrial membranes are dependent on the concentrations of oxygen and glucose in which the cells are grown. Recent studies have confirmed that products of both the cytoplasmic and mitochondrial protein-synthesizing systems are required for the assembly of some complexes of the inner membrane of yeast mitochondria, e.g., cytochrome c oxidase (4, 10, 20) and adenosine triphosphatase (18, 20). A significant proportion of the mitochondrial protein products can be extracted selectively from rat liver (1, 6-9) and yeast (13, 19) mitochondria with organic solvents. Kadenbach and Hadvary (7), Sierra and Tzagoloff (17), and Michel and Neupert (12) have discussed the possible role of small lipophilic molecules in the synthesis and assembly of mitochondrial membranes. Previous studies showed that the nature of the membrane proteins made by the mitochondria in Saccharomyces cerevisiae is regulated by environmental factors such as oxygen tension and glucose concentration (5, 11, 15). Whether this control extends to the synthesis of the lipophilic proteins and, therefore, reflects fundamental changes in mitochondrial membrane structure is the subject of this communication. MATERIALS AND METHODS Cell culture. A diploid strain of S. cerevisiae and a petite cell derived from the parent strain by euflavine treatment were studied. Cells were grown in continuous cultures in a New Brunswick Microferm, fitted with ph and oxygen controls, on a semisynthetic medium (16) at ph 5.5 with a dilution rate of 0.06/h. Labeling, extraction, and electrophoresis of lipophilic proteins from mitochondria. The protein products synthesized by the mitochondria were labeled in vivo with [4, 5-3H]leucine as described previously (15). The concentration of tritiated leucine used during the labeling period was varied in some experiments and is referred to in the legends. Lipophilic proteins were extracted essentially by the method of Tzagoloff and Akai (19). Pellets of submitochondrial particles were extracted twice with methanol (90%, vol/vol) at room temperature (1 ml of solvent per 2 mag of mitochondrial protein), and the extract was dried at 50 C under dry nitrogen and then dissolved in sodium dodecyl sulfate solubilizer by heating at 70 C for 30 min as described previously (15). The residue recovered by centrifugation after methanol extraction was extracted twice (1 ml of solvent per 2.5 mg of protein) with chloroformmethanol (2:1, vol/vol) at 50 C for 20 min. The extracts were pooled, dried, and dissolved as above. The residue from this extraction was extracted twice (2 ml of solvent per 2.5 mg of protein) with acidified chloroform-methanol (2: 1, vol/vol; 10 mm HCl) for 20 min at 50 C, and the extracts were pooled, dried under nitrogen, and solubilized. The residue remaining after these solvent extractions was also dried and solubilized in sodium dodecyl sulfate. Samples of the different solvent extracts were removed, before drying, for the determination of protein and radioactivity (15). Gel electrophoresis was performed on 7.5% polyacrylamide gels in the presence of sodium dodecyl 653

2 654 ROGERS AND STEWART sulfate as described by Weber and Osborn (21), except that the buffer was tris(hydroxymethyl)aminomethane (ph 8). Gels were sliced and analyzed for radioactivity as described earlier (15). Results are plotted as radioactivity per slice versus slice number without subtraction of background, and are plotted up to the point on the gel after which no further significant amounts of radioactivity are found. Recovery in the slices of radioactivity applied to the gel was in the range 75 to 110% provided that solubilization of the applied material was complete. In some experiments the gels were stained with Coomassie blue (2%, vol/vol, in methanol-acetic acid-water, 5:1:4). Destaining was carried out at 40 C in methanol-acetic acid-water (5:1:4). Stained gels were scanned at 650 nm in a Gilford spectrophotometer fitted with a scanning accessory and set with a full scale absorbance value of 1.0. Amino acid identification. Dried extracts were incubated for 30 to 32 h at 110 C in a mixture of 6 M HC1 and concentrated acetic acid (2:1, vol/vol). The lipids were extracted with petroleum ether, and the water-soluble components were separated by twodimensional paper chromatography (14). Amino acids were identified with ninhydrin; other substances were identified with iodine. When radioactivity was to be quantitated, the spot corresponding to leucine was cut out and leached in 1 ml of water overnight, and radioactivity in the extract was determined. RESULTS Glucose-limited aerobic cells. Cells were grown in glucose-limited cultures at an oxygen TABLE 1. concentration of 30 gm. Mitochondrial membranes labeled in vivo in the presence of cycloheximide were extracted first with methanol (90%) and then successively with neutral and acidified chloroform-methanol. Approximately 5 to 10% of the radioactivity initially present in the mitochondrial membranes was recovered in the methanol extract (Table 1). Neutral chloroform-methanol solubilized about one-quarter of the radioactive products (Table 1). With acidic chloroform-methanol a further 15% of the original radioactivity was recovered, and the residue after the extractions accounted for more than 50% of the initial activity. However, the specific activity of the neutral chloroform-methanol fraction had increased about 25-fold compared to that of the original submitochondrial particles. The radioactivity in the solvent extracts was shown to be primarily due to incorporated leucine by identification of the amino acid after acid hydrolysis of the extracted material. More than 85% of the incorporated radioactivity was recovered as leucine by two-dimensional paper chromatography. The mitochondrial products extracted with the solvents were examined by gel electrophoresis. Resolution of individual protein species on gels run under these conditions is more or less inversely related to the amount of protein Recoverv of protein and radioactivity in organic solvent extracts of mitochondrial membranes from glucose-limited aerobic and microaerobic cultures of S. cerevisiaea Culture conditions Labeling conditions Solvent extracts Oxygen Glucose Oxygen 1 Glucose 1 Total Total protein Spact (AM) (pm) (pm) (mm) Fractionb (counts/min) radioactivity (mg) (counts/min/ mg of protein) I. Wild-type, aerobic, derepressed II. Wild-type, microaerobic, derepressed j Membranes Methanol C:M C: M: HCl Residue Membranes Methanol C:M C: M: HCl Residue 285,000 10,700 77,100 38, ,800 72,500 2,000 14,400 17,600 36, w 4 J. BACTERIOL. 10,600 53, ,800 25,700 7,600 10,400 10,000 57,600 10,400 9,000 amitochondrial membranes were labeled, isolated, and extracted as described. The concentration of [3Hjleucine used to label mitochondria in cells from aerobic and microaerobic cultures was 2.5 and 10 pci/ml, respectively, in experiments I and II. b C: M, Chloroform-methanol (2:1, vol/vol); C: M: HCl, acidic chloroform-methanol (2:1, vol/vol; 10 mm HCl). Radioactivity and protein values indicate amounts extracted by the solvents.

3 VOL. 119, 1974 LIPOPHILIC PROTEINS OF MITOCHONDRIA 655 applied to the gels. On the other hand, the amount of radioactivity may become unsatisfactorily low as gel loading is decreased. A compromise must therefore be made between resolution and workable levels of radioactivity. The significant comparisons to be made lie primarily in differences between cell types; hence, loading has been kept comparable within a given type of extract or residue. Nevertheless, some of the gels are loaded with relatively large amounts of protein (> 100,ug), and the number of individual species identified must therefore be considered a minimal estimate. In Fig. 1 the electrophoretic patterns of the extracted fractions are shown. Two major bands of radioactivity, both with slightly lower mobilities than cytochrome c, were observed in the methanol fraction (Fig. la). The disperse leading edge of the radioactivity trace is associated with opaque material which does not stain significantly with Coomassie blue. This material migrates with similar mobility as phospholipid under these conditions; its properties remain to be examined. Components of similar mobilities to those in the methanol extracts were also detected after electrophoresis of the neutral chloroformmethanol extracts (Fig. lb). In this case, most of the label was associated with a discrete band of radioactivity corresponding to a molecular weight of 14,000; some activity was also present in the region close to the cytochrome c marker. A minor component with mobility greater than that of cytochrome c was also detected. In the acidified chloroform-methanol extract (Fig. lc), almost all the radioactivity was present as a discrete band of apparent molecular weight 16,000. Label also migrated with the same mobility as the cytochrome c marker. High-molecular-weight products of mitochondrial translation were detected in the residue remaining after each successive solvent extraction (Fig. 2). After extraction of the particles with 90% methanol, most of the label associated with 90,000-molecular weight components in untreated membranes was removed; an increase in the relative amount of radioactivity associated with products of apparent molec- FIG. 1. Mitochondrial membranes from aerobic, glucose-limited cells were labeled with [3H]leucine under similar conditions and extracted sequentially with organic solvents. The extracts were then analyzed on sodium dodecyl sulfate-acrylamide gels: (a) methanol extract (40 ug of protein); (b) neutral chloroform-methanol extract (20 /ug of protein); and (c) acidic chloroform-methanol extract (120 ug of slice number protein). The arrows in the figures indicate the position of the cytochrome c marker. The solid line represents the absorbance scan (650 nm) of the gel after staining with Coomassie blue. Background for slice count was 12 counts/min.

4 656 ROGERS AND STEWART J. BACTERIOL. I ular weight 16,000 also occurred. When the mol wi i residue after chloroform-methanol extraction 100l a was subjected to electrophoresis, a decrease in llithe amount of components of approximate molecular weight 35,000 was evident, and a further -ṙ t Rincrease in the proportion of label in the 16, ll molecular weight region had occurred (Fig. 2c). The increased sharpness of this band suggested Vt5 J*. \ \ that another component had been removed by neutral chloroform-methanol. The residue re- 20 Jita "\s>; '" maining after the final extraction with acidic chloroform-methanol contained both 16,000 and, I A 35,000-molecular weight products (Fig. 2d). 20 H '.l, z 11 t 100 ' Since the parallel analysis of the solvent ex- I lt tracts had not demonstrated the selective exb 7W1 t traction of 30,000 to 35,000 and 90,000-molecuj 1l llar weight products (Fig. 1), it is possible that 60O'7' ) t * tthe 16,000 and 14,000-molecular weight components seen in the extracts are subunits of these m" *rva 1 Ihigh-molecular-weight species. 20c L Microaerobic glucose-limited cultures (0.1 E um oxygen). Earlier studies (15) showed that the protein products of the mitochondrial sys- Item were significantly different in microaerobic : 100. cells compared to those from aerobic cultures, D provided the labeling was carried out O at the c same oxygen concentration as that prevailing during steady-state growth (cf. Fig. 2a and 3a). 60l The relative contribution of mitochondrial protein synthesis to organelle membrane in the former cells was 10% of the total protein as \il tv t 4 tb * T \ compared to about 20% in aerobic cells (15) The distribution of radioactivity in mitochonit s drial membranes from microaerobic, glucoselimited cells is shown in Fig. 3a. Submitochondrial particles were extracted 1009I sequentially with organic solvents in the same d order as for aerobic cultures (Table 1). Methanol (90%, vol/vol) solubilized less than 5% of the 60 + radioactive products. Extraction with neutral pilsn*tilk ta.n^lkt.ditional 20 and 24% of the original counts,.v.t;i..respectively. Approximately half of the initial 20 8 radioactivity was recovered in the residue after the solvent extractions. A significant enrichment of mitochondrial products was obtained with neutral chloroform-methanol extraction; an almost sixfold increase in the specific radislice n u m be r oactivity of the solubilized protein was evident compared with the untreated mitochondrial FIG. 2. Mitochondrial membranes from cells membranes grown in aerobic, glucose-limited conditions were labeled under similar conditions with [3Hlleucine and sequentially extracted with organic solvents. The neutral chloroform-methanol extraction (150 pg of residues remaining after each solvent extraction were protein); and (d) residue remaining after acidified analyzed on sodium dodecyl sulfate-acrylamide gels: chloroform-methanol extraction (200 pg of protein). (a) untreated mitochondrial particles (200 pg of The arrows in the figures indicate the position of the protein); (b) residue remaining after methanol extrac- cytochrome c marker. Background: (a and c), 13 tion (250 Ag of protein); (c) residue remaining after counts/min; (b and d), 32 counts/min.

5 VOL. 119, 1974 LIPOPHILIC PROTEINS OF MITOCHONDRIA 657 C 0 **,,,i, 4) 00 c oo 200 :, 0 200k slice number FIG. 3. Mitochondrial membranes from cells grown in microaerobic, glucose-limited conditions were labeled under similar conditions with [3H]leucine and sequentially extracted with organic solvents. The solvent extracts and the residues were analyzed on sodium dodecyl sulfate-acrylamide gels: (a) untreated mitochondrial membranes (350 Ag of protein); (b) residue after extraction with methanol (200 Mg of protein); (c) neutral chloroform-methanol extract (100 Ag of protein); (d) residue after neutral chloroform-methanol extraction (350 ug of protein); (e) acidic, chloroform-methanol extract (180 Mg of protein); and (f) residue after acidic chloroform-methanol extraction (100 Mg of protein). The arrows in the figures indicate the position of the cytochrome c marker. Background: (a and f), 31 counts/min; (b, c, and d), 54 counts/min; (e) 20 counts/min. When the extracted products were characterized by electrophoresis, the residue after methanol extraction showed a significantly different radioactivity profile compared with the untreated particles (cf. Fig. 3a and b). Firstly, the label associated with 90,000-molecular weight products had disappeared; secondly, a decline had occurred in the relative proportion of products of approximately 35,000 molecular weight; thirdly, radioactivity was now associated with products of molecular weight 12,000 and 14,000; and finally, some label now migrated faster than the cytochrome c marker. Neutral chloroform-methanol extracted 12,000 and 14,000- molecular weight products and also components of molecular weight less than 12,000 (Fig. 3c and 3d). Acidified chloroform-methanol extracted mainly 14,000-molecular weight material and a small proportion of higher-molecular-weight species (Fig. 3e). The 14,000-molecular weight material could still be detected in the final residue (Fig. 3f) and, in addition, label was still associated with 20,000- to 30,000-molecular weight products. Glucose-repressed aerobic cultures. When cells are grown under glucose repression in continuous culture, mitochondrial respiratory development is repressed (15). In this regard, the effects of anaerobiosis and glucose repression in S. cerevisiae are similar. Furthermore, the fraction of mitochondrial protein made by mitochondrial ribosomes declines by almost half in repressed cells compared with derepressed cells (15). Whether or not the nature of the mitochondrial protein products formed under these conditions is also altered was examined by extracting proteolipid fractions and analyzing them by gel electrophoresis. Practically all of the mitochondrial membrane synthesized by mitochondrial ribosomes was recovered in the 20,000- to 45,000-molecular weight region of the gel (Fig. 4a). The results of solvent extraction are summarized in Table 2. In comparison to glucose-limited cultures, most of the p

6 658 ROGERS AND STEWART J. BACTERIOL. C E L- CL c 0 u a C '- i.. N l.il Ir' 11 6 *'1 20 AO slice number FIG. 4. Mitochondrial membranes from cells grown in glucose-repressed, aerobic continuous culture were labeled under similar conditions and extracted sequentially with organic solvents. The solvent extracts and the residues were analyzed on sodium dodecyl sulfate-acrylamide gels: (a) untreated radioactivity (>80%) was solubilized with neutral chloroform-methanol, although once again the extracted label consisted predominantly of 12,000, 14,000, and 16,000-molecular weight products; label was also present near the origin of the gels (Fig. 4b). Acidic chloroformmethanol solubilized most of the remaining radioactivity, which was found to be primarily 14,000- and 16,000-molecular weight material (Fig. 4c). Absence of significant amounts of lipophilic proteins of mitochondrial origin in a petite strain of S. cerevisiae. Protein products of yeast cells labeled in vivo in the presence of cycloheximide are usually classified as mitochondrial in origin, provided the following criteria are satisfied. Firstly, the label must be associated with the mitochondria; secondly, the labeling should be inhibited by specific inhibitors of mitochondrial protein synthesis; and thirdly, the labeling should not occur in petite mutants in which a specific deletion of the mitochondrial protein-synthesizing system has occurred. The first and second criteria have been confirmed and are described elsewhere (15). Petite cells from aerobic, glucose-limited continuous cultures were labeled in the presence of cycloheximide and a high concentration of radioactive leucine (Table 2); a mitochondrial membrane preparation was prepared and examined by gel electrophoresis. A single small peak of radioactivity migrating with the cytochrome c marker could be identified in mitochondrial membranes prepared from this mutant (Fig. 5). Most of this radioactivity (80%) was extracted with 90% (vol/vol) methanol (Table 2). The fact that none of major species labeled in wild-type cells was labeled in the petite mutant indicates that these are products in one form or another of the mitochondrial system. When the methanol extract was examined by gel electrophoresis, all the radioactivity migrated faster than the cytochrome c marker (Fig. 5b) at a position comparable to the bromophenol blue dye front. Similarly, when the neutral chloroform-methanol extract was subjected to electrophoresis, traces of label were detected (Fig. 5c) at the same position. These extracts thus give an indication of the nonmitochondrial membranes (300 ug of protein); (b) neutral chloroform-methanol extract (25 ug of protein); and (c) acidic chloroform-methanol extract (300,gg of protein). The arrows in the figures indicate the position of the cytochrome c marker. Background: (a and c), 25 counts/min; (b) 9 counts/min.

7 VOL. 119, 1974 LIPOPHILIC PROTEINS OF MITOCHONDRIA 659 TABLE 2. Recovery of protein and radioactivity in organic solvent extracts of mitochondrial membranes from glucose-repressed, aerobic cultures of S. cerevisiae and a petite mutant grown in aerobic, glucose-limited conditionsa III. IV. Culture conditions Labeling conditions Solvent extracts Oxygen lucose Oxygen Glucose Total Total Sp act (ymn) (GMl) ymn) (GmMl) Fraction radioactivity protein (counts/min/ (counts/min) (mg) mg of protein) Wild-type, aerobic, repressed Membranes 374, ,400 Methanol 27, ,800 C: M 254, ,800 C:M:HCI 24, ,600 Residue 6, Petite Membrane 9, ,200 Methanol 7, ,700 C:M 1, ,000 C:M:HCI 0.7 Residue 6.0 afor explanation of abbreviations and methods refer to Table 1. The concentration of [3H]leucine used to label glucose-repressed cells and petite cells was 5 and 20 ACi/ml, respectively. specific amounts of label that may be associated with, and extracted from, mitochondrial membranes from the wild-type cells. DISCUSSION The existence of hydrophobic membrane proteins soluble in organic solvents is now well established (3, 20). Lipophilic proteins of this type may be the primary products of the mitochondrial protein-synthesizing system (1, 2, 6-9, 13, 17, 19, 20), and recent evidence indicates that they include components of the rutamycin-sensitive adenosine triphosphatase complex of yeast (17, 20). The apparent molecular weights of the mitochondrial lipophilic proteins have been determined from polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. Under these conditions, the molecular weights of a large number of soluble proteins are directly related to their electrophoretic mobilities. Whether or not a similar relationship applies to the mobility of lipophilic proteins is not known, and the molecular weights reported for these species must therefore be regarded as tentative. The apparent molecular weights of the lipophilic proteins identified in this study generally lie in the range of 12,000 to 16,000, in agreement with the observations of Ebner et al. (4). These values are higher than those reported for yeast (19) and for rat liver (7). The apparent differences between these results and those of Tzagoloff and Akai (19) may be due, in part, to c E 0. c :3 0 LI o a slice number FIG. 5. Mitochondrial membranes from a petite mutant of S. cerevisiae grown in glucose-limited, aerobic culture were labeled under similar conditions and extracted sequentially with organic solvents. The solvent extracts and the residues were analyzed on sodium dodecyl sulfate-acrylamide gels: (a) untreated mitochondrial membranes (350 pg of protein); (b) methanol extract (100 pg of protein); and (c) neutral chloroform-methanol extract (300 pg of protein). The arrows in the figures indicate the position of the cytochrome c marker. Background: 49 counts/min. 1C b_ 4. b _\ >/ c C L differences in the culture and labeling conditions and, as such, would emphasize the control exercised by the cultural and intracellular environments. It is also clear from the electrophoretic patterns of corresponding extracts and residues

8 660 ROGERS AND STEWART J. BACTERIOL. that solvent extraction causes a substantial depolymerization of the high-molecular-weight components recovered from intact mitochondrial membranes. This is in agreement with the findings of Tzagoloff and co-workers (20). Sodium dodecyl sulfate alone is thus insufficient to bring about a complete depolymerization of these lipophilic membrane proteins; at this point it is difficult to specify their molecular status within the intact membrane. In aerobic repressed and derepressed cells, and in microaerobic cells, the major lipophilic proteins have nominal molecular weights of 16,000, 14,000, and 12,000. However, the relative proportions of label extractable by organic solvent systems differed markedly for the three membrane types examined. Thus, label from aerobic, repressed cells was almost completely extractable and most of this was recovered in neutral chloroform-methanol. About one-half of the labeled material could be extracted with organic solvents from derepressed cells, whether aerobic or microaerobic. Most of this was extractable with neutral chloroform-methanol in the case of aerobic cells, and with acidic chloroform-methanol in the case of microaerobic cells. The significance of these differences is not immediately evident; they could be due to intrinsic differences in the solubility of the membrane proteins (i.e., to chemical differences), to differences in accessibility to solvents resulting from altered interactions within the membranes, or to altered relative synthesis in the mitochondria of different products, lipophilic and hydrophilic. In the case of altered patterns of synthesis, this could result directly from changes in the amounts of different proteins formed or, alternatively, to regulation by the availability of and binding (covalent or otherwise) to products of the cytoplasmic protein-synthesizing system located in the mitochondrial membrane. Differences in relative proportions of 12,000-, 14,000-, and 16,000- molecular weight components extracted from mitochondria from the three cell types might be explicable in similar terms. A conclusive demonstration that these represent real differences in mitochondrial translation control in cells grown in different environments must await the isolation and chemical characterization of these lipophilic components of the mitochondrial membrane. LITERATURE CITED 1. Burke, J. P., and D. S. Beattie The synthesis of proteolipid protein by isolated rat liver mitochondria. Biochem. Biophys. Res. Commun. 51: Cattell, K. J., C. R. Lindop, I. G. Knight, and R. B. Beechey The identification of the site of action of NN'-dicyclohexyl carbodi-imide as a proteolipid in mitochondrial membranes. Biochem. J. 125: De Robertis, E Molecular biology of synaptic receptors. Science 171: Ebner, E., T. L. Mason, and G. Schatz Mitochondrial assembly in respiratory-deficient mutants of Saccharomyces cerevisiae. II. Effect of nuclear and extrachromosomal mutations on the formation of cytochrome c oxidase. J. Biol. Chem. 248: Groot, G. S. P., W. Rouslin, and G. Schatz Promitochondria of anaerobically grown yeast. VI. Effect of oxygen on promitochondrial protein synthesis. J. Biol. Chem. 247: Kadenbach, B Isolation and characterization of a peptide synthesized in rat-liver mitochondria. Biochem. Biophys. Res. Commun. 44: Kadenbach, B., and P. Hadvary Demonstration of two types of proteins synthesized in isolated rat-liver mitochondria. Eur. J. Biochem. 32: Koppikar, S. V., P. Fatterpaker, and A. Sreenivasan Study of soluble lipoprotein in rat liver mitochondria. Biochem. J. 121: Kuzela, S., J. Kolarov, and V. Krempasky Solubility of the intramitochondrially synthesized protein and other membrane proteins of rat liver mitochondria in acidic or neutral chloroform-methanol. Biochem. Biophys. Res. Commun. 54: Mason, T. L., and G. Schatz Cytochrome c ouidase from bakers yeast. II. Site of translation of the protein components. J. Biol. Chem. 248: Mian, F. A., M. J. Kuenzi, and H. 0. Halvorson Studies on mitochondrial membrane proteins in Saccharomyces cerevisiae under different degrees of glucose repression. J. Bacteriol. 115: Michel, R., and W. Neupert Mitochondrial translation products before and after integration into the mitochondrial membrane in Neurospora crassa. Eur. J. Biochem. 36: Murray, D. R., and A. W. Linnane Synthesis of proteolipid protein by yeast mitochondria. Biochem. Biophys. Res. Commun. 49: Rockland, L. B., and J. C. Underwood Small scale filter paper chromatography. A rapid two-dimensional procedure. Anal. Chem. 26: Rogers, P. J., and P. R. Stewart Effects of oxygen tension and glucose repression on mitochondrial protein synthesis in continuous culture of Saccharomyces cerevisiae. J. Bacteriol. 118: Rogers, P. J., and P. R. Stewart Mitochondrial and peroxisomal contributions to the energy metabolism of Saccharomyces cerevisiae in continuous culture. J. Gen. Microbiol. 79: Sierra, M. F., and A. Tzagoloff Assembly of the mitochondrial membrane system. Purification of a mitochondrial product of the ATPase. Proc. Nat. Acad. Sci. U.S.A. 70: Tzagoloff, A Assembly of the mitochondrial membrane system. IV. Role of mitochondrial and cytoplasmic protein synthesis in the biosynthesis of the rutamycin-sensitive adenosine triphosphatase. J. Biol. Chem. 246: Tzagoloff, A., and A. Akai Assembly of the mitochondrial membrane system. VII. Properties of the products of mitochondrial protein synthesis in yeast. J. Biol. Chem. 247: Tzagoloff, A., M. S. Rubin, and M. F. Sierra Biosynthesis of mitochondrial enzymes. Biochim. Biophys. Acta 301: Weber, K., and M. Osborn The reliability of molecular weight determination by dodecyl sulphatepolyacrylamide gel electrophoresis. J. Biol. Chem. 244: