cell-free synthesis of a putative precursor to the rat liver

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 263, No. 36, Issue of December 25, pp. 19821-19826,1988 Printed in U.S.A.

Cell-free Synthesis of a Putative Precursor to the Rat Liver Mitochondrial Glycerol-3-phosphate Dehydrogenase*

(Received for publication, September 14, 1987, and in revised form, August 17, 1988)

Alam Garrib$ and William C. McMurrayS From the Deuartment of Biochemistrv. Medical Science Building, University of Western Ontario, “, London, Ontario, Canda N6A 5Cl

Antibodies to purified glycerol-3-phosphate dehydrogenase were raised in rabbits and purified from serum by affinity chromatography on enzyme-bound Sepharose columns. RNA from membrane-free polyribosomes, or poly(A)+ RNA (total cellular RNA) of rat liver, was translated in a rabbit reticulocyte protein- synthesizing system in the presence of [36S]methionine, and the glycerol-3-phosphate dehydrogenase synthesized was isolated by immunoprecipitation using the antibody. The in vitro product moved on sodium dodecyl sulfate-polyacrylamide gels as a polypeptide that was about 5,000 daltons larger than the subunit of the mature enzyme (74,000 daltons). Digestion of both the mature and the in vitro newly synthesized forms of the enzyme yielded respective sets of peptide fragments which had similar patterns upon sodium dodecyl sulfate-gel electrophoresis. When the presumptive precursor that had been synthesized in vitro was incubated with isolated intact rat liver mitochondria, it was converted to “mature” subunits that were no longer susceptible to externally added proteases. Im- port of the presumptive precursor is dependent upon an electrochemical potential across the inner mitochondrial membranes. The mature form of the protein is assembled in its native location (the outer surface of the inner mitochondrial membrane).

Flavin-linked glycerol-3-phosphate dehydrogenase (EC 1.1.99.5) is located on the outer surface of the inner mitochondrial membrane (1) and converts cystolic L-a-glycero- phosphate to dihydroxyacetone phosphate. This enzyme is found in mitochondria from mammals (2-6), fungus (7), or yeast (8), and in bacteria (9) and provides a mechanism for transfer of reducing equivalents across the mitochondrial membranes (10). The dehydrogenase constitutes less than 1% of the total mitochondrial proteins (ll), but its level can be increased by an order of magnitude upon daily triiodothyronine injection (2, 11). The enzyme consists of four identical subunits of M , 74,000 (11-13).

Most mitochondrial proteins are coded by nuclear genes, synthesized in the cytoplasm, and imported into the mitochondria. Some secretory proteins are transferred into the plasma membrane or across the lipid bilayer by a co-transla-

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Present address: Department of Immunology and Neurobiology, Mount Sinai Hospital, 600 University Ave., Toronto M5GlX7, On- tario, Canada

§ T O whom correspondence and reprint requests should be ad- dressed.

tional mechanism (14). However, this mechanism is not uni- versal, since various studies have shown that the transfer of other proteins particularly those destined for the mitochon- drion proceeds by a post-translational modification process. In such cases, a cytoplasmically made protein may be synthesized in a cell-free system as a large precursor that can be transported into the membrane and then cleaved to the “mature” form in the absence of protein synthesis (15-24). This study deals with the isolation of mRNA coding for mitochondrial flavin-linked glycerol-3-phosphate dehydrogenase. In SDS’-polyacrylamide gels, the primary translation product has an apparent M, 79,000, which is 5000 daltons in excess of the mature glycerol-3-phosphate dehydrogenase.

MATERIALS AND METHODS

Isolation of Glycerol-3-phosphate Dehydrogenase-Rats (male Wis- tar strain) were fed ad libitum and were given daily subcutaneous injections of triiodothyronine as described before (11). The animals were killed by decapitation, and livers were removed and mitochondria isolated (11). Glycerol-3-phosphate dehydrogenase was purified to homogeneity according to the procedure of Garrib and McMurray (11-13). The enzyme activity was determined as described (11, 25).

Preparation of Antibodies-Antibody against rat liver glycerol-3- phosphate dehydrogenase was raised in rabbits by subdermal injection of 100 fig of glycerol-3-phosphate dehydrogenase emulsified in Freund’s complete adjuvant. Four weeks later, the animal was injected with 50 pg of glycerol-3-phosphate dehydrogenase in Freund’s incom- plete adjuvant. This booster injection was repeated 2 weeks later. Anti-glycerol-3-phosphate dehydrogenase antibody was immobilized on a CNBr-activated Sepharose CL-GB column to which purified glycerol-3-phosphate dehydrogenase was linked. The antibody was eluted with 0.17 M glycine, pH 2.3, neutralized, dialyzed against NaCl/ phosphate buffer, pH 7.4, and stored in small aliquots at -70 “C.

Isolation of Membrane-bound and Membrane-free Polyribosomes- Membrane-bound and membrane-free polyribosomes were isolated from rat liver by the method of Ramsey and Steele (26).

Isolation of RNA-Male Wistar rats (-200-250 g) were given daily subcutaneous injections of triiodothyronine (11) in an attempt to increase the glycerol-3-phosphate dehydrogenase mRNA level. Total liver RNA was isolated by the guanidinium/hot phenol method as described by Maniatis et al. (27). The RNA preparation gave sharp 18 and 28 S ribosomal RNA bands and mRNA smear after denatu- ration with glyoxal and dimethyl sulfoxide and electrophoresis on 1.4% agarose gels (27). Electrophoresis running conditions were according to Maniatis et al. (27).

Selection of Poly(A)+ RNA-Poly(A)+ RNA was isolated on an oligo(dT)-cellulose column exactly as described by Maniatis et al. (27) except that at step 7 SDS was omitted from the eluting buffer. The RNA was suspended in sterile distilled water and stored at

Cell-free Tramlation-Polysomal RNA or total cellular poly(A)+ RNA was translated in a nuclease-treated rabbit reticulocyte lysate system (28) without dog pancreas microsomal membranes (29). When in vitro cell-free translation of rat liver mRNA poly(A)+ was carried

The abbreviations used are: SDS, sodium dodecyl sulfate; HEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; FCCP, carbonyl cyanide p-trifluoromethoxyphenylhydrazone.

-70 “C.

19821

19822 Precursor to Glycerol-3-phosphate Dehydrogenase

out at 37 "C, protein synthesis was linear only up to 10 min, and [35S] methionine incorporation in trichloroacetic acid-precipitable material was very low and ended abruptly without the appearance of immu- noprecipitate to anti-glycerol-3-phosphate dehydrogenase. On incubation at 25 "C, total radioactivity incorporated into trichloroacetic acid-precipitable protein was high and rapid until 60 min and contin- ued at a reduced rate up to 120 min. In the translation assays, 1 pg of RNA was added to 24 pl of the translation mixture containing 50 pCi of ~-[~~S]methionine, and incubated at 25 "C for 90 min.

Enzyme-linked Zmmunosorbent Assays-A modification of a stand- ard method (30) was used. This procedure makes use of a sodium carbonate buffer at elevated pH (pH 9.6). The enzyme-linked immunosorbent assay plates were read with a Dynatech MR 600 microplate reader in dual mode, using a 410-nm filter for sample determination and a 740-nm filter for reference. The plate was read periodically, and rates of color development were calculated. An enzyme-linked immunosorbent assay unit was defined as that amount of antibody which would give an increase of 1 absorbance unit per min.

Immunological Detection of Proteins on Nitrocellulose Blots-Pro- teins were electrophoretically transferred from gels to nitrocellulose (Schleicher & Schuell, 0.22-micron pore size) as described by Towbin et al. (31). The nitrocellulose was stained with Amido black for protein band visualization and destained with 45% methanol, 9% acetic acid. Precipitation of glycerol-3-phosphate dehydrogenase on nitrocellulose using affinity purified anti-glycerol-3-phosphate dehydrogenase antibodies, skimmed milk powder, and fixed Staphylococcus aureus cells gave better results than using a second antibody and bovine serum albumin as a blocking reagent. In the presence of nonimmune serum and bovine serum albumin there was a high background level without appearance of the glycerol-3-phosphate dehydrogenase band. During both Western blots and in vitro translation, maximum glycerol-3-phosphate dehydrogenase precipitation was obtained with 5 pg of purified anti-glycerol-3-phosphate dehydrogenase and 50 pl of 10% (w/v) Staphylococcus aureus cells. Nonspecific binding to the nitrocellulose strip was blocked by incubation at 4 "C in blot rinsing buffer containing 5% skimmed milk powder. Antibody binding and immunological detection procedures were as described elsewhere (32).

Liver polyribosomes (5-15 pl) a t a final concentration of 20 Azso units/ml were incubated in the nuclease-treated rabbit reticulocyte lysate system as described above.

Immunoprecipitations-After 90 min at 25 'C, the in vitro translation mixture was transferred on ice and then centrifuged at 160,000 X g for 15 min at 4 "C in a Beckman TL-100 ultracentrifuge. The

with 50 mM Tris-HC1, pH 7.4,150 mM NaCl, and 5 mM EDTA. The supernatant containing the synthesized proteins was diluted 1:10

mixture was boiled for 5 min, and 25 pl of anti-glycerol-3-phosphate dehydrogenase antibody was added. After 12 h at 4 "C, 50 pl of fixed Staphylococcus aureus cells (from a 10% stock suspension) was added, and the suspension was incubated for a further 2 h at room temper- ature with gentle agitation. The Staphylococcus aureus cells were recovered by centrifugation at 4 'C, washed four times with the same buffer but without Triton X-100, and subsequently boiled for 3 min in SDS-Laemmli sample buffer (33).

Trichloroacetic acid precipitation of insoluble polypeptides follow- ing in vitro translation and peptide mapping by partial proteolysis were carried out according to published methods (34-35).

Polyacrylamide Gel Electrophoresis-Electrophoresis on 9% and 5- 20% linear gradient polyacrylamide in 0.1% SDS was carried out according to the procedure of Laemmli (33).

In Vivo Labeling of Mitochondrial Proteins-Rat liver mitochondrial proteins were labeled with [35S]methionine according to the procedure of Shore et al. (22).

Isolation of Mitochondria-Rats (male Wistar 2200 g) were allowed free access to chow and water. They were killed by head decapitation, livers were removed quickly, minced, and mitochondria isolated as described by Greenawalt (36) except that bovine serum albumin was omitted from the isolation medium. Mitochondria were washed sev- eral times and finally resuspended in a medium containing 10 mM HEPES, pH 7.4, and 0.25 M sucrose. For import studies mitochondria (2 mg of protein/ml) were mixed with an equal volume (10-50 pl) of reticulocyte lysate containing freshly made translation products under the direction of liver mRNA.

Rat hind quarter skeletal muscle was trimmed, minced with a pair of scissors, ground, and suspended in ice-cold sucrose-EDTA buffer (11). The mixture was homogenized with three 30-s bursts at a 5 setting using a Brinkmann Polytron interspersed by 90-s rest periods. The homogenate was centrifuged at 2000 X g for 10 min. The pellet was rehomogenized in 5 volumes of the same medium and the com-

bined supernatants centrifuged at 6000 X g for 10 min to collect the mitochondrial pellet.

EDTA buffer (11). The homogenate was centrifuged at 850 X g for Rat brain was minced and homogenized in 10 volumes of sucrose-

15 min. The supernatant was discarded and the pellet rehomogenized in 5 volumes of sucrose-EDTA buffer and centrifuged at 6000 X g for 15 min to collect the mitochondrial pellet.

Rat heart was excised, minced, and homogenized in 5 volumes of sucrose-EDTA buffer, 0-4 "C. The suspension was homogenized with a Brinkmann Polytron as described above. Mitochondria were isolated by the procedure of Garrib and McMurray (11) and resuspended in the import buffer as described above.

Osmotic Shock-Mitochondria were pelleted at 15,000 rpm (29,000 X g) for 10 min and resuspended in 0.15 M KC1, 1 mM dithiothreitol using a ground-glass homogenizer and kept at 0 "C for 10 min with stirring. The homogenate was centrifuged at 10,000 X g, and the pellet was homogenized in the same buffer. The above procedure was repeated, and finally the pellets were resuspended in a minimal volume of 0.25 M sucrose, 0.01 M Tris-HC1, pH 8.5. The suspension was diluted to 10 times its volume with 1 mM dithiothreitol in water and stirred for 1 h at 0-4 "C to osmotically shock the outer membranes. The homogenate was centrifuged at 14,000 X g for 15 min to collect the intact inner membranes. The supernatant contained the outer membrane. The two fractions were washed twice with the medium containing 10 mM HEPES, pH 7.4, and 0.25 M sucrose.

Isolation of Soluble Matrix Contents-All steps were performed at 0-4 "C. The mitoplasts (as prepared above) were resuspended in 10 mM Tris-HC1, pH 7.4, using a Dounce homogenizer. The suspension was frozen and thawed twice and disrupted as described by Bonhi et al. (37) and centrifuged at 115,000 X g for 2 h. The pellet was discarded, and the supernatant representing the matrix fraction was stored frozen at -70 "C.

RESULTS

Effect of Triiodothyronine on Levels of Translatable mRNA for Glycerol-3-phosphate Dehydrogenase-Rat liver mitochondrial glycerol-3-phosphate dehydrogenase (flavin-linked) is inducible with daily subcutaneous injections of triiodothyronine (2, 11). In order to reinforce this observation, mRNA levels for this enzyme were estimated by means of cell-free translation. Liver mRNA was isolated from rats that had received triiodothyronine injections for 7 days and from the controls as described under "Materials and Methods" and translated in a messenger-dependent rabbit reticulocyte lysate translation system. Proteins were immunoprecipitated, electrophoresed, fluorographed, and radioautographed as described earlier. There was approximately a &fold increase in p-glycerol-3-phosphate dehydrogenase (where p indicates putative precursor) content in triiodothyronine-treated animals as opposed to those in controls.

When rat hepatic free and membrane-bound polyribosomes were isolated by the procedure described above (26), it was found that the membrane-free polyribosomal fraction had a higher level of glycerol-3-phosphate dehydrogenase mRNA activity than the membrane-bound polyribosomal fraction (-8-fold/unit of poly(A)+ RNA). Due to the low levels of glycerol-3-phosphate dehydrogenase mRNA in normal rat liver, membrane-free polyribosomal mRNA or total liver mRNA-poly(A)+ from rats treated with triiodothyronine were used in vitro in translation mixtures.

Hepatic poly(A)+ mRNA was translated in a messenger- dependent rabbit reticulocyte lysate system as described under "Materials and Methods." The translation products were immunoprecipitated, electrophoresed, fluorographed, and radioautographed. Fig. 1, A and B, shows that the polypeptide synthesized in vitro had an altered electrophoretic mobility compared with the mitochondrial glycerol-3-phosphate dehydrogenase. When a similar experiment was carried out but in the presence of nonimmune rabbit serum no radioactivity band corresponding to the glycerol-3-phosphate dehydrogenase band was apparent. The putative precursor has an appar-

Precursor to Glycerol-3-phosphate Dehydrogenase 19823

FIG. 1. A, identification of p-glycerol-3-phosphate dehydrogenase. In vitro translation products were immunoprecipitated with anti- glycerol-3-phosphate dehydrogenase antibody ( l a n e B ) or nonimmune serum ( l a n e C ) . [35S]Methionine-labeled rat liver mitochondrial glycerol-3-phosphate dehydrogenase immunoprecipitated with antibody ( l a n e A ) . GDPH, glycerol-3-phosphate dehydrogenase. B, in vitro translation products. Lane A, 35S-labeled mitochondria were lysed, immunoprecipitated with anti-glycerol-3-phosphate dehydrogenase, electrophoresed, and fluorographed; lane €3, immunoprecipitated product of in vitro translation using total hepatic RNA, poly(A)'; lane c, in vitro translation carried out using membrane-bound RNA, and lane D, in vitro translation carried out with membrane-free polysomes.

ent size in SDS gels equivalent to 5,000 daltons in excess of the molecular weight of the mitochondrial form of glycerol-3- phosphate dehydrogenase; its labeling accounted for only 0.7% of the radioactivity incorporated into total trichloroacetic acid-precipitable product.

In order to ascertain if the larger polypeptide was a precursor form of the glycerol-3-phosphate dehydrogenase after in Vitro translation, immunoprecipitation was performed with a limited amount of anti-glycerol-3-phosphate dehydrogenase antibody in the presence of excess unlabeled mitochondrial glycerol-3-phosphate dehydrogenase. Under these conditions, p-glycerol-3-phosphate dehydrogenase was the only polypeptide that competed for binding of the antibody to glycerol-3- phosphate dehydrogenase (Fig. 2). This result suggests that the antibody used is very specific for glycerol-3-phosphate dehydrogenase and does not precipitate any other polypeptide.

Additional evidence concerning the coidentities of the im- munoprecipitates was derived by digestion of the mature subunit (glycerol-3-phosphate dehydrogenase) and its larger form (p-glycerol-3-phosphate dehydrogenase) with Staphylo- coccus aweus V8 protease. The proteolytic fragments were analyzed on a 12% SDS-polyacrylamide gel. The two peptide maps were similar but not identical (Fig. 3). This demon- strates that the larger polypeptide cross-reacts immunologi- cally with the mitochondrial glycerol-3-phosphate dehydrogenase subunit and that it also exhibits a considerable sequence homology.

The Precursor Is Processed and Imported into Mitochondria in Absence of Protein Synthesis-Protein synthesis was carried out in a rabbit reticulocyte lysate system in the presence of rat liver poly(A)+ RNA and [35S]methionine for 90 min. Polypeptide synthesis was stopped by addition of cycloheximide. A 100-pl portion of the translation mixture was added to intact mitochondria (0.2-0.3 mg in mitochondrial isolation buffer) and incubated at 25 "C. The reaction was terminated by chilling and mitochondria separated by centrifugation. Both mitochondria and supernatant were analyzed for labeled

FIG. 2. Mitochondrial proteins were labeled by injecting rats intraperitoneally with 5 mCi of ['"SJmethionine at 0 time and then 4 mCi again after 90 min. Lane A, mitochondria were isolated from liver, lysed, immunoprecipitated with anti-glycerol-3- phosphate dehydrogenase antibody, and electrophoresed in a 5-15% linear gradient SDS-polyacrylamide gel; lane B, membrane-free po- lyribosome-dependent in vitro translation product after immunoprecipitation; lane C, poly(A)' mRNA added to translation mixture; lanes D and E, in vitro translation products were immunoprecipitated in the presence of excess unlabeled glycerol-3-phosphate dehydrogenase.

A B C D

a 0

FIG. 3. Peptide mapping of p-glycerol-3-phosphate dehydrogenase and mature glycerol-3-phosphate dehydrogenase. Radioactive mature glycerol-3-phosphate dehydrogenase was isolated by immunoprecipitation from mitochondria of rat liver that was labeled as described in Fig. 2. Labeled p-glycerol-3-phosphate dehydrogenase was isolated from an in vitro translation system as described under "Materials and Methods." Samples were digested with V8 protease from Staphylococcus aureus according to Cleveland (35) and electrophoresed in 5-15% linear gradient SDS gel. Lanes A and C, glycerol-3-phosphate dehydrogenase and p-glycerol-3-phosphate dehydrogenase, respectively; lanes B and D, glycerol-3-phosphate dehydrogenase and digest with 15 pg of protease.

proteins by immunoprecipitation and SDS/polyacrylamide gel electrophoresis (Fig. 4). Where mitochondria were omitted only the larger form of glycerol-3-phosphate dehydrogenase was present, whereas the mitochondrial pellets initially contained both the larger and the mature form which became predominant afterward.

Intact rat liver mitochondria were incubated with proteinase K (1 mg/ml final) for 15 min at 4 "C. The mixture was further incubated with phenylmethylsulfonyl fluoride and mRNA-dependent rabbit reticulocyte lysate translation product at 30 "C as described above. Under these conditions the

19824

1 2 3

Precursor to Glycerol-3-phosphate Dehydrogenase

4 5 6 0 2 4 5 - -

- 0 - 0 " 0 .

FIG. 4. In vitro import and post-translational processing of glycerol-3-phosphate dehydrogenase by isolated rat liver mitochondria. Lane l, mitochondria were omitted from the reticulocyte lysate system; Lanes 2-6, mitochondria were added to reticulocyte lysate and the mixture incubated for 0, 15, 30, 50, and 60 min, respectively. At times indicated, mitochondrial products were immunoprecipitated with antibody against glycerol-3-phosphate dehydrogenase, electrophoresed, and fluorographed.

precursor and mature forms of glycerol-3-phosphate dehydrogenase failed to appear in the mitochondria. One immediate explanation for such an observation is that there may be receptors on the surface of mitochondria which either specif- ically or nonspecifically bind the precursor form of glycerol- 3-phosphate dehydrogenase and help its transport inside the outer mitochondrial membrane. Treatment of intact mitochondria with proteinase K leads to rapid loss of such membrane-surface receptors. Thus, the precursor polypeptide could not be internalized by the mitochondria.

Rat muscle, heart, and brain mitochondria were isolated as described elsewhere. When mitochondria from such tissues were incubated with the mRNA-dependent rabbit reticulocyte lysate translated products, only one band of radioactive polypeptide with a molecular mass of 74,000 daltons was apparent upon immunoprecipitation with anti-glycerol-3-phosphate dehydrogenase antibody, SDS-polyacrylamide gel electrophoresis, and fluorography. This band had the same electrophoretic mobility as the mature glycerol-3-phosphate dehydrogenase. The rate of import of p-glycerol-3-phosphate dehydrogenase by mitochondria from various organs was similar. Mitochon- dria from various sources when pretreated with proteinase K failed to bind and internalize the in vitro rabbit reticulocyte lysate-translated product. These results suggest that mitochondria from various tissues import p-glycerol-3-phosphate dehydrogenase by a similar mechanism.

Mature glycerol-3-phosphate dehydrogenase failed to be imported by intact mitochondria from various sources. One immediate explanation for such an observation is that the mature polypeptide which has lost its targeting signal sequence (which is crucial for import) cannot be internalized by intact mitochondria.

In Vivo Transport ofp-Glycerol-3-phosphate Dehydrogenase from the Cytosol to Mitochondria-Using rat liver explants for pulse and chase experiments, it is evident that p-glycerol- 3-phosphate dehydrogenase disappears completely from the cytosol within 3 min (Fig. 5). We deduce that transfer of p- glycerol-3-phosphate dehydrogenase from cytosol to mitochondria occurs very rapidly and that accumulated p-glycerol- 3-phosphate dehydrogenase in the cytosolic compartment is highly unstable. A similar time-dependent disappearance of p-glycerol-3-phosphate dehydrogenase was observed, when L6 rat myoblasts were pulsed and chased with [35S]methionine and cold methionine, respectively.

" " - 0

FIG. 5. Disappearance of labeled p-glycerol-3-phosphate dehydrogenase from the cytosolic compartment of liver explants. Rat liver explants were incubated with ["S]methionine for 40 min in modified Eagle's medium. The explants were then incubated in fresh medium containing 2 mM unlabeled methionine and 15 pg/ ml cycloheximide for various lengths of time. At times indicated the explants were removed, chilled, homogenized, and postmicrosomal cytosol fraction prepared by centrifugation. The cytosolic fractions were immunoprecipitated with antibody against glycerol-3-phosphate dehydrogenase, and the products were electrophoresed and fluorographed.

FIG. 6. Processing of the glycerol-3-phosphate dehydrogenase precursor by isolated rat liver mitochondria in the presence of inhibitors of mitochondrial electron transport or uncouplers of mitochondrial oxidative phosphorylation. Aliquots of mRNA-directed translation mixture containing [3sS]methionine were incubated with intact mitochondria (2 mg/ml protein concentration) and mitochondrial inhibitors. After 1 h of incubation at 27 "C, samples were immunoprecipitated, electrophoresed, and fluorographed. Lune I, no mitochondria added; lane 2, no inhibitor present; lane 3, 12 p~ rotenone; lane 4, 1 p~ antimycin; lane 5, 10 pM FCCP; lane 6,5 pM valinomycin; lane 7, 50 pM 2,4-dinitrophenol; lane 8, 12 p~ rotenone + 1 p~ antimycin; lane 9, 2 p~ oligomycin; lane 10,50 p~ atractyloside.

Inhibition of Maturation of p-Glycerol-3-phosphate Dehy- drogenase by Uncouplers and Inhibitors of Electron Trans- port-It is well documented in the literature that import of many cytoplasmically synthesized precursor proteins is dependent upon an electrochemical potential across the mitochondrial inner membranes. With this point in mind, we investigated the maturation process of p-glycerol-3-phosphate dehydrogenase in the presence of uncouplers or electron transport inhibitors (Fig. 6). In the absence of mitochondria only the precursor was present. Mitochondria, in the absence of inhibitors, were able to import and process p-glycerol-3-phosphate dehydrogenase ( l a n e 2) to the mature form. Rotenone, antimycin, and KCN impair the processing of the precursor

Precursor to Glycerol-3-phosphate Dehydrogenase 19825

to the mature subunit by intact mitochondria. FCCP, valinomycin, and 2,4-dinitrophenol, uncouplers of oxidative phosphorylation, each blocked the formation of the mature form of glycerol-3-phosphate dehydrogenase. However, oligomycin, an inhibitor of oxidative phosphorylation at the level of F1- ATPase, or atractyloside, an inhibitor of the adenine nucleo- tide translocation system, did not prevent maturation of the p-glycerol-3-phosphate dehydrogenase by mitochondria. Ro- tenone and antimycin, in combination (lane 9 ) , blocked processing of p-glycerol-3-phosphate dehydrogenase completely, although when used individually neither totally obliterated the appearance of the mature subunit into the mitochondria (compare with lanes 3 and 4 ) . Preliminary studies showed that maturation of p-glycerol-3-phosphate dehydrogenase was also prevented by treatment of mitochondria with O-phen- anthroline. This suggests that the precursor form of the enzyme is processed by a chelator-sensitive protease. Incu- bation of an in vitro translated product (mRNA-dependent) with mitochondrial matrix isolate converted the p-glycerol-3- phosphate dehydrogenase into the mature form. This sequence of events did not happen in the presence of O-phen- anthroline. One can conclude that the precursor form of glycerol-3-phosphate dehydrogenase is internalized by the inner membranes of mitochondria into the matrix where it is processed.

The Mature Form of Glycerol-3-phosphate Dehydrogenase Is Bound to the Outer Surface of the Inner Membrane-In order to identify the intracellular location of mature glycerol- 3-phosphate dehydrogenase, rat liver mitochondrial proteins were labeled in vivo by intraperitoneal injection of rats with [35S]methionine as described under “Materials and Methods.” The labeled cells were then fractionated into mitochondria and postmitochondrial supernatant. The mitochondrial fraction was further subfractionated into outer membrane and mitoplasts. One-half of each fraction was treated with trypsin to digest exposed proteins, and each fraction was then analyzed for glycerol-3-phosphate dehydrogenase by immunoprecipitation in the presence of an excess of antibody to glycerol- 3-phosphate dehydrogenase, SDS-gel electrophoresis, and fluorography. The mature form of glycerol-3-phosphate dehydrogenase was found exclusively (100%) in the mitochondria and was resistant to proteolysis by externally added trypsin since it was protected by the outer membrane of mitochondria. When mitochondria were converted to mitoplasts, mature glycerol-3-phosphate dehydrogenase (92%) re- mained associated with the mitoplasts but was degraded by added trypsin. These data suggest that the mature form of glycerol-3-phosphate dehydrogenase is located mainly on the outer surface of the inner mitochondrial membrane. This observation is supported by the fact that purified glycerol-3- phosphate dehydrogenase is accessible to externally added trypsin when it is reconstituted into diphosphatidylglycerol liposomes.

DISCUSSION

This paper describes the isolation and characterization of a novel precursor protein destined for the outer surface of the inner mitochondrial membrane. This protein, which is precip- itated by anti-glycerol-3-phosphate dehydrogenase antibody after in vitro synthesis in a rabbit reticulocyte lysate translation system, has a slightly slower mobility than the holoenzyme on a SDS gel. The unlabeled holoenzyme competed with the de nouo-synthesized product for binding sites on the anti- glycerol-3-phosphate dehydrogenase antibody. Peptide mapping of both the presumptive precursor and the mature subunit proteins showed that the two polypeptides are similar.

Only one polypeptide band with a molecular mass of 74,000 daltons is apparent when rat liver mitochondria were labeled in vivo with [35S]methionine and subsequently excised, fractionated, immunoprecipitated, electrophoresed, and fluorographed. All these data indicate that the larger form of the enzyme with a molecular mass of about 79,000 daltons is a precursor form of the holoenzyme, i.e. it is synthesized in the cytosolic compartment of the cell in a larger form and converted into the smaller subunit with the mitochondrial compartment. Over the years there have been documentations of protein trafficking between the cytosol and mitochondria of yeasts (19-21), Neurospora craSsa (38), and rat liver (22-24). Based on labeling and genetic studies, the amount of protein contributed by the nuclear and mitochondrial genomes has been evaluated. It has been conservatively estimated that the nucleocytoplasmic system contributes greater than 90% of the protein mass to the organelle, glycerol-3-phosphate dehydrogenase being one of the extramitochondrial products.

This report indicates that import of the presumptive precursor by intact mitochondria is energy-dependent. The res- piratory inhibitors rotenone and antimycin partially impair processing of the precursor to the mature subunit by intact mitochondria; however, complete inhibition is observed when the electron transport inhibitors are used in combination. On the other hand, uncouplers of oxidative phosphorylation, e.g. 2,4-dinitrophenol, valinomycin, and FCCP, but not oligomycin nor atractyloside prevent the appearance of the mature subunit. The findings suggest that a transmembrane electrochemical gradient, but not ATP itself, is an essential energetic requirement for uptake of the precursor form of glycerol-3- phosphate dehydrogenase. A similar energy dependence for processing of cytoplasmically made precursors of mitochondrial proteins in rat liver and yeast has been demonstrated by Kolansky et al. (39) and Nelson and Schatz (40).

The localization of the mitochondrial glycerol-3-phosphate dehydrogenase on the outer surface of the inner membrane has been deduced from various observations (41): (a) glycerol phosphate does not permeate the inner membrane of mitochondria and (6) the mitochondrial glycerol-3-phosphate dehydrogenase reacts with the membrane-impermeable electron acceptor ferricyanide. (This is in contrast to the situation with succinate dehydrogenase, which can only interact with ferricyanide after breakage of the membrane.) Our studies with the labeled p-glycerol-3-phosphate dehydrogenase and fractionation of the mitochondrial membranes suggest that the labeled glycerol-3-phosphate dehydrogenase is imported into its correct locale in the outer phase of the inner mitochondrial membrane.

Once precursor proteins are synthesized, they are rapidly cleared from the cytosol and imported into mitochondria. Precursors can be accumulated in uiuo by growing cells in the presence of uncouplers of oxidative phosphorylation; under these conditions import is blocked. However, the accumulated precursors are rapidly degraded in contrast to the more stable mature forms (42). Pulse-labeling experiments reveal that accumulated precursor proteins have differential rates of deg- radation. The precursor to the @-subunit of F1-ATPase is stable with a half-life of about 50 min. On the other hand, precursors of cytochrome c1, aspartate aminotransferase, and carbamyl-phosphate synthetase are very unstable and are degraded with half-lives of about 10, 5, and 2-3 min, respectively. Pulse-labeling experiments carried out in our laboratory indicate that the precursor to glycerol-3-phosphate dehydrogenase is very unstable with a half-life of 3 min.

Most of the precursor proteins are imported into mitochondria where they undergo maturation. Maturation does not

19826 Precursor to Glycerol-3-phosphate Dehydrogenase

take place in the cytosolic compartment because the matura- 19. tion process is an integral part of the import pathway and can 20, only happen after certain import steps have occurred. The nature of these steps and details of the role of putative matrix 21. protease in the processing await sequence determinations of the mature and precursor forms of the glycerol-3-phosphate 22. dehydrogenase. 23.

1. 2.

3.

4. 5.

6.

7. 8.

9.

10.

11.

12.

13.

14.

15.

16. 17.

18.

REFERENCES Klingenberg, M. (1970) Eur. J. Biochem. 13, 247-252 Lee, Y. P., and Lardy, H. A. (1965) J. Biol. Chem. 240 , 1427-

Schenkman, J. B., Richert, D., and Westerfeld, W. W. (1965)

Hess, R., and Pearse, A. G. E. (1961) Nature 191 , 718-719 Salganicoff, L., and Fukami, M. H. (1972) Arch. Biochem. Bio-

Swierczynski, J., Scislowski, P., and Aleksandrowicz, Z. (1976)

Courtright, J. B. (1975) Arch. Biochem. Biophys. 167, 21-33 Von Jagow, G., and Klingenberg, M. (1960) Eur. J. Biochem. 12 ,

Kistler, N. S., Hirsch, C. A., Cozzarelli, N. R., and Lin, E. C.

Haddock, B. A., and Jones, C. W. (1977) Bacteriol. Rev. 41,47-

Garrib, A., and McMurray, W. C. (1986) J. Biol. Chem. 261 ,

Garrib, A., and McMurray, W. C. (1985) Proc. Can. Fed. Biol. SOC. Annu. Meet. 28, 227

Garrib, A., and McMurray, W. C. (1986) Proc. Can. Fed. Biol. SOC. Annu. Meet. 2 9 , 106

Blobel, G. (1977) in International Cell Biology, 1976-1977 (Brink- ley, B. R., and Porter, K. R., e&) pp. 318-325, The Rockefeller University Press, New York

Dobberstein, B., Blobel, G., and Chua, N. H. (1977) Proc. Natl. Acud. Sci. U. S. A. 74,1082-1085

Highfield, P. E., and Ellis, R. J. (1978) Nature 271,420-424 Goldman, B. M., and Blobel, G. (1978) Proc. Natl. Acud. Sci.

Ohashi, A., Gibson, J., Gregor, I., and Schatz, G. (1982) J. Biol.

1436

Endocrinology 76 , 1055-1061

phys. 153 , 726-735

Biochim. Biophys. Actu 429,46-54

583-592

(1969) J. Baeteriol. 100 , 1133-1135

99

8042-8048

U. S. A. 75,5066-5070

Chem. 257.13042-13047

24.

25.

26.

27.

28.

29.

30. 31.

32.

33. 34.

35.

36. 37.

38.

39.

40.

41. 42.

Reid, G. A., and Schatz, G. (1982) J. Biol. Chem. 257, 13056-

Reid, G. A., and Schatz, G. (1982) J. Biol. Chem. 2 5 7 , 13062-

Daum, G., Bohni, P. C., and Schatz, G. (1982) J. Biol. Chem.

Shore, G. C., Carignan, P., and Raymond, Y. (1979) J. Biol. Chem.

Argan, C., and Shore, G. C. (1985) Biochem. Biophys. Res. Com-

Raymond, Y., and Shore, G. C. (1979) J. Biol. Chem. 254,9335-

Garrib, A., and McMurray, W. C. (1984) Anal. Biochem. 139 ,

Ramsey, J. C., and Steele, W. J. (1976) Biochemistry 15 , 1704- 1712

Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular ClonirwA Laboratory Manual, Cold Spring Harbor Laboratory.

13061

13067

257,13028-13033

254,3141-3144

mun. 131,289-298

9338

319-321

"

Cold Siring Harbor,-NY - .

Pelham. H. R. B.. and Jackson. R. J. (1976) Eur. J. Biochern. 67, 247-256

. .

Jackson, R. C., and Blobel, G. (1977) Proc. Natl. Acud. Sci.

Engvall, E. (1980) Methods Enzymol. 70,419-439 Towbin, J., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acud.

Sci. U. S. A. 7 6 , 4350-4354 Macceccluni, M. L., Rudin, Y., Blobel, G., and Schatz, G. (1979)

Proc. Natl. Acud. Sci. U. S. A. 7 6 , 343-347 Laemmli, U. K. (1970) Nature 227,680-685 Mori, M., Mlura, S., Tatibana, M., and Cohen, P. P. (1981) J.

Cleveland, D. W., Fisher, S. G., Kirschner, M. W., and Laemmli,

Greenawalt, J. W. (1974) Methods Enzymol. 31,310-323 Boehni, P., Gasser, S., haver , C., and Schatz, G. (1980) in

Organization and Expression of the Mitochondrial Genome (Kroon, A. M., and Saccone, C., eds) p. 423, Elsevier/North- Holland, Amsterdam

Hallermayer, G., Zimmermann, R., and Neupert, W. (1977) Eur.

u. s. A. 74, m a 5 6 0 2

Biol. Chem. 256,4127-4132

U. K. (1977) J. Biol. Chem. 252, 1102-1106

J. Biochem. 81,523-532 Kolansky, D. M.,.Conboy, J. G., Fenton, W. A., and Rosenberg,

L. E. (1982) J. Biol. Chem. 257, 8461-8471 Nelson, N., and Schatz, G. (1979)-Proc. Natl. Acud. Sci. U. S. A.

Klingenberg, M. (1970) Eur. J. Biochern. 13 , 247-252 Reid, G. A., and Schatz, G. (1982) J. Biol. Chem. 257 , 13062-

76,4365-4369

13067

cell-free synthesis of a putative precursor to the rat liver

Documents