Portal de Periódicos da CAPES

Mss51p and Cox14p jointly regulate mitochondrial Cox1p expression in Saccharomyces cerevisiae

2004; Springer Nature; Volume: 23; Issue: 17 Linguagem: Inglês

10.1038/sj.emboj.7600358

1460-2075

Antoni Barrientos, Andrea Zambrano, Alexander Tzagoloff,

Plant biochemistry and biosynthesis

Article12 August 2004free access Mss51p and Cox14p jointly regulate mitochondrial Cox1p expression in Saccharomyces cerevisiae Antoni Barrientos Corresponding Author Antoni Barrientos Department of Biological Sciences, Columbia University, New York, NY, USA Department of Neurology, The John T Macdonald Foundation Center for Medical Genetics, University of Miami School of Medicine, Miami, FL, USA Search for more papers by this author Andrea Zambrano Andrea Zambrano Department of Neurology, The John T Macdonald Foundation Center for Medical Genetics, University of Miami School of Medicine, Miami, FL, USA Search for more papers by this author Alexander Tzagoloff Alexander Tzagoloff Department of Biological Sciences, Columbia University, New York, NY, USA Search for more papers by this author Antoni Barrientos Corresponding Author Antoni Barrientos Department of Biological Sciences, Columbia University, New York, NY, USA Department of Neurology, The John T Macdonald Foundation Center for Medical Genetics, University of Miami School of Medicine, Miami, FL, USA Search for more papers by this author Andrea Zambrano Andrea Zambrano Department of Neurology, The John T Macdonald Foundation Center for Medical Genetics, University of Miami School of Medicine, Miami, FL, USA Search for more papers by this author Alexander Tzagoloff Alexander Tzagoloff Department of Biological Sciences, Columbia University, New York, NY, USA Search for more papers by this author Author Information Antoni Barrientos 1,2, Andrea Zambrano2 and Alexander Tzagoloff1 1Department of Biological Sciences, Columbia University, New York, NY, USA 2Department of Neurology, The John T Macdonald Foundation Center for Medical Genetics, University of Miami School of Medicine, Miami, FL, USA *Corresponding author. Department of Neurology, The John T Macdonald Foundation Center for Medical Genetics, University of Miami School of Medicine, Miami, FL 33136, USA. Tel.: +1 305 243 7683; Fax: +1 305 243 3914; E-mail: [email protected] The EMBO Journal (2004)23:3472-3482https://doi.org/10.1038/sj.emboj.7600358 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Figures & Info Mutations in SURF1, the human homologue of yeast SHY1, are responsible for Leigh's syndrome, a neuropathy associated with cytochrome oxidase (COX) deficiency. Previous studies of the yeast model of this disease showed that mutant forms of Mss51p, a translational activator of COX1 mRNA, partially rescue the COX deficiency of shy1 mutants by restoring normal synthesis of the mitochondrially encoded Cox1p subunit of COX. Here we present evidence showing that Cox1p synthesis is reduced in most COX mutants but is restored to that of wild type by the same mss51 mutation that suppresses shy1 mutants. An important exception is a null mutation in COX14, which by itself or in combination with other COX mutations does not affect Cox1p synthesis. Cox14p and Mss51p are shown to interact with newly synthesized Cox1p and with each other. We propose that the interaction of Mss51p and Cox14p with Cox1p to form a transient Cox14p–Cox1p–Mss51p complex functions to downregulate Cox1p synthesis. The release of Mss51p from the complex occurs at a downstream step in the assembly pathway, probably catalyzed by Shy1p. Introduction Assembly of yeast cytochrome oxidase (COX) requires the assistance of at least 20 nuclear gene products (McEwen et al, 1986; Tzagoloff and Dieckmann, 1990). These proteins act at all stages of the assembly process, including processing (Seraphin et al, 1989) and translation (Costanzo and Fox, 1993) of the mitochondrially encoded mRNAs for subunits 1, 2, and 3, membrane insertion of the hydrophobic subunits (Hell et al, 2001), and maturation of the heme and copper centers (Glerum et al, 1996a; 1996b; Barros et al, 2001). Although much progress has been made in understanding some of the events leading to the assembly of COX, the functions of a number of gene products essential for this process have yet to be clarified (Barrientos et al, 2002a; 2002b). A case in point is Shy1p, a yeast mitochondrial protein needed for full expression of COX (Mashkevich et al, 1997; Barrientos et al, 2002b). The function of this protein is of considerable interest because mutations in its human homologue, Surf1p, have been shown to be responsible for most diagnosed cases of Leigh's syndrome (Tiranti et al, 1998; Zhu et al, 1998), a neuromuscular disease presenting a COX deficiency (DiMauro and De Vivo, 1996). Extragenic suppressors of shy1 null mutants have been mapped to MSS51, a nuclear gene coding for a Cox1p-specific translational activator (Barrientos et al, 2002b), suggesting that unassembled Cox1p may downregulate its own translation by competitively trapping a translational activator complex in which Mss51p is one of the components. According to this model, the translational block is relieved by Shy1p-dependent assembly of Cox1p making Mss51p available for Cox1p synthesis (Barrientos et al, 2002b). In the present study, we show that Cox1p synthesis is reduced in most COX assembly mutants, cox14 mutants being an exception. Mutant forms of Mss51p are able to restore Cox1p expression in strains carrying null alleles of either COX structural genes or genes coding for COX assembly factors. We also present evidence for an interaction of newly synthesized Cox1p with Mss51p and Cox14p, but not Shy1p. We propose that the formation and turnover of Cox1p/Mss51p/Cox14p couple Mss51p-dependent translation of Cox1p to its utilization for COX assembly. Results Synthesis of Cox1p is repressed in assembly-arrested mutants To ascertain if the Cox1p labeling defect previously noted in shy1 mutants (Barrientos et al, 2002b) is general to all mutants blocked in COX assembly, a wide range of strains with lesions in COX subunits or assembly factors were pulsed with [35S]methionine in vivo in the presence of cycloheximide. Labeling of Cox1p, but not Cox2p or Cox3p, was visibly reduced in most mutants (Figures 1, 2 and 3). The only exception was the cox14 mutant (Glerum et al, 1995). The precise function of Cox14p is not known at present. Figure 1.In vivo labeling of mitochondrial gene products in COX mutants. Wild-type (W303-1A) and mutant cells (described in Table I) were labeled with [35S]methionine at 30°C for the indicated times in the presence of cycloheximide. One-half of each culture was incubated in the presence of 2 mg/ml chloramphenicol during the last 2 h of growth prior to labeling (+CAP). Samples were removed after the indicated times of labeling and processed as detailed in Materials and methods. The mitochondrial translation products are identified in the margin. Cox2p is not processed in ΔIMP1 and ΔIMP2 mutants. The Cox2p precursor (p) in these strains migrates slower that the mature Cox2p (m). Download figure Download PowerPoint Figure 2.In vivo labeling of mitochondrial gene products in COX mutants expressing different alleles of MSS51. MSS51 and the suppressor mss51T167R, which partially suppress the respiratory defect of shy1 mutants, were cloned in YIp351. The resultant constructs pSG91/ST9 and pSG91/ST6, respectively, were integrated at the chromosomal LEU2 locus of the indicated mutants (see Table I for description of mutants). The mutants (−) and transformants were labeled with [35S]methionine at 30°C for 15 min in the presence of cycloheximide. Download figure Download PowerPoint Figure 3.In vivo labeling of mitochondrial gene products in COX14 mutants and effect of overexpression of COX14p on Cox1p labeling. (A) Wild type (W303), different COX mutants, and the same mutants carrying an additional null mutation in COX14 were labeled with [35S]methionine at 30°C for 15 min in the presence of cycloheximide. With the exception of CYC3, which codes for a cytochrome c-specific heme lyase, the functions affected in the different strains are described in Table I. (B) The wild-type W303-1A, the indicated COX mutants, and the same strains harboring COX14 on a multicopy plasmid (pG93/T1) were pulse-labeled with [35S]methionine and equivalent amounts of protein were separated by SDS–PAGE on a 17.5% polyacrylamide gel. Download figure Download PowerPoint The reduced Cox1p labeling in the mutants was seen even with 5 min pulses, the shortest time at which incorporation of [35S]methionine into the mitochondrial translation products could be consistently detected in the wild type. This phenotype is easy to explain in the case of the mss51 and pet309 mutants, both of which have mutations in Cox1p-specific translational factors (Decoster et al, 1990; Manthey and McEwen, 1995). Similarly, the poor labeling of Cox1p in the oxa1 mutant (Figure 3) may be the result of rapid turnover when Oxa1p-dependent membrane insertion of Cox1p is blocked (Hell et al, 2001). The Cox1p deficit in the other strains, however, is difficult to rationalize since the functions affected are unrelated to translation of this protein. With the exception of the mss51, pet309, and oxa1 mutants, Cox1p labeling was increased in mutants and in wild type when cells were preincubated in chloramphenicol prior to the pulse (Figure 1A and B). The improved translation of mitochondrial products following chloramphenicol treatment is presumed to occur as a result of the larger pools of nuclear-encoded subunits available for assembly of intermediates and/or because of the accumulation of nuclear-encoded factors required for mitochondrial gene expression. Methionine incorporation into Cox1p in the mutants after the chloramphenicol incubation was comparable to that seen in wild type under normal pulse-labeling conditions, indicating that the translation apparatus is fully functional in the mutants and that the phenotype stems from a decreased rate of synthesis and/or increased rate of turnover of Cox1p. Turnover seemed less likely in view of the ability of cox14 mutants to synthesize Cox1p at rates similar to wild type even though they are also blocked in COX assembly and displays low steady-state concentrations of Cox1p and Cox2p (Glerum et al, 1995). This and other observations discussed below favor decreased translation as the more likely explanation for the observed deficit of newly synthesized Cox1p in the various COX-deficient mutants. Cox1p synthesis defect in COX assembly mutants in suppressed by the mss51T167R allele A single copy of the mss51T167R allele or an extra copy of wild-type MSS51 was shown to suppress partially the respiratory defect of shy1 mutants by increasing Cox1p translation (Barrientos et al, 2002b). This suggested that the shy1 mutation may inactivate or reduce the effective concentration of Mss51p as a translational activator of the COX1 mRNA. These observations have been extended to other COX-deficient mutants. Mitochondrial translation products were labeled in vivo with [35S]methionine in an assortment of COX mutants, with and without an extra copy of the wild-type MSS51 or the mss51T167R suppressor integrated at the leu2 locus of nuclear DNA. In all the strains except the cox14 mutant examined, synthesis of Cox1p, but not of the other COX subunits, was markedly increased by the suppressor and to a lesser extent also by the extra copy of wild-type MSS51 (Figure 2). It is significant that the higher expression of Cox1p leads to a partial rescue of respiration and COX activity in the shy1 mutant (Barrientos et al, 2002b), but not in any of the other mutants examined. This suggests that the functions of Shy1p and Mss51p are related. In other COX mutants, however, restoration of normal rates of Cox1p synthesis by the mss51T167R suppressor (see ΔCOX18 in Figure 2) is not a sufficient condition to compensate for the assembly defect because the impaired function is unrelated to expression of this subunit. Increased Cox1p synthesis in cox14 mutants is epistatic in mutants with lesions in other COX assembly factors The lack of effect of the cox14 mutation on Cox1p synthesis (Figures 1, 2, 3 and 4) suggested that Cox14p might negatively regulate translation of this subunit. This was tested by measuring Cox1p synthesis in strains carrying mutations in COX14 and other COX-specific genes. The cox14 mutation restored normal Cox1p synthesis in all the COX mutants except the mss51, pet309 (not shown), and oxa1 mutants (Figure 3A). Despite being epistatic with respect to Cox1p expression, the cox14 mutation did not rescue either respiration or the ability of the mutants, including the of shy1 mutant, to assemble COX (not shown). Figure 4.Turnover of in vivo-labeled mitochondrial translation products and steady-state concentration of Cox1p in wild type and COX mutants. (A) Wild type (W303-1A) and mutants (described in Table I) were grown and labeled for 20 min at 30°C with [35S]methionine. Labeling was terminated by addition of 80 μmol cold methionine and 12 μg/ml puromycin (0 time). Samples of the cultures were collected after the indicated times of incubation at 30°C and processed as in Figure 1. (B) The wild-type strain W303-1A and the cox14 and cox17 mutants were labeled and chased for the indicated times as in panel A. One-half of each culture was incubated in the presence of chloramphenicol as in Figure 1 prior to labeling. (C) The wild-type W303-1A and D273-10B and mutant strains were grown in 2% galactose, 1% yeast extract, and 2% peptone to stationary phase. Mitochondria were prepared and 10 μg of protein was separated by SDS–PAGE on a 12% polyacrylamide gel. The proteins were transferred to nitrocellulose and probed with a polyclonal antibody against yeast Cox1p. The antibody–antigen complexes were visualized by a secondary reaction with [125I]protein A. Download figure Download PowerPoint Overexpression of Cox14p does not alter synthesis of Cox1p Normal labeling of Cox1p in cox14 single and double mutants could indicate that Cox14p acts to increase degradation of unassembled or incompletely assembled Cox1p. If this were the case, overexpression of Cox14p in a wild-type or mutant background might be expected to affect the amount of newly synthesized Cox1p. This was examined by transforming the wild-type strain and several COX assembly-deficient mutants with the episomal plasmid pG93/T1 (Glerum et al, 1995), which contains a wild-type COX14 gene. Overexpression of Cox14p in these strains did not affect in vivo labeling of Cox1p (Figure 3B) or restore respiration (data not shown). Stability of newly synthesized Cox1p in wild type and COX mutants The stability of unassembled Cox1p was assessed in wild type and different COX mutants by pulse-chase. Most of the translation products, including the three COX subunits, were stable during 2 h of chase in wild type but not in the mutants in which a significant fraction of Cox2p and Cox3p were degraded (Figure 4A). In contrast, the small amount of labeled Cox1p detected in the mutants was stable during the chase (Figure 4A). The exception was the cox14 mutants, in which most of the Cox1p was degraded (Figure 4A and B). The kinetics of Cox1p turnover was also examined during longer periods of chase of cells pulse-labeled with and without a prior incubation in the presence of chloramphenicol (Figure 4B). Cox1p was stable even after 20 h of chase, but under the same conditions most of Cox1p in the cox14 mutant was degraded after 2 h of chase. The small amount of Cox1p detected in the cox17 mutant appears to be as stable as in wild type. The greater lability of Cox1p in the cox14 mutant is also supported by the results of Western analysis of the steady-state concentrations of this subunit in different mutants (Figure 4C). Cox1p, Cox2p, and Cox3p synthesized in cycloheximide-inhibited wild-type cells are not incorporated into the holoenzyme, even following a 2 h period of chase (data not shown). The relatively high stability of these COX subunits, however, suggests that they are in a protease-protected environment either as monomers or partially assembled intermediates, or that they are complexed to a 'stabilizing’ factor(s). Do Shy1p and Cox14p act post-translationally? Manthey and McEwen (1995) have shown that ρ− genomes in which COX1 is fused to the 5′ leader of the mitochondrial COB (SUP2; see Figure 5A) or COX3 (SUP1) genes are able to suppress pet309 mutants. This argues strongly against a post-translational role of Pet309p in expression of COX. In this study, we have constructed strains heteroplasmic for wild type and the SUP2 ρ− suppressor in the context of shy1, mss51, or cox14 null mutations. Figure 5.The cox14 defect is not rescued by a ρ− genome with COX1 fused to the 5′-UTR of COB. (A) Map of the mitochondrial bypass suppressor ρSUP2. The rearrangement in the suppressor leads to a fusion of the 5′-UTR of COB to nucleotide −174 of COX1 (Manthey and McEwen, 1995). (B) ρSUP2 was transferred by cytoduction to a kar1 mutant (Conde and Fink, 1976) lacking mitochondrial DNA (ρ°). The SUP2 suppressor was transferred from the kar1 donor to ρ° derivatives of null mutants of PET309, MSS51, COX14, and SHY1. The different mutants with the SUP2 genome (ρSUP2) were then crossed to the isogenic mutants with wild-type mitochondrial DNA (ρ+) to obtain the heteroplasmic diploid mutants of PET309 (a/α-ΔPET309/ρ+ρSUP2), MSS51 (a/α-ΔMSS51/ρ+ρSUP2), COX14 (a/α-ΔCOX14309/ρ+ρSUP2), and SHY1 (a/α-ΔSHY1/ρ+ρSUP2). Serial dilutions of the haploid mutants with wild-type mitochondrial DNA and of the diploid strains with the wild-type and suppressor genomes were spotted on YPD and YPEG plates and incubated at 30°C for 2.5 days. Download figure Download PowerPoint The failure of the SUP2 suppressor to rescue cox14 and shy1 mutants (Figure 5B) indicates that Cox14p and Shy1p are required at a post-translational stage of COX assembly, although it does not exclude a role in translation as well. In agreement with the findings of Perez-Martinez et al (2003) who used SUP1 in their studies, SUP2 does not restore the COX deficiency of mss51 mutants (Figure 5B). Cox14p interacts with newly synthesized Cox1p Restoration of normal Cox1p translation in mutants that have a second mutation in COX14 suggested that Cox14p might be regulating Cox1p synthesis. A physical interaction of Cox1p with Cox14p was tested by expressing the latter as a GST fusion protein from a chromosomally integrated gene. The GST-tagged Cox14p was able to fully complement the respiratory defect of the cox14 mutant (data not shown). Mitochondria from aW303ΔCOX14/ST32 expressing the Cox14p-GST fusion protein were labeled in organello with [35S]methionine, extracted with lauryl maltoside, and adsorbed onto glutathione–Sepharose beads. The proteins that were recovered from the beads indicated a selective and virtually quantitative enrichment of labeled Cox1p (Figure 6A). A similar enrichment of Cox1p was seen when mitochondria were isolated from a strain expressing Mss51p-GST but not from wild type or from a strain expressing a Shy1p-GST fusion protein. Like Cox14p-GST, the latter two fusions also complemented the respective null mutants. An association of HA-tagged Mss51p with newly synthesized Cox1p was also reported by Perez-Martinez et al (2003). Trace amounts of cytochrome b, Cox2p, and Cox3p were also adsorbed to the beads but the signals varied in different experiments. The enrichment of the Cox2p precursor seen in the pull-down of the strain expressing Cox14p-GST was consistent but was investigated further. Figure 6.Cox14p and Mss51p interact with Cox1p. (A) Mitochondria were prepared from the wild-type W303-1A, a shy1 null mutant (ΔSHY1/ST62) with a chromosomally integrated plasmid expressing the Shy1p-GST fusion protein, an mss51 null mutant (ΔMSS51/ST13) with a chromosomally integrated plasmid expressing Mss51p-GST, and a cox14 null mutant (ΔCOX14/ST32) with a chromosomally integrated plasmid expressing Cox14p-GST. Mitochondria were labeled with [35S]methionine for 30 min and extracted with 1% lauryl maltoside, 1 M KCl, and 1 mM PMSF. The extract was clarified by centrifugation at 50 000 gav for 30 min and incubated with glutathione–Sepharose beads for 4 h at 4°C. After centrifugation at 1500 rpm for 5 min, the supernatant was collected and the beads were washed three times with PBS. Mitochondria (M) corresponding to 2 μg protein, equivalent volumes of the membrane pellet (P) after lauryl maltoside extraction and of the supernatant from the glutathione–Sepharose beads (S) were separated on a 17.5% polyacrylamide gel by SDS–PAGE. The amount of washed beads (B), however, corresponded to ∼500 μg of the starting mitochondria. (B) Mitochondria from W303-1A, the cox14 null mutant (ΔCOX14), and a cox14 point mutant transformed with a high-copy plasmid containing COX14 (C179/L1/ST1) were labeled for 30 min at 30°C in the presence of [35S]methionine. After a 5 min pulse, the samples were treated with the crosslinker DSP (+) or were mock-treated (−) as described (Hell et al, 2000). Immunoprecipitation of crosslinked adducts was performed using antiserum specific for Cox14p (+) and preimmune serum (−). Immunoprecipitates were analyzed by SDS–PAGE and autoradiography as in Figure 1. Download figure Download PowerPoint To test if the binding of Mss51p to newly synthesized Cox1p requires the presence of Cox14p, an mss51 and cox14 double mutant was transformed with an integrative plasmid expressing Mss51p-GST. Co-precipitation of newly synthesized Cox1p with Mss51p-GST in the cox14 null background (data not shown) indicates that the interaction of Mss51p and Cox1p is not Cox14p dependent. The interaction of Cox14p with newly synthesized Cox1p was also studied by labeling mitochondria in organello with [35S]methionine in the absence or presence of the cleavable crosslinker dithio-bis-succinimidyl propionate (DSP) to trap transient complexes that might be formed early after completion of Cox1p synthesis. Detergent extracts containing the labeled translation products were treated with antibody to Cox14p and analyzed by SDS–PAGE under conditions causing cleavage of the crosslinker. A small fraction of newly synthesized Cox1p was present in the immunoprecipitate obtained with the Cox14p antibody in a strain overexpressing Cox14p but not in a wild-type strain (Figure 6B). The co-immunoprecipitation of Cox1p did not depend on the inclusion of DSP during translation. The poor recovery of Cox1p in this procedure is probably due to the low efficiency of immunoprecipitation with the Cox14p antibody. Cox14p interacts with Mss51p The interaction of Cox14p with Mss51p was examined in strains of yeast, expressing either Cox14p-GST or Mss51p-GST in cox14 and mss51 null backgrounds, respectively. Pull-down assays of mitochondria extracted with lauryl maltoside indicated that approximately 61% of Mss51p-GST and 48% of Cox14p were adsorbed onto the beads (Figure 7A). Likewise, when crude mitochondrial extracts containing Cox14p-GST were adsorbed onto glutathione–Sepharose beads, more than 75% of Cox14p-GST and 50% of Mss51p were pulled down (Figure 7B). Figure 7.Cox14p interacts with Mss51p. (A) Mitochondria (M) from an mss51 null mutant with a chromosomally integrated plasmid expressing Mss51p-GST fusion protein (ΔMSS51/ST13) were extracted with 1% lauryl maltoside, 1 M KCl, and 1 mM PMSF. The pellet (P) after centrifugation at 50 000 gav for 30 min was suspended in the starting volume of buffer and the extract (E) was mixed and incubated for 4 h at 4°C with glutathione–Sepharose. The supernatants (Es) from the beads were collected and the beads (Eb) were washed three times with PBS. The different fractions adjusted for volume were separated by SDS–PAGE. The lane labeled (Eb2) was loaded with two times the amount of beads. Cox14p and Mss51p-GST were detected by Western blot analysis using specific antibodies against each protein. The proteins were visualized by a secondary reaction with [125I]protein A and the radiolabeled bands were detected with a PhosphorImager (Molecular Dynamics). (B) Same as (A) except that the mitochondria were prepared from a cox14 null mutant with an integrated plasmid expressing a Cox14p-GST fusion protein (ΔCOX14/ST32). (C) The bands shown in (A, B) were quantified with the PhosphorImager. The open bars represent the percentage of the corresponding protein bound to the beads, and the filled bars represent the percentage of unbound protein recovered in the supernatant fraction. Download figure Download PowerPoint Part of the lauryl maltoside extracts was centrifuged on a linear 7.5–25% sucrose gradient. Gradient fractions were analyzed for the distributions of Mss51p and Cox14p and the peak fractions with Mss51p-GST and Cox14p or Cox14p-GST and Mss51p were treated with glutathione–Sepharose beads. These pull-down assays confirmed that the Cox14p cosedimenting with Mss51p-GST and vice versa were complexed to each other (data not shown). Our data, however, do not discriminate between a direct interaction of the two proteins and an interaction mediated by other proteins that may constitute the complex. The GST pull-down assays suggested that only a fraction of Mss51p is complexed to Cox14p. This was supported by the results of sucrose gradient sedimentations of mitochondrial detergent extracts. Mss51p and Cox14p sedimented similarly in a gradient loaded with a lauryl maltoside extract of wild-type mitochondria. Both proteins peaked only a fraction behind lactate dehydrogenase with estimated masses of 130 kDa (Figure 8A). The distribution of Mss51p and Cox14p in this gradient, however, was not symmetrical, indicating the presence of higher molecular weight specie(s) (Figure 8A). The gradient of the lauryl maltoside extract of mitochondria from the mss51 deletion mutant showed a symmetrical distribution of Cox14p with a peak at approximately the same position as Cox14p in the wild-type extract (Figure 8B). The symmetrical Cox14p peak in the gradient of the mutant extract suggests that the faster sedimenting component(s) in the wild-type extract may represent that fraction of Cox14p complexed to Mss51p. All of the Cox14p in the mutant and most of Cox14p in wild type were estimated to have a mass at least 10 times larger than the monomer. This indicates that in addition to interacting with Mss51p, Cox14p also exists as part of a larger homo or hetero-oligomeric complex. Figure 8.Sedimentation of Mss51p and Cox14p in sucrose gradients. (A) Mitochondria of the wild-type strain W303-1A were extracted at a protein concentration of 10 mg/ml with 1% lauryl maltoside, 20 mM Tris–HCl (pH 7.5), and 0.5 M KCl. The extract (0.4 ml) was mixed with 2.5 mg of hemoglobin and 60 μg lactate dehydrogenase and applied to 4.6 ml of a linear 7–25% sucrose gradient containing 10 mM Tris–HCl and 0.1% Triton X-100. Following centrifugation at 65 000 rpm in a Beckman SW65Ti rotor for 6 h, 14.5 fractions were collected, separated on a 12% polyacrylamide gel, transferred to nitrocellulose and probed with rabbit antiserum against Mss51p or Cox14p followed by a secondary goat peroxidase-conjugated antibody against rabbit IgG. Antibody–antigen complexes were visualized with the Super Signal reagent (Pierce Chemical Co., Rockford, IL). Hemoglobin (○- - -○) was estimated from absorbance at 410 nm and lactate dehydrogenase (⧫- -⧫) was assayed by measuring oxidation of NADH at 340 nm with pyruvate as the substrate. (B) Same as (A) except that the mitochondria were isolated from the mss51 null mutant aW303ΔMSS51 and the gradient was collected in 15 fractions. Download figure Download PowerPoint Unlike Cox14p and Mss51p, Shy1p was not adsorbed onto the glutathione beads from mitochondrial extracts containing either Mss51p or Cox14p fused to GST (data not shown). COX14p is a mitochondrial inner membrane protein facing the matrix Cox14p and Mss51p were previously shown to be associated with the inner membrane of mitochondria (Glerum et al, 1995; Siep et al, 2000). To see if the localization of the two proteins is consistent with their proposed functions, we have determined their topology and solubility properties. Sonic irradiation of wild-type mitochondria solubilized cytochrome b2, a soluble protein of the intermembrane space, but not Cox14p or Mss51p (Figure 9A). Cox14p and Mss51p, however, were solubilized with alkaline carbonate, suggesting that they are peripheral proteins (Figure 9A). In this experiment, Shy1p was recovered in the membrane fraction, confirming earlier evidence that it is an intrinsic protein of the inner membrane (Barrientos et al, 2002b). Figure 9.Localization and topology of Cox14p and Mss51p. (A) Mitochondria and the post-mitochondrial supernatant fractions were prepared from the wild-type strain W303-1A. A sample of mitochondria at 4 mg/ml was sonically irradiated and centrifuged at 50 000 gav for 30 min. The membrane pellet was suspended in the starting volume of buffer. To 500 μl of the membranes at a protein concentration of 1 mg/ml was added 50 μl of 1 M Na2CO3 (pH 11.3) and 50 mM EDTA. After 30 min on ice, the sample was centrifuged at 100 000 gav for 15 min at 4°C to separate the soluble from the insoluble intrinsic membrane proteins. Equivalent volumes of mitochondria (Mt), membranes (P), the supernatant obtained after centrifugation of the sonicated mitochondria (S), the carbonate supernatant (CS), and pellet (CP) were separated on a 12% polyacrylamide gel, transferred to nitrocellulose, and treated with antiserum against Cox14p as in Figure 8. Antibodies against Mss51p, Shy1p, and cytochrome b2 (Cyt b2) were used to monitor the conversion of mitochondria to mitoplasts and th