MTO1 Codes for a Mitochondrial Protein Required for Respiration in Paromomycin-resistant Mutants of Saccharomyces cerevisiae*

Mutations in MTO1 express a respiratory defect only in the context of a mitochondrial genome with a paromomycin-resistance allele. This phenotype is similar to that described previously for mss1 mutants by Decoster, E., Vassal, A., and Faye, G. (1993) J. Mol. Biol. 232, 79–88. We present evidence that Mto1p and Mss1p are mitochondrial proteins and that they form a heterodimer complex. In a paromomycin-resistant background, mss1 andmto1 mutants are inefficient in processing the mitochondrial COX1 transcript for subunit 1 of cytochrome oxidase. The mutants also fail to synthesize subunit 1 and show a pleiotropic absence of cytochromes a,a 3, and b. In vivo pulse labeling of an mto1 mutant, however, indicate increased rates of synthesis of other mitochondrial translation products. The respiratory defective phenotype of mto1 and mss1 mutants is not seen in a paromomycin-sensitive genetic background. The visible absorption spectra of such strains indicate a higher ratio of cytochromes b/a and elevated NADH- and succinate-cytochromec reductase activities. To explain these phenotypic characteristics, we proposed that the Mto1p·Mss1p complex plays a role in optimizing mitochondrial protein synthesis in yeast, possibly by a proofreading mechanism.

Respiratory competence in Saccharomyces cerevisiae is controlled by upward of several hundred nuclear genes referred to as PET genes (1). Mutations in PET genes are generally recessive and are recognized by an inability of the mutant to utilize nonfermentable carbon sources. In some instances, loss of mitochondrial functions depends on the genotype of mitochondrial DNA (mtDNA). For example, lesions in gene products that promote splicing of specific mitochondrial introns elicit a respiratory defect only in strains whose mitochondrial genome contains the cognate intron (2,3). Another interesting situation has been reported in mss1 mutants which become respiratorydeficient only when their mtDNA has a point mutation in the 15 S rRNA conferring resistance to paromomycin (4). In the paromomycin-resistant background, mss1 mutants fail to splice the COX1 pre-mRNA for subunit 1 of cytochrome oxidase (4). The mutant phenotype was interpreted to indicate that Mss1p normally interacts with the 15 S rRNA. This interaction was proposed to be weakened or to not occur in the presence of the paromomycin-resistance allele, resulting in suppression of translation of the intron encoded maturases of the COX1 gene (4). According to this interpretation, the processing block is secondary to a translational defect. The absence in mss1 mutants of an obvious phenotype in the context of a wild type 15 S rRNA gene further suggests that Mss1p may be necessary for optimization of mitochondrial translation but is not essential for this activity (4).
The pet phenotype expressed in the presence of the paromomycin-resistance allele is not unique to mss1 mutants. In the present communication, we show that mutations in another gene, here designated as MTO1 (mitochondrial translation optimization), result in properties similar to those described for mss1 mutants. Evidence is also presented that Mto1p and Mss1p are complexed to one another and therefore are likely to carry out a common function.

MATERIALS AND METHODS
Yeast Strains and Media-The genotypes and sources of the strains of S. cerevisiae used in this study are listed in Table I. The media used to grow yeast have been described elsewhere (7).
Cloning of MTO1 and MSS1-MTO1 was cloned by transformation of E39/U1 (␣ ura3-1 mts1-1) with a yeast genomic plasmid library by the method of Schiestl and Gietz (8). The library used for the transformation was made from partial Sau3A fragments of nuclear DNA (averaging 7-15 kb) 1 cloned into the BamHI site of the shuttle vector YEp24 (9). This library was kindly provided by Dr. Marian Carlson, Department of Genetics and Development, Columbia University. Approximately 1 ϫ 10 9 cells were transformed with 50 g of plasmid DNA. The transformation mixtures were plated on minimal glucose medium to select Uraϩ clones. After 3 days, the transformation plates containing 10 4 colonies were replicated to YEPG medium to identify respiratory-competent transformants.
MSS1 was cloned by transformation of E221/L1 (␣ leu2-3,112 mss1) with a library similar to that described above except that the shuttle vector was YEp13 (10). Following transformation of the mutant by the method of Beggs (11), respiratory-competent clones were selected on minimal glucose. Approximately 5000 Leu ϩ clones were obtained from the transformation of 10 8 cells. Of these, one clone was ascertained to be respiratory-competent. This transformant E221/L1/T1 contained plasmid pG121/T1.
Preparation of Yeast Mitochondria and Enzyme Assays-Wild type and mutant yeast were grown to stationary phase in YPGal (2% galactose, 1% yeast extract, and 2% peptone), and mitochondria were prepared by the procedure of Faye et al. (12), except that Glusulase was replaced by Zymolyase 20,000 (ICN Biomedicals, Inc.). Mitochondria were assayed for NADH-cytochrome c reductase, succinate-cytochrome-c reductase, and cytochrome oxidase as described previously (13) Construction of W303⌬MTO1 and W303⌬MSS1-A null allele of MTO1 was constructed by polymerase chain reaction amplification of the 5Ј and 3Ј sequences adjacent to the gene using the divergent primers 5Ј-gccgggtaccgtatgtaaccatcaaattcg and 5Ј-cgccggtaccgttctttggggcgtttagc. The template for the amplification consisted of the 3.5-kb BamHI-SacI fragment cloned in pUC18. The polymerase chain reaction product was digested with KpnI and ligated to a 1.1-kb KpnI fragment containing the yeast URA3 gene.
The null allele, recovered as a linear BamHI-SacI fragment, was introduced into the diploid strain W303 by the one-step gene replacement procedure (14). Ura ϩ transformants were sporulated and subjected to tetrad dissections. Four independent Ura ϩ spores were verified by Southern analysis of their genomic DNA to harbor the null allele. The disrupted allele of MSS1 was obtained by insertion of a 1.8-kb BamHI fragment with HIS3 at the BglII site inside the gene. The linear KpnI-EcoRI fragment containing the disrupted allele was substituted for the wild type gene in the respiratory-competent haploid strain W303-1A. Transformation with the linear fragment yielded seven His ϩ transformants of which one (aW303⌬MSS1) was respiratory-deficient and was not complemented by a o tester, indicating that it lacked or had a deletion in mitochondrial DNA.
Preparation of Antibodies to Mto1p and Mss1p-Antibodies against Mto1p were obtained by immunizing rabbits with a fusion protein expressed in Escherichia coli. The gene for the fusion protein was made by ligating the 900-base pair PstI-EcoRI fragment coding for residues 154 -474 to the amino-terminal half of the E. coli trpE gene in the pATH23 (15). The resultant plasmid expressed a 68-kDa protein consisting of the amino-terminal half of anthranylate synthetase component I fused in-frame to the Mto1p sequence. This protein was recovered in the insoluble fraction of E. coli cells harboring the pATH23 construct. The insoluble protein fraction was dissolved in a 10 mM Tris-HCl, 1 mM EDTA, pH 7.5, buffer containing 2% SDS, 5 mM ␤-mercaptoethanol, and 20 g/ml phenylmethylsulfonyl fluoride and was further purified on a Bio-Gel A-0.5 column developed with a buffer containing 10 mM Tris-HCl, 0.1 mM EDTA, 5 mM ␤-mercaptoethanol. Fractions enriched for the fusion protein were pooled, concentrated by acetone precipitation, and used to raise antibodies in rabbits.
The above method was also used to obtain an antibody against an Mss1p fusion protein. The gene was made by ligating the 1.1-kb BglII-EcoRI fragment containing the MSS1 sequence starting from codon 195 to pATH23 (15) linearized with BamHI and EcoRI. The 70-kDa fusion protein expressed in E. coli transformants was partially purified as above and used to immunize rabbits.
Miscellaneous Procedures-Standard procedures were used for the preparation and ligation of DNA fragments and for transformation and recovery of plasmid DNA from E. coli (16). The preparation of yeast nuclear DNA and the conditions for the Southern hybridizations, were as described by Myers et al. (7). DNA probes were labeled by random priming (17). DNA was sequenced either by the method of Maxam and Gilbert (18) or Sanger et al. (19) with T7 sequencing kit (United States Biochemical). Proteins were separated by polyacrylamide gel electrophoresis in the buffer system of Laemmli (20), and Western blots were treated with antibodies against the fusion proteins followed by a second reaction with 125 I-protein A (21). Protein concentrations were determined by the method of Lowry et al. (22).

Phenotype of Mutants from Complementation Group G158 -
E39 is one of two independent isolates assigned to complementation group G158 of a pet mutant collection (1). Both mutants were derived from the respiratory-competent haploid strain D273-10B/A21 (6). This strain has a mitochondrial genome with the three drug-resistant markers oli1 R , ery R , and par R . The respiratory defect of both G158 mutants is complemented by o testers, indicating that they have recessive mutations in a nuclear gene. The mutants are genetically unstable and convert readily to Ϫ and o derivatives.
The absorption spectrum of an extract of E39 mitochondria indicates the absence of cytochrome b and cytochromes a, a 3 ( Fig. 1). This pleiotropic phenotype is usually found in mutants defective in mitochondrial protein synthesis. Pulse labeling of whole cells with [ 35 S]methionine in the presence of cycloheximide, however, indicated that the mutant is able to synthesize most of the mitochondrial gene products (Fig. 2A). The excep-FIG. 1. Spectra of mitochondrial cytochromes in wild type and mutant yeast. Mitochondria at a protein concentration of 5 mg/ml (except E39 mitochondria whose concentration was 12 mg/ml) were extracted with potassium deoxycholate under conditions that quantitatively solubilize all the cytochromes (23). Difference spectra of the reduced versus oxidized extracts were recorded at room temperature. The genotypes of the parental strain W303-1A and of the mto1 (aW303⌬MTO1(P R ) and E39) and mss1 (aW303⌬MSS1(P R )) mutants are described in Table I. The ␣ absorption bands corresponding to cytochromes a and a 3 have maxima at 603 nm. The corresponding maximum for cytochrome b is 560 nm and for cytochrome c is 550 nm. The percentage of Ϫ cells in the cultures used to obtained the spectra were 50% for E39, 40% for aW303⌬MTO1, and 60% for aW303⌬MSS1.  tion is subunit 1 of cytochrome oxidase that is not detected among the labeled proteins. It is interesting that incorporation of [ 35 S]methionine is more efficient in the mutant than in wild type yeast. The exact increase is difficult to quantitate but must be more than indicated by the autoradiogram in view of the fact that the mutant culture used in the experiment consisted of 50% Ϫ and o cells which do not contribute to the activity. The deficit in the cytochrome oxidase subunit 1 was consistent with the results of Northern blot analysis of the mitochondrial COX1 transcripts for cytochrome oxidase subunit 1. Probes specific for exon regions of these RNAs indicated lower concentrations of the mature mRNAs and accumulation of partially processed precursors (Fig. 2B). Processing of cytochrome b is also retarded in the mutant. It is not clear from the gel of the radioactively labeled proteins whether synthesis of this protein is affected in the mutant because of its comigration with subunit 2 of cytochrome oxidase (Fig. 1A).
Cloning of MTO1 and MSS1-A gene capable of complementing the respiratory defect of E39/U1 was cloned by transformation of the mutant with a yeast genomic library based in the episomal plasmid YEp24 (9). The transformation yielded five respiratory-competent and uracil-independent clones. The plasmids were isolated from the transformants and amplified in E. coli, and their nuclear DNA inserts were characterized. Restriction mapping indicated that the nuclear DNA inserts of the different plasmids were related. The end points of one of the plasmids (pG158/T1) were sequenced and located in the genomic sequence of chromosome VII. This plasmid was also used to subclone the complementing gene by transferring different regions of the insert to the yeast/E. coli shuttle vector YEp352 (24) and testing the ability of the new constructs to confer respiration in E39/U1 (Fig. 3). The results of these transformations indicated that the gene is located between the unique BamHI and SacI sites of the pG158/T1 insert. This region encompasses HAP2 (25) and reading frame YGL236C

FIG. 2. Mitochondrial translation products and transcripts in wild type and a mto1 mutant.
A, the wild type parental strain D273-10B/A21 and the mto1 mutant E39 were grown in 10 ml of YPD (yeast extract-peptone-dextrose) overnight. The cells were harvested, washed with sterile water, transferred to 25 ml of minimal 0.8% glucose medium (Difco, nitrogen base without amino acids), and incubated with shaking for 1 h. Cycloheximide was then added to a final concentration of 1 mM. After 10 min, 1 mCi of [ 35 S]methionine was added and incorporation allowed to proceed for 1.5 h. Protein synthesis was quenched by addition of 1.2 M sorbitol containing 2 mg/ml chloramphenicol and 1 mM cycloheximide. The cells were collected, washed several times in the same buffer, and digested with Zymolase for 5 min in 2 ml of 1.2 M sorbitol, 20 mM potassium phosphate, pH 7.5, 1 mM EDTA, 140 mM ␤-mercaptoethanol, 20 mg/ml Zymolyase. The spheroplasts were lysed in 0.5 M sorbitol and 0.2 mg/ml phenylmethylsulfonyl fluoride, and mitochondria were isolated by differential centrifugation of the lysate. Total mitochondrial proteins (90 g) were separated on a 12% polyacrylamide gel. The gel was stained, dried, and exposed to x-ray film. The migration of size standards is indicated in the right-hand margin. The known mitochondrial translation products Var1, subunit 1 (Cox1), 2 (Cox2), and 3 (Cox3) of cytchrome oxidase, cytochrome b (Cob), subunits 6 (Atp6), and 9 (Atp9) of ATPase are identified in the left-hand margin. B, total mitochondrial RNA (2 g) of the parental strain D273-10B/A21 and of the mto1 mutant E39 was separated on a 1% nondenaturing gel and transferred to DBM (diazobenzyloxymethyl) paper. The blot was hybridized to a COB probe consisting of 104 nucleotides of 5Ј untranslated and 67 nucleotides from the first exon of the gene. A comparable blot was hybridized to a 350-base pair HinfI fragment from the fourth exon of the COX1 gene. The bands corresponding to the precursors and the fully spliced mRNAs are identified in the legend.
coding for a 74-kDa protein of unknown function. Neither pG158/ST3 or pG158/ST2 containing the entire HAP2 gene restored respiration, indicating that the complementing gene is YGL23C. We propose MTO1 (mitochondrial translation optimization) as a name for this gene for reasons discussed later in this paper. The sequence of Mto1p is homologous to the gdiA proteins encoded in the genomes of various bacteria (26,27) and of Caenorhabditis elegans. 2 The sequence similarity extends from approximately residue 40 to the very carboxyl terminus of the protein (Fig. 4). The function of this highly conserved protein is not known.
MSS1 was cloned by transformation of E221/L1, a representative of complementation group G122, with a YEp13 library. This mutant is phenotypically similar to E39/U1. A single respiratory-competent clone was obtained in the transformation. Respiration of the transformant was found to be dependent on a segregating plasmid (pG122/T1) whose nuclear DNA insert was determined to span the region of chromosome XIII from nucleotides 314,900 to 321,400. The smallest subclone (pG122/ST4) capable of complementing E221/L1 contained MSS1 (Fig. 5).
Properties of mto1 Null Mutants-To determine whether E39/U1 has a mutation in MTO1, a null mutation (⌬mto1::URA3) was introduced in the respiratory-competent diploid strain W303. The construction of the null allele is described under "Materials and Methods" and depicted in Fig. 6. Uracil-independent clones obtained from a transformation of the diploid strain W303 with the null allele on a linear fragment of DNA were subjected to tetrad dissections. The auxotro-phic requirement of the meiotic spore progeny indicated a 2:2 segregation of the URA3 disrupter. The Ura ϩ haploid strains were confirmed by Southern analysis of their genomic DNA to harbor the ⌬mto1::URA3 null mutation in their chromosomal DNA (Fig. 6). Unlike the original mutant, however, the Ura ϩ clones (W303⌬MTO1(P S )) were respiratory-competent as judged by growth on glycerol (Fig. 7). The presence of a functional respiratory chain in the null mutant was also evident from the visible absorption spectrum of mitochondria, which indicated the presence of cytochromes b and cytochrome oxidase (Fig. 8). The transformants, however, have a higher concentration of cytochrome b relative to cytochrome a. The aberrant ratio of the two cytochromes is also reflected in the increased NADH-cytochrome c reductase and reduced cytochrome oxidase activity of mitochondria from the transformant (Fig. 9). The increased ratio of cytochromes b/a is also seen in an mto1/msss1 double mutant (W303⌬MTO1/MSS1(P S )) ( Fig.  7).
The ability of the mto1 null mutant, W303⌬MTO1(P S ), to respire could indicate either that MTO1 is a suppressor of E39 or that the respiratory defect of mto1 mutants is affected by the mitochondrial genetic background. The mitochondrial genomes of W303 and D279 -10B/A21 used to obtain E39 differ in their intron composition and in the presence in the latter of mutations in the OLI1, ERY, and PAR loci conferring resistance to these antibiotics (6). To test the possible influence of mitochondrial genetic background on the phenotype of mto1 mutants, W303⌬MTO1(P S ) was cured of mtDNA by treatment with ethidium bromide. The o derivative was crossed to D273-10B/ A21, whose mtDNA carries the drug-resistance mutations oli1 R , ery R , and par R , and to D273-10B/A1, which has an identical genome except for the absence of the drug-resistant mutations. Diploid cells were sporulated for tetrad analysis. Meiotic progeny issued from the cross to D273-10B/A21 showed a 2:2 segregation of the respiratory-competent phenotype. In all cases, the respiratory-defective phenotype coincided with the uracil-independence. In contrast, all four meiotic progeny from the cross to the drug-sensitive strain D273-10B/A1 were respiratory-competent even though the Ura ϩ phenotype showed the expected 2:2 segregation patterns. These results excluded 2 A. Gardner (1995), GenBank TM accession number Z66512. an effect of intron composition difference on the phenotype but rather suggested that expression of respiratory deficiency in the mto1 mutant was dependent on one of the drug-resistance mutations (Fig. 5).
In a previous study mutations in MSS1 were shown to elicit respiratory deficiency only in strains carrying the paromomycin-resistance allele in mitochondrial DNA (4). To determine whether this was also true of the mto1 mutant, meiotic progeny were analyzed from a cross of the o derivative of W303⌬MTO1(P S ) to a strain containing only the paromomycinresistance marker. Tetrad analysis of meiotic products of this cross confirmed that the presence of mtDNA with the paromomycin-resistance allele was sufficient for the expression of respiratory deficiency in the mto1 mutant.
In addition to its respiratory deficiency, W303⌬MTO1(P R ) containing the paromomycin-resistance marker displayed the same pleiotropic absence of cytochromes b, a, and a 3 as E39 (Fig. 1). The similar phenotypes of W303⌬MTO1(P R ) and E39 suggested linkage of the two mutations. This was confirmed genetically. For these allelism tests, MTO1 was transferred to the integrative vector YIp352 (24) yielding pG158/ST5. This plasmid was linearized at the unique SmaI site of MTO1. The linear plasmid was used to transform E39/U1. Ura ϩ clones obtained from the transformation of the mutants were respiratory-competent, suggesting that the integration had occurred at the locus of the mutation. This was verified by crosses of a respiratory-competent clone to the wild type haploid strain W303-1B and to W303⌬MTO1(P R ) containing the ⌬mto1::URA3 allele. Diploid cells issued from each cross FIG. 7. Effect of the paromomycin-resistance marker on growth of mto1 mutants. A respiratory-competent strain (W303-1A), the null mutant in a paromomycin-resistant (aW303⌬MTO1(P R )) and paromomycin-sensitive background (aW303⌬MTO1(P S )), and the point mutant E39/U1 were grown on rich glucose medium and replicated on rich glycerol medium (YEPG). The photograph was taken after overnight incubation of the YEPG plate at 30 o C. were sporulated for tetrad analysis. The four meiotic products from the back-cross to the wild type W303 strain were respiratory-competent (nine complete tetrads). Analysis of the tetrads obtained from the cross to the mto1 null mutant showed a 2:2 segregation of the respiratory-competent phenotype (nine complete tetrads). In all cases, respiratory competence cosegregated with uracil independence. These results provide direct evidence of genetic linkage of the E39 and mto1 null allele.
Mto1p Is a Mitochondrial Protein-To facilitate the localization of Mto1p, part of the gene was fused in-frame to a segment of trpE coding for the amino-terminal half of anthranylate synthase component I. Antibodies raised against the fusion protein were used to test for the presence of Mto1p in the mitochondrial and postmitochondrial supernatant fractions of wild type yeast, a mutant transformant with the gene on a high copy plasmid, and the mto1 null mutant. The antibody detected a protein of the expected size (approximately 70 kDa) in mitochondria of wild type and of the transformant but not in mitochondria from the null mutant. The signal was weak in wild type mitochondria and was greatly enhanced in the transformant (Fig. 10).
Antibodies were also raised against an Mss1p fusion protein. This antibody was used to probe for Mss1p in the mitochondria and postmitochondrial supernatant fractions from wild type and mutant yeast strains. The results of the Western blot analysis confirmed the presence of a protein of approximately 55 kDa in the mitochondrial but not in the soluble fraction of wild type and the transformant (E221/L1/ST1). The stronger signal seen in the transformant and its absence in the mutant confirmed that this band corresponds to Mss1p (Fig. 10).
Do Mto1p and Mss1p Form a Complex?-The similarity in the phenotypes of mto1 and mss1 mutants raised the possibility that the two proteins are functionally related and therefore may exist in a physical complex. This was tested by sedimentation analysis of a yeast transformed with a multicopy plasmid (pG158/ST12) containing a copy of MTO1 and MSS1 in a tandem orientation. An extract obtained by sonic disruption of mitochondria was applied to a 5-20% sucrose gradient. Hemoglobin and lactate dehydrogenase were added to the extract as molecular weight standards. The gradient was analyzed for the distribution of Mto1p, Mss1p, and for the two molecular weight standards, hemoglobin and lactate dehydrogenase. The sedimentation properties of Mto1p and Mss1p are consistent with their existence in a complex. Mto1p and Mss1p cosediment and peak in a region of the gradient closer to lactate dehydrogenase than to hemoglobin (Fig. 11). Based on its position relative to the two marker proteins, the average molecular weight of the complex was estimated as 124,000. This would suggest a heterodimer consisting of a single copy of each protein. DISCUSSION The respiratory deficiency of mto1 mutants is contingent on the presence in mitochondrial DNA of a single base change at nucleotide 1477 of the 15 S ribosomal RNA gene (28,29). The mutation, a C 3 G transposition, confers resistance to paromomycin, an inhibitor of procaryotic protein synthesis. This unusual phenotype has also been reported for mss1 mutants FIG. 8. Spectra of mto1 and mto1/mss1 double mutants with the paromomycin-sensitive allele. Mitochondria of W303⌬MTO1(P S ) and W303⌬MTO1/MSS1(P S ) at a protein concentration of 5 mg/ml were extracted and their spectra recorded as described in the legend to Fig. 1.   FIG. 9. Respiratory chain activities in mto1. Mitochondria were prepared from the parental wild type strain W303-1A and W303⌬MTO1(P S ) and assayed for NADH-cytochrome c reductase, succinate-cytochrome c reductase, and cytochrome oxidase. (4). These and other properties shared by mto1 and mss1 mutants prompted us to examine if the products of the two genes might be subunits of a single complex.
As expected, immunochemical studies confirmed Mto1p and Mss1p to be constituents of mitochondria. When a mitochondrial extract obtained from a yeast transformant overexpressing Mto1p and Mss1p was centrifuged through a sucrose gradient the two proteins displayed identical sedimentation properties indicative of their presence in a complex. The molecular weight of the complex was estimated to be 121,000 based on its sedimentation relative to the dehydrogenase and 128,000 relative to hemoglobin. These values are 4 -11 kDa lower than the 132 kDa expected for a complex consisting of one copy of each subunit. The calculated molecular weight of the heterodimer, however, does not take into account the loss of amino-terminal signal sequences that are likely to be present in both proteins.
The phenotype of mss1 and mto1 mutants suggests that the C at nucleotide 1477 is necessary for interaction of the Mto1p⅐Mss1p complex with the 15 S rRNA. This explanation would require that the interaction is weakened or prevented by the mutation conferring paromomycin resistance. Western blot analysis of wild type mitochondrial ribosomes and of the subunits fractionated on sucrose gradients, however, failed to reveal any association of either protein with the large or small subunit of the ribosomes. Even though this constitutes strong evidence that neither protein is a subunit of mitochondrial ribosomes, it does not exclude a transient interaction of the Mto1p⅐Mss1p complex with ribosomes during translation.
The properties of the mutants are indicative of a role of Mss1p and Mto1p in some aspect of mitochondrial translation. In a paromomycin-resistant background, mutations in MSS1 result in an increase of the steady-state concentration of partially spliced COX1 transcripts (4). Decoster et al. (4) concluded that the processing block was a consequence of the failure of the mutant to translate the maturases encoded in some of the intronic regions of the COX1 pre-mRNA, thereby causing accumulation of incompletely spliced transcripts. Another hallmark of mss1 mutants is the complete absence of subunit 1 of cytochrome oxidase, the product of COX1. Both of the above properties are also true of mto1 mutants. The results of in vivo labeling of mitochondrial translation products indicate that other translation products including subunits 2, ATP6, and subunit 9 (the status of cytochrome b is difficult to ascertain from the radioautogram because of its comigration with subunit 2 of cytochrome oxidase) are more efficiently translated in the mto1 mutant. A possible explanation for this observation is that the translation rate in the mutant is less regulated. Although we have not examined mitochondrial translation in mto1 mutants with a paromomycin-sensitive genome, their spectra are also indicative of an unbalanced production of the respiratory chain components. The mutants have more cytochrome b relative to cytochromes a and a 3 and higher NADHand succinate-cytochrome c reductase activity. The latter is probably a consequence of an increase in coenzyme QH 2 -cyto- Mitochondria were isolated from W303-1A transformed with pG158/ST12. This plasmid contains MTO1 and MSS1 in a tandem arrangement in YEp352. The transformant was ascertained to overexpress both gene products. Mitochondria (6 mg) were sonically irradiated and centrifuged at 220,000 ϫ g av for 15 min. The clear supernatant (0.3 ml) was mixed with 2 mg of hemoglobin and 0.15 mg of canine muscle lactate dehydrogenase and layered on a 7-20% sucrose gradient containing 10 mM Tris-HCl, pH 7.5, 0.5 mM EDTA, and 0.05% Triton X-100. The gradient was centrifuged at 64,000 rpm for 7 h in a Beckman SW 65 rotor. The gradient was collected by gravity in 15 fractions. Each fraction was assayed for hemoglobin by absorption at 410 nm (E-E) and for lactate dehydrogenase (q-q) by measuring NADH-dependent conversion of pyruvate to lactate. The sedimentation of Mto1p and Mss1p in the gradient was determined by Western blot analysis. Proteins were separated on a 9% polyacrylamide gel, electrophoretically transferred to nitrocellulose, and probed for Mto1p. The same blot was subsequently treated with antibody to Mss1p. chrome c reductase, which contains cytochrome b as one of its electron carriers.
The absence of subunit 1 in mto1 and mss1 mutants, despite 30 -50% of mature COX1 mRNA, could be because of a defect in initiation of translation. Although the involvement of the Mto1p⅐Mss1p complex in translational initiation cannot be excluded, it seems unlikely for the following reasons. With the exception of the COX1, the mutants are able to translate most other mRNAs even more efficiently than the wild type strain. The argument would therefore need to be made that only translational initiation of selected transcripts such COX1 depend on Mto1p⅐Mss1p. This is difficult to reconcile with the widespread occurrence of homologous proteins in bacteria, which implies general rather than a specialized function of the proteins. A requirement in initiation of translation also fails to explain the almost 2-fold increase in the coenzyme QH 2 -cytochrome c reductase relative to wild type in the context of the paromomycin-sensitive allele.
An alternative explanation is that the Mto1p⅐Mss1p complex has a proofreading function in mitochondrial and perhaps more generally in eucaryotic protein synthesis. According to this interpretation, Mto1p⅐Mss1p interact with the small ribosomal subunit probably at a site in the 15 S rRNA close to or inclusive of the paromomycin locus. In the presence of the paromomycinresistance mutation, this interaction is preempted, resulting in a high rate of mutations in mitochondrially translated proteins. As a result, a large percentage of the proteins are no longer functional. The accumulation of partially spliced COX1 and COB intermediates in the mutant according to this interpretation is a consequence not of a defect in translation of the intron-encoded maturases but rather of mutations in proteins of this class that impair their functions. The complete absence of the subunit 1 despite the presence of some COX1 mRNA protein may be explained by a more rapid turnover of mutant copies of subunit 1. Although the other proteins are present in higher than normal concentrations, the steady-state levels may give a false impression of the actual amounts synthesized. Thus, depending on their intrinsic stability there may be differences in the rates of degradation. It is interesting to note that there appears to be an inverse relationship between the size of the protein and the apparent increase in its concentration in the mutant. For example, Atp9, the smallest product of mitochondrial protein synthesis shows the largest increase in the mutant.
The ability of mto1 and mss1 mutants to respire in paromomycin-sensitive strains, according to this model, implies that translation is more accurate under these circumstances. The fewer mutations in this background may be tolerated in the mutants without grossly affecting respiration. The increase of cytochrome c reductase (cytochrome b) relative to cytochrome oxidase (cytochromes a, a 3 ) may be indicative of some loss of subunit 1 even in the paromomycin-sensitive background. This loss, however, would need to be much less than in the paromomycin-resistant situation where the protein is undetectable. Finally, proofreading by Mto1p⅐Mss1p may entail a slower overall rate of translation. This would explain the observed higher efficiency of incorporation of radioactive precursor into most of the mitochondrial translation products in the mto1 mutant.