Mutations in the Saccharomyces cerevisiae ATP22 gene were previously shown to block assembly of the F0 component of the mitochondrial proton-translocating ATPase. Further inquiries into the function of Atp22p have revealed that it is essential for translation of subunit 6 of the mitochondrial ATPase. The mutant phenotype can be partially rescued by the presence in the same cell of wild-type mitochondrial DNA and a ρ− deletion genome in which the 5′-UTR, first exon, and first intron of COX1 are fused to the fourth codon of ATP6. The COX1/ATP6 gene is transcribed and processed to the mature mRNA by splicing of the COX1 intron from the precursor. The hybrid protein translated from the novel mRNA is proteolytically cleaved at the normal site between residues 10 and 11 of the subunit 6 precursor, causing the release of the polypeptide encoded by the COX1 exon. The ability of the ρ− suppressor genome to express subunit 6 in an atp22 null mutant constitutes strong evidence that translation of subunit 6 depends on the interaction of Atp22p with the 5′-UTR of the ATP6 mRNA.
STUDIES of respiratory-deficient mutants of Saccharomyces cerevisiae have led to the discovery of at least a dozen nuclear gene products that are necessary for biogenesis of the mitochondrial proton-translocating ATPase (Ackerman and Tzagoloff 2005). Three genes ATP11 (Ackerman and Tzagoloff 1990), ATP12 (Ackerman and Tzagoloff 1990), and FMC1 (Lefebvre-Legendre et al. 2001) code for mitochondrial proteins that aid assembly of the α- and β-subunits of F1 ATPase into a catalytically active oligomer. The other nine genes affect expression of the F0 unit of the ATPase. Their products target the mitochondrially encoded subunits 6 and 9 of F0 by promoting processing, stability, and translation of the mRNAs (Payne et al. 1991; Camougrand et al. 1995; Paul et al. 2000; Helfenbein et al. 2003; Ellis et al. 2004). One such protein, encoded by ATP22 and located in the inner membrane of mitochondria, was shown to be necessary for expression of F0 (Helfenbein et al. 2003). Mitochondria of atp22 null and point mutants have a catalytically active F1 ATPase but are deficient in oligomycin-sensitive ATPase activity (Helfenbein et al. 2003). The lack of inhibition of ATPase activity by oligomycin is characteristic of mutants defective in F0. Like other ATPase-deficient mutants of yeast, atp22 mutants are difficult to study because of their propensity to undergo deletions in mitochondrial DNA (mtDNA) (Helfenbein et al. 2003).
Although earlier studies pointed to a requirement of Atp22p for F0 assembly, its specific role in this process remained unclear. The present study was undertaken to gain a better insight into the function of Atp22p. The key to understanding the function of this protein came from studies of partial revertants of an atp22 null mutant. Such revertants were found to have mitochondrial deletion (ρ−) genomes, which when present together with wild-type mtDNA, partially restored oligomycin-sensitive ATPase in the atp22 null mutant. The sequence of one such suppressor genome, combined with Northern analysis of mitochondrial transcripts in the revertant, has led us to conclude that Atp22p is required for translation of subunit 6. Thus, Atp22p is a mitochondrial translation factor (Costanzo and Fox 1990) specific for subunit 6 of the ATPase.
MATERIALS AND METHODS
Yeast strains and growth media:
The genotypes and sources of the S. cerevisiae strains used in this study are listed in Table 1. The compositions of the media for growth of yeast were: 2% glucose, 1% yeast extract, 2% peptone (YPD); 2% galactose, 1% yeast extract, 2% peptone (YPGal); and 2% ethanol, 3% glycerol, 1% yeast extract, 2% peptone (YEPG).
Preparation of yeast mitochondria and ATPase assays:
Mitochondria were prepared by the method of Faye et al. (1974) except that Zymolyase 20,000 (MP Biochemicals, Aurora, OH) instead of glusulase was used to convert cells to spheroplasts. ATPase activity was assayed by measuring release of inorganic phosphate (King 1932) from ATP at 37° in the presence and absence of oligomycin.
In vivo and in organello labeling of mitochondrial gene products:
Cells grown overnight in YPGal were inoculated into 10 ml of medium containing 0.67% yeast nitrogen base without amino acids, the appropriate auxotrophic requirements, and 2% galactose. Cells equivalent to an A600 of 0.5 were harvested at a growth density of 1–2 A600. After centrifugation and washing with 40 mm potassium phosphate pH 6 containing 2% galactose, the cells were suspended in 500 μl of the same buffer and 10 μl of a freshly prepared aqueous solution of cycloheximide (7.5 mg/ml) were added. The cells were incubated at 24° for 5 min prior to addition of 4 μl of [35S]methionine (10 Ci/ml). The reaction was terminated after 30 min with 500 μl of 20 mm methionine and 75 μl of 1.8 m NaOH, 1 m β-mercaptoethanol, and 0.01 m phenylmethylsulfonyl fluoride. An equal volume of 50% TCA was added and the mixture was centrifuged. The precipitated proteins were washed once with 0.5 m Tris (free base) and two times with water and were suspended in 50 μl of sample buffer (Laemmli 1970).
Mitochondria prepared by the method of Herrmann et al. (1994) were labeled at 24° with [35S]methionine for 30 min (Hell et al. 2000). The radiolabeled proteins were separated on a 12.5% polyacrylamide gel containing 4 m urea and 25% glycerol, transferred to nitrocellulose, and exposed to Kodak X-OMAT film.
Extraction and Northern blot analysis of mitochondrial RNAs:
Mitochondria were solubilized in 1% sodium dodecyl sulfate (SDS), 10 mm Tris–Cl, pH 7.5, 0.1 m NaCl at a concentration of ∼10 mg/ml and immediately extracted with an equal volume of water-saturated phenol. After separation of the two phases by centrifugation at 2000 × g for 10 min, the upper phase was collected and nucleic acids were precipitated with 3 vol of ethanol in the presence of 0.25 m NaCl. The precipitate was rinsed twice with 80% ethanol and dried under vacuum.
Equivalent amounts of RNA were separated on a 1% agarose gel, stained with ethidium bromide, and blotted to a nylon membrane (Nytran, SuPerCharge; Schleicher & Schuell, Keene, NH). Following crosslinking with UV light, the blot was prehybridized at 43° for 4 hr with 125 μg of salmon sperm DNA in 5× SSC, 5× Denhardt's, 0.5% SDS. The blot was hybridized overnight at 43° with probes from the first exon of COX1, first intron of COX1, and ATP6. The COX1 intron and ATP6 probes were labeled with [α-32P]dATP by random priming (Feinberg and Vogelstein 1983). The COX1 exon probe was labeled at the 5′ end with [γ-32P]ATP and T4 polynucleotide kinase.
Mapping of mitochondrial suppressors:
Revertant cells were converted to ρ− mutants (respiratory-deficient mutants with a partially deleted mitochondrial genome) with ethidium bromide (Slonimski and Tzagoloff 1976). Purified mutants were crossed to the atp22 mutant and to testers with mutations in the mitochondrial COX1, COX2, COX3, COB, or ATP6 genes. After prototrophic selection on minimal glucose, the diploid cells were replicated on rich glycerol/ethanol medium (YEPG) and scored for growth after incubation for 2–3 days at 30°.
Standard methods were used for the preparation and ligation of DNA fragments and for transformation and recovery of plasmid DNA from Escherichia coli (Maniatis et al. 1989). Proteins were separated by SDS–polyacrylamide gel electrophoresis (PAGE) (Laemmli 1970). Western blots were treated with a rabbit polyclonal antibody against yeast subunit 6 followed by a second reaction with peroxidase-coupled anti-rabbit IgG. The antibody complexes were visualized with the Super Signal chemiluminescent substrate kit (Pierce Chemical, Rockford, IL). Protein concentrations were determined by the method of Lowry et al. (1951).
In vivo and in organello labeling of mitochondrial translation products:
Complementation group G99 consists of nine independent atp22 point mutants (Tzagoloff and Dieckmann 1990). Western analysis of ATPase subunits indicated that atp22 point mutants are grossly deficient in subunit 6 (Helfenbein et al. 2003). A subunit 6 deficit is not necessarily related to translation of the protein or absence of the mRNA (Ellis et al. 2004) but can also be caused by its turnover in mutants defective in F0 assembly (Paul et al. 1989, 2000; Arselin et al. 1996). In our earlier studies of the atp22 mutant, N417, some incorporation of 35S-methionine into subunit 6 was observed when cells were labeled in vivo in the presence of cycloheximide to arrest cytoplasmic protein synthesis (Helfenbein et al. 2003). In subsequent studies we found that with the exception of N417, other atp22 mutants tested failed to synthesize subunit 6 when assayed either in vivo with whole cells (Figure 1A) or in organello with isolated mitochondria (Figure 1C). The atp22 mutants are able to translate subunits 8 and 9, the other two mitochondrially encoded constituents of F0 (Figure 1B). The slow growth of N417 on rich glycerol/ethanol medium suggests that the genetic lesion does not abolish expression of F0 totally, which would explain why this strain is able to synthesize subunit 6. Like other ATPase mutants that display a pleiotropic deficiency of cytochrome oxidase (Marzuki et al. 1989; Paul et al. 1989), most of the atp22 mutants showed a substantial decrease in translation of subunit 1 of this respiratory complex.
Revertants of the atp22 null mutant:
Revertants of the atp22 null mutant with extragenic suppressors were isolated by spreading W303ΔATP22 at high density on rich glycerol/ethanol medium. Revertant colonies became discernible after 3–4 days and continued to appear over a period of several weeks. Three revertants were purified and their growth on glycerol/ethanol was compared to the mutant and the wild type parental strain. Unlike the null mutant, which has a very clear growth defect on glycerol and ethanol, the three revertants grew on the nonfermentable carbon sources but at a rate considerably slower than that of the parental wild type (Figure 2A).
Several criteria were used to ascertain that the revertants assemble a functional F0. Sensitivity of mitochondrial ATPase to oligomycin is a simple and reliable test for the presence of a functional F1–F0 complex. The partial restoration of the F1–F0 complex in the revertants is supported by the 10–15% inhibition of the mitochondrial ATPase activity by oligomycin (Table 2). The presence of F0 is also supported by the detection of subunit 6 in mitochondria of the revertant (Figure 2B), although in much lower concentrations than in wild type.
Suppression depends on the coexistence in the atp22 null mutant of wild type and a ρ− mitochondrial genome:
The suppressors in the three atp22 revertants behaved as dominant mutations in mtDNA. Diploid cells issued from crosses of the revertants to the mutants displayed the growth phenotype of the haploid revertants. When the crosses to the atp22 mutant were repeated with revertants that had been converted to ρo mutants (respiratory-deficient mutants lacking mitochondrial DNA), none of the diploid progeny grew on glycerol/ethanol, indicating that the suppressors were dominant mutations in mtDNA.
Unlike normal mtDNA mutations, which are transmitted to vegetative progeny, the atp22 suppressors were very unstable and rapidly lost when revertants were grown under nonselective conditions (Table 3). Following overnight growth of the revertant W303ΔATP22/R3 in rich glucose medium, 94% of the cells had lost the ability to grow on nonfermentable substrates. When grown in selective medium (glycerol/ethanol), the percentage of respiratory-competent cells was increased to 25%. The rapid loss of respiratory function, even under selective conditions, suggested that the revertants might be heteroplasmic cells containing both the wild-type and suppressor ρ− genomes. This type of mitochondrial suppressor has been reported for a number of mitochondrial protein-coding genes that fail to be expressed because of mutations in nuclear gene products that promote translation of the mRNAs by interacting with their 5′-UTRs (Dieckmann et al. 1984; Muller et al. 1984; Rodel and Fox 1987; Manthey and McEwen 1995). The presence in W303ΔATP22/R3 of a suppressor ρ− genome was confirmed by the mitochondrial genotypes of 104 respiratory-deficient segregants (Table 3). Approximately 72% of the cells analyzed were not complemented by either the ρo tester or the atp22 null mutant. These correspond to revertants that lost the suppressor and because of the instability of wild-type mtDNA in the atp22 null background had degraded to ρo and ρ− mutants. A second genotype, represented by 24% of the segregants corresponded to cells that had lost the suppressor but retained wild-type mtDNA. These cells were complemented by the ρo tester but did not confer respiratory growth when they were backcrossed to the null mutant. The least frequent segregants represented by only five of the colonies tested were not complemented by the ρo tester but were able to restore respiratory growth to the mutant. These segregants had lost the wild-type genome but had retained the ρ− suppressor genome.
A more complete genotype of the ρ− suppressor mtDNA was obtained from crosses of one of the segregants that complemented the apt22 mutant but was not complemented by the ρ− testers, to different mit− mutants (mit− mutants have loss-of-function point mutations or deletions in mitochondrial genes). Crosses of the mutant with the ρ− suppressor to mit− testers indicated the presence in the suppressor genome of the OLI2/ATP6 and COB but not the COX1, COX2, and COX3 markers. The cox1 tester (aM10-150-4D) used in these crosses contained a large deletion in COX1.
Mapping and sequence analysis of the ρ− suppressor:
The locus of mtDNA responsible for suppressing the atp22 null mutation was assessed by ρ− deletion mapping. A library of ρ− clones was obtained by mutagenesis of the revertant W303ΔATP22/R3 with ethidium bromide. Crosses of the purified clones to the atp22 null mutant and to several mit− testers representing different regions of mtDNA indicated that the suppressor activity was coretained with the ATP6 but not the COB marker (data not shown). This result pointed to ATP6 as the target of Atp22p.
The ρ− genome of one of the clones, containing only the ATP6 marker, was partially sequenced. The sequence indicated that the suppressor genome had the wild-type ATP6 sequence except for the absence of the normal 5′-UTR and the first three codons of the gene. In their place the second nucleotide of the fourth codon of ATP6 was fused to the 3′ end of the first intron of COX1 (Figure 3A). Although the ρ− mtDNA was not sequenced in its entirety, the continuous sequence obtained indicated that the entire COX1 5′-UTR together with its first exon and the entire first intron was fused to the ATP6 coding sequence (Figure 3B). The same gene fusion was found in the mtDNA isolated from the revertants W303ΔATP22/R1 and W303ΔATP22/R2. The sequence of the ρ− mtDNA also indicated the presence of ENS2 (Nakagawa et al. 1991), a gene that lies downstream of ATP6 in the mitochondrial genome of W303.
Analysis of mitochondrial transcripts in the atp22 revertant:
The presence of subunit 6 in the atp22 revertants indicates that this mitochondrial gene product is translated from a transcript emanating from the ATP6/COX1 fusion gene in the ρ− suppressor. In wild type, COX1 is cotranscribed with ATP6, ATP8, and ENS2 (Pelissier et al. 1995). This primary polycistronic transcript is cleaved after ENS2 and at two sites 600 bp apart located between COX1 and ATP8 (Pelissier et al. 1995). These endonucleolytic cleavages produce two distinct ATP6 mRNAs of 4.6 and 5.2 kb, the latter having a 5′ terminus coincident with the 3′ end of COX1 mRNA. On the basis of the structure of COX1/ATP6 in the suppressor and assuming transcription of the gene is initiated from the normal COX1 promoter, the primary transcript should be 6.6 kb. This transcript should be reduced by 2.5 kb after splicing of the COX1 intron. Mitochondrial RNAs were prepared from wild type, the atp22 point mutant C208, and the revertant W303ΔATP22/R3. The point mutant was used instead of the null mutant because of the extreme instability of wild-type mtDNA in the latter. To maximize the number of heteroplasmic cells with both wild-type and suppressor genomes, the revertant was grown in rich glycerol/ethanol medium. Even under these selective growth conditions, only 24% of the cells were ascertained to be heteroplasmic following overnight growth. Northern blots of the different RNAs from the three strains were hybridized to an ATP6 probe, a COX1 exonic probe, and a probe from the first intron of COX1. The ATP6 probe hybridized to the expected 4.6- and 5.2-kb transcripts in the wild type and the point mutant. (Since the agarose gel was run under nondenaturing conditions the true sizes of all the RNAs should be less than those estimated from their positions relative to the DNA size standards.) The two ATP6 transcripts were less abundant in the mutant because 75% of the cells used to prepare mitochondria consisted of ρ− and ρo mutants. The 4.6- and 5.2-kb transcripts were present in the revertant but in addition two novel transcripts of ∼4 and 7 kb were also detected (Figure 4B). Hybridization of the Northern blot with a COX1 exon aE1 probe revealed the expected COX1 mRNA of 2.2 kb and a less abundant 4-kb transcript corresponding to a partially spliced intermediate in wild type and the mutant (Figure 4C). The COX1 exonic probe detected the same transcripts in the revertant but the larger 4-kb transcript was more abundant relative to the 2.2-kb mRNA, implying that more than one transcript contribute to the 4-kb signal in this strain. The similar migration of the 4-kb transcripts detected by the COX1 aE1 and the ATP6 probes in the revertant suggests that the ATP6 transcript is responsible for the stronger signal. This was confirmed by hybridization of the aE1 probe to mitochondrial RNA from a revertant stain (aW303ΔATP22/R3/Io) in which the ρ+ genome of W303 was substituted by the ρ+ genome of aMY375/Io, a strain devoid of all mitochondrial introns. In this background there was no contribution of the partially spliced COX1 transcript in the 4-kb region (not shown). The different RNAs were also hybridized to a COX1 aI1 intron probe. This probe detected the aI1 circular group II intron in all three strains (Figure 4D). The probe also revealed a 6.6-kb transcript in the revertant. This transcript was similar to the 6.6-kb RNA seen in the Northern of the revertant hybridized to the ATP6 and COX1 exon aE1 probes (Figure 4, B and C).
On the basis of the results of these hybridizations we conclude that the 6.6-kb ATP6 transcript in the revertant contains a hybrid gene in which the aE1 exon and aI1 intron of COX1 are joined to the fourth codon of ATP6 as shown in Figure 3. The novel 4-kb transcript is a spliced product of the 6.6-kb transcript from which the intron has been excised, presumably at the same splice sites as in the native COX1 gene, resulting in a transcript (COX/ATP6) with an in-frame fusion of the COX1 aE1 exonic sequence to the ATP6 sequence.
Unprocessed subunit 6 in the atp22 revertant:
The primary translation product expressed from the spliced novel COX1/ATP6 mRNA of the revertant should be a hybrid protein starting with the entire sequence encoded by the first exon of COX1 followed by the subunit 6 sequence beginning with the fourth codon of the ATPase sequence and differing only at that position where the leucine is substituted by a valine (Figure 3A). The valine codon is formed as a result of splicing of the aI1 intron. This protein has a predicted mass of 35 kDa or ∼6 kDa larger than that of the normal mature subunit 6. In vivo labeling of mitochondrial gene products disclosed the presence in the revertants but not in wild type of a novel protein that migrated to a position midway between the 31- and 21-kDa markers (not shown). A protein with the same migration was also detected by Western analysis of mitochondria with an antibody to subunit 6 (Figure 2C), confirming the identity of this protein as the primary translation product from the fusion gene.
Since the primary translation product of the COX/ATP6 fusion gene retains the proteolytic cleavage site of the subunit 6 precursor (Figure 3A), processing would be expected to remove the sequence encoded by the first exon of COX1 together with the remaining seven residues of the subunit 6 presequence. The protein resulting from this cleavage has a sequence identical to that of wild-type mature subunit 6. This is supported by the identical electrophoretic behavior of subunit 6 in wild type and the revertants (Figure 2, B and C).
The proton-translocating ATPase is composed of the two functionally distinct sectors, F1 and F0. The yeast nuclear ATP22 gene was previously shown to code for a mitochondrial protein essential for assembly of F0 (Helfenbein et al. 2003). The earlier studies, however, fell short of defining the function of Atp22p. Here we present evidence that Atp22p is required for translation of ATPase subunit 6 encoded by the mitochondrial ATP6 gene.
Both in vivo and in organello labeling of mitochondrial gene products indicate that mutations in ATP22 elicit a deficit of subunit 6. The failure of atp22 mutants to synthesize subunit 6 can stem from instability/impaired processing of the ATP6 transcript or inability of the mutants to translate the protein. Subunit 6 is translated from two transcripts of 4.6 and 5.2 kb, both of which contain ATP6, ATP8, and ENS2. The latter gene is not present in all yeast strains but is a feature of the mitochondrial genome of W303. ATP8 and ENS2 code for subunit 8 of the ATPase (MacReadie et al. 1983) and an endonuclease (Nakagawa et al. 1991), respectively. Three nuclear genes have been implicated in processing and translation of the transcript. Mutations in NCA2 and NCA3 when present in the same strain produce an increased ratio of the 4.6-kb relative to the 5.2-kb transcript and decreased translation of subunits 6 and 8 (Camougrand et al. 1995; Pelissier et al. 1995). Mutations in AEP3 render the ATP6/ATP8/ENS2 transcripts unstable (Ellis et al. 2004). Northern analysis of ATP6 transcripts in an atp22 point mutant indicates the presence of both the 4.6- and the 5.2-kb transcripts, although processing of the larger 5.2- to the 4.6-kb transcript is less efficient than in wild type. The presence of the two mRNAs suggests that the absence of subunit 6 is unlikely to be caused by instability of the ATP6 mRNAs or a defect in processing of the primary transcript, which in addition to the aforementioned ATP6, ATP8, and ENS2 also contains COX1 (Pelissier et al. 1992). The Northern results instead imply that the subunit 6 deficit is caused by a translation defect.
Translation of subunit 6, oligomycin sensitivity, and growth of the atp22 null mutant on nonfermentable carbon sources are partially restored by mutations in mtDNA. The mitochondrial suppressors in the three revertants studied have been mapped to ATP6 and shown to consist of ρ− genomes with rearrangements such that the 5′-UTR and first exon and intron of COX1 are fused to ATP6 starting with the fourth codon of the gene. The partial rescue of the atp22 mutant by the ρ− suppressor depends on its coexistence with a ρ+ genome in the same cell. This heteroplasmic condition is very unstable and under nonselective conditions the revertant phenotype is rapidly lost as a result of rapid segregation of the two mtDNAs. The requirement of both genomes for respiratory growth suggested that the subunit 6 detected in the revertant is derived from the ρ− suppressor while all the other gene products are expressed from the ρ+ genome. Translation of subunit 6 in the revertant, therefore, is a function of the substitution of the COX1 sequence for the normal 5′-UTR of ATP6.
Mitochondrial ρ− genomes with heterologous 5′ upstream sequences have been shown to suppress mutations in translation factors for COX1 (Manthey and McEwen 1995), COX2 (Poutre and Fox 1987), COX3 (Muller et al. 1984), and COB (Rodel and Fox 1987). These transcript-specific translation factors interact with the 5′-UTRs of their cognate mRNAs. In each case the presence of the suppressor obviates the requirement for the missing or mutated translation factor by recruitment of an alternate factor specific for the new 5′-UTR (Costanzo and Fox 1988, 1990). In this study translation of subunit 6 in the atp22 revertants is probably mediated by interaction of COX1 translation factors with the COX1 5′-UTR in the novel ATP6 transcript. Our results are, therefore, consistent with Atp22p being a translation factor specific for the ATP6 mRNA. The specificity of Atp22p for ATP6 is also supported by the ability of atp22 mutants to translate subunits 8 and 9. This indicates that even though ATP6 and ATP8 are located on the same transcript, their translation is controlled by different translation factors.
The ρ− suppressor in the revertant is of some interest in itself. Unlike other previously described suppressors with novel 5′-UTR fusions, the COX1/ATP6 gene contains not only the 5′-UTR but also the first exon and intron of COX1. The primary transcript resulting from the COX1/ATP6 gene must be spliced to remove the intronic sequence and to achieve an in-frame fusion of the COX1 aE1 exon to the ATP6 coding sequence. This is supported by the presence in the atp22 revertant of two new transcripts, only the larger of which hybridizes to the aI1 intron probe. Splicing of the first COX1 intron depends on a splicing factor (maturase) encoded by the first intron of COB (De La Salle et al. 1982). The presence of COB in the ρ− genome of W303ΔATP22/R3 accounts for its ability to function as a suppressor even in the background of intronless mtDNA.
The subunit 6 synthesized in the revertant has an electrophoretic migration identical to that of the wild-type protein, indicating that the protease responsible for processing the amino-terminal 10 residues of the primary translation product (Michon et al. 1988) can also process the new hybrid protein, most likely at its normal site despite the presence of the N-terminal sequence of cytochrome oxidase subunit 1. That the primary translation product of the COX1/ATP6 fusion gene contains the sequence encoded by aE1 is supported by the presence in the revertant of a novel translation product of a size commensurate with the extra residues contributed by the aE1 sequence. This protein is detected both by pulse labeling of mitochondrial translation products in the revertant cells and by Western blot analysis of mitochondria with an antibody against subunit 6.
This study identifies a new translational factor specific for subunit 6 of the ATPase. Currently ATP8 and VAR1 are the only genes of the yeast mitochondrial genome for which translational factors are not known.
We thank Jean Velours (Institut de Biochimie et Genetique Cellulaires du Centre National de la Recherche Scientifique, Bordeaux, France) for his generous gift of the polyclonal antibody against subunit 6 of the yeast ATPase. This research was supported by National Institutes of Health research grants HL2274.
Communicating editor: M. D. Rose
- Received September 13, 2006.
- Accepted October 16, 2006.
- Copyright © 2007 by the Genetics Society of America