The Dominant PNM 2 - Mutation Which Eliminates the q 5 Factor of Saccharomyces cerevisiae Is the Result of a Missense Mutation in the SUP 35 Gene

The PNM2mutation of Saccharomyces cereuisiaeeliminates the extrachromosomal element 9. PNM2 is closely linked to the omnipotent suppressor gene SUP35 (also previously identified as SUP2, SUFl2, SAL3 and G S T l ) . We cloned PNM2and showed that PNM2 and SUP35 are the same gene. We sequenced the P N M Z mutant allele and found a single G --f A transition within the N-terminal domain of the protein. We tested the effects of various constructs of SUP35 and PNM2on $ inheritance and on allosuppressor and antisuppressor functions of the gene. We found that the C-terminal domain of SUP35 protein (SUP35p) could be independently expressed; expression produced dominant antisuppression. Disruption of the N-terminal domain of PNMZdestroyed the ability to eliminate 9. These results imply that the domains of SUP35p act in an antagonistic manner: the N-terminal domain decreases chain-termination fidelity, while the Cterminal domain imposes fidelity. Two transcripts were observed for SUP35, a major band at 2.4 kb and a minor band at 1.3 kb; the minor band corresponds to 3‘ sequences only. We propose a model for the function of SUP35, in which comparative levels of Nand C-terminal domains of SUP35p at the ribosome modulate translation fidelity. T HE 1cf factor of Saccharomyces cereuisiae affects translation fidelity by acting to enhance the efficiency of weak tRNA suppressors of nonsense (COX 1965, 1971; LIEBMAN et al. 1975; LIEBMAN and SHERMAN 1979; ONO et al. 1979a,b). 1cf also has some effect on certain frameshift suppressors and on phenotypic suppression of nonsense mutations by paromomycin (see review by Cox et al. 1988). In vivo, t/J is most commonly assayed by its ability to enhance the activity of the serineinserting ochre suppressor SUQP(SUP16), by which phenotype it was first identified; SUQ5”‘ can suppress efficiently only in a [+‘I background. Suppressor tRNA-mediated read-through of nonsense codons in in vitro translation experiments can be observed only in +’ cell-free extracts (TUITE et al. 1983, 1987). The experiments suggest that [$‘I and [$-I strains differ in the presence (in [$-I) or absence (in [$‘I) of a protein fidelity factor that loosely binds to ribosomes. The 1cf “gene” is not thought to code for that protein, but to control its presence or absence (Cox et al. 1988). 1cf is different from other factors that affect translational suppression (except for q, LIEBMAN and ALL-ROBW 1984) in that it is inherited as an extrachromosomal geEngland. ’ Present address: British Biotechnology plc, Watlington Road, Oxford, chemistry, Oxford University, South Parks Road, Oxford OX1 3QU, England, ’To whom correspondence should be addressed at: Department of BicMedicine, Bronx, New York 10461. Present address: Department of Cell Biology, Albert Einstein College of Present address: Cold Spring Harbor Laboratory, P.O. Box 100, Cold Spring Harbor, New York 11724. Genetics 137: 659-670 uuly, 1994) netic factor: in a cross between a [ 14’1 strain and a [+-I, all spores are [$+I. It appears as if a 1cf’ gene is carried on a plasmid or other extrachromosomal nucleic acid. However, in spite of numerous searches (MCCREADY and MCLAUCHLIN 1977; LUND 1982; TUITE et al. 1982), no extrachromosomal location has been identified as the carrier of the 1cf gene. At least two nuclear genes, PNMl and PNM2, are known to affect 1cf inheritance (YOUNG and COX 1971). The PNMmutations which have been analyzed genetically are dominant. They appear to have two effects: they inactivate ++, and they also appear to block its replication. Thus in a cross between a P N M l or PNM2strain and a [$‘I strain, the diploid has the [$-I phenotype, but if sporulated immediately will give rise to some [++I spores. 1cf’ determinants are only diluted out by subsequent cell divisions of the diploid, and eventually sporulation produces only 4:O 1cf-:t/J+ tetrads (YOUNG and Cox 1971; MCCREADY et al. 1977). Identification of the PNM genes and their products should help us understand 1cf. Recessive nuclear mutations also occur which enhance the efficiency of tRNA suppressors. They define five “allosuppressor” loci, SALl-SAL5 (COX 1977). PNM2 is closely linked to one of these loci, SAL3. In this report we show that a sal3 mutation cannot be separated from PNM2 by genetic mapping. Nevertheless, it is possible to isolate sal3 mutations in PNM2mutant strains without affecting the PNM2 phenotype. The wild-type SAL3 gene, cloned by CROUZET and TUITE (1987), is identical to the gene identified as SUP35 (H.~WTHORNE 660 S. M. Doe1 et al.