The product of the RADI gene of the yeast Saccharomyces cerevisiae is believed to function at the incision step of excision repair of DNA damage (9). Thus, it is not surprising that defects in this gene increase sensitivity to a variety of DNA-damaging agents, including UV, the UV-mimetic chemical 4-nitroquinoline-1-oxide, mono-and bifunctional alkylating agents,.and photoactivated psoralens (6,8,9,40,41,49). As well, the RADI gene product is required for the repair of NM-methyladenine (13). Not only do RADI deficiencies sensitize cells to the lethal effects of certain genotoxic agents, but also they enhance UV and 4-nitroquinoline-1-oxide-induced mutagenesis (27, 40) and UV-induced mitotic interchromosomal recombination (45). Recent findings suggest that the RADI gene product also plays a role in mitotic intrachromosomal recombination and in integration of linear DNA molecules into homologous genomic sequences (2,17,44). In addition to these various properties, defects in RADl confer a mutator phenotype (35,43,47). Both enhanced locus reversion to prototrophy and forward mutation to suppression and canavanine resistance have been reported, but neither the precise mutational changes involved nor the specificity of the mutator effect has been elucidated.To account for the association between repair defects and enhanced spontaneous mutagenesis in yeast cells, von Borstel and colleagues (12, 42) have invoked the repair channeling hypothesis (5, 10). According to this hypothesis, impairment of a specific repair pathway results in channeling of damage normally repaired by that pathway along other, competitive pathways. In a similar fashion, shunting of spontaneous DNA lesions through mutagenic repair pathways might account for elevated spontaneous tnutation in repair-deficient strains. However, the nature of these errorprone pathways has remained obscure, and generally the magnitudes of the mutator effects in yeast cells are relatively small. Consequently, it has been suggested that repair defects might have a more indirect influence on spontaneous * Corresponding author. mutation, perhaps by somehow reducing the fidelity. of DNA replication (43) or the efficiency of correcting replication errors (26). Alternatively, the products of some repair genes might play a regulatory role (7,16) so that modulation of processes other than DNA repair could be responsible for the mutator phenotypes of certain repair-deficient mutants.Characterization of the locations and types of DNA sequence alteration occurring spontaneously within a single gene in a radl background would provide valuable infoi-mation about the specificity of the radl mutator effect. In turn, this could yield important clues about the mechanism(s) of enhanced spontaneous mutagenesis in excision repair-deficient strains. To examine mutational specificity, we previously developed a system for the DNA sequence analysis of forward mutations in the yeast tRNA suppressor gene . In this system, all types of base pair substitution as well as deletions, duplications, in...

Inactivation of the Saccharomyces cerevisiae RAD18 gene confers a mutator phenotype. To determine the specificity of this effect, a collection of 212 spontaneous SUP4-o mutations arising in a radl8 strain was characterized by DNA sequencing. Comparison of the resulting mutational spectrum with that for an isogenic wild-type (RAD18) strain revealed that the radl8 mutator specifically enhanced the frequency of single base pair substitutions. Further analysis indicated that an increase in the frequency of G.C-*T-A transversions accounted for the elevated SUP4-o mutation frequency. Thus, radl8 is the first eucaryotic mutator found to generate only a particular base pair substitution. The majority of G.C pairs that were not mutated in the radl8 background were at sites where G-C-*T.A events can be detected in SUP4-o, suggesting that DNA sequence context influences the radl8 mutator effect. Transformation of heteroduplex plasmid DNAs into the two strains demonstrated that the radl8 mutator did not reduce the efficiency of correcting G-A or C-T mismatches to G.C pairs or preferentially correct the mismatches to A.T pairs. We propose that the RAD18 gene product might contribute to the fidelity of DNA replication in S. cerevisiae by involvement in a process that serves to limit the formation of G-A and C-T mismatches at template guanine and cytosine sites during DNA synthesis.The RAD18 gene of the yeast Saccharomyces cerevisiae is believed to function in the repair of DNA damage (14). Originally isolated on the basis of increased sensitivity to the lethal effects of UV and X-rays (50), radl8 mutations also sensitize cells to killing by gamma rays (34), the UV mimetic chemical 4-nitroquinoline-1-oxide (46), the folate antagonist trimethoprim (11), the antitumor antibiotic bleomycin (37), and a number of mono-and bifunctional alkylating agents (5,8,30,46). In addition, defects in RAD18 confer a mutator phenotype (48, 61), increase spontaneous, UV-, and gamma ray-induced mitotic recombination (4, 33, 52), enhance the induction of reverse mutation by nitrous acid (46) and decrease UV-, ethylmethanesulfonate-, and 4-nitroquinoline-1-oxide-induced reversion of certain cycl alleles (29,30,46). Furthermore, RAD18 has been found to be semidominant for UV and trimethoprim sensitivity and for enhanced spontaneous and UV-induced mitotic recombination (33), but not for methylmethanesulfonate sensitivity or the mutator phenotype (4, 61).Despite the radiation sensitivity of radl8 strains, they are capable of excising UV-induced pyrimidine dimers and appear to repair gamma ray-induced DNA single-and double-strand breaks (40, 51). However, the efficiency of strand break repair is obscured by a progressive decrease in DNA size during incubation after gamma irradiation (40). There is evidence that a component of postreplication repair in S. cerevisiae requires the RAD18 gene product (9, 47), and it has been suggested that a deficiency in base excision repair subsequent to endonuclease action might account for the cross-sensitivity of radl8...

SUP4-o, a suppressor tRNA gene, is the target in a system for characterizing mutational specificity in the yeast Saccharomyces cerevisiae. To date, screening for loss of suppression has detected 172 of the 267 base-pair substitutions possible in the exons and intron of the SUP4-o gene. Although many of the remaining 95 changes might not be detected by this screen, some might occur spontaneously, or be induced, only at very low frequencies. For the purpose of analyzing mutational specificity, it would be valuable to determine which of these substitutions can be detected with the genetic screen employed in this system and which cannot. Thus we used in vitro mutagenesis to generate the 95 substitutions not yet detected in SUP4-o in vivo. Assessment of the phenotypes conferred by these 95 directed mutations revealed that 50 would completely escape detection and only 45 would pass through the first stage of the screen. Of these 45 substitutions, 2 are detectable but have not yet been found among more than 5,000 characterized SUP4-o mutations that arose in vivo. In addition, 4 others should be detected by slightly relaxing the current criteria for selection of SUP4-o mutants. The results indicate that with these modifications the system can detect 174/225 substitutions possible in the SUP4-o exons and 4/42 in the intron.

Inactivation of the Saccharomyces cerevisiae RAD18 gene confers a mutator phenotype. To determine the specificity of this effect, a collection of 212 spontaneous SUP4-o mutants arising in a rad18 strain was characterized by DNA sequencing. Comparison of the resulting mutational spectrum with that for an isogenic wild-type (RAD18) strain revealed that the rad18 mutator specifically enhanced the frequency of single base pair substitutions. Further analysis indicated that an increase in the frequency of G.C----T.A transversions accounted for the elevated SUP4-o mutation frequency. Thus, rad18 is the first eucaryotic mutator found to generate only a particular base pair substitution. The majority of G.C pairs that were not mutated in the rad18 background were at sites where G.C----T.A events can be detected in SUP4-o, suggesting that DNA sequence context influences the rad18 mutator effect. Transformation of heteroduplex plasmid DNAs into the two strains demonstrated that the rad18 mutator did not reduce the efficiency of correcting G-A or C-T mismatches to G.C pairs or preferentially correct the mismatches to A.T pairs. We propose that the RAD18 gene product might contribute to the fidelity of DNA replication in S. cerevisiae by involvement in a process that serves to limit the formation of G-A and C-T mismatches at template guanine and cytosine sites during DNA synthesis.

No abstract