Dna Repair Dependence of Somatic Mutagenesis Drosophila Melanogaster after Treatment with Alkylating Agents of Transposon-caused White Alleles I N

DNA repair-defective alleles of the mei-9, mei-41, mus-104 and mus-101 loci of Drosofihila melanogaster were introduced into stocks bearing the UZ and SZ marker sets. Males with the UZ marker set, z1 (zesle allele) and w + ( ' ~ ) (genetically unstable white allele presumably caused by a transposable element), or the SZ marker set, z ' and w + ~ (semistable white allele caused by partial duplication of the w+ locus plus transposon insert), were exposed to EMS at the first instar. After emergence, adult males bearing red spots on lemon-yellow eyes were scored as flies with somatic reversions of w + ( ' ~ ) or w + ~ . T h e relative mutabilities (relative values of reversion frequency at an equal EMS dose) of either w + ( ' ~ ) or w + ~ in a repair-proficient strain and in mei-9, mei-41, mus-104 and mus-101 strains were 1:-1.2:0.3:0.3:0.7, despite the fact that w+(TE) reverted two to three times as frequently as w + ~ under both the repair-proficient and repair-deficient genetic conditions. Similarly, after treatment with MMS, MNNG and ENNG, wYTE) was somatically more mutable in the mei-9 strain and less mutable in the mei-41 and mus-104 strains than in the repair-proficient strain. From these results, we propose that mutagenic lesions produced in DNA by treatment with these chemicals a r e converted to mutant DNA sequences via the error-prone repair mechanisms dependent o n the products of the genes mei-41+ (mei-41 and mus-104 being alleles of the same locus) and mus-l01+, whereas they a re eliminated by mei-9+dependent excision repair. In contrast to the approximately linear responses of induced reversions of w+('") with ENNG in the repair-proficient, mei-9, and mei41 strains, seemingly there were dosage insensitive ranges for induced reversion with MNNG in the repair-proficient and mei-41 strains, but not for reversion in the mei-9 strain; w + ( ' ~ ) in the mus-104 strain was virtually nonmutable with MNNG and ENNG. These results suggest that O'-methylguanine (O'MeG) produced in DNA with MNNG, but not O'-ethylguanine produced with ENNG, is almost completely repaired in a low dose range by constitutive activity of DNA 06MeG transmethylase. From the distribution of clone sizes of spontaneous revertant spots and other data, we propose that both w+(") and w + ~ have a similar tendency to spontaneously revert more frequently at early rather than a t late Abbreviations used: EMS, ethyl methanesulfonate; ENNG, N-ethyl-A"-nitro-N-nitrosoguanidine; ENU, ethyl nitrosourea; MMS, methyl methanesulfonate; MNNC, N-methyl-N'-nitro-N-nitrosoguanidine; MNU, methyl nit rosourea. ' PreFent address: Central Research Division, Tdkeda Chemical Industries Ltd., Osaka 532 , Japan. Genetics 112: 505-522 March, 1986. 506 K . FUJIKAWA AND S. KONDO developnierital stages, probably reflecting a common property of their inserted transposons. N both prokaryotic and lower eukaryotic organisms, it has been shown that I mutability by various alkylating agents is dependent on repair capacities of target cells. The recApostreplication-repair defective mutation reduces the mutability of Escherichia coli by EMS slightly, by MNNG moderately, and by MMS completely (KONDO et al. 1970; ISHII and KONDO 1975), whereas the uvrAor uvrBexcision-repair defective mutations enhance the mutability by EMS and ENU (KONDO et al. 1970; GARNER, PICKERING and MARTIN 1979; WARREN and LAWLEY 1980; TODD, BROUWER and GLICKMAN 1981). A strain of E. coli unable adaptively to produce DNA 06-methylguanine methyltransferase is more mutable when treated with EMS, ENU, MMS, ENNG and MNNG than is a wild-type strain with the normal ability to produce adaptively the transmethylase UEGGO 1979; SEDGWICK and LINDAHL 1982). The rad6 or rad9 postreplication-repair defective mutations reduce the mutability of Saccharomyces cerevisiae by EMS, MMS, MNNG and other alkylating agents (PRAKASH 1974). Several mutants belonging to the rad52 epistasis group, whose members are defective in repairing double-strand breaks, also reduce EMSinduced mutations (PRAKASH and HIGGINS 1982). A major question posed in the present study is whether mutation induced in the complex eukaryote D. melanogaster by EMS, MMS, MNNG and ENNG depends on DNA repair. WURGLER and his co-workers (GRAF, GREEN and WURGLER 1979; GRAF and WURGLER 1982) have recently reported that males treated with EMS, MMS, ENU and MNU produce lower yields of recessive Xchromosome lethals when crossed to females homozygous for the mei-4ID5 mutation defective in postreplication repair than when crossed to normal females. Furthermore, there is suggestive evidence that the introduction of the excision-repair deficient mutations mei-9 or mus(2)201 enhances sex-linked recessive lethals induced by treating germ cells of both sexes with EMS, MMS, MNNG and MNU (GRAF, GREEN and WURGLER 1979; GRAF and WURGLER 1982; VOCEL 1982; SMITH et al. 1982; SMITH, BAUMEN and DUSENBERY 1983; BADARUDDIN, DUSENBERY and SMITH 1984; VOGEL, DUSENBERY and SMITH 1985). To obtain a precise relationship between mutability and DNA repair deficiency, we have sought mutant marker genes which are somatically revertible at a high frequency because somatic mutagenesis can be studied more easily and somatic cells have been employed to study the mechanism of DNA repair at the molecular level (BOYD et al. 1983). In the present experiments, we used two types of w+ loci, each of which evokes the expression of the zeste eye-color mutant in males. We monitored changes at the w+ locus after treating larvae with mutagens by scoring phenotypic reversions of zeste to wild type in the eyes of eclosing males. Into chromosomes carrying each locus, we introduced either a postreplication-repair deficient mutation ( mei-#In5, mus(l)lO#”’ or mus(1)lOl”’) or an excision-repair deficient mutation mei-9” (e.g., see review of BOYD et al. 1983). Previously, BAKER and his colleagues (BAKER et al. 1976; BAKER, CARPENTER and RIPOLL SOMATIC MUTAGENESIS IN DROSOPHILA 507 1978; BAKER and SMITH 1979; GATTI 1979) reported that the mei-Y, mei-4ID5 and mus(1)lOl"' mutations greatly enhanced spontaneous induction of chromosome breaks and somatic chromosome mutations. Here, we report that (1) somatic reversions of zeste to wild type, which measure genetic changes at the w+ locus, after treatment of larvae with EMS increased in the mei-9" strains but decreased in the strains carrying mei-4ID5, m u ~ ( l ) l 0 4 ~ ' and mus(1)lOl"'; (2) there is a dosage insensitive range for the aforementioned phenotypic reversion induced by MNNG, but not by ENNG; and (3) spontaneous phenotypic reversions occur more frequently at early rather than at late developmental stages. These results are discussed in relation to the role of DNA repair in mutagenesis. MATERIALS AND METHODS Strains for the assay of somatic reversions: Strains of D. melanogaster with UZ and SZ marker sets used for assaying reversions were obtained from B. RASMUSON. Both marker sets had the recessive zeste marker z L (1-1.0) and a white allele (1-1.5). The white allele w+(~') in the UZ marker set is genetically unstable, probably because a transposable genetic element (TE) is inserted in or near the rightmost part of the white locus (RASMUSON et al. 1981). The white allele in the SZ marker set contains a tandem duplication of the proximal part of the white locus (GREEN 1963) and a transposon insert between the duplicates (GOLDBERG et al. 1983). This duplication denoted as D p (W+R61 e ) (GREEN 1963; RASMUSON et al. 1978) will be named w + ~ (see GOLDBERC et al. 1983) here. Thus, symbols UZ and SZ stand for the marker sets z1 w+(~') and zl w + ~ , respectively. Both w + ( ~ ~ ) and w + ~ promote phenotypic expression of the recessive mutant z ' giving rise to a lemon-yellow eye in the adult fly. A single, normal white (w+) gene does not promote the expression of the allele z' [for zeste-white interaction, see JUDD (1976)l. Loss of the zeste-promoting activity by germinal mutation in the white loci of w+(Tb:) or w + ~ in the tester strains results in the production of offspring with red eyes (RASMUSON et al. 1978). Somatic reversion of or w + ~ in stem cells of an eyeimaginal disc at a larval stage leads to the production of a red spot against the lemonyellow background of the eye in an adult male. Construction of reversion-tester strains with DNA repair-deficient mutations: The stocks used had X-chromosomes marked with y mei-9" (I-6), w mei-4ZD5 (I-53.8), w mu~( l )Z04~ ' (I-53), w mus(Z)ZOZD1 (I-9), and w (obtained from J. B. BOYD). The map positions of these repair-deficient mutations are taken from SMITH, SNYDER and DUSENBERG (1 980). The mei-9" strain is unable to remove UV-induced pyrimidine dimers from DNA (BOYD, GOLINO and SETLOW 1976; HARRIS and BOYD 1981) and has reduced apurinic endonuclease activity (OSGOOD and BOYD 1982). The mei-4ZD5, mu~(Z)Z04~' and mus(Z)lOZ"' strains are deficient in postreplication repair of damage in DNA (BOYD and SETLOW 1976). The deficiency is thought to be an impaired ability to fill the gaps produced in DNA newly synthesized from UV damage-bearing templates (BROWN and BOYD 1981; BOYD and SHAW 1982). The mei-9", mei-4ZD5 and mus(I)ZOZD' strains exhibit reduced meiotic recombination (BAKER et al. 1976; BOYD et al. 1976) and enhanced chromosome instabilities in mitotic cells (BAKER, CARPENTER and RIPOLL 1978; BAKER and SMITH 1979; GATTI 1979; GATTI, PIMPINELLI and BAKER 1980). The four repair-deficient mutants used all have normal ability to repair double-strand breaks produced in DNA by X-irradiation (DEZZANI, HARRIS and BOYD 1982). Each of the four repair-deficient mutations was introduced into the UZor SZ-bearing chromosomes in the following way: First, single males with the genotype se U2 spl or SZ spl were mated to Base females with the genotype In(l)s21L seaR +S, s 8 ' se8 w' B. Resulting virgin F1 females were mated to males with each of the following genotypes: y mei-9", w mei-41D5 w m ~ s ( Z ) l 0 4 ~ ' , w mus(l)lOZD' or w. F2 non-Base females were crossed 508 K . FLJJIKAWA AND S. KONDO