Mutation Research 451 Ž2000. 277–293 www.elsevier.comrlocatermolmut Community address: www.elsevier.comrlocatermutres
The Saccharomyces repair genes at the end of the century John C. Game Life Sciences DiÕision, Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley, CA 94720, USA Received 23 November 1999; received in revised form 20 December 1999; accepted 23 December 1999
1. Introduction It is now more than 30 years since the isolation of radiation sensitive mutants of Saccharomyces cereÕisiae was first reported by Nakai and Matsumoto w1x and more extensively in the following year by Cox and Parry w2x. As described elsewhere in this book, during the intervening decades, Saccharomyces quickly became the key system for understanding DNA repair, and remains pre-eminent as a model organism in this area today. Here, I discuss the historical development of the Saccharomyces repair field, and give an overview of current knowledge about the repair loci. I include discussion of allelic relationships, epistasis groups, and phenotypes. There is a large body of additional and more detailed knowledge about specific pathways and enzymatic functions that is outside the scope of this chapter. Here, I have tried to cite references that are representative of the field, but there are many other excellent reviews and original papers that are too numerous to list, hence the reference list should not be considered comprehensive.
attempt to identify all the loci in the yeast genome that could mutate to confer sensitivity to UV. Although they identified 22 loci defined by one or more of 96 mutants isolated, they used Poisson statistics to argue that they had failed to saturate the genome. The mean number of mutants identified per locus and the distribution around this mean implied that there were probably several additional loci that by chance were not represented by any mutants in the collection. This was borne out later by the identification elsewhere of several more genes controlling UV resistance. In addition, collections of mutants conferring sensitivity to IR became available. Some mutants initially isolated on the basis of mild UV sensitivity w2x, for example rad50-1 and rad50-2, were found to be highly sensitive to IR, and a number of mutants first defined by other phenotypes were shown to confer slight UV sensitivity. The number of independent mutant collections available made it important to establish standard nomenclature, and to do this, it was essential to determine allelic relationships between mutants from different laboratories and among those isolated on the basis of different phenotypes.
2. Historical review At least nine laboratories published accounts of yeast mutants sensitive to ultraviolet light ŽUV. or ionising radiation ŽIR. between 1967 and 1970. An early surprise was the large number of loci found to be involved. Cox and Parry w2x in particular made an
3. Allelic relationships and nomenclature At the International Yeast Genetics Conference at Chalk River, Ontario, in 1970, it was agreed that genes which mutate to confer significant sensitivity
0027-5107r00r$ - see front matter q 2000 Published by Elsevier Science B.V. PII: S 0 0 2 7 - 5 1 0 7 Ž 0 0 . 0 0 0 5 5 - 5
J.C. Game r Mutation Research 451 (2000) 277–293
to radiation should be designated by the symbol RAD. Those primarily affecting UV repair or both UV and IR repair were to be numbered from RAD1 upwards, with numbers 1–49 set aside for them, whereas genes controlling IR sensitivity with only a minimal role in UV repair were to be numbered from RAD50 upwards. In 1971, Game and Cox w3x published a standardized set of RAD locus numbers from RAD1 to RAD22, resulting from allelism tests based primarily on complementation analysis with mutants from eight laboratories. This was followed in 3 years by a study of X-ray sensitive mutants by Game and Mortimer w4x that designated eight loci numbered from RAD50 to RAD57. While almost all the allelic relationships reported in these papers remain valid today, many other mutants have since been isolated and some of the original UV-sensitive mutants have been lost or can no longer be shown to segregate. In addition, observations by C. Lawrence Žpersonal communication. imply that the reÕ3 mutant isolated by Lemontt w5x is probably allelic with RAD8. To date, the additional loci given RAD numbers are as follows.
4. Additional RAD loci In the UV-repair series, the RAD23 locus was identified by McKnight et al. w6x while characterizing mutations that over-expressed the CYC7 gene product. They found that one such mutation consisted of a deletion of two adjacent loci so that the CYC7 structural gene became placed near a more actively transcribed site. One of these loci conferred UV sensitivity when deleted, and was named RAD23. McKnight et al. w6x argued that it was probably derived from RAD7 by an ancient duplication and transposition, since a three-gene cluster including RAD7 and CYC1 closely resembles the three-gene cluster including RAD23 and CYC7. Subsequent studies, reviewed in Ref. w7x, confirmed that rad7 and rad23 mutants each confer a partial defect in the incision step of nucleotide excision repair ŽNER., but do not completely substitute for each other’s function, since the double mutant is more sensitive than either single mutant w8x.
RAD24 was used by Eckardt-Schupp et al. w9x to denote the locus for the r1s mutant isolated by Averbeck w10x. Its mutants display moderate UV and IR sensitivity w10,11x and fail to undergo meiosis w9x. The gene maps within 2 cM of RAD3 on Chr. V w9x and is involved in mediating the G2 checkpoint control after DNA damage w12x. The mammalian homologue of the Sac. cereÕisiae RAD24 gene is also probably involved in mitotic G2 checkpoint control w13x and has been found to be most highly expressed in testis tissue w14x, probably implying a meiotic or premeiotic role, as in yeast. It is referred to as hRAD17 w13,14x based on its homology to Schizosaccharomyces pombe RAD17, which is also homologous to Sac. cereÕisiae RAD24, and it may be implicated in some human tumors w13,14x. RAD25 was identified as a repair gene by Park et al. w15x, who cloned the yeast homologue of the human xeroderma pigmentosum group B gene ŽERCC3.. The yeast locus was cloned independently based on suppression of blocked translation initiation and named SSL2 by Gulyas and Donahue w16x. Park et al. w15x found that deletions in this locus are inviable. However, a mutation in the part encoding the C-terminus was viable and conferred UV sensitivity via a defect in excision repair, hence they renamed the gene RAD25. More recently, Lee et al. w17x found that RAD25 Žwhich they identified from an independently isolated mutant named rtf4 . encodes a DNA helicase present in NER complexes. Lee et al. w17x also discuss parallels between RAD25 and RAD3. RAD26 was identified as a new locus in Saccharomyces by van Gool et al. w18x by searching for a gene with homology to the human ERCC6 locus, whose mutant alleles are responsible for complementation group B of Cockayne’s syndrome. Surprisingly, yeast disruption alleles of rad26 did not confer sensitivity to UV or IR, but the authors argued that the RAD designation was appropriate because the mutants showed a defect in transcription-coupled repair and a delay in recovery of growth after UV exposure w18x. RAD27 is the locus name given by Reagan et al. w19x for a gene identified earlier w20x as a novel sequence showing substantial homology to two regions at the N and C termini of the RAD2 gene. The sequence is immediately adjacent to the APN1 locus
J.C. Game r Mutation Research 451 (2000) 277–293
w20x. Disruptions of RAD27 created by Reagan et al. w19x were found to confer lethality at 378C. At lower temperatures, the same disruptions were viable but conferred increased sensitivity to UV light and methylmethane sulfonate ŽMMS., but not to ionizing radiation w19x. The RAD27 gene has been involved in many recent studies and its product is the yeast homologue of the mammalian flap endonuclease, FEN1 w21x. It has been shown to be required for base excision repair ŽBER. of apurinic sites in DNA by Wu and Wang w22x. Its homology to RAD2 probably reflects common endonuclease motifs rather than a role in NER, since double mutant analysis places it in the RAD6 UV epistasis group w19x. RAD28 was used by Bhatia et al. w23x for the Saccharomyces homologue of the human Cockayne syndrome A ŽCSA. gene. The null mutant resembled rad26 mutants in showing no sensitivity to radiation as a single mutant, in being hypermutable following UV exposure and in being mildly sensitive to UV in the presence of rad7 or rad16 mutations. However, unlike rad26, it did not show a defect in transcription-coupled repair w23x. RAD29 was defined by a new mutation that conferred hypersensitivity to UV in a RAD2 background w24x. The mutant was reported to show increased sensitivity to IR as a diploid but not as a haploid, and to show sensitivity to MMS and nitrous acid. The locus was mapped genetically to the right arm of Chr. II, but was not cloned or identified with an open reading frame. The map position excludes it from allelism with available named RAD loci with the possible exception of RAD16, but RAD29 could still be identical with a known locus named for a different phenotype. It probably belongs in either the RAD6 or the RAD50 series epistasis groups. RAD30, currently the highest number in this series, was identified by McDonald et al. w25x as a Saccharomyces gene with homology to the E. coli dinB and umuC and the Saccharomyces REV1 loci. A disruption allele of rad30 showed moderate UV sensitivity mediated via the error-free branch of the RAD6 epistasis group. McDonald et al. w25x found no defects in UV-induced mutability, although Roush et al. w26x did find subtle alterations in mutability. The gene was subsequently shown to code for DNA polymerase eta, responsible for error-free translesion synthesis w27x, and mutations in the human homo-
logue ŽhRAD30. have recently been shown to be responsible for the variant form of xeroderma pigmentosum ŽXP-V. w28x Žsee also Ref. w29x.. In the RAD50 upwards genes, Game and Mortimer w4x noted that two mutants, xrs2 and xrs4 described in 1970 by Suslova and Zakharov w30x, complemented mutants in 17 genes Žincluding those in the RAD6 epistasis group. and suggested that they could represent additional loci in the RAD50 to RAD57 series. This has been confirmed, and both genes are major players in the RAD50 subset of recombinational repair loci Žsee below.. XRS4 was shown in 1998 by Tsubouchi and Ogawa w31x to be allelic to a gene named MRE11 identified on the basis of a meiotic recombination defect by Ajimura et al. w32x in 1993. The designation RAD58 was proposed in 1995 for the XRS4 locus by Chepurnaya et al. w33x Žsee also Ref. w34x., but has not gained wide acceptance, and the locus is most commonly referred to as MRE11. RAD59 was used by Bai and Symington w35x for a new gene whose mutant allele greatly lowered the residual intrachromosomal recombination seen in rad51 strains. In a RAD51 wild-type background, the rad59 mutant is significantly X-ray sensitive but less so than the major rad52-group mutants. It lowers intrachromosomal recombination moderately while raising interchromosomal recombination frequencies. The RAD59 gene is about half the length of the RAD52 gene, and shows significant homology with RAD52, especially with the amino-terminal end, which is the region most conserved between different eukaryote species. The X-ray sensitivity of rad59 mutants can be partially alleviated by overexpressing RAD52 on a plasmid, suggesting some duplication of enzymatic function between the two gene products w35x. The XRS2 gene w30x remains distinct from other named and available RAD loci, and I propose that the locus number RAD60 be set aside for authors who wish to refer it using the RAD nomenclature. It encodes a protein forming part of a conserved complex including RAD50 and MRE11 w36,37x. Although the latter two genes have mammalian homologues, no true homologue has been identified for XRS2. However, a human protein similar in size to yeast Xrs2p is known to form a complex with the human Rad50p and Mre11p proteins. This protein,
J.C. Game r Mutation Research 451 (2000) 277–293
p95, is probably the functional equivalent of yeast Xrs2p, and is defective in the human disease Nijmegen breakage syndrome w38,39x. XRS2 represents the first of the RAD50 series genes found to show a probable relationship to a specific human disease, although mutations in mammalian homologues of several other members confer severe phenotypes in animals or in cell culture Žreviewed in Ref. w40x.. 5. The ‘‘lost’’ RAD loci Although 40 loci have been given RAD designations, a number of them have never been mapped, cloned or assigned to open reading frames. Their mutant alleles may no longer be available or may fail to show Mendelian segregation. Some of these ‘‘lost’’ loci may have been alleles of the more recently designated loci listed above, or alleles of other genes now known to have a minor role in repair, discussed below. The dubious loci are RAD11, 12, 13, 15, 19, 20, 21, and 22. There are several reasons for their neglected status. Apart from RAD15, each is known from only a single allele, and several of these, especially rad19-1 and rad21-1, conferred only minor sensitivity to UV w2x with no other known phenotype. This made them difficult to score in crosses, in contrast to mutants such as the rad50 series, which were also of limited UV sensitivity but could be readily monitored by their high X-ray sensitivity. In the mid-1980s, Game Žunpublished observations. out-crossed stocks of several of these mutants maintained in storage for 15 years. While some UV-sensitive spores were observed in the progeny, no clear 2:2 segregations were obtained. The results implied that suppresser or modifier mutations were present in the stocks themselves or in the wild-type parent, but surprisingly, back-crosses to this parent still failed to yield clear segregations. In addition, the original allele of RAD8 Ž rad8-1. is difficult to work with and may be lost, although reÕ3 mutant alleles, which are probably allelic with rad8 Žsee above., are readily available. 6. HDF1 and YKU80 Two other repair genes of great interest that have been identified recently are the Saccharomyces ho-
mologues of the mammalian KU70 and KU80 loci, named Žin yeast. HDF1 w41x and YKU80 w42x or KU80 w43x, respectively. In mammals, these gene products form a complex with the catalytic subunit of DNA protein kinase ŽDNA-PKcs.. This complex mediates a major route for DSB repair via an endjoining mechanism that is different from the recombinational repair mechanism mediated by RAD51 and related genes Žsee Ref. w44x for a review.. The yeast HDF1 gene was identified by Feldmann and Winnacker w41x from the sequence of the 70 kDa subunit of a heterodimeric DNA binding protein, designated high affinity DNA-binding factor ŽHDF. that was identified as binding to the ends of doublestranded DNA. The amino acid sequence showed significant homology with the p70 subunit of the human Ku autoantigen, and disruptions of the yeast gene encoding it conferred temperature-sensitive lethality at 378C w41x. The gene has also been referred to as YKU70 w42x. The yeast KU80 homologue was identified by screening the yeast genome for homology with the relevant mammalian and Caenorhabditis sequences w42,43x. Yeast yku80 disruption mutants resemble hdf1 mutants in being inviable at 378C w42x. At 308C, null mutants in each gene show moderate sensitivity to MMS but not to UV or IR, either as haploids or as homozygous diploids, and these diploids undergo a normal meiosis w41–43x. However, a sharp defect in end-joining repair of linearised plasmids is observed in hdf1 and in yku80 mutants w42,43x, and in one study w42x, a moderate increase in IR sensitivity has been reported in double mutants involving either of them with a rad52null mutation. In addition, they appear to be required for the maintenance of normal telomere length w42x. The DNA end-binding complex encoded by the associated products of these genes is absent in hdf1null and yku80null single mutants and in hdf1null yku80null double mutants w43x. The double mutants also resemble the single mutants in each of several phenotypes studied, implying epistasis between the two genes w42,43x, as expected from their known role in mammals. The existence of the Saccharomyces HDF1 and YKU80 genes implies that at least some aspects of this major mammalian DSB end-joining repair mechanism are also present in yeast. Currently, no yeast homologue for the mammalian catalytic subunit ŽPKcs. of the DNA-PK
J.C. Game r Mutation Research 451 (2000) 277–293
complex has been identified, although in mammals, the KU70 and KU80 gene products interact strongly with the DNA-PK catalytic subunit and are believed to guide the protein complex to damaged DNA. They may also have other functions that are independent of their interaction with DNA-PKcs w44x. In future, to conform with Saccharomyces nomenclature, the three-letter prefix YKU seems appropriate for the YKU80 gene in yeast, rather than the KU80 name sometimes used w43x.
7. Other Saccharomyces repair genes Many other yeast genes are known to play at least a minor role in repair but have not been designated as RAD loci. These include genes with an incidental role in repair as a consequence of a largely unrelated primary activity, and those with a role in more closely related processes such as mutagenesis or resistance to chemical mutagens. These are often named for the phenotypes by which they were identified, for example mutations in MMS genes confer sensitivity to MMS w45x and MUT genes w46x control rates of spontaneous mutation. Among such loci, some of the more significant players in repair are the REV genes, defined as required for wild-type levels of UV-induced mutagenesis. Their mutant alleles block or reduce reversion of auxotrophic signal mutations and confer mild to moderate radiation sensitivity through a defect in error-prone RAD6-mediated post replication repair Žsee below.. REV1, REV2, and REV3 were identified by Lemontt w5x. REV2 is allelic with RAD5 w3x and is usually referred to by the latter name. REV3 is probably allelic with the former RAD8 locus Žsee above., and encodes the catalytic subunit of DNA polymerase zeta. This enzyme is responsible for error-prone replication past DNA damage, and is a necessary prerequisite for most damage-induced and spontaneous mutagenesis in yeast w47x and probably in humans w48x. REV6 w49x and REV7 w50x are additional genes whose mutant alleles confer the reÕ phenotype and some sensitivity to UV or IR, while a mutant Ž ngm2-1. in the NGM2 locus w51x decreases nitrosoguanidine and UV mutagenesis but is scarcely sensitive to killing by UV.
Mutants isolated on the basis of MMS sensitivity were described by Prakash and Prakash w45x. They included many alleles of known RAD genes as well as some radiation-sensitive mutants in new loci, and some mutants sensitive only to MMS. The MMS19 gene represents a distinct locus whose mutants are inviable at 378C and are moderately sensitive to UV at lower temperatures. The MMS19 gene product appears to be involved in transcription and in excision repair, but its role in repair may be an indirect one w52x. There is no true congruence between repair of MMS damage and IR damage, although MMS is often used in place of IR by researchers who do not have access to radiation sources. rad mutants sensitive to IR are usually also sensitive to MMS, but in addition to its radiation-mimicking effect, MMS induces methylation damage to DNA bases, especially a 3-methyladenine lesion. A specific pathway known as BER operates to repair these and related chemical-induced lesions Žsee Xiao et al. w53x.. It involves removal of the damaged bases by DNA glycosylases, followed by action of an endonculease encoded by the APN1 gene w54x. This cleaves DNA strands at abasic sites, allowing subsequent repair. The APN1 locus is a repair gene that falls outside the major RAD groups but has a major role in the repair of spontaneous damage and the consequent control of spontaneous mutation rates w53,55x. The BER pathway is now thought to interact with the RAD gene pathways and other repair systems that provide partially overlapping functions in the removal of spontaneous damage w56x. DNA mismatch correction genes including the PMS, MSH, and MLH loci can be considered as repair genes since they function to repair mismatched base pairs and some of their products also play a role in the repair of DNA double-strand breaks w57,58x. Homologues of these genes predispose humans to several types of cancer Žsee Refs. w59,60x for reviews. hence mismatch repair genes Žreviewed in Ref. w61x. have been extensively studied and are of major importance. Mutations in mismatch repair genes usually confer sharply elevated spontaneous mutation frequencies in yeast w62x, but they do not confer significant radiation sensitivity by themselves and are not considered further here. In contrast, a role in repair of radiation damage is apparent for some genes that are required for viability, as
J.C. Game r Mutation Research 451 (2000) 277–293
shown by conditional mutant alleles that confer some radiation sensitivity at semi-permissive temperatures. These include loci such as CDC7, CDC9, CDC21, and several others whose products catalyse reactions that occur both in repair and as part of normal metabolism w63,64x. These essential genes, whose mis-sense alleles can encode temperature-sensitive products, should not be confused with RAD26, RAD27, MMS19, HDF1, and YKU80, whose null alleles confer temperature-conditional lethality. In addition, null alleles of two RAD genes, RAD3 and RAD53, confer lethality unconditionally but in these cases, the loci were given RAD designations based on radiation-sensitive mis-sense mutants prior to the finding that null mutants were inviable.
by classifying genes into distinct subgroups. Within the RAD50 upwards genes, two subgroups stand out, as discussed later, with some loci remaining separate from each. In excision repair, two partially overlapping sub-pathways have been distinguished, sometimes termed transcription-coupled repair and global genome repair Žsee Verhage et al. w70x.. The former primarily acts on damage in the coding strand of actively transcribed genes, whereas the latter acts on both strands of non-transcribed DNA. There are additional overlaps between NER functions and other pathways such as BER, mismatch repair, and recombinational repair. However, at present the RAD6 group remains the least well understood and most complex of the three main UV epistasis families, as discussed below.
8. Epistasis relationships in yeast repair genes 9. Epistasis relationships within the RAD6 group The facile genetic system of Saccharomyces is well suited to the construction of double and multiple mutants, and this opened the way for detailed epistasis analysis of repair mutants. The rationale of epistasis analysis and its results especially with respect to UV sensitivity have been reviewed elsewhere w65,66x. Broadly speaking, most repair genes in yeast can still be classified into one of three main families, controlling NER, post-replicational repair, or recombinational repair, respectively. In most cases mutants in the same epistasis group show a similarity in the phenotypes they influence, and have also been shown by other methods to affect the same repair processes. More recently, other genes have been studied that function in more than one group or are outside the three original groups. In addition, complexity within the groups has emerged. Genes in more than one group include some that code for essential functions but whose conditional alleles confer some radiation sensitivity in semi-permissive environments. An example is the CDC9 gene that codes for DNA ligase w67x. Ligation is clearly required to complete both excision repair and recombinational repair, as well as for DNA replication w64,68,69x. Genes outside the original groups may include HDF1, YKU80 and others that function in DSB repair by mechanisms independent of the major recombinational repair pathway. Complexity within the three main groups can be substantially addressed
The RAD6 gene in yeast occupies a central role in repair and a large group of other loci are often referred to as belonging to the RAD6 epistasis group. This remains a useful classification, since it differentiates these genes from either the excision repair loci or the RAD50 series, and rad6null alleles are truly epistatic to all other tested mutants in the RAD6 group. However, there is greater complexity and heterogeneity within the RAD6 group than is present in the other groups. First, many null alleles of loci hypostatic to RAD6 confer much less sensitivity than rad6null itself, and secondly, double mutant combinations of these null alleles among themselves frequently show additive or synergistic interactions rather than epistasis, despite falling individually under the RAD6 epistasis umbrella. The maximum sensitivity reported in these combinations is no greater than that of rad6null, suggesting that the other loci mediate separate branches of RAD6-dependent repair. This is supported by observations of differences among the mutants in other phenotypes, especially UV-induced mutagenesis. Such observations as well as biochemical data have led to the recognition of separate error-prone and error-free branches of RAD6 dependent post-replicational repair w71x. Major genes in the error-prone branch include the REV loci discussed above, whereas more recently identified genes required in error-free
J.C. Game r Mutation Research 451 (2000) 277–293
RAD6-dependent repair include MMS2 w72x, RAD30 w25x, POL30 w73x, and probably RAD27 w19x. Errorprone and error free functions of the RAD6 locus itself can be differentiated by mutants such as rad6 D1-9, which blocks error-free but not errorprone repair Ždiscussed in Ref. w72x.. Some loci such as RAD5 do not fit easily into either sub-branch. rad5 mutants show increased UV sensitivity in combination with reÕ3 w74x, but are allelic with reÕ2 w5x and are at least partially defective in UV-induced mutation w5,75x. Hence RAD5 probably plays at least some role in both error-free and error-prone RAD6dependent repair. However, rad5null mutants do not have the sensitivity of rad6null mutants, hence clearly only a subset of post-replicational repair is affected. rad5 and rad18 single mutants each elevate spontaneous mutation rates, but paradoxically this effect, which is REV3-dependent, is reversed in rad5 rad18 double mutants, which reduce spontaneous mutation to well below wild-type rates w76x. This can be explained in pathway terms if each gene plays an alternative role and can substitute for the other in error-prone repair, while acting in error-free repair in such a way that either gene alone, when mutant, channels more damage via the REV3 pathway w76x. RAD5 is also thought to play a role in recombinational repair of DSBs w77x, probably enhancing recombinational repair at the expense of non-homologous end-joining w78x. Despite this, both rad5null and rad18null mutants increase some types of spontaneous recombination, as does the rad5null rad18 null double mutant, in contrast to the situation for mutagenesis w76x.
10. Subgroups within the RAD50 upwards series Double or multiple mutants within the RAD50 upwards genes seldom if ever show an IR sensitivity greater than that seen in null single mutants of rad51, rad52, or rad54. Hence, all of these genes are in the same major epistasis group. ŽPossible exceptions are RAD56, which has been little studied and may have been inadequately tested, and RAD53, which is required for viability, so that only alleles with partial defects can be tested for epistasis.. Most of the RAD50 upwards genes are required for successful completion of meiosis w4x and have specific
defects in meiotic recombination w79,80x Žsee below.. In contrast, in the other repair gene families, RAD6 and RAD24 are the only major non-essential loci that are required for meiosis w2,9x. Within this major grouping, however, two quite distinct gene subgroups are evident. First, the loci RAD50, RAD58rMRE11, and XRS2 form a separate subgroup based on the phenotypes of their mutants, the functions of their products and the fact that their products interact physically in the cell to form one or more protein complexes. The differences between mutants in the RAD50 versus RAD52 subgroups are most clearly seen in meiosis, especially in the spo13 genetic background, where two diploid spores are formed from a single division rather than four haploid ones from the usual two divisions w81x. In repair-proficient strains, the spo13 division is preceded by a normal round of meiotic recombination, but the division itself is predominantly equational in nature, that is, the centromere and chromosome disjunction usually seen at the first meiotic division is largely absent w81,82x. As a consequence, meiotic recombination, which is usually an essential prerequisite to successful meiotic chromosome segregation, is not essential in a spo13 genetic background w83x. Therefore, some classes of recombination-defective mutants that normally produce dead meiotic spores are able to produce viable diploid spores showing no recombination in spo13 strains. null mutants of rad50, mre11 and xrs2 are of this type w83–85x, as are several mutants not affected in repair, for example rec mutants that function early in meiosis w86,87x. However, mutants in rad51, rad52, rad55, and rad57 are not ‘‘rescued’’ in meiosis by the spo13 genetic background and few if any viable spores are formed in these mutants either in spo13 strains w66,83x or in normal meiosis w4,79x. This is probably because recombination is initiated in these mutants but cannot be fully completed Žsee Refs. w79,80,88x., leaving unrepaired DSBs or other lethal intermediates. Based on other criteria, RAD54 also clearly belongs in the latter subgroup and resembles RAD51 and RAD52, but its mutants confer at most a minimal defect in meiosis w4x and so it cannot be classified by the spo13 criterion. At the molecular level, the meiotic DSBs that are believed to initiate recombination in wild-type occur in the rad52 subgroup mutants w89x, and some recombinant molecules
J.C. Game r Mutation Research 451 (2000) 277–293
are seen, but at lower frequency than in wild-type w80x. In contrast, no meiotic breaks are formed in null mutants in the rad50 subgroup w84,85,90,91x, and recombinant molecules are not seen w80,90x. However, some point mutants in the RAD50 and MRE11 genes confer an intermediate level of IR sensitivity and these mutants do form meiotic breaks. These breaks are not resected and processed into recombinant molecules as they are in wild-type, leading to an accumulation of DSBs to levels above what is seen at any one time in wild-type w31,91–93x and, in contrast to null alleles, an inability to form live gametes even in a spo13 background w33,91,93x. These point mutants are sometimes designated by the letter ‘‘s’’ for ‘‘separation of function’’ in repair and meiosis w31,91,93x. It is not clear that this is appropriate, since they usually also confer significant sensitivity to radiation or MMS Žw31,93x; Game, unpublished observations., although within meiosis the functions of DSB induction and DSB resection do appear to be separate w93x. They are useful for studying meiotic DSBs w91,92x and especially for mapping their positions w94,95x. However, inferences about meiotic DSB frequencies in wild-type based on mutant data may be risky, since normal controls on break induction could be over-ridden in the absence of the subsequent steps of recombination. Mitotically, strains mutant in the RAD50, RAD58rMRE11, or XRS2 genes resemble rad51, rad52, or rad54 mutants in IR sensitivity as haploids, but their homozygous diploids may be somewhat more resistant than the equivalent haploids or than rad51, rad52, or rad54 diploids Žw33,85,96x; Game, unpublished observations.. The two mutant subgroups resemble each other in showing lowered radiation-induced recombination rates compared to wild-type w96,97x Žsee Ref. w98x for review.. However, there is a sharp difference between the two subgroups with respect to spontaneous recombination. This occurs at lower than wild-type frequency in rad51 subgroup mutants, but surprisingly is above wild-type in the rad50 subgroup w32,85,96x Žsee Ref. w98x for review.. The two subgroups discussed above can also be defined by protein interactions inferred genetically and observed biochemically and with the yeast twohybrid system w99–106x. Eight loci are now known to fall into two subsets whose products form separate
protein complexes. The subsets are identical to those defined by the meiotic analysis described above. The RAD50, MRE11, and XRS2 gene products form a single protein complex each of whose subunits are required in DSB repair and in the initiation of meiotic recombination Žreviewed in Ref. w36x.. In human cells, one or more additional proteins is also associated with this complex w38x. The RAD51, 52, 54, 55 and 57 gene products are implicated in one or more protein complexes amongst themselves but are not found in association with the Rad50rMre11rXrs2 proteins. As demonstrated by the yeast two hybrid system w99x and other methods, Rad51p interacts with itself and with Rad52p, Rad54p, and Rad55p; Rad55p interacts with Rad57p w100–106x. However, it is not yet known if these interactions all occur together to form a single complex, or whether several smaller complexes each comprising two or more of the interacting proteins are involved. Three of these genes, RAD51, RAD55, and RAD57 are also related to each other by DNA sequence homology. RAD51 is regarded as the yeast counterpart to the E. coli RECA gene w100,107,108x, and is highly conserved in eukaryotes, whereas RAD55 and RAD57 share weaker homology with each other and with RAD51 w109,110x. Double mutants involving members of the two subgroups reveal that, as expected from phenotypic data, in meiosis, the RAD50 subgroup functions before the RAD51 subgroup in a dependent sequence. rad50null mutants will restore meiotic viability to rad52 mutants in spo13 background w83x and the same is true for several other tested combinations between the groups Žw33,85x; J. Game and R. E. Esposito, unpublished data; see Ref. w66x.. Mitotically, however, diploid double mutants resemble the more sensitive rad51 subgroup single mutants rather than the slightly less sensitive rad50 subgroup mutants, at least when classical point mutants are studied Žrigorous data for null alleles are still not available for some combinations. ŽGame and Mortimer, unpublished data.. The mitotic data confirm epistasis but define the rad51 subgroup genes as epistatic to the hypostatic rad50 subgroup, in reverse of the situation in spo13-mediated meiosis. This difference is not surprising, since survival of the double mutants in meiosis is determined by lack of initiation of DSBs, whereas survival of irradiated mitotic cells is
J.C. Game r Mutation Research 451 (2000) 277–293
determined by their ability to repair breaks already induced by IR. The two subgroups may also differ in their double mutant interactions with hdf1 and yku80 Žsee below..
11. RAD55 and RAD57 These two loci stand out within the RAD51 subgroup in each having a cold-dependent mutant phenotype that is much more pronounced at 238C than at 378C. In addition, the phenotype can be substantially reversed even at low temperatures by high osmotic strength in the growth medium w111x. The temperature-dependence reflects an increased requirement for the gene products themselves at low temperatures rather than cold-sensitive proteins in the mutants, since null alleles as well as classical point mutants show the same cold-enhanced IR sensitivity w111x. This is unique amongst Saccharomyces repair genes, and stands in contrast to repair genes listed above that are required more stringently at higher temperatures, where null mutants confer inviability, than at lower temperatures where only repair is affected. Possibly, Rad55p and Rad57p act to stabilize recombination complexes in less than ideal conditions. This is perhaps supported by observations that over-expressing Rad51p or Rad52p, and especially Rad51p and Rad52p together, will suppress the phenotype of rad55null or rad57null mutants even at 238C w105x. Recently, a RAD51-family gene has been identified in Sch. pombe whose deletion mutant confers IR sensitivity that is enhanced at low temperatures w112x. Possibly, this is an analogue of Sac. cereÕisiae RAD55 w112x. However, RAD55 and RAD57 differ from most of the other RAD50 upwards genes in thus far having no true mammalian homologues identified by sequence homology, although they belong to the hRAD51 family of human sequences w40,113x. Their function may be provided in mammals by more distantly related genes in the same family. Alternatively, there could be a reduced need for such gene functions in the warmer, more stable environment of mammalian cells. It would be interesting to search for these sequences in the genomes of higher eukaryotes lacking temperature homeostasis, such as plants.
12. Epistasis analysis and X-ray sensitivity There is much current interest in yeast genes controlling sensitivity to IR. Characterization of IR repair has in many ways lagged behind that of UV repair, and mammalian cell researchers have only recently demonstrated the relevance of the RAD50 upwards yeast genes to mammalian repair. Early work on epistasis relationships in Sac. cereÕisiae was based on UV repair, and the success of this approach makes it important also to define epistasis groups based on IR sensitivity. There is no a priori reason to assume that epistasis relationships based on the two phenotypes will exactly coincide, since the primary damage is different and might be channeled in different ways through cellular repair pathways. In addition, IR repair studies are complicated by the differing radiation biology of haploid and diploid cells, and especially the fact that recombinational repair cannot operate in G1 haploids. Sister-chromatid repair in G2 haploids may also differ from inter-chromosomal repair in G1 diploids. However, several earlier reports have indicated that the
Fig. 1. X-ray survival of inbred haploid yeast strains carrying single, double or triple mutations in the RAD1, RAD18 and RAD51 genes.
J.C. Game r Mutation Research 451 (2000) 277–293
RAD6rRAD18 group and the RAD50 upwards genes do form separate IR epistasis groups resembling those seen for UV-sensitivity w114,115x. Data presented here in Figs. 1 and 2 confirm this in two sets of inbred strains, and more importantly, they indicate that the excision repair loci play a significant role in X-ray repair that is independent of the other two groups. This role is exposed in triple mutants where the other two pathways are also blocked. Fig. 1 ŽJ. Game and D. Schild, unpublished data. presents haploid X-ray survival data for some single, double and triple mutant strains that are mutant for rad1-1 ŽUV-excision repair., rad18-2 Žpost-replicational UV repair., and rad51-1 Žrecombinational repair.. It can be seen that the two triple mutant strains show good agreement with each other and are significantly more sensitive than the two rad18-2 rad51-1 double mutants, which are in turn more sensitive than either single mutant. Hence, the three loci probably represent three epistasis groups for X-ray sensitivity, just as they do for UV sensitivity. This result is supported by survival curves in Fig. 2 ŽJ. Game and D. Schild, unpublished data.. Here, mutant alleles of three loci different from those in Fig. 1 were used to block the same three UV-repair
pathways. It can be seen that the same additive or synergistic relationships obtain as are seen in Fig. 1. The excision-blocking mutation rad3-2 adds significant extra X-ray sensitivity in a triple mutant combination with rad6null Žpost-replicational repair. and rad54null Žrecombinational repair.. Three triple mutant spore-clones from a single inbred diploid show excellent agreement and are marginally more sensitive than the triples in Fig. 1. This probably reflects the extra sensitivity of the rad6null mutant compared to rad18-2. While Figs. 1 and 2 demonstrate additivity or synergism for X-ray sensitivity between members of the three main UV-epistasis groups, it remains to be formally demonstrated that there is epistasis between different mutants within the excision group. It is unlikely but possible that blocking more than one excision gene in a rad6null rad54null background would add even more sensitivity than seen in the triple mutants shown. Figs. 1 and 2 seem to delineate three X-ray epistasis groups congruent with those for UV, but epistasis relationships for IR sensitivity are also currently of interest in light of findings with the Saccharomyces HDF1 and YKU80 genes w42,43x. An apparent paradox exists concerning the role of the RAD50, MRE11, and XRS2 genes, which fall into the rad52 epistasis group but represent a distinct subset Žsee above.. Unlike the rad52 subset, they appear to be required for end-joining DSB repair w116,117x as well as playing a major role in recombinational repair Žsee Ref. w118x.. However, no consistent increase in sensitivity has been observed when mutants in the rad50 subset are combined with members of the rad52 subset Žw4x; Game, unpublished observations.. One would expect these doubles to have the same increased sensitivity as double mutants involving rad52 and hdf1 or yku80, which also involve blocks in both HRR and end-joining and have been reported to show increased sensitivity w42x. Further investigation here would be useful. 13. The IR repair gene groups may act on different types of initial damage
Fig. 2. X-ray survival of inbred haploid yeast strains carrying single, double or triple mutations in the RAD3, RAD6 and RAD54 genes.
While there is congruence for the three major epistasis groups for both UV and IR sensitivity, the degree of synergism seen when two or more groups
J.C. Game r Mutation Research 451 (2000) 277–293
are blocked is less for IR than for UV sensitivity. This may reflect the more heterogeneous nature of the initial damage after IR, which would favor additive rather than synergistic interactions. While there has been much focus on DSBs as a key lethal lesion after IR, in fact other types of damage are induced with much greater frequency. The observation that one or two unrepaired DSBs are lethal in repair-defective mutants w119–121x is often taken as evidence that DSBs are the critical lethal lesions at higher doses in wild-type cells. This assumes that DSB repair becomes saturated at the doses that kill wild-type cells, or at least that some DSBs provoke misrepair or lead to other lesions that are lethal. An alternative hypothesis is that non-DSB lesions are also substantially responsible for killing in wild-type, and that DSBs only become the predominant lethal damage when their normally efficient repair is blocked by mutation. On this hypothesis, wild-type cells would still be able to repair many DSBs even at doses where substantial killing occurs due to other lesions. The strong radiation sensitivity of some mutants in the RAD6rRAD18 group supports this, since these mutants probably remain proficient at DSB repair w122x. A priori, there is no reason to suppose that wild-type cells are killed by saturation of the RAD52 DSB repair pathway rather than the RAD6rRAD18 pathway that operates on other damage. A further complication is that in wild-type, recombinational repair can probably act both on DSB and non-DSB damage, but recombinational repair of non-DSB damage may still involve a DSB intermediate. This raises the question of whether induced DSBs are essentially the only lethal lesion even in rad52 single mutants, since that fraction of non-DSB damage that provokes recombination in wild-type cells is likely to provoke lethality in rad52 mutants. It may be more logical to regard rad52group mutants as defective in recombination, however provoked, rather than simply defective in DSB repair. In this case, DSBs would still be the main lethal lesion induced by IR in rad52-group mutants, since non-recombinational repair of such DSBs is inefficient in yeast, whereas non-recombinational repair of other IR damage would still occur efficiently via the intact RAD6rRAD18 pathway. The nature of the initial lesions, including the type of DSBs induced, probably also substantially
determines the role of the YKU80rHDF1 genes versus the RAD loci in their repair. Lewis et al. w123x found that EcoRI endonuclease expressed in yeast in vivo produced much greater lethality in hdf1 mutants than in wild type. In contrast, rad52 mutants were not killed by these DSBs, which have complementary ends, although they differed from wild-type in being growth-arrested in the G2 phase while EcoRI was expressed. In wild-type, the EcoRI breaks could be repaired in G1 haploid cells, in contrast to IR-induced DSBs, providing further evidence that the Ku complex mediates an end-joining response that does not require homologous DNA. The repair was accurate, differentiating it from single-strand annealing repair leading to deletions or insertions. However, interchromosomal recombination was stimulated in wild-type yeast by EcoRI breaks, implying that when both pathways are intact, at least some breaks are repaired by recombination. In rad9 mutants, G2 checkpoint arrest did not occur and cells lost viability rapidly when EcoRI was expressed. Hence RAD9 appears to be needed for checkpoint arrest for DSBs of either type, but the YKU genes and the RAD52 pathway appear to discriminate in yeast based on whether the breaks have complementary ends w123x. This is consistent with earlier work by Boulton and Jackson w42x, who found that Ku-dependent end-joining of an EcoRI cut site in a plasmid was error-free, but in the absence of Ku function, end-joining repair of the same site could still occur by other mechanisms, but involved deletions from 1 to ; 800 bases.
14. Repair pathways in yeast and humans In summarizing the repair field in Saccharomyces, perhaps the most remarkable things to stand out are the complexity of the processes, the degree of homology throughout eukaryotes, and the importance of repair in maintaining life. Complexity is evident not only from the large number of gene products that are required, but from the frequent need for these products to associate into multi-enzyme complexes and the many interactions now known to occur between pathways. These factors, in addition to the complexity of the DNA damage substrates themselves and their repair intermediates, were significant
J.C. Game r Mutation Research 451 (2000) 277–293
obstacles in the way of earlier understanding of molecular reactions in repair. Great progress is currently being made in all areas of repair, especially with mammalian cells, and yeast continues to be an excellent paradigm for repair in humans. Almost all known repair pathways and most of the individual loci are significantly homologous from Saccharomyces to mammals. Amongst the few differences, perhaps the most striking is the apparently greater importance of recombinational repair compared to the Ku-mediated end-joining pathway in processing IR-induced DSBs in yeast compared to mammals. Surprisingly, this apparent difference does not seem to affect the overall efficiency of IR repair in the two systems, since there is a good correspondence between yeast and mammalian cells in survival per unit dose per unit of DNA, i.e., when the much smaller DNA target in yeast is taken into account w124,125x. A possible reason for the greater emphasis on recombinational repair in Saccharomyces may be that yeast has little non-coding DNA, hence almost all effective DSB repair must be accurate, whereas mammalian cells may be better able to tolerate short deletions or additions in noncoding DNA arising from end-joining mechanisms that are error-prone for breaks with complex ends. Another factor may be the different life forms involved. Yeast cells must be capable of ongoing division, hence repair must be accurate, and sisterchromatids as well as homologous chromosomes are available for recombinational repair in G2. Less accurate repair may be tolerated in non-dividing tissues of complex eukaryotes, and recombinational repair between large non-replicating chromosome homologues may be harder to effect. It would be interesting to study DSB repair in complex haploid eukaryotes with long-lived non-dividing tissue, such as bryophytes and fern gametophytes. In these tissues, recombinational repair would seem difficult, yet error-prone end-joining could be unsuitable for haploid cells. Finally, the availability of so many yeast mutants in different groups and the ease with which mutations can be combined permits a demonstration of the extreme necessity for some kind of DNA damage tolerance in cells. In an attempt to demonstrate the consequences of having no UV repair in yeast, Game and Mortimer Žunpublished data. tested the sensitiv-
ity to sunlight of a haploid quadruple mutant blocked in all three major UV epistasis groups as well as being defective in photoreactivation through a mutation in the phr1 gene. The sensitivity of this strain was even more extreme than the authors expected and it is shown in Fig. 3. It can be seen that survival versus exposure to early afternoon July sunlight in Berkeley, CA shows an exponential relationship with a D37 of less than 3 s, i.e. this short exposure produces an average of one lethal event per cell. In contrast, the survival of the wild-type strain remains close to 100% after 30 min of exposure to the same sunlight. The experiment was repeated at 8000 ft elevation in the Californian Sierra Nevada, where the
Fig. 3. Sensitivity of a quadruple mutant repair-deficient Saccharomyces strain to natural sunlight at latitude 378N in July 1986. Percent survival is shown against seconds of exposure Žlower scale. for the quadruple mutant, and against minutes of exposure Župper scale. for the wild-type control.
J.C. Game r Mutation Research 451 (2000) 277–293
D37 was reduced to about 2 s of exposure ŽFig. 3.. The quadruple mutant in California shows very much greater sensitivity than a rad1 phr1 double mutant exposed to English sunlight in July by Resnick w126x, which had an average D37 of about 4 min. Clearly, these experiments do not address mechanisms and embrace several uncontrolled variables including meteorological conditions. However, such multiple mutant strains may be useful for testing the genotoxicity of artificial lighting as well as sunlight at different times and conditions. They demonstrate the potential toxicity of sunlight and the high vulnerability of DNA to damage. One can view this damage as lethal in the absence of repair. Alternatively, one can argue that the damage may provoke lethality more incidentally through activating cellular mechanisms such as recombination that themselves lead to lethal intermediates when blocked by mutations. Either way, it is clear that the cell’s elaborate methods for coping with DNA damage are essential for life.
Acknowledgements While writing this chapter, John Game was supported in part by US Department of Energy funds administered through the Lawrence Berkeley National Laboratory.
References w1x S. Nakai, S. Matsumoto, Two types of radiation-sensitive mutants in yeast, Mutation Research 4 Ž1967. 129–136. w2x B.S. Cox, J. Parry, The isolation, genetics, and survival characteristics of ultraviolet light-sensitive mutants in yeast, Mutation Research 6 Ž1968. 37–55. w3x J.C. Game, B.S. Cox, Allelism tests of mutants affecting sensitivity to radiation in yeast and a proposed nomenclature, Mutation Research 20 Ž1971. 35–44. w4x J.C. Game, R.K. Mortimer, A genetic study of X-ray sensitive mutants in yeast, Mutation Research 24 Ž1974. 281– 292. w5x J.F. Lemontt, Mutants of yeast defective in mutation induced by ultraviolet light, Genetics 68 Ž1971. 21–33. w6x G.L. McKnight, T.S. Cardillo, F. Sherman, An extensive deletion causing overproduction of yeast iso-2-cytochrome c, Cell 25 Ž1981. 409–419. w7x L. Prakash, S. Prakash, Excision repair genes of Saccharomyces cereÕisiae, Annali dell Istituto Superiore di Sanita 25 Ž1989. 99–113.
w8x G. Perozzi, S. Prakash, RAD7 gene of Saccharomyces cereÕisiae: transcripts, nucleotide sequence analysis, and functional relationship between the RAD7 and RAD23 gene products, Molecular and Cellular Biology 6 Ž1986. 1497– 1507. w9x F. Eckardt-Schupp, W. Siede, J.C. Game, The RAD24 Ž Rsr1. gene product of Saccharomyces cereÕisiae participates in two different pathways of DNA repair, Genetics 115 Ž1987. 83–90. w10x D. Averbeck, Isolierung und charakterisierung von drei UV-sensiblen Saccharomyces-mutanten, PhD Thesis, Free University Berlin, 1970. w11x D. Averbeck, W. Laskowski, F. Eckardt, E. LehmannBrauns, Four radiation sensitive mutants of Saccharomyces, survival after UV- and X-ray-irradiation as well as UV-induced reversion rates from isoleucine-valine dependence to independence, Molecular & General Genetics 107 Ž1970. 117–127. w12x T.A. Weinert, Dual cell cycle checkpoints sensitive to chromosome replication and DNA damage in the budding yeast Saccharomyces cereÕisiae, Radiation Research 132 Ž1992. 141–143. w13x S. Bao, M.S. Chang, D. Auclair, Y. Sun, Y. Wang, W.K. Wong, J. Zhang, Y. Liu, X. Qian, R. Sutherland et al., hRAD17, a human homologue of the Schizosaccharomyces pombe checkpoint gene rad17, is overexpressed in colon carcinoma, Cancer Research 59 Ž1999. 2023–2028. w14x F. von Deimling, J.M. Scharf, T. Liehr, M. Rothe, A.R. Kelter, P. Albers, W.F. Dietrich, L.M. Kunkel, N. Wernert, B. Wirth, Human and mouse RAD17 genes: identification, localization, genomic structure and histological expression pattern in normal testis and seminoma, Human Genetics 105 Ž1999. 17–27. w15x E. Park, S.N. Guzder, M.H. Koken, I. Jaspers-Dekker, G. Weeda, J.H. Hoeijmakers, S. Prakash, L. Prakash, RAD25 Ž SSL2 ., the yeast homolog of the human xeroderma pigmentosum group B DNA repair gene, is essential for viability, Proceedings of the National Academy of Sciences of the United States of America 89 Ž1992. 11416–11420. w16x K.D. Gulyas, T.F. Donahue, SSL2, a suppressor of a stemloop mutation in the HIS4 leader encodes the yeast homolog of human ERCC-3, Cell 69 Ž1992. 1031–1042. w17x B.S. Lee, C.P. Lichtenstein, B. Faiola, L.A. Rinckel, W. Wysock, M.J. Curcio, D.J. Garfinkel, Posttranslational inhibition of Ty1 retrotransposition by nucleotide excision repairrtranscription factor TFIIH subunits Ssl2p and Rad3p, Genetics 148 Ž1998. 1743–1761. w18x A.J. van Gool, R. Verhage, S.M. Swagemakers, P. van de Putte, J. Brouwer, C. Troelstra, D. Bootsma, J.H. Hoeijmakers, RAD26, the functional S. cereÕisiae homolog of the Cockayne syndrome B gene ERCC6, EMBO Journal 13 Ž1994. 5361–5369. w19x M.S. Reagan, C. Pittenger, W. Siede, E.C. Friedberg, Characterization of a mutant strain of Saccharomyces cereÕisiae with a deletion of the RAD27 gene, a structural homolog of the RAD2 nucleotide excision repair gene, Journal of Bacteriology 177 Ž1995. 364–371.
J.C. Game r Mutation Research 451 (2000) 277–293
w20x A. Jacquier, P. Legrain, B. Dujon, Sequence of a 10.7 kb segment of yeast Chromosome XI identifies the APN1 and BAF1 loci and reveals one tRNA gene and several new ORFs including homologs to RAD2 and kinases, Yeast 8 Ž1992. 121–132. w21x A. Klungland, T. Lindahl, Second pathway for completion of human DNA base excision-repair: reconstitution with purified proteins and requirement for DNase IV ŽFEN1., EMBO Journal 16 Ž1997. 3341–3348. w22x X. Wu, Z. Wang, Relationships between yeast Rad27 and Apn1 in response to apurinicrapyrimidinic ŽAP. sites in DNA, Nucleic Acids Research 27 Ž1999. 956–962. w23x P.K. Bhatia, R.A. Verhage, J. Brouwer, E.C. Friedberg, Molecular cloning and characterization of Saccharomyces cereÕisiae RAD28, the yeast homolog of the human Cockayne syndrome A ŽCSA. gene, Journal of Bacteriology 178 Ž1996. 5977–5988. w24x S.A. Kozhin, T.N. Kozhina, V.G. Korolev, The RAD29 gene: map position on the right arm of chromosome II, Yeast 13 Ž1997. 489–490. w25x J.P. McDonald, A.S. Levine, R. Woodgate, The Saccharomyces cereÕisiae RAD30 gene, a homologue of Escherichia coli dinB and umuC, is DNA damage inducible and functions in a novel error-free postreplication repair mechanism, Genetics 147 Ž1997. 1557–1568. w26x A.A. Roush, M. Suarez, E.C. Friedberg, M. Radman, W. Siede, Deletion of the Saccharomyces cereÕisiae gene RAD30 encoding an Escherichia coli DinB homolog confers UV radiation sensitivity and altered mutability, Molecular & General Genetics 257 Ž1998. 686–692. w27x R.E. Johnson, S. Prakash, L. Prakash, Efficient bypass of a thymine–thymine dimer by yeast DNA polymerase, Pol eta, Science 283 Ž1999. 1001–1004. w28x R.E. Johnson, C.M. Kondratick, S. Prakash, L. Prakash, hRAD30 mutations in the variant form of xeroderma pigmentosum, Science 285 Ž1999. 263–265. w29x C. Masutani, R. Kusumoto, A. Yamada, N. Dohmae, M. Yokoi, M. Yuasa, M. Araki, S. Iwai, K. Takio, F. Hanaoka, The XPV Žxeroderma pigmentosum variant. gene encodes human DNA polymerase eta, Nature 399 Ž1999. 700–704. w30x N.G. Suslova, I.A. Zakharov, Genetic control of radiosensitivity in yeast: VII. Identification of the genes determining the sensitivity to X-rays, Genetika 6 Ž1970. 571–575. w31x H. Tsubouchi, H. Ogawa, A novel mre11 mutation impairs processing of double-strand breaks of DNA during both mitosis and meiosis, Molecular and Cellular Biology 18 Ž1998. 260–268. w32x M. Ajimura, S.H. Leem, H. Ogawa, Identification of new genes required for meiotic recombination in Saccharomyces cereÕisiae, Genetics 133 Ž1993. 51–66. w33x O.V. Chepurnaya, S.A. Kozhin, V.T. Peshekhonov, V.G. Korolev, RAD58 (XRS4) — a new gene in the RAD52 epistasis group, Current Genetics 28 Ž1995. 274–279. w34x S.A. Kozhin, O.V. Chepurnaya, V.G. Korolev, The RAD58 (XRS4) gene: map position on the right arm of chromosome XIII, Yeast 11 Ž1995. 1211–1213. w35x Y. Bai, L.S. Symington, A RAD52 homolog is required for
RAD51-independent mitotic recombination in Saccharomyces cereÕisiae, Genes and Development 10 Ž1996. 2025–2037. H. Ogawa, K. Johzuka, T. Nakagawa, S.H. Leem, A.H. Hagihara, Functions of the yeast meiotic recombination genes, MRE11 and MRE2, Advances in Biophysics 31 Ž1995. 67–76. T. Usui, T. Ohta, H. Oshiumi, J. Tomizawa, H. Ogawa, T. Ogawa, Complex formation and functional versatility of Mre11 of budding yeast in recombination, Cell 95 Ž1998. 705–716. J.P. Carney, R.S. Maser, H. Olivares, E.M. Davis, M. Le Beau, J.R. Yates III, L. Hays, W.F. Morgan, J.H. Petrini, The hMre11rhRad50 protein complex and Nijmegen breakage syndrome: linkage of double-strand break repair to the cellular DNA damage response, Cell 93 Ž1998. 477–486. M. Digweed, A. Reis, K. Sperling, Nijmegen breakage syndrome: consequences of defective DNA double strand break repair, BioEssays 21 Ž1999. 649–656. L.H. Thompson, D. Schild, The contribution of homologous recombination in preserving genome integrity in mammalian cells, Biochimie 81 Ž1999. 87–105. H. Feldmann, E.L. Winnacker, A putative homologue of the human autoantigen Ku from Saccharomyces cereÕisiae, Journal of Biological Chemistry 268 Ž1993. 12895–12900. S.J. Boulton, S.P. Jackson, Identification of a Saccharomyces cereÕisiae Ku80 homologue: roles in DNA double strand break rejoining and in telomeric maintenance, Nucleic Acids Research 24 Ž1996. 4639–4648. G.T. Milne, S. Jin, K.B. Shannon, D.T. Weaver, Mutations in two Ku homologs define a DNA end-joining repair pathway in Saccharomyces cereÕisiae, Molecular and Cellular Biology 16 Ž1996. 4189–4198. C. Featherstone, S.P. Jackson, Ku, a DNA repair protein with multiple cellular functions?, Mutation Research 434 Ž1999. 3–15. L. Prakash, S. Prakash, Isolation and characterization of MMS-sensitive mutants of Saccharomyces cereÕisiae, Genetics 86 Ž1977. 33–55. P.J. Hastings, S.K. Quah, R.C. von Borstel, Spontaneous mutation by mutagenic repair of spontaneous lesions in DNA, Nature 264 Ž1976. 719–722. C.W. Lawrence, D.C. Hinkle, DNA polymerase zeta and the control of DNA damage induced mutagenesis in eukaryotes, Cancer Surveys 28 Ž1996. 21–31. P.E. Gibbs, W.G. McGregor, V.M. Maher, P. Nisson, C.W. Lawrence, A human homolog of the Saccharomyces cereÕisiae REV3 gene, which encodes the catalytic subunit of DNA polymerase zeta, Proceedings of the National Academy of Sciences of the United States of America 95 Ž1998. 6876–6880. C.W. Lawrence, B.R. Krauss, R.B. Christensen, New mutations affecting induced mutagenesis in yeast, Mutation Research 150 Ž1985. 211–216. C.W. Lawrence, G. Das, R.B. Christensen, REV7, a new gene concerned with UV mutagenesis in yeast, Molecular & General Genetics 200 Ž1985. 80–85.
J.C. Game r Mutation Research 451 (2000) 277–293 w51x P.E. Nisson, C.W. Lawrence, The isolation and characterization of ngm2, a mutation that affects nitrosoguanidine mutagenesis in yeast, Molecular & General Genetics 204 Ž1986. 90–97. w52x M. Lombaerts, M. Tijsterman, R.A. Verhage, J. Brouwer, Saccharomyces cereÕisiae mms19 mutants are deficient in transcription-coupled and global nucleotide excision repair, Nucleic Acids Research 25 Ž1997. 3974–3979. w53x W. Xiao, B.L. Chow, L. Rathgeber, The repair of DNA methylation damage in Saccharomyces cereÕisiae, Current Genetic 30 Ž1996. 461–468. w54x S.C. Popoff, A.I. Spira, A.W. Johnson, B. Demple, Yeast structural gene ŽAPN1. for the major apurinic endonuclease: homology to Escherichia coli endonuclease IV, Proceedings of the National Academy of Sciences of the United States of America 87 Ž1990. 4193–4197. w55x D. Ramotar, S.C. Popoff, E.B. Gralla, B. Demple, Cellular X role of yeast Apn1 apurinic endonucleaser3 -diesterase: repair of oxidative and alkylation DNA damage and control of spontaneous mutation, Molecular and Cellular Biology 11 Ž1991. 4537–4544. w56x R.L. Swanson, N.J. Morey, P.W. Doetsch, S. Jinks-Robertson, Overlapping specificities of base excision repair, nucleotide excision repair, recombination, and translesion synthesis pathways for DNA base damage in Saccharomyces cereÕisiae, Molecular and Cellular Biology 19 Ž1999. 2929–2935. w57x M. Saparbaev, L. Prakash, S. Prakash, Requirement of mismatch repair genes MSH2 and MSH3 in the RAD1– RAD10 pathway of mitotic recombination in Saccharomyces cereÕisiae, Genetics 142 Ž1996. 727–736. w58x N. Sugawara, F. Paques, M. Colaiacovo, J.E. Haber, Role of Saccharomyces cereÕisiae Msh2 and Msh3 repair proteins in double-strand break-induced recombination, Proceedings of the National Academy of Sciences of the United States of Amercia 94 Ž1997. 9214–9219. w59x P. Peltomaki, DNA mismatch repair gene mutations in human cancer, Environmental Health Perspectives 105 ŽSuppl. 4. Ž1997. 775–780. w60x T.A. Prolla, DNA mismatch repair and cancer, Current Opinion in Cell Biology 10 Ž1998. 311–316. w61x R.D. Kolodner, G.T. Marsischky, Eukaryotic DNA mismatch repair, Current Opinion in Genetics and Development 9 Ž1999. 89–96. w62x M.S. Williamson, Mutational analysis of gene conversion in Saccharomyces cereÕisiae, PhD Thesis, University of California Berkeley, 1984. w63x H. Baranowska, Z. Swietlinska, D. Zaborowska, J. Zuk, cdc and prt mutants of Saccharomyces cereÕisiae with increased sensitivity to diepoxybutane and ultraviolet, Acta Microbiologica Polonica 31 Ž1982. 119–128. w64x L.H. Johnston, The DNA repair capability of CDC9, the Saccharomyces cereÕisiae mutant defective in DNA ligase, Molecular & General Genetics 170 Ž1979. 89–92. w65x B.S. Cox, J.C. Game, Repair systems in Saccharomyces, Mutation Research 26 Ž1974. 257–264.
w66x J.C. Game, Radiation-sensitive mutants and repair in yeast, in: J.F.T. Spencer, D. Spencer, A.R.W. Smith ŽEds.., Yeast Genetics, Fundamental and Applied Aspects, Springer Verlag, New York, 1983, pp. 109–137. w67x L.H. Johnston, K.A. Nasmyth, Saccharomyces cereÕisiae cell cycle mutant cdc9 is defective in DNA ligase, Nature 274 Ž1978. 891–893. w68x C.W. Moore, Ligase-deficient yeast cells exhibit defective DNA rejoining and enhanced gamma ray sensitivity, Journal of Bacteriology 150 Ž1982. 1227–1233. w69x D.R. Wilcox, L. Prakash, Incision and postincision steps of pyrimidine dimer removal in excision-defective mutants of Saccharomyces cereÕisiae, Journal of Bacteriology 148 Ž1981. 618–623. w70x R.A. Verhage, A.J. van Gool, N. de Groot, J.H. Hoeijmakers, P. van de Putte, J. Brouwer, Double mutants of Saccharomyces cereÕisiae with alterations in global genome and transcription-coupled repair, Molecular and Cellular Biology 16 Ž1996. 496–502. w71x R.H. McKee, C.W. Lawrence, Genetic analysis of gammaray mutagenesis in yeast: I. Reversion in radiation-sensitive strains, Genetics 93 Ž1979. 361–373. w72x S. Broomfield, B.L. Chow, W. Xiao, MMS2, encoding a ubiquitin-conjugating-enzyme-like protein, is a member of the yeast error-free postreplication repair pathway, Proceedings of the National Academy of Sciences of the United States of America 95 Ž1998. 5678–5683. w73x C.A. Torres-Ramos, B.L. Yoder, P.M. Burgers, S. Prakash, L. Prakash, Requirement of proliferating cell nuclear antigen in RAD6-dependent postreplicational DNA repair, Proceedings of the National Academy of Sciences of the United States of America 93 Ž1996. 9676–9681. w74x R.E. Johnson, S.T. Henderson, T.D. Petes, S. Prakash, M. Bankmann, L. Prakash, Saccharomyces cereÕisiae RAD5encoded DNA repair protein contains DNA helicase and zinc-binding sequence motifs and affects the stability of simple repetitive sequences in the genome, Molecular and Cellular Biology 12 Ž1992. 3807–3818. w75x C.W. Lawrence, R.B. Christensen, Ultraviolet-induced reversion of cyc1 alleles in radiation sensitive strains of yeast: II. reÕ2 mutant strains, Genetics 90 Ž1978. 213–226. w76x B. Liefshitz, R. Steinlauf, A. Friedl, F. Eckardt-Schupp, M. Kupiec, Genetic interactions between mutants of the ‘‘error-prone’’ repair group of Saccharomyces cereÕisiae and their effect on recombination and mutagenesis, Mutation Research 407 Ž1998. 135–145. w77x L.K. Lewis, J.W. Westmoreland, M.A. Resnick, Repair of endonuclease-induced double-strand breaks in Saccharomyces cereÕisiae: essential role for genes associated with nonhomologous end-joining, Genetics 152 Ž1999. 1513– 1529. w78x F. Ahne, B. Jha, F. Eckardt-Schupp, The RAD5 gene product is involved in the avoidance of non-homologous end-joining of DNA double strand breaks in the yeast Saccharomyces cereÕisiae, Nucleic Acids Research 25 Ž1997. 743–749.
J.C. Game r Mutation Research 451 (2000) 277–293
w79x J.C. Game, T.J. Zamb, R.J. Braun, M.A. Resnick, R.M. Roth, The role of radiation Ž rad . genes in meiotic recombination in yeast, Genetics 94 Ž1980. 51–68. w80x R.H. Borts, M. Lichten, J.E. Haber, Analysis of meiosis-defective mutations in yeast by physical monitoring of recombination, Genetics 113 Ž1986. 551–567. w81x S. Klapholz, R.E. Esposito, Isolation of spo12-1 and spo131 from a natural variant of yeast that undergoes a single meiotic division, Genetics 96 Ž1978. 567–588. w82x S. Klapholz, R.E. Esposito, Recombination and chromosome segregation during the single division meiosis in spo12-1 and spo13-1 diploids, Genetics 96 Ž1978. 589–611. w83x R. Malone, R.E. Esposito, Recombinationless meiosis in Saccharomyces cereÕisiae, Molecular and Cellular Biology 1 Ž1981. 891–901. w84x K. Johzuka, H. Ogawa, Interaction of Mre11 and Rad50: two proteins required for DNA repair and meiosis-specific double-strand break formation in Saccharomyces cereÕisiae, Genetics 139 Ž1995. 1521–1532. w85x E.L. Ivanov, V.G. Korolev, F. Fabre, XRS2, a DNA repair gene of Saccharomyces cereÕisiae, is needed for meiotic recombination, Genetics 132 Ž1992. 651–664. w86x R.E. Malone, S. Bullard, M. Hermiston, R. Rieger, M. Cool, A. Galbraith, Isolation of mutants defective in early steps of meiotic recombination in the yeast Saccharomyces cereÕisiae, Genetics 128 Ž1991. 79–88. w87x J.M. Gardiner, S.A. Bullard, C. Chrome, R.E. Malone, Molecular and genetic analysis of REC103, an early meiotic recombination gene in yeast, Genetics 146 Ž1997. 1265–1274. w88x R.E. Malone, Multiple mutant analysis of recombination in yeast, Molecular & General Genetics 189 Ž1983. 405–412. w89x A. Høgset, T.B. Øyen, Correlation between supressed meiotic recombination and the lack of DNA strand-breaks in the rDNA genes of Saccharomyces cereÕisiae, Nucleic Acids Research 12 Ž1984. 7199–7212. w90x J.C. Game, K.C. Sitney, V.E. Cook, R.K. Mortimer, Use of a ring chromosome and pulsed-field gels to study interhomolog recombination, double-strand DNA breaks and sister-chromatid exchange in yeast, Genetics 123 Ž1989. 695–713. w91x L. Cao, E. Alani, N. Kleckner, A pathway for generation and processing of double-strand breaks during meiotic recombination in S. cereÕisiae, Cell 61 Ž1990. 1089–1101. w92x H. Sun, D. Treco, J.W. Szostak, Extensive 3X-overhanging single-stranded DNA associated with the meiosis-specific double-strand breaks at the ARG4 recombination initiation site, Cell 64 Ž1991. 1155–1161. w93x K. Nairz, F. Klein, mre11S-a yeast mutation that blocks double-strand-break processing and permits nonhomologous synapsis in meiosis, Genes and Development 11 Ž1997. 2272–2290. w94x J.C. Game, Pulsed-field gel analysis of the pattern of DNA double-strand breaks in the Saccharomyces genome during meiosis, Developmental Genetics 13 Ž1992. 485–497. w95x D. Zenvirth, T. Arbel, A. Sherman, M. Goldway, S. Klein,
G. Simchen, Multiple sites for double-strand breaks in whole meiotic chromosomes of Saccharomyces cereÕisiae, EMBO Journal 11 Ž1992. 3441–3447. I.A. Zakharov, G.V. Kasinova, S.V. Koval’tsova, Intragenic mitotic recombination induced by ultraviolet and gamma rays in radiosensitive mutants of Saccharomyces cereÕisiae yeasts, Genetika 19 Ž1983. 49–57, ŽRussian Language.. T. Saeki, I. Machida, S. Nakai, Genetic control of diploid recovery after g-irradiation in the yeast Saccharomyces cereÕisisiae, Mutation Research 73 Ž1981. 251–265. T.D. Petes, R.E. Malone, L.S. Symington, Recombination in yeast, in: J.R. Broach, E.W. Jones, J.R. Pringle ŽEds.., The Molecular and Cellular Biology of the Yeast Saccharomyces: Genome Dynamics, Protein Synthesis, and Energetics, Cold Spring Harbor Laboratory Press, New York, 1991, pp. 407–521. C.T. Chien, P. Bartel, R. Sternglanz, S. Fields, The two-hybrid system: a method to identify and clone genes for proteins that interact with a protein of interest, Proceedings of the National Academy of Sciences of the United States of America 88 Ž1991. 9578–9582. A. Shinohara, H. Ogawa, T. Ogawa, Rad51 protein involved in repair and recombination in S. cereÕisiae is a RecA-like protein, Cell 69 Ž1992. 457–470. G.T. Milne, D.T. Weaver, Dominant negative alleles of RAD52 reveal a DNA repairrrecombination complex including Rad51 and Rad52, Genes and Development 7 Ž1993. 1755–1765. H. Jiang, Y. Xie, P. Houston, K. Stemke-Hale, U.H. Mortensen, R. Rothstein, T. Kodadek, Direct association between the yeast Rad51 and Rad54 recombination proteins, Journal of Biological Chemistry 271 Ž1996. 33181– 33186. P. Sung, Yeast Rad55 and Rad57 proteins form a heterodimer that functions with replication protein A to promote DNA strand exchange by Rad51 recombinase, Genes and Development 11 Ž1997. 1111–1121. B. Clever, H. Interthal, J. Schmuckli-Maurer, J. King, M. Sigrist, W.D. Heyer, Recombinational repair in yeast: functional interactions between Rad51 and Rad54 proteins, EMBO Journal 16 Ž1997. 2535–2544. S.L. Hays, A.A. Firmenich, P. Berg, Complex formation in yeast double-strand break repair: participation of Rad51, Rad52, Rad55, and Rad57 proteins, Proceedings of the National Academy of Sciences of the United States of Amercia 92 Ž1995. 6925–6929. R.D. Johnson, L.S. Symington, Functional differences and interactions among the putative RecA homologs Rad51, Rad55, and Rad57, Molecular and Cellular Biology 15 Ž1995. 4843–4850. A. Aboussekhra, R. Chanet, A. Adjiri, F. Fabre, Semidominant suppressors of Srs2 helicase mutations of Saccharomyces cereÕisiae map in the RAD51 gene, whose sequence predicts a protein with similarities to procaryotic RecA proteins, Molecular and Cellular Biology 12 Ž1992. 3224–3334.
J.C. Game r Mutation Research 451 (2000) 277–293 w108x G. Basile, M. Aker, R.K. Mortimer, Nucleotide sequence and transcriptional regulation of the yeast recombinational repair gene RAD51, Molecular and Cellular Biology 12 Ž1992. 3235–3246. w109x J.A. Kans, R.K. Mortimer, Nucleotide sequence of the RAD57 gene of Saccharomyces cereÕisiae, Gene 105 Ž1991. 139–140. w110x S.T. Lovett, Sequence of the RAD55 gene of Saccharomyces cereÕisiae: similarity of RAD55 to prokaryotic RecA and other RecA-like proteins, Gene 142 Ž1994. 103– 106. w111x S.T. Lovett, R.K. Mortimer, Characterization of null mutants of the RAD55 gene of Saccharomyces cereÕisiae: effects of temperature, osmotic strength, and mating type, Genetics 116 Ž1987. 547–553. w112x F.K. Khasanov, G.V. Savchenko, E.V. Bashkirova, V.G. Korolev, W.D. Heyer, V.I. Bashkirov, A new recombinational DNA repair gene from Schizosaccharomyces pombe with homology to Escherichia coli RecA, Genetics 152 Ž1999. 1557–1572. w113x J. Thacker, The role of homologous recombination processes in the repair of severe forms of DNA damage in mammalian cells, Biochimie 81 Ž1999. 77–85. w114x M. Brendel, R.H. Haynes, Interactions among genes controlling sensitivity to radiation and and alkylation in yeast, Molecular & General Genetics 125 Ž1973. 197–216. w115x R.H. McKee, C.W. Lawrence, Genetic analysis of g-ray mutagenesis in yeast III. Double mutant strains, Mutation Research 70 Ž1980. 37–48. w116x R.H. Schiestl, J. Zhu, T.D. Petes, Effect of mutations in genes affecting homologous recombination on restriction enzyme-mediated and illegitimate recombination in Saccharomyces cereÕisiae, Molecular and Cellular Biology 14 Ž1994. 4493–4500. w117x Y. Tsukamoto, J. Kato, H. Ikeda, Effects of mutations of RAD50, RAD51, RAD52, and related genes on illegitimate
recombination in Saccharomyces cereÕisiae, Genetics 142 Ž1996. 383–391. J.C. Game, DNA double-strand breaks and the RAD50 – RAD57 genes in Saccharomyces, Seminars in Cancer Biology 4 Ž1993. 73–83. K. Ho, Induction of DNA double strand breaks by X-rays in a radiosensitive strain of the yeast Saccharomyces cereÕisiae, Mutation Research 30 Ž1975. 327–334. M.A. Resnick, P. Martin, The repair of double-strand breaks in the nuclear DNA of Saccharomyces cereÕisiae and its genetic control, Molecular & General Genetics 143 Ž1976. 119–129. D. Frankenberg, M. Frankenberg-Schwager, D. Blocher, R. Harbich, Evidence for DNA double-strand breaks as the critical lesions in yeast cells irradiated with sparsely or densely ionizing radiation under oxic or anoxic conditions, Radiation Research 88 Ž1981. 524–532. E.M. Geigl, F. Eckardt-Schupp, Repair of gamma ray-induced S1 nuclease hypersensitive sites in yeast depends on homologous mitotic recombination and a RAD18-dependent function, Current Genetics 20 Ž1991. 33–37. L.K. Lewis, J.M. Kirchner, M.A. Resnick, Requirement for end-joining and checkpoint functions, but not RAD52-mediated recombination, after EcoRI endonuclease cleavage of Saccharomyces cereÕisiae DNA, Molecular and Cellular Biology 18 Ž1998. 1891–1902. M.A. Resnick, Similar responses to ionizing radiation of fungal and vertebrate cells and the importance of DNA doublestrand breaks, Journal of Theoretical Biology 71 Ž1978. 339–346. M. Frankenberg-Schwager, D. Frankenberg, DNA doublestrand breaks: their repair and relationship to cell killing in yeast, International Journal of Radiation Biology 58 Ž1990. 569–575. M.A. Resnick, Sunlight-induced killing in Saccharomyces cereÕisiae, Nature 226 Ž1970. 377–378.