Molecular Mechanisms of Pyrimidine Dimer Excision in Saccharomyces cerevisiae: Incision of Ultraviolet-Irradiated Deoxyribonucleic Acid In Vivo

A group of genetically related ultraviolet (UV)-sensitive mutants of Saccharomyces cerevisiae has been examined in terms of their survival after exposure to UV radiation, their ability to carry out excision repair of pyrimidine dimers as measured by the loss of sites (pyrimidine dimers) sensitive to a dimer-specific enzyme probe, and in terms of their ability to effect incision of their deoxyribonucleic acid (DNA) during post-UV incubation in vivo (as measured by the detection of single-strand breaks in nuclear DNA). In addition to a haploid RAD⁺ strain (S288C), 11 different mutants representing six RAD loci (RAD1, RAD2, RAD3, RAD4, RAD14, and RAD18) were examined. Quantitative analysis of excision repair capacity, as determined by the loss of sites in DNA sensitive to an enzyme preparation from M. luteus which is specific for pyrimidine dimers, revealed a profound defect in this parameter in all but three of the strains examined. The rad14-1 mutant showed reduced but significant residual capacity to remove enzyme-sensitive sites as did the rad2-4 mutant. The latter was the only one of three different rad2 alleles examined which was leaky in this respect. The UV-sensitive strain carrying the mutant allele rad18-1 exhibited normal loss of enzyme-sensitive sites consistent with its assignment to the RAD6 rather than the RAD3 epistatic group. All strains having mutant alleles of the RAD1, RAD2, RAD3, RAD4, and RAD14 loci showed no detectable incubation-dependent strand breaks in nuclear DNA after exposure to UV radiation. These experiments suggest that the RAD1, RAD2, RAD3, RAD4 (and probably RAD14) genes are all required for the incision of UV-irradiated DNA during pyrimidine dimer excision in vivo.     

Radiosensitive and mitotic recombination phenotypes of the Saccharomyces cerevisiae dun1 mutant defective in DNA damage-inducible gene expression.

The biological significance of DNA damage-induced gene expression in conferring resistance to DNA-damaging agents is unclear. We investigated the role of DUN1-mediated, DNA damage-inducible gene expression in conferring radiation resistance in Saccharomyces cerevisiae. The DUN1 gene was assigned to the RAD3 epistasis group by quantitating the radiation sensitivities of dun1, rad52, rad1, rad9, rad18 single and double mutants, and of the dun1 rad9 rad52 triple mutant. The dun1 and rad52 single mutants were similar in terms of UV sensitivities; however, the dun1 rad52 double mutant exhibited a synergistic decrease in UV resistance. Both spontaneous intrachromosomal and heteroallelic gene conversion events between two ade2 alleles were enhanced in dun1 mutants, compared to DUN1 strains, and elevated recombination was dependent on RAD52 but not RAD1 gene function. Spontaneous sister chromatid exchange (SCE), as monitored between truncated his3 fragments, was not enhanced in dun1 mutants, but UV-induced SCE and heteroallelic recombination were enhanced. Ionizing radiation and methyl methanesulfonate (MMS)-induced DNA damage did not exhibit greater recombinogenicity in the dun1 mutant compared to the DUN1 strain. We suggest that one function of DUN1-mediated DNA damage-induced gene expression is to channel the repair of UV damage into a nonrecombinogenic repair pathway.     

Stabilization of yeast artificial chromosome clones in a rad54-3 recombination-deficient host strain.

The cloning and propagation of large fragments of DNA on yeast artificial chromosomes (YACs) has become a routine and valuable technique in genome analysis. Unfortunately, many YAC clones have been found to undergo rearrangements or deletions during the cloning process. The frequency of transformation-associated alterations and mitotic instability can be reduced in a homologous recombination-deficient yeast host strain such as a rad52 mutant. RAD52 is one member of an epistatic group of genes required for the recombinational repair of double-strand breaks in DNA. rad52 mutants grow more slowly and transform less efficiently than RAD + strains and are therefore not ideal hosts for YAC library construction. We have investigated the ability of both null and temperature-sensitive alleles of RAD54 , another member of the RAD52 epistasis group, to prevent rearrangements of human YAC clones containing tandemly repeated DNA sequences. Our results show that the temperature-sensitive rad54-3 allele blocks mitotic recombination between tandemly repeated DYZ3 satellite sequences and significantly stabilizes a human DYZ5 satellite-containing YAC clone. Yeast carrying the rad54-3 mutation can undergo meiosis, have growth and transformation rates comparable with RAD + strains, and therefore represent improved YAC cloning hosts.     

Induction of pure and sectored mutant clones in excision-proficient and deficient strains of yeast.

We have found that UV-induced mutation frequency in a forward non-selective assay system (scoring white adex ade2 double auxotroph mutants among the red pigmented ade2 clones) increases linearly with dose up to a maximum frequency of about 3 X 10(-3) mutants per survivor and then declines in both RAD wild-type and rad2 excision deficient strains of Saccharomyces cerevisiae. Mutation frequencies of the RAD and the rad2 strains plotted against survival are nearly identical over the entire survival range. On this basis we conclude that unexcised pyrimidine dimers are the predominant type of pre-mutational lesions in both strains. In the RAD wild-type strain pure mutant clones outnumber sectors in a 10:1 ratio at all doses used; in rad2 this ratio varies from 1:1 at low doses up to 10:1 at high doses. As others have concluded for wild-type strains we find also in the rad2 strain that pure clone formation cannot be accounted for quantitatively by lethal sectoring events alone. We conclude that heteroduplex repair is a crucial step in pure mutant clone formation and we examine the plausibility of certain macromolecular mechanisms according to which heteroduplex repair may be coupled with replication, repair and sister strand exchange in yeast mutagenesis.     

Production and Characterization of Radiation-Sensitive Meiotic Mutants of Coprinus Cinereus

We have isolated four gamma-ray-sensitive mutants of the basidiomycete Coprinus cinereus. When homozygous, two of these (rad 3-1 and rad 9-1) produce fruiting bodies with very few viable basidiospores, the products of meiosis in this organism. A less radiation-sensitive allele of RAD 3, rad 3-2, causes no apparent meiotic defect in homozygous strains. Quantitative measurements of oidial survival of rad 3-1; rad 9-1 double mutants compared to the single mutants indicated that rad 3-1 and rad 9-1 mutants are defective in the same DNA repair pathway. In the few viable basidiospores that are produced by these two strains, essentially normal levels of meiotic recombination can be detected. None of the mutants exhibits increased sensitivity to UV radiation. Cytological examination of meiotic chromosomes from mutant and wild-type fruiting bodies showed that rad 3-1 homozygous strains fail to condense and pair homologous chromosomes during prophase I. Although rad 9-1 strains are successful at chromosome pairing, meiosis is usually not completed in these mutants.     

Genetic control of the bypass of mono-adducts and of the repair of cross-links photoinduced by 8-methoxypsoralen in yeast

A large UVA dose by itself induces lethal damage revealed in some repair-deficient strains of Saccharomyces cerevisiae. Following photoaddition of a monofunctional psoralen derivative, 3-carbethoxypsoralen, an extra killing effect is observed by applying a second high UVA dose, in conditions where a fraction of 8-methoxypsoralen (8-MOP) plus UVA-induced monoadducts are transformed into DNA cross-links. In an excision-repair-deficient context, the bypass of 8-MOP plus UVA-induced monoadducts is under the control of the RAD6+ gene product. However, when other steps of the mutagenic pathway are blocked by the rad18-2 or the pso1-1 mutations, bypass occurs. This is also true when in excision-deficient strains the recombinogenic pathway is blocked by the rad52-1 mutation. The recombinogenic pathway may be an alternative to the mutagenic pathway for bypass of monoadducts. The repair of the lesions induced by a second UVA dose applied after a first treatment by 8-MOP plus UVA [i.e. cross-links and other putative lesion(s)] is controlled by at least the RAD2+, RAD6+, RAD52+, PSO2+ and PSO1+ gene products. The role of the pathways involved is discussed according to the nature of the secondarily induced lesions.     

Molecular cloning of eucaryotic genes required for excision repair of UV-irradiated DNA: isolation and partial characterization of the RAD3 gene of Saccharomyces cerevisiae.

We describe the molecular cloning of a 6-kilobase (kb) fragment of yeast chromosomal DNA containing the RAD3 gene of Saccharomyces cerevisiae. When present in the autonomously replicating yeast cloning vector YEp24, this fragment transformed two different UV-sensitive, excision repair-defective rad3 mutants of S. cerevisiae to UV resistance. The same result was obtained with a variety of other plasmids containing a 4.5-kb subclone of the 6-kb fragment. The UV sensitivity of mutants defective in the RAD1, RAD2, RAD4, and RAD14 loci was not affected by transformation with these plasmids. The 4.5-kb fragment was subcloned into the integrating yeast vector YIp5, and the resultant plasmid was used to transform the rad3-1 mutant to UV resistance. Both genetic and physical studies showed that this plasmid integrated by homologous recombination into the rad3 site uniquely. We conclude from these studies that the cloned DNA that transforms the rad3-1 mutant to UV resistance contains the yeast chromosomal RAD3 gene. The 4.5-kb fragment was mapped by restriction analysis, and studies on some of the subclones generated from this fragment indicate that the RAD3 gene is at least 1.5 kb in size.     

Synergism between base excision repair,mediated by the DNA glycosylases Ntg1 and Ntg2, and nucleotide excision repair in the removal of oxidatively damaged DNA bases in Saccharomyces cerevisiae

In Saccharomyces cerevisiae, inactivation of the two DNA N-glycosylases Ntg1p and Ntg2p does not result in a spontaneous mutator phenotype, whereas simultaneous inactivation of Ntglp, Ntg2p and Radlp or Rad14p, both of which are involved in nucleotide excision repair (NER), does. The triple mutants rad1 ntg1 ntg2 and rad14 ntg1 ntg2 show 15- and 22-fold increases, respectively, in spontaneous forward mutation to canavanine resistance (CanR) relative to the wild-type strain (WT). In contrast, neither of these triple mutants shows an increase in the incidence of Lys+ revertants of the lys1-1 ochre allele. Furthermore, the rad1 ntg1 ntg2 mutant is hypersensitive to the lethal effect of H2O2 relative to WT, rad1 and ntg1 ntg2 mutant strains. Moreover, the rad1 ntg1 ntg2 strain is hypermutable (CanR and Lys+) upon exposure to H2O2, relative to WT, rad1 and ntg1 ntg2 strains. Mutagen sensitivity and enhanced mutagenesis in the rad1 ntg1 ntg2 triple mutant, relative to the other strains tested, were also observed upon exposure to oxidizing agents such as tertbutylhydroperoxide and menadione. In contrast, the sensitivity of the rad1 ntg1 ntg2 triple mutant to gamma-irradiation does not differ from that of the WT. However, the triple mutant shows an increase in the frequency of Lys+ revertants recovered after gamma-irradiation. The results reported in this study demonstrate that base excision repair (BER) mediated by Ntglp and Ntg2p acts synergistically with NER to repair endogenous or induced lethal and mutagenic oxidative DNA damage in yeast. The substrate specificity of Ntg1 p and Ntg2p, and the spectrum of lesions induced by the DNA-damaging agents used, strongly suggest that oxidized DNA bases, presumably oxidized pyrimidines, represent the major targets of this repair pathway.     

Incision and postincision steps of pyrimidine dimer removal in excision-defective mutants of Saccharomyces cerevisiae.

cdc9, a temperature-sensitive mutant defective in polynucleotide deoxyribonucleic acid (DNA) ligase activity, accumulates low-molecular-weight DNA fragments (as measured by sedimentation of DNA in alkaline sucrose gradients) at the nonpermissive temperature after irradiation with ultraviolet light. This phenotype of cdc9 is a sensitive indicator of successful incision during excision repair of dimers. In strains containing excision-defective mutations in any of nine genes in combination with the cdc9 mutation, the absence of low-molecular-weight DNA at the nonpermissive temperature after ultraviolet treatment suggests that these mutants are incision defective, whereas the presence of low-molecular-weight DNA indicates that the mutants are defective in a step after incision. With rad1, rad2, rad3, rad4, and rad10 mutants, the molecular weight of the DNA remained unchanged after ultraviolet irradiation and incubation at the restrictive temperature, despite the presence of the cdc9 mutation; these mutants are therefore incision defective. Low-molecular-weight DNA was observed in rad14 cdc9 and rad16 cdc9 strains. With the rad16 strain, the accumulation of low-molecular-weight DNA correlated with the amount of excision taking place, whereas in the rad14 mutant strain, no evidence of dimer removal was obtained. Therefore, rad14 is likely to be defective in a step after incision.     

Repair of 8-methoxypsoralen photoinduced cross-links in yeast

Summary The repair of interstrand cross-links induced by 8-methoxypsoralen plus UVA (365 nm) radiation DNA was analyzed in diploid strains of the yeast Saccharomyces cerevisiae. The strains employed were the wild-type D₇ and derivatives homozygous for the rad18-1 or the rad3-12 mutation. Alkaline step-elution and electron microscopy were performed to follow the process of induction and removal of photoinduced crosslinks. In accordance with previous reports, the D₇ rad3-12 strain failed to remove the induced lesions and could not incise cross-links. The strain D₇ rad18-1 was nearly as efficient in the removal of 8-MOP photoadducts after 2 h of post-treatment incubation as the D₇ RAD⁺ wild-type strain. However, as demonstrated by alkaline step-elution and electron microscopic analysis, the first incision step at DNA cross-links was three times more effective in D₇ rad18-1 than in D₇ RAD⁺. This is consistent with the hypothesis that the RAD18 gene product is involved in the filling of gaps resulting from persistent non-informational DNA lesions generated by the endonucleolytic processing of DNA cross-links. Absence of this gene product may lead to extensive strand breakage and decreased recognition of such lesions by structural repair systems.     

Unrepaired Heteroduplex DNA in Saccharomyces Cerevisiae Is Decreased in Rad1 Rad52-Independent Recombination

A direct repeat recombination assay between SUP4 heteroalleles detects unrepaired heteroduplex DNA (hDNA) as sectored colonies. The frequency of unrepaired heteroduplex is dependent on the mismatch and is highest in a construct that generates C:C or G:G mispairs and lowest in one that generates T:G or C:A mispairs. In addition, unrepaired hDNA increases for all mismatches tested in pms1 mismatch repair-deficient strains. These results support the notion that hDNA is formed across the SUP4 repeats during the recombination event and is then subject to mismatch repair. The effects of various repair and recombination defective mutations on this assay were examined. Unrepaired heteroduplex increases significantly only in rad52 mutant strains. In addition, direct repeat recombination is reduced 2-fold in rad52 mutant strains, while in rad51, rad54, rad55 and rad57 mutants direct repeat recombination is increased 3-4-fold. Mutations in the excision repair gene, RAD1, do not affect the frequency of direct repeat recombination. However, the level of unrepaired heteroduplex is slightly decreased in rad1 mutant strains. Similar to previous studies, rad1 rad52 double mutants show a synergistic reduction in direct repeat recombination (35-fold). Interestingly, unrepaired heteroduplex is reduced 4-fold in the double mutants. Experiments with shortened repeats suggest that the reduction in unrepaired heteroduplex is due to decreased hDNA tract length in the double mutant strain.     

RAD1, an excision repair gene of Saccharomyces cerevisiae, is also involved in recombination.

The RAD1 gene of Saccharomyces cerevisiae is required for the incision step of excision repair of damaged DNA. In this paper, we report our observations on the effect of the RAD1 gene on genetic recombination. Mitotic intrachromosomal and interchromosomal recombination in RAD+, rad1, rad52, and other rad mutant strains was examined. The rad1 deletion mutation and some rad1 point mutations reduced the frequency of intrachromosomal recombination of a his3 duplication, in which one his3 allele is deleted at the 3' end while the other his3 allele is deleted at the 5' end. Mutations in the other excision repair genes, RAD2, RAD3, and RAD4, did not lower recombination frequencies in the his3 duplication. As expected, recombination between the his3 deletion alleles in the duplication was reduced in the rad52 mutant. The frequency of HIS3+ recombinants fell synergistically in the rad1 rad52 double mutant, indicating that the RAD1 and RAD52 genes affect this recombination via different pathways. In contrast to the effect of mutations in the RAD52 gene, mutations in the RAD1 gene did not lower intrachromosomal and interchromosomal recombination between heteroalleles that carry point mutations rather than partial deletions; however, the rad1 delta mutation did lower the frequency of integration of linear plasmids and DNA fragments into homologous genomic sequences. We suggest that RAD1 plays a role in recombination after the formation of the recombinogenic substrate.     

A role for histone H2B during repair of UV-induced DNA damage in Saccharomyces cerevisiae

To investigate the role of the nucleosome during repair of DNA damage in yeast, we screened for histone H2B mutants that were sensitive to UV irradiation. We have isolated a new mutant, htb1-3, that shows preferential sensitivity to UV-C. There is no detectable difference in bulk chromatin structure or in the number of UV-induced cis-syn cyclobutane pyrimidine dimers (CPD) between HTB1 and htb1-3 strains. These results suggest a specific effect of this histone H2B mutation in UV-induced DNA repair processes rather than a global effect on chromatin structure. We analyzed the UV sensitivity of double mutants that contained the htb1-3 mutation and mutations in genes from each of the three epistasis groups of RAD genes. The htb1-3 mutation enhanced UV-induced cell killing in rad1Delta and rad52Delta mutants but not in rad6Delta or rad18Delta mutants, which are defective in postreplicational DNA repair (PRR). When combined with other mutations that affect PRR, the histone mutation increased the UV sensitivity of strains with defects in either the error-prone (rev1Delta) or error-free (rad30Delta) branches of PRR, but did not enhance the UV sensitivity of a strain with a rad5Delta mutation. When combined with a ubc13Delta mutation, which is also epistatic with rad5Delta, the htb1-3 mutation enhanced UV-induced cell killing. These results suggest that histone H2B acts in a novel RAD5-dependent branch of PRR.     

Molecular cloning of the RAD10 gene of Saccharomyces cerevisiae

We have cloned the RAD10 gene of Saccharomyces cerevisiae and physically mapped it to a 1.0-kb DNA fragment. Strains containing disruptions of the RAD10 gene were found to show enhanced UV sensitivity compared with the previously characterized rad10-1 or rad10-2 mutants. The UV sensitivity of the disruption mutant is comparable to the highly UV sensitive rad1-19, rad2-delta, and rad3-2 mutants.     

The Saccharomyces cerevisiae checkpoint genes RAD9, CHK1 and PDS1 are required for elevated homologous recombination in a mec1 (ATR) hypomorphic mutant

Specific ataxia telangiectasia and Rad3-related (ATR) mutations confer higher frequencies of homologous recombination. The genetic requirements for hyper-recombination in ATR mutants are unknown. MEC1, the essential yeast ATR/ATM homolog, controls S and G2 checkpoints and the DNA damage-inducibility of genes after radiation exposure. Since the mec1-D (null) mutant is defective in both S and G2 checkpoints, we measured spontaneous and DNA damage-associated sister chromatid exchange (SCE), homolog (heteroallelic) recombination, and homology-directed translocations in the mec1-21 hypomorphic mutant, which is defective in the S phase checkpoint but retains some G2 checkpoint function. We observed a sixfold, tenfold and 30-fold higher rate of spontaneous SCE, heteroallelic recombination, and translocations, respectively, in mec1-21 mutants compared to wild type. The mec1-21 hyper-recombination was partially reduced in rad9, pds1, and chk1 mutants, and abolished in rad52 mutants, suggesting the hyper-recombination results from RAD52-dependent recombination pathway(s) that require G2 checkpoint functions. The HU and UV sensitivities of mec1-21 rad9 and mec1-21 rad52 were synergistically increased, compared to the single mutants, indicating that mec1-21, rad52 and rad9 mutants are defective in independent pathways for HU and UV resistance. G2-arrested mec1-21 rad9 cells exhibit more UV resistance than non-synchronized cells, indicating that one function of RAD9 in conferring UV resistance in mec1-21 is by triggering G2 arrest. We suggest that checkpoint genes that function in the RAD9-mediated pathway are required for either homologous recombination or DNA damage resistance in the S phase checkpoint mutant mec1-21.     

S. cerevisiae has three pathways for DNA interstrand crosslink repair.

Yeast mutants, snm1 (pso2-1), rev3 (pso1-1), and rad51, which display significant sensitivity to interstrand crosslinks (ICLs) have low relative sensitivity to other DNA damaging agents. SNM1, REV3, and RAD51 were disrupted in the same haploid strain, singly and in combination. The double mutants, snm1 Delta rev3 Delta, snm1 Delta rad51 Delta and rev3 Delta rad51 Delta were all more sensitive to ICLs than any of the single mutants, indicating that they are in separate epistasis groups for survival. A triple mutant displayed greater sensitivity to ICLs than any of the double mutants, with one ICL per genome being lethal. Therefore, Saccharomyces cerevisiae appears to have three separate ICL repair pathways, but no more. S-phase delay was not observed after ICL damage introduced by cisplatin (CDDP) or 8-methoxypsoralen (8-MOP) during the G1-phase, in any of the above mutants, or in an isogenic rad14 Delta mutant deficient in nucleotide excision repair. However, the psoralen analog angelicin (monoadduct damage) induced a significant S-phase delay in the rad14 Delta mutant. Thus, normal S-phase in the presence of ICLs does not seem to be due to rapid excision repair. The results also indicate that monoadduct formation by CDDP or 8-MOP at the doses used is not sufficient to delay S-phase in the rad14 Delta mutant. While the sensitivity of a rev3 Delta mutant indicates Pol zeta is needed for optimal ICL repair, isogenic cells deficient in Pol eta (rad30 Delta cells) were not significantly more sensitive to ICL agents than wild-type cells, and have no S-phase delay.     

Mutational inactivation of the Saccharomyces cerevisiae RAD4 gene in Escherichia coli.

The RAD4 gene of Saccharomyces cerevisiae is required for the incision of damaged DNA during nucleotide excision repair. When plasmids containing the wild-type gene were transformed into various Escherichia coli strains, transformation frequencies were drastically reduced. Most plasmids recovered from transformants showed deletions or rearrangements. A minority of plasmids recovered from E. coli HB101 showed no evidence of deletion or rearrangement, but when they were transformed into S. cerevisiae on centromeric vectors, little or no complementation of the UV sensitivity of rad4 mutants was observed. Deliberate insertional mutagenesis of the wild-type RAD4 allele before transformation of E. coli restored transformation to normal levels. Plasmids recovered from these transformants contained an inactive rad4 allele; however, removal of the inserted DNA fragment restored normal RAD4 function. These experiments suggest that expression of the RAD4 gene is lethal to E. coli and show that lethality can be prevented by inactivation of the gene before transformation. Stationary-phase cultures of some strains of E. coli transformed with plasmids containing an inactivated RAD4 gene showed a pronounced delay in the resumption of exponential growth, suggesting that the mutant (and, by inference, possibly wild-type) Rad4 protein interferes with normal growth control in E. coli. The rad4-2, rad4-3, and rad4-4 chromosomal alleles were leaky relative to a rad4 disruption mutant. In addition, overexpression of plasmid-borne mutant rad4 alleles resulted in partial complementation of rad4 strains. These observations suggest that the Rad4 protein is relatively insensitive to mutational inactivation.     

Homologous recombination is required for the viability of rad27 mutants.

The RAD27/RTH1 gene of Saccharomyces cerevisiae encodes a structural and functional homolog of the 5'-3' exonuclease function of Escherichia coli DNA polymerase I. Four alleles of RAD27 were recovered in a screen for hyper-recombination, a phenotype also displayed by polA mutants of E.coli. All four rad27 mutants showed similar high levels of mitotic recombination, but varied in their growth rate at various temperatures, and sensitivity to the DNA damaging agent methyl methane sulfonate. Dependence of viability of rad27 strains on recombination was determined by crossing a strain containing a null allele of RAD27 to strains containing a mutation in either the RAD1, RAD50, RAD51, RAD52, RAD54, RAD55, RAD57, MRE11, XRS2 or RAD59 gene. In no case were viable spore products recovered that contained both mutations. Elimination of the non-homologous end-joining pathway did not affect the viability of a rad27 strain. This suggests that lesions generated in the absence of RAD27 must be processed by homologous recombination.     

Multiple Pathways for Homologous Recombination in Saccharomyces Cerevisiae

The genes in the RAD52 epistasis group of Saccharomyces cerevisiae are necessary for most mitotic and meiotic recombination events. Using an intrachromosomal inverted-repeat assay, we previously demonstrated that mitotic recombination of this substrate is dependent upon the RAD52 gene. In the present study the requirement for other genes in this epistasis group for recombination of inverted repeats has been analyzed, and double and triple mutant strains were examined for their epistatic relationships. The majority of recombination events are mediated by a RAD51-dependent pathway, where the RAD54, RAD55 and RAD57 genes function downstream of RAD51. Cells mutated in RAD55 or RAD57 as well as double mutants are cold-sensitive for inverted-repeat recombination, whereas a rad51 rad55 rad57 triple mutant is not. The RAD1 gene is not required for inverted-repeat recombination but is able to process spontaneous DNA lesions to produce recombinant products in the absence of RAD51. Furthermore, there is still considerably more recombination in rad1 rad51 mutants than in rad52 mutants, indicating the presence of another, as yet unidentified, recombination pathway.