Role of Mismatch Repair in the Fidelity of RAD51- and RAD59-Dependent Recombination in Saccharomyces cerevisiae

Corresponding author: Sue Jinks-Robertson, Emory University, 1510 Clifton Rd., Atlanta, GA 30322., jinks{at}biology.emory.edu (E-mail)

Communicating editor: A. NICOLAS

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

To prevent genome instability, recombination between sequences that contain mismatches (homeologous recombination) is suppressed by the mismatch repair (MMR) pathway. To understand the interactions necessary for this regulation, the genetic requirements for the inhibition of homeologous recombination were examined using mutants in the RAD52 epistasis group of Saccharomyces cerevisiae. The use of a chromosomal inverted-repeat recombination assay to measure spontaneous recombination between 91 and 100% identical sequences demonstrated differences in the fidelity of recombination in pathways defined by their dependence on RAD51 and RAD59. In addition, the regulation of homeologous recombination in rad51 and rad59 mutants displayed distinct patterns of inhibition by different members of the MMR pathway. Whereas the requirements for the MutS homolog, MSH2, and the MutL homolog, MLH1, in the suppression of homeologous recombination were similar in rad51 strains, the loss of MSH2 caused a greater loss in homeologous recombination suppression than did the loss of MLH1 in a rad59 strain. The nonequivalence of the regulatory patterns in the wild-type and mutant strains suggests an overlap between the roles of the RAD51 and RAD59 gene products in potential cooperative recombination mechanisms used in wild-type cells.

GENOMIC instability, a hallmark of aging and carcinogenesis, is manifested as mutation and inappropriate recombination (BISHOP and SCHIESTL 2002 ; BOHR 2002 ). An important pathway that promotes genomic stability is the mismatch repair (MMR) pathway, which not only prevents mutation by removing mismatches in newly replicated DNA but also monitors the fidelity of homologous recombination. Although homologous recombination is critical for the repair of DNA double strand breaks (DSBs), it is strictly regulated such that a single mismatch is sufficient to inhibit recombination in the budding yeast Saccharomyces cerevisiae (DATTA et al. 1996 ). In the absence of MMR, recombination between sequences that are similar but not identical (called homeologous recombination) increases dramatically. Although the role of MMR in the repair of replication errors has been well studied, the mechanism of MMR suppression of homeologous recombination is not yet fully understood.

Eukaryotic MMR proteins are classified by homology to their Escherichia coli counterparts, MutS and MutL. In S. cerevisiae, nuclear mitotic mismatch recognition is carried out by heterodimers of MutS homologs (Msh2:Msh3 or Msh2:Msh6; JOHNSON et al. 1996 ; MARSISCHKY et al. 1996 ). Heterodimers of the MutL homologs (Mlh1:Pms1, Mlh1:Mlh2, or Mlh1:Mlh3) then interact with the MutS heterodimer to mediate steps that lead to the removal of the mismatch (CROUSE 1998 ; WANG et al. 1999 ; HARFE et al. 2000 ). Studies of mutator phenotypes and mutation spectra suggest that Msh2:Msh6 heterodimers recognize small insertions/deletions and base-base mismatches, while Msh2:Msh3 heterodimers detect only insertion/deletion loops (KOLODNER and MARSISCHKY 1999 ). The mutator phenotypes of msh2, msh3 msh6, mlh1, or pms1 mutant strains are roughly equivalent, indicating that MMR is primarily carried out by interaction of Mlh1:Pms1 heterodimers with Msh2:Msh3 or Msh2:Msh6 complexes (FLORES-ROZAS and KOLODNER 1998 ; HARFE and JINKS-ROBERTSON 2000A ; HARFE et al. 2000 ). The Msh2:Msh3 complex, together with the Rad1:Rad10 complex, is also involved in the removal of 3' nonhomologous tails generated during recombination processes (SUGAWARA et al. 1997 ).

Although loss of the MutS or the MutL complex has similar effects on mutation rates, previous studies suggest that the different MMR complexes perform differently in the regulation of homeologous recombination (CHEN and JINKS-ROBERTSON 1999 ; NICHOLSON et al. 2000 ). Loss of the MutS homologs, for example, causes a greater increase in homeologous recombination than does the loss of either Mlh1 or Pms1. In addition, the absence of Msh3 causes an increase in recombination between sequences containing only base-base mismatches, despite the conclusions from previous studies that Msh3 has no role in the repair of base-base mismatches (MARSISCHKY et al. 1996 ; EARLEY and CROUSE 1998 ). The observed differences may arise from unique activities or interactions required in postreplication repair vs. the maintenance of recombination fidelity. In the repair of replication errors, the binding of the MutS and then the MutL homologs leads to the removal of the newly synthesized strand, resynthesis across the resulting gap, and ligation. In the inhibition of homeologous recombination, binding of MMR proteins to mismatches present in heteroduplex recombination intermediates presumably prevents the completion of recombination by an as-yet-undefined mechanism. Whether this mechanism is by rejection of the invading strand, blocks in Holliday junction migration, or inhibition of the extension of heteroduplex DNA is unknown (ALANI et al. 1994 ; CHEN and JINKS-ROBERTSON 1998 ). Characterization of the relationship of MMR-dependent regulation of homeologous recombination and the components of recombination pathways may shed light on the mechanisms by which this regulation takes place.

Genes involved in the DSB repair pathway in S. cerevisiae comprise the RAD52 epistasis group. Numerous genetic and biochemical analyses of the members of this epistasis group have clarified their roles in DSB repair (PAQUES and HABER 1999 ; SUNG et al. 2000 ; SYMINGTON 2002 ). The Mre11, Rad50, and Xrs2 proteins form the Mre11-Rad50-Xrs2 (MRX) complex, which is thought to have an end-processing role in the creation of the 3' single-stranded tails used to invade duplex DNA and a structural role in promoting recombination between sister chromatids. Replication protein A (RPA) binds the 3' single-stranded tails that result from DSB processing, preventing the formation of secondary structure. Rad52 promotes the replacement of RPA by Rad51, the yeast homolog of the bacterial RecA protein, and the resulting Rad51 nucleoprotein filament is stabilized by the Rad55:Rad57 complex. Rad54 has homology to the Swi/Snf family of chromatin-remodeling proteins and facilitates the invasion of a homologous duplex by the nucleoprotein filament. Following strand invasion, repair synthesis is primed from the 3' end of the invading strand. If the newly extended invading strand is unwound from the invaded duplex, it can reanneal with the other end of the broken molecule in a process called synthesis-dependent strand annealing (SDSA), resulting in a noncrossover event. If, on the other hand, the strand displaced from the duplex by the strand invasion process is captured by the 3' tail of the other end of the broken molecule, a double Holliday junction will be formed that can be resolved as either a crossover or a noncrossover (gene conversion) event. The proteins that regulate the steps subsequent to strand invasion have not yet been fully characterized.

An alternative recombination pathway that can occur in the absence of RAD51 but that requires RAD59 was identified in a screen for mutants defective in inverted repeat recombination in a rad51 background (BAI and SYMINGTON 1996 ). This RAD59-dependent alternative pathway involves break-induced replication (BIR) with single-strand annealing (SSA). BIR occurs when DNA synthesis primed from the 3' invading tail of a broken chromosome proceeds down the entire arm of the intact chromosome (MALKOVA et al. 1996 ). SSA takes place between direct repeats after resection of the ends of a DSB reveals complementary single-stranded tails that can base pair, resulting in the deletion of the region between the repeats as well as one of the repeats. Rad59 has partial homology to Rad52 and has some inherent strand-annealing activity. It has been suggested that the functions of Rad52 and Rad59 may overlap to provide strand invasion capability in the absence of Rad51 (BAI et al. 1999 ).

We have utilized an intron-based inverted-repeat assay system (Fig 1) to explore the regulation of homeologous recombination. This system detects recombination events that reorient the region between homologous or homeologous inverted repeats, leading to the expression of a selectable marker. To elucidate potential mechanisms for the inhibition of recombination between homeologous substrates, we examined the effects of the loss of individual proteins involved in the yeast recombination pathways. In this study, the mutants fall into two classes defined by rad51 and rad59 mutant phenotypes with respect to their effect on homeologous recombination relative to homologous recombination. The difference in the phenotypes suggests that the regulation of homeologous recombination is different in recombination pathways dependent on RAD51 or RAD59. To dissect the role of the MMR machinery in that regulation, we examined the phenotypes of strains defective in MMR and in either RAD51 or RAD59. The effect of MMR defects on the corresponding pathways suggests that the regulation of homeologous recombination in different pathways is distinct and is affected by separate activities of MMR complexes.

View larger version (15K): In this window In a new window Download PPT slide   Figure 1. Intron-based inverted-repeat recombination assay. Inverted repeats (shaded and open arrows) were fused to intron splice sites (solid boxes) and placed next to the 5' and 3' halves of the coding sequence for HIS3 (hatched boxes). The direction of transcription of HIS3 is indicated by dashed arrows. Recombination between the repeats that leads to the inversion of the sequence between the repeats (invertible segment) allows expression of the full-length HIS3 gene and growth on plates lacking histidine (see Fig 4 for recombination models of the reorientation process). To measure homeologous recombination, repeats of 91% identity were used; to measure homologous recombination, 100% identical repeats were used.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Media and growth conditions:
Yeast strains were grown nonselectively in YEP medium (1% yeast extract, 2% Bacto-peptone, 250 mg/liter adenine; 2% agar for plates) supplemented with either 2% dextrose (YEPD) or 2% glycerol and 2% ethanol (YEPGE). Selective growth was done on synthetic complete (SC) media lacking the appropriate nutrient (SHERMAN 1991 ) and supplemented with 2% dextrose (SCD) or with 2% galactose, 2% glycerol, and 2% ethanol (SCGGE). Ura^- derivatives were isolated on SCD-uracil plates containing 1 g/liter 5-fluoroorotic acid and 12 mg/liter uracil (AUSUBEL et al. 1994 ). Geneticin- and hygromycin-resistant transformants were isolated on YEPD supplemented with 200 mg/liter geneticin (G418) or 300 mg/liter hygromycin B, respectively. Methyl methanesulfonate (MMS) sensitivity was tested on YEPD containing 0.016% MMS (Kodak). Mutator phenotype was assessed by forward mutation at CAN1 on SC-arginine containing 60 mg/liter canavanine. All incubations were done at 30°.

Strain construction:
All strains used in this study (Table 1) were derived from the congenic strains SJR1486 and SJR1487, which contain 100 or 91% identical inverted repeats, respectively (WELZ-VOEGELE et al. 2002 ). The repeats are 783 bp in length, and the sequence differences in the inverted repeats of SJR1487 are exclusively base substitutions.

Disruptions of the genes of the mismatch repair and recombination pathways were created in both parent strains using standard genetic techniques. To disrupt RAD50, MRE11, RAD55, RAD57, and RAD59, homology to the targeted genes was added to the kanMX2 cassette by PCR, and transformants were selected on YEPD containing G418 (WACH et al. 1994 ). RAD51, RAD52, and RAD54 were disrupted using restriction fragments containing rad51::URA3, rad52::hisG-URA3-hisG, and rad54::URA3 alleles, respectively (FREEDMAN and JINKS-ROBERTSON 2002 ). MSH2 was deleted by transformation with AatII-XbaI-digested pmsh2 (EARLEY and CROUSE 1998 ); MSH3 was disrupted by transformation with EcoRI-digested pEN33 (DATTA et al. 1996 ); RAD1 was deleted using SalI-EcoRI-digested pR1.6 (HIGGINS et al. 1983 ); and MSH6 was disrupted by transformation with EcoRI-SacI-digested Msh6pHUH (KRAMER et al. 1996 ). MLH1 was disrupted using a PCR-generated hygMX2 cassette (GOLDSTEIN and MCCUSKER 1999 ). Double mutants defective in both MMR and recombination were created by first disrupting the relevant MMR gene and then the relevant recombination gene. The presence of each targeted disruption was inferred using phenotypic tests (MMS sensitivity and mutator phenotype for recombination-defective and MMR-defective strains, respectively) and confirmed by PCR. Primer sequences are available upon request.

Determination of recombination rates:
Yeast colonies grown 2 days on YEPD were inoculated into 5 ml YEPGE and grown for 3 days on a roller drum. Two or more isolates of each strain were tested, using a minimum of six colonies per isolate. The cultures were washed with 5 ml sterile H₂O and resuspended in 1 ml sterile H₂O. Appropriate dilutions were plated on YEPD and SCGGE-His to calculate the total number of viable cells and number of recombinants per culture. Colonies on YEPD and SCGGE-His were counted after 2 and 5 days growth, respectively. Recombination rates (recombination per cell per generation) were determined by the method of the median or the method of p₀ (LEA and COULSON 1949 ; SPELL and JINKS-ROBERTSON 2004 ). To calculate the 95% confidence interval for each recombination rate, the numbers of recombinants per culture were first ranked in ascending order. Table B11 of ALTMAN 1991 was then used to identify which ranked cultures should provide the number of recombinants for calculation of the upper and lower rate limits that define the confidence interval.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The regulation of homeologous recombination is affected by mutations in RAD52 epistasis group genes:
To determine the genetic dependence of the regulation of homeologous recombination on the members of the RAD52 epistasis group, we examined the recombination rates in null mutants of RAD50, RAD51, RAD52, RAD54, RAD55, RAD57, and RAD59. The levels of spontaneous recombination between homeologous and homologous substrates were assayed using inverted repeats of 91 or 100% identity, respectively, contained within introns fused to the coding sequence of a selectable marker, HIS3. Recombination between the repeats reorients the region between them such that HIS3 expression is restored (Fig 1).

The recombination rates in wild-type and recombination-defective strains are given in Table 2. In wild-type cells, the rate of homeologous recombination is suppressed 40-fold relative to homologous recombination (4.2 x 10^-8 and 1.7 x 10^-6, respectively), and as expected, homeologous recombination and homologous recombination rates are severely reduced in strains lacking the RAD52 gene that defines the epistasis group. Recombination is elevated in rad50 mutants, with the rate of homeologous recombination being elevated more than that of homologous recombination (4.3-fold vs. 1.9-fold, respectively). A similar increase is seen in mutants of another member of the MRX complex, MRE11 (4.6-fold increase in homeologous recombination rates vs. 3.2-fold increase in homologous recombination). Recombination is generally decreased in the other recombination mutants tested, but the degree of the decrease varies, depending on whether the substrates are homologous or homeologous.

The differential effect of a given rad mutation on homeologous and homologous recombination is evident in the graph of recombination levels relative to wild type in Fig 2 and as a change in the relative homeologous to homologous recombination ratio in Table 2. In Fig 2, the relative homeologous recombination rate is greater than the relative homologous recombination rate for each of the rad strains tested except the rad59 mutant. Compared to wild type, the relative homeologous to homologous recombination ratio is elevated in strains defective in RAD50, RAD51, RAD54, RAD55, or RAD57 (Table 2). This increase in levels of homeologous recombination relative to homologous recombination levels suggests that either the suppression of homeologous recombination is dependent on these gene products or the loss of the gene products shifts recombination into a pathway that is more tolerant of mismatches. In contrast, strains defective in RAD59 have a greater reduction in homeologous recombination than in homologous recombination, indicating that either the mechanism of homeologous recombination is dependent on RAD59 or loss of RAD59 shifts recombination into a more stringent pathway.

View larger version (15K): In this window In a new window Download PPT slide   Figure 2. The recombination rate of homeologous and homologous substrates in recombination-defective strains relative to wild-type strains. Recombination rates were determined for wild-type and rad mutant strains for homologous substrates (open bars) and homeologous substrates (shaded bars). The relative rate was determined by dividing the recombination rate of mutant strains by that of the same substrates in the wild-type strain. Error bars based on the confidence intervals for the rates are indicated.

Homeologous recombination requires either RAD51 or RAD59:
Previous studies have shown that strains defective in either RAD51 or RAD59 have small decreases in recombination, whereas strains defective in both suffer severe decreases (BAI and SYMINGTON 1996 ). To confirm that these two pathways account for the bulk of the homeologous recombination measured in our inverted-repeat assay, double-mutant strains defective in both genes were constructed. The levels of homeologous recombination and homologous recombination in the rad51rad59 double mutant approach those of a rad52 mutant (Table 2 and Fig 2), and the residual recombination levels seen in the double mutant resemble those seen in other studies (BAI and SYMINGTON 1996 ). The synergistic phenotype of the double mutant generally has been interpreted as a competition or functional redundancy between distinct pathways requiring either RAD51 or RAD59, with one pathway compensating for the loss of the other (BAI and SYMINGTON 1996 ). It should be noted, however, that the level of recombination provided by the remaining pathway in the single mutants may not necessarily reflect the levels performed by that pathway in a wild-type cell.

Loss of MSH2 has a differential effect on homeologous recombination occurring by the RAD51- and RAD59-dependent pathways:
Previous studies have shown that most of the suppression of homeologous recombination results from the activity of the MMR machinery (DATTA et al. 1996 ; CHEN and JINKS-ROBERTSON 1999 ; NICHOLSON et al. 2000 ). In agreement with these studies, homeologous recombination and homologous recombination rates are statistically equivalent in strains defective in the major MMR gene, MSH2 (Table 3). To determine the role of MMR in the suppression of homeologous recombination in the RAD51 vs. RAD59 recombination pathways, one recombination pathway was eliminated, and the MMR-dependent regulation of the other pathway was assessed. For example, in a rad51 mutant where only the RAD59-dependent pathway functions, the role of MMR proteins in the regulation of RAD59-dependent recombination becomes apparent. The recombination rates in the relevant single- and double-mutant strains are given in Table 3. Homeologous recombination catalyzed by the RAD59-dependent pathway increases 3.6-fold in an MMR-defective strain (msh2 rad51) relative to an MMR-proficient strain (rad51; Fig 3). In contrast to this small increase, loss of MMR in a msh2 rad59 strain in which only the RAD51-dependent pathway is intact shows a much larger, 27-fold increase relative to the rad59 single mutant. These differences demonstrate that Msh2 has a stronger suppressive effect on RAD51-mediated homeologous recombination than on RAD59-mediated homeologous recombination.

View larger version (62K): In this window In a new window Download PPT slide   Figure 3. The effect of MMR defects in separate recombination pathways. Homologous (open bars) and homeologous (shaded bars) recombination rates were determined for RAD(A), rad51(B), and rad59(C) mutant strains in MMR-proficient and MMR-defective backgrounds. The relative rate was determined by dividing the recombination rate of double-mutant strains by that of the same substrates in the corresponding RAD or rad single-mutant strain. Error bars based on the confidence intervals for the rates are indicated.

Mutations in MLH1 and MSH6 have similar effects on homeologous recombination occurring by the RAD51- and RAD59-dependent pathways:
It has been shown that mutations in the MutS homolog MSH2 and in the MutL homologs MLH1 or PMS1 affect homeologous recombination differently, with msh2 mutants exhibiting higher levels of homeologous recombination than mlh1 or pms1 strains exhibit (CHEN and JINKS-ROBERTSON 1999 ; NICHOLSON et al. 2000 ). To test the role of MLH1 in regulating homeologous recombination in the RAD51- and RAD59-dependent pathways, homeologous recombination rates were measured in mlh1, rad51 mlh1, and rad59 mlh1 mutants (Table 3). It should be noted that the complete deletion of the MLH1 locus used here (mlh1::hyg) causes slightly higher increases in the homeologous to homologous recombination ratio in a RAD background than that seen in studies using the mlh1::URA3 allele (R. SPELL and S. JINKS-ROBERTSON, unpublished results). Because the mlh1::URA3 allele deletes only the amino terminus of the protein (PROLLA et al. 1994 ; NICHOLSON et al. 2000 ; WELZ-VOEGELE et al. 2002 ), it is possible that the carboxy terminus alone or as part of a fusion protein may confer residual antirecombination activity. However, the stronger phenotype caused by the complete loss of MLH1 is still below that of a msh2 mutant (NICHOLSON et al. 2000 ; WELZ-VOEGELE et al. 2002 ). In a rad51 background, the deletion of MLH1 has the same effect as the deletion of MSH2 (Fig 3). This suggests that the MMR-dependent regulation in the RAD59-dependent pathway requires both Mlh1 and Msh2. In contrast, in a rad59 background, elimination of MLH1 elevates homeologous recombination much less (4-fold vs. 27-fold) than the deletion of MSH2 does (Fig 3). Therefore, most of the homeologous recombination inhibition in the RAD51 pathway is effected by MSH2 and does not require MLH1.

The effect of the loss of Msh6, Msh3, and Rad1 on the regulation of homeologous recombination:
The difference in the suppression of homeologous recombination by Msh2 and Mlh1 in the RAD51-dependent pathway could be due to different roles of MutS and MutL homolog proteins in MMR-related processes or could be due to functions the proteins play in other processes. In addition to its role in mismatch recognition, for example, Msh2 together with Msh3 and Rad1:Rad10 are also involved in the removal of 3' nonhomologous tails in recombination intermediates (SUGAWARA et al. 1997 ). To determine if the differential effect of loss of MSH2 and MLH1 in the RAD51-dependent pathway was attributable to the role of Msh2 in MMR or in nonhomologous tail removal, we also examined the role of the proteins that form the clippase complex and the role of a MutS homolog not required for the clippase activity, Msh6 (Table 3). Because all of the sequence nonidentities in the homeologous recombination substrates are base substitutions, we would predict that all potential mismatches in the recombination intermediates should be recognized by Msh2:Msh6. The increase in the homeologous recombination rate in the msh6 rad59 double mutant relative to the rad59 single mutant demonstrates that the loss of Msh6 has a much smaller effect on homeologous recombination than does the loss of Msh2 and is comparable to the effect of Mlh1 loss (Fig 3). Although this result would be consistent with clippase activity causing the 10-fold difference in the relative homeologous to homologous recombination ratio in rad59 msh2 vs. rad59 mlh1 mutants, neither loss of Msh3 nor loss of Rad1 in a rad59 background leads to significant changes in the relative level of homeologous to homologous recombination (Table 3, Fig 3). Together, these results suggest that either the Msh2:Msh3 complex inhibits homeologous recombination in a clippase-independent fashion that is redundant with the Msh2:Msh6 complex or Msh2 acts independently of the Msh2:Msh3 and Msh2:Msh6 complexes to regulate homeologous recombination. These two possibilities can be distinguished by comparing the recombination phenotype of the rad59 msh2 double mutant to that of the rad59 msh3 msh6 triple mutant. As shown in Table 3, the relative homeologous to homologous recombination ratio is similar in the double and triple mutants, indicating that the Msh2:Msh6 and Msh2:Msh3 complexes have overlapping or competing roles in the regulation of homeologous recombination in the RAD51-dependent pathway. These roles are only partially dependent on the Mlh1 protein.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The importance of homologous recombination for genome stability is indicated by the many levels of regulation and by the disease processes that result from a breakdown of that regulation (HARFE and JINKS-ROBERTSON 2000B ; BISHOP and SCHIESTL 2002 ). A crucial level of regulation is the homology restrictions enforced by the MMR pathway, such that the level of recombination between substrates of 91% identity (homeologous recombination) is reduced 40-fold relative to recombination between 100% identical substrates (homologous recombination) in the assay system used here. To determine if interaction with specific factors in recombination pathways is required for that suppression, the MMR-dependent inhibition of recombination between nonidentical substrates was examined in several recombination mutants. We found that the level of homeologous recombination suppression is distinct in different recombination pathways and dependent on different activities of the MMR proteins.

Potential mechanisms for the recombination detected by the inverted repeat assay used in this study are depicted in Fig 4. First, reorientation of the region between the inverted repeats can occur by intramolecular recombination that requires crossing over between the repeats on a single chromosome (intrachromatid crossing over; Fig 4A). A second method is gene conversion without crossing over between sister chromatids after unequal alignment of the sisters (sister-chromatid gene conversion; Fig 4B). Pairing of the misaligned repeats makes possible a long conversion tract that extends from one set of paired repeats, through the invertible segment, into the other set of paired repeats such that the segment between the inverted repeats is reoriented on the repaired sister (Fig 4B). Because gene conversion can occur via SDSA or via an intermediate that involves Holliday junction formation/resolution (see Introduction), the intermediate steps of this mechanism are not depicted. Both of these processes in Fig 4A and Fig B, are completely dependent on RAD51 (PAQUES and HABER 1999 ).

View larger version (24K): In this window In a new window Download PPT slide   Figure 4. Recombination mechanisms for reorientation of the invertible segment in an inverted repeat assay system. Homeologous inverted repeats (shaded and open arrows) are depicted flanking the invertible segment (solid arrow, see Fig 1). (A) Intrachromatid crossover. Intramolecular crossing over between flanking inverted repeats reorients the HIS3 sequences to create a His+ prototroph. (B) Sister-chromatid gene conversion without crossing over. Long conversion tracts between misaligned sister chromatids will lead to inversion of the intervening sequence, allowing expression of the full-length HIS3 gene. Following alignment of the flanking repeats, the intervening sequence can loop around to create a continuous tract of homology. Conversion of the intervening region will reorient the HIS3 coding sequence to allow His+ prototrophy. (C) BIR with SSA. The 3' tail of a DSB in one repeat invades the intact repeat and primes DNA synthesis to the other end of the break. Reannealing of the single-stranded repeat sequences can occur such that the intervening sequence is inverted 50% of the time, allowing HIS3 expression.

A third mechanism, originally described for inverted repeats on plasmids, is BIR followed by SSA (see Introduction and Fig 4C; KANG and SYMINGTON 2000 ). BIR is the one-ended invasion that can prime DNA synthesis down an entire chromosome arm. In the case of a DSB in an inverted repeat, the broken end of one repeat can invade the intact repeat on the same chromatid and prime replication through the intervening sequence to the other end of the break. Thus, BIR not only partially duplicates the repeats but also inverts the region between them. SSA between the repeats can then occur such that HIS3 can be expressed (Fig 4C). For simplicity, an intramolecular event is depicted, but BIR with SSA between sister chromatids could also lead to His⁺ prototrophy, although this would require an interruption in replication of the sister-chromatid arm to allow SSA. Studies of BIR followed by SSA with plasmids containing inverted repeats have shown that such events can occur in the absence of RAD51 yet are dependent on RAD59 and RAD50 (KANG and SYMINGTON 2000 ; IRA and HABER 2002 ). RAD52 is required for both RAD51- and RAD59-dependent recombination mechanisms (BAI et al. 1999 ).

In our assay system, recombination between homeologous substrates or homologous substrates is severely reduced in the absence of RAD52 or of both RAD51 and RAD59. The slightly higher level of homeologous and homologous recombination in the rad51 rad59 double mutant than in the rad52 single mutant may reflect the ability of the cells to undertake some BIR coupled with SSA (see Fig 4) or nonhomologous end joining in the absence of rad59 (RICHARDSON and JASIN 2000 ; SUGAWARA et al. 2000 ; SIGNON et al. 2001 ). The similar genetic dependencies of homologous and homeologous recombination on RAD52 and on either RAD51 or RAD59 suggest that the recombination mechanism is the same for homeologous recombination as for homologous recombination.

Loss of each of the members of the RAD52 epistasis group, with the exception of RAD50, caused decreases in the level of homologous recombination. RAD50 has long confused geneticists because its mutant phenotype is assay dependent. For example, the level of inverted-repeat recombination on chromosomes and plasmids typically decreases in a rad50 mutant (RATTRAY and SYMINGTON 1995 ; IRA and HABER 2002 ), whereas the level of allelic recombination between homologous chromosomes or ectopic recombination between nonhomologous chromosomes increases (MALONE et al. 1990 ; FREEDMAN and JINKS-ROBERTSON 2002 ). Increases in recombination between different chromosomes in rad50 strains have led to the suggestion that Rad50 plays a role in fostering sister-chromatid interactions. Because of the nature of the recombination mechanisms proposed for the assay system described here, it was assumed that rad50 mutants would also display a decrease in recombination. The opposite occurred, however; rad50 as well as mre11 mutants displayed a mild hyperrecombination phenotype in this assay. In addition, measurement of the levels of recombination on other chromosomes in rad50 or mre11 mutants has demonstrated strong chromosome context effects with our assay system (R. SPELL and S. JINKS-ROBERTSON, unpublished results). Our results indicate that the sister-chromatid interaction model for Rad50 may be too simplistic, and this view is supported by the finding that the genetic requirement for RAD50 differs for spontaneous and DNA-damage-induced recombination (DONG and FASULLO 2003 ). Context and assay dependence suggests that the type and location of the recombination-initiating lesion may affect the requirement for RAD50 and other members of the MRX complex.

Strikingly, mutations in the RAD52 epistasis group genes affected the level of homeologous recombination differently than the level of homologous recombination, as if the regulation of homeologous recombination was altered in the mutants. The mutants fell into two groups, defined by the phenotypes of the rad51 and rad59 single mutants. For clarity, we discuss the recombination that occurs in the rad51 and rad59 single mutants as RAD59 dependent (i.e., RAD51 independent) and RAD51 dependent (i.e., RAD59 independent), respectively. In the first class, containing RAD50, RAD51, RAD54, RAD55, and RAD57, loss of the gene caused a larger decrease in homologous recombination than in homeologous recombination, indicating a loss of mismatch-associated inhibition of recombination. For example, in the rad51 mutant, homeologous recombination remained at wild-type levels, while homologous recombination decreased to one-third of the wild-type level. The increase in the homeologous to homologous recombination ratio in the rad50 mutant may result from less efficient end processing, perhaps forcing the use of shorter homologies and thereby lowering the chance of mismatches in heteroduplex DNA. Explanations for the observed increase in the ratio of homeologous to homologous recombination in the rad51 mutant include the slight possibility that Rad51 is partially required for homologous recombination, while homeologous recombination is completely independent of Rad51. However, the evidence that the mechanism for homeologous recombination and homologous recombination is similar makes that possibility unlikely. Alternatively, the increase in the homeologous recombination to homologous recombination ratio in a rad51 mutant may indicate that Rad51 is involved in the normal suppression of homeologous recombination. This possibility is made more interesting by the reports of physical interactions between Rad51 and the MMR factor Mlh1, between Rad51 and the BLM helicase homolog Sgs1, and between Mlh1 and Sgs1 (PEDRAZZI et al. 2001 ; WU et al. 2001 ; HO et al. 2002 ). The potential for the recruitment of Sgs1 to homeologous recombination intermediates by interaction with Rad51 and Mlh1 suggests that one mechanism for rejecting recombination intermediates containing heterology occurs by unwinding heteroduplex DNA. Such a model is consistent with the observation that mutation in SGS1 causes an increase in homeologous recombination (MYUNG et al. 2001 ; R. M. SPELL and S. JINKS-ROBERTSON, unpublished data). Finally, in the absence of the RAD51-dependent pathway, homeologous recombination intermediates may be shunted into a recombination pathway that is less sensitive to mismatches or involves formation of fewer mismatches.

The loss of the RAD59-dependent recombination pathway defines the second phenotypic class with regard to the change in the ratio of homeologous to homologous recombination. In the rad59 mutant, there was a larger decrease in homeologous recombination than in homologous recombination, as if these mutants had lost a pathway that is more tolerant of mismatches. One potential explanation for lower stringency of the RAD59-dependent pathway is the ability of RAD59-dependent recombination to use shorter lengths of homology, as described by Ira and Haber using an HO-induced intraplasmid recombination assay (IRA and HABER 2002 ). The use of shorter homologies could lower the number of mismatches in the heteroduplex recombination intermediates formed between homeologous substrates, making them a less-efficient target for the antirecombination activity of the MMR machinery.

To explore the roles of different MMR proteins in the suppression of homeologous recombination in different potential recombination pathways described above, we examined several MMR-defective backgrounds. Strains lacking Msh2 lack both the Msh2:Msh3 and Msh2:Msh6 mismatch recognition complexes and, in addition, are deficient in the clippase function of Msh2:Msh3 together with Rad1:Rad10. It should be noted that loss of Msh6 specifically removes the recognition of base-base mismatches, the only type of mismatch present in heteroduplex recombination intermediates of the repeats used here. In the absence of Mlh1, mismatch recognition presumably cannot be coupled to the downstream processing events, which may be different for the antimutator vs. the antirecombination activity of the MMR machinery (WELZ-VOEGELE et al. 2002 ). It has been shown previously that the homeologous recombination phenotypes of mutants defective in MSH2, MSH6, or MLH1 are not equivalent (NICHOLSON et al. 2000 ; this study), with loss of MSH2 causing the greatest increase in homeologous recombination. Despite the fact that only Msh6 should be involved in the recognition of potential mismatches between the recombination substrates used here, the msh6 phenotype was much weaker than that of msh2 mutants (NICHOLSON et al. 2000 ; this study). As reported previously, the ratio of homeologous to homologous recombination in a msh3 msh6 double mutant was the same as that in a msh2 single mutant. Because loss of Msh3 alone or of Rad1 did not affect the regulation of homeologous recombination in the current assays, we conclude that the antirecombination activity of the Msh2:Msh3 complex is not related to its clippase function. Finally, the level of homeologous recombination in mlh1 strains was only approximately one-half of that in a msh2 mutant. The distinction between phenotypes in msh2 and mlh1 backgrounds suggests that mismatch recognition alone can contribute to the inhibition of homeologous recombination.

Analysis of the impact of MMR defects when only RAD51 or RAD59 is functional reveals the distinct roles that individual MMR proteins play in the corresponding pathways. In a rad51 background (RAD59-dependent recombination), the pairing of single DNA strands in recombination intermediates would presumably occur via an annealing mechanism, whereas in a rad59 background (RAD51-dependent recombination) pairing would involve invasion of a duplex DNA molecule. In a rad51 background, loss of Msh2, Msh6, or Mlh1 resulted in an approximately threefold increase in the ratio of homeologous to homologous recombination, while elimination of either Msh3 or Rad1 had no effect. This difference suggests that the MMR-associated suppression of homeologous recombination in the RAD59-dependent pathway involves complexes of Msh2 with Msh6 and also requires the Mlh1:Pms1 complex (Fig 5). Despite the increase in the homeologous to homologous recombination ratio observed upon elimination of MMR components in a rad51 background, the homeologous recombination rate was still fourfold below the homologous recombination rate. Therefore, although homeologous recombination in the RAD59-dependent annealing pathway is regulated by the MMR machinery, most of the regulation occurs via an MMR-independent mechanism.

View larger version (25K): In this window In a new window Download PPT slide   Figure 5. Relationship between recombination pathways and MMR-dependent regulation of homeologous recombination. A homeologous recombination intermediate can enter pathways dependent on Rad51, Rad59, or both. Depending on the pathway, MMR proteins inhibit homeologous recombination to different degrees. Regulation of homeologous recombination in wild-type cells suggests that recombination intermediates utilize overlapping pathways.

In contrast to the 3-fold increase in homeologous recombination observed upon loss of Msh2 in a rad51 background, there was an 30-fold increase in homeologous recombination associated with loss of Msh2 in a rad59 background (RAD51-dependent pathway). The phenotypes of msh3, msh6, rad1, and msh3 msh6 mutants demonstrate that regulation in the RAD59-dependent pathway is mediated by both the Msh2:Msh6 and the Msh2:Msh3 complexes but does not significantly involve clippase activity. Also, in the rad59 mutant background, the 10-fold difference in homeologous recombination rates in msh2 vs. mlh1 strains suggests that Mlh1 has relatively little antirecombination activity in the RAD51-dependent strand invasion pathway. Therefore, in the RAD51-dependent recombination pathway, most suppression of homeologous recombination is effected by a Msh2-containing complex, with a MutL-like complex being largely dispensable (Fig 5). Finally, in contrast to the minor role of Msh2 in regulation of the RAD59-dependent pathway, Msh2 plays a major role in the enforcement of recombination fidelity in the RAD51-dependent pathway (Fig 5).

In addition to advancing the understanding of the role of MMR in homeologous recombination regulation, the pattern of regulation revealed in the wild-type, rad51, and rad59 mutant backgrounds suggests that neither pathway remaining in the single mutants accurately represents the mechanism utilized in wild-type cells (Fig 5). This is especially evident when Mlh1 is also absent. Whereas loss of Mlh1 causes a large increase in homeologous recombination in a RAD background, the effect of mlh1 mutation in rad51 or rad59 backgrounds is much weaker (29-fold vs. 2.6-fold or 3.8-fold, respectively). Thus, the greater effect of the loss of Mlh1 in the presence of both Rad51 and Rad59 suggests that the regulation of recombination in wild type is not fully described by the separate regulation of either the RAD51-dependent pathway or the RAD59-dependent pathway. Since the phenotype of the rad51 rad59 mutants suggests that Rad51 and Rad59 promote the bulk of all mitotic recombination, one need not invoke a pathway involving novel proteins to explain these results. Instead, it is much more reasonable to imagine a pathway that involves the activities of both Rad51 and Rad59 (Fig 5). It is possible, for example, that Rad59 may facilitate the reannealing of 3' ends following the extension of one of those ends during Rad51-mediated SDSA. Alternatively, Rad51 could facilitate strand invasion for BIR that is completed by Rad59-mediated SSA (see Fig 4C). Indeed, the potential overlap of RAD51- and RAD59-dependent processes has been noted for inverted repeat recombination on plasmids (BARTSCH et al. 2000 ; KANG and SYMINGTON 2000 ; IRA and HABER 2002 ). This study provides further support for such a model of cooperation between Rad51 and Rad59 for spontaneous recombination between repeats in chromosomal DNA.

In summary, the study of spontaneous recombination between homeologous repeats in chromosomal DNA has revealed the fidelity of recombination in different recombination pathways, defined by their dependence on RAD51 and RAD59. Examination of the loss of suppression of inappropriate recombination between mismatched sequences in strains lacking different members of the MMR pathway suggests distinct roles of these proteins in the different pathways. These studies highlight the complexity of the regulatory mechanisms that affect recombination fidelity and genome stability.

	ACKNOWLEDGMENTS

We thank Caroline Welz-Voegele for her advice and members of the S.J.R. lab and Gray Crouse for critical reading of this manuscript. This work was supported by National Research Service Award grant no. GM-20753 (to R. M. Spell) and grant no. GM-38464 (to S. Jinks-Robertson) from the National Institutes of Health.

Manuscript received May 23, 2003; Accepted for publication September 2, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALANI, E., R. A. G. REENAN, and R. D. KOLODNER, 1994  Interaction between mismatch repair and genetic recombination in Saccharomyces cerevisiae.. Genetics 137:19-39.[Abstract]

ALTMAN, D. G., 1991 Practical Statistics for Medical Research. Chapman & Hall, New York.

AUSUBEL, F. M., R. BRENT, R. E. KINGSTON, D. D. MOORE, J. G. SEIDMAN et al., 1994 Current Protocols in Molecular Biology. John Wiley & Sons, New York.

BAI, Y. and L. S. SYMINGTON, 1996  A Rad52 homolog is required for RAD51-independent mitotic recombination in Saccharomyces cerevisiae.. Genes Dev. 10:2025-2037.[Abstract/Free Full Text]

BAI, Y., A. P. DAVIS, and L. S. SYMINGTON, 1999  A novel allele of RAD52 that causes severe DNA repair and recombination deficiencies only in the absence of RAD51 or RAD59.. Genetics 153:1117-1130.[Abstract/Free Full Text]

BARTSCH, S., L. E. KANG, and L. S. SYMINGTON, 2000  RAD51 is required for the repair of plasmid double-stranded DNA gaps from either plasmid or chromosomal templates. Mol. Cell. Biol. 20:1194-1205.[Abstract/Free Full Text]

BISHOP, A. J. and R. H. SCHIESTL, 2002  Homologous recombination and its role in carcinogenesis. J. Biomed. Biotechnol. 2:75-85.[Medline]

BOHR, V. A., 2002  Human premature aging syndromes and genomic instability. Mech. Ageing Dev. 123:987-993.[Medline]

CHEN, W. and S. JINKS-ROBERTSON, 1998  Mismatch repair proteins regulate heteroduplex formation during mitotic recombination in yeast. Mol. Cell. Biol. 18:6525-6537.[Abstract/Free Full Text]

CHEN, W. and S. JINKS-ROBERTSON, 1999  The role of the mismatch repair machinery in regulating mitotic and meiotic recombination between diverged sequences in yeast. Genetics 151:1299-1313.[Abstract/Free Full Text]

CROUSE, G. F., 1998 Mismatch repair systems in Saccharomyces cerevisiae, pp. 411–448 in DNA Damage and Repair, Vol. 1: DNA Repair in Prokaryotes and Lower Eukaryotes, edited by J. A. NICKOLOFF and M. F. HOEKSTRA. Humana Press, Totowa, NJ.

DATTA, A., A. ADJIRI, L. NEW, G. F. CROUSE, and S. JINKS-ROBERTSON, 1996  Mitotic crossovers between diverged sequences are regulated by mismatch repair proteins in Saccharomyces cerevisiae.. Mol. Cell. Biol. 16:1085-1093.[Abstract]

DONG, Z. and M. FASULLO, 2003  Multiple recombination pathways for sister chromatid exchange in Saccharomyces cerevisiae: role of RAD1 and the RAD52 epistasis group genes. Nucleic Acids Res. 31:2576-2585.[Abstract/Free Full Text]

EARLEY, M. C. and G. F. CROUSE, 1998  The role of mismatch repair in the prevention of base pair mutations in Saccharomyces cerevisiae.. Proc. Natl. Acad. Sci. USA 95:15487-15491.[Abstract/Free Full Text]

FLORES-ROZAS, H. and R. D. KOLODNER, 1998  The Saccharomyces cerevisiae MLH3 gene functions in MSH3-dependent suppression of frameshift mutations. Proc. Natl. Acad. Sci. USA 95:12404-12409.[Abstract/Free Full Text]

FREEDMAN, J. A. and S. JINKS-ROBERTSON, 2002  Genetic requirements for spontaneous and transcription-stimulated mitotic recombination in Saccharomyces cerevisiae.. Genetics 162:15-27.[Abstract/Free Full Text]

GOLDSTEIN, A. L. and J. H. MCCUSKER, 1999  Three new dominant drug resistance cassettes for gene disruption in Saccharomyces cerevisiae.. Yeast 15:1541-1553.[Medline]

HARFE, B. D. and S. JINKS-ROBERTSON, 2000a  DNA mismatch repair and genetic instability. Annu. Rev. Genet. 34:359-399.[Medline]

HARFE, B. D. and S. JINKS-ROBERTSON, 2000b  Mismatch repair proteins and mitotic genome stability. Mutat. Res. 451:151-167.[Medline]

HARFE, B. D., B. K. MINESINGER, and S. JINKS-ROBERTSON, 2000  Discrete in vivo roles for the MutL homologs Mlh2p and Mlh3p in the removal of frameshift intermediates in budding yeast. Curr. Biol. 10:145-148.[Medline]

HIGGINS, D. R., S. PRAKASH, P. REYNOLDS, and L. PRAKASH, 1983  Molecular cloning and characterization of the RAD1 gene of Saccharomyces cerevisiae. Gene 26:119-126.[Medline]

HO, Y., A. GRUHLER, A. HEILBUT, G. D. BADER, and L. MOORE et al., 2002  Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry. Nature 415:180-183.[Medline]

IRA, G. and J. E. HABER, 2002  Characterization of RAD51-independent break-induced replication that acts preferentially with short homologous sequences. Mol. Cell. Biol. 22:6384-6392.[Abstract/Free Full Text]

JOHNSON, R. E., G. K. KOVVALI, L. PRAKASH, and S. PRAKASH, 1996  Requirement of the yeast MSH3 and MSH6 genes for MSH2-dependent genomic stability. J. Biol. Chem. 271:7285-7288.[Abstract/Free Full Text]

KANG, L. E. and L. S. SYMINGTON, 2000  Aberrant double-strand break repair in rad51 mutants of Saccharomyces cerevisiae.. Mol. Cell. Biol. 20:9162-9172.[Abstract/Free Full Text]

KOLODNER, R. D. and G. T. MARSISCHKY, 1999  Eukaryotic DNA mismatch repair. Curr. Opin. Genet. Dev. 9:89-96.[Medline]

KRAMER, W., B. FARTMANN, and E. C. RINGBECK, 1996  Transcription of mutS and mutL-homologous genes in Saccharomyces cerevisiae during the cell cycle. Mol. Gen. Genet. 252:275-283.[Medline]

LEA, D. E. and C. A. COULSON, 1949  The distribution of the numbers of mutants in bacterial populations. J. Genet. 49:264-285.

MALKOVA, A., E. L. IVANOV, and J. E. HABER, 1996  Double-strand break repair in the absence of RAD51 in yeast: a possible role for break-induced DNA replication. Proc. Natl. Acad. Sci. USA 93:7131-7136.[Abstract/Free Full Text]

MALONE, R. E., T. WARD, S. LIN, and J. WARING, 1990  The RAD50 gene, a member of the double strand break repair epistasis group, is not required for spontaneous mitotic recombination in yeast. Curr. Genet. 18:111-116.[Medline]

MARSISCHKY, G. T., N. FILOSI, M. F. KANE, and R. KOLODNER, 1996  Redundancy of Saccharomyces cerevisiae MSH3 and MSH6 in MSH2-dependent mismatch repair. Genes Dev. 10:407-420.[Abstract/Free Full Text]

MYUNG, K., A. DATTA, C. CHEN, and R. D. KOLODNER, 2001  SGS1, the Saccharomyces cerevisiae homologue of BLM and WRN, suppresses genome instability and homeologous recombination. Nat. Genet. 27:1-4.[Medline]

NICHOLSON, A., M. HENDRIX, S. JINKS-ROBERTSON, and G. F. CROUSE, 2000  Regulation of mitotic homeologous recombination in yeast: functions of mismatch repair and nucleotide excision repair genes. Genetics 154:133-146.[Abstract/Free Full Text]

PAQUES, F. and J. E. HABER, 1999  Multiple pathways of recombination induced by double-strand breaks in Saccharomyces cerevisiae.. Microbiol. Mol. Biol. Rev. 63:349-404.[Abstract/Free Full Text]

PEDRAZZI, G., C. PERRERA, H. BLASER, P. KUSTER, and G. MARRA et al., 2001  Direct association of Bloom's syndrome gene product with the human mismatch repair protein MLH1. Nucleic Acids Res. 29:4378-4386.[Abstract/Free Full Text]

PROLLA, T. A., D.-M. CHRISTIE, and R. M. LISKAY, 1994  Dual requirement in yeast DNA mismatch repair for MLH1 and PMS1, two homologs of the bacterial mutL gene. Mol. Cell. Biol. 14:407-415.[Abstract/Free Full Text]

RATTRAY, A. J. and L. S. SYMINGTON, 1995  Multiple pathways for homologous recombination in Saccharomyces cerevisiae.. Genetics 139:45-56.[Abstract]

RICHARDSON, C. and M. JASIN, 2000  Coupled homologous and nonhomologous repair of a double-strand break preserves genomic integrity in mammalian cells. Mol. Cell. Biol. 20:9068-9075.[Abstract/Free Full Text]

SHERMAN, F., 1991  Getting started with yeast. Methods Enzymol. 194:3-20.[Medline]

SIGNON, L., A. MALKOVA, M. L. NAYLOR, H. KLEIN, and J. E. HABER, 2001  Genetic requirements for RAD51- and RAD54-independent break-induced replication repair of a chromosomal double-strand break. Mol. Cell. Biol. 21:2048-2056.[Abstract/Free Full Text]

SPELL, R. M., and S. JINKS-ROBERTSON, 2004 Determination of mitotic recombination rates by fluctuation analysis in Saccharomyces cerevisiae, pp. 3–12 in Genetic Recombination Protocols, edited by A. S. WALDMAN. Humana Press, Totowa, NJ.

SUGAWARA, N., F. PAQUES, M. COLAIACOVO, and J. E. HABER, 1997  Role of Saccharomyces cerevisiae Msh2 and Msh3 repair proteins in double-strand break-induced recombination. Proc. Natl. Acad. Sci. USA 94:9214-9219.[Abstract/Free Full Text]

SUGAWARA, N., G. IRA, and J. E. HABER, 2000  DNA length dependence of the single-strand annealing pathway and the role of Saccharomyces cerevisiae RAD59 in double-strand break repair. Mol. Cell. Biol. 20:5300-5309.[Abstract/Free Full Text]

SUNG, P., K. M. TRUJILLO, and S. VAN KOMEN, 2000  Recombination factors of Saccharomyces cerevisiae.. Mutat. Res. 451:257-275.[Medline]

SYMINGTON, L. S., 2002  Role of RAD52 epistasis group genes in homologous recombination and double-strand break repair. Microbiol. Mol. Biol. Rev. 66:630-670.[Abstract/Free Full Text]

WACH, A., A. BRACHAT, R. POHLMANN, and P. PHILIPPSEN, 1994  New heterologous modules for classical or PCR-based gene disruption in Saccharomyces cerevisiae.. Yeast 10:1793-1808.[Medline]

WANG, T.-F., N. KLECKNER, and N. HUNTER, 1999  Functional specificity of MutL homologs in yeast: evidence for three Mlh1-based heterocomplexes with distinct roles during meiosis in recombination and mismatch correction. Proc. Natl. Acad. Sci. USA 96:13914-13919.[Abstract/Free Full Text]

WELZ-VOEGELE, C., J. E. STONE, P. T. TRAN, H. M. KEARNEY, and R. M. LISKAY et al., 2002  Alleles of the yeast PMS1 mismatch-repair gene that differentially affect recombination- and replication-related processes. Genetics 162:1131-1145.[Abstract/Free Full Text]

WU, L., S. L. DAVIES, N. C. LEVITT, and I. D. HICKSON, 2001  Potential role for the BLM helicase in recombinational repair via a conserved interaction with RAD51. J. Biol. Chem. 276:19375-19381.[Abstract/Free Full Text]

This article has been cited by other articles:

				C. Welz-Voegele and S. Jinks-Robertson Sequence Divergence Impedes Crossover More Than Noncrossover Events During Mitotic Gap Repair in Yeast Genetics, July 1, 2008; 179(3): 1251 - 1262. [Abstract] [Full Text] [PDF]

				Y. Wu, N. Kantake, T. Sugiyama, and S. C. Kowalczykowski Rad51 Protein Controls Rad52-mediated DNA Annealing J. Biol. Chem., May 23, 2008; 283(21): 14883 - 14892. [Abstract] [Full Text] [PDF]

				J. E. Stone, R. G. Ozbirn, T. D. Petes, and S. Jinks-Robertson Role of Proliferating Cell Nuclear Antigen Interactions in the Mismatch Repair-Dependent Processing of Mitotic and Meiotic Recombination Intermediates in Yeast Genetics, March 1, 2008; 178(3): 1221 - 1236. [Abstract] [Full Text] [PDF]

				Y. W. Kow, G. Bao, J. W. Reeves, S. Jinks-Robertson, and G. F. Crouse Oligonucleotide transformation of yeast reveals mismatch repair complexes to be differentially active on DNA replication strands PNAS, July 3, 2007; 104(27): 11352 - 11357. [Abstract] [Full Text] [PDF]

				R. L. Barnes and R. McCulloch Trypanosoma brucei homologous recombination is dependent on substrate length and homology, though displays a differential dependence on mismatch repair as substrate length decreases Nucleic Acids Res., May 11, 2007; 35(10): 3478 - 3493. [Abstract] [Full Text] [PDF]

				K. H. Schmidt, J. Wu, and R. D. Kolodner Control of Translocations between Highly Diverged Genes by Sgs1, the Saccharomyces cerevisiae Homolog of the Bloom's Syndrome Protein. Mol. Cell. Biol., July 1, 2006; 26(14): 5406 - 5420. [Abstract] [Full Text] [PDF]

				Y. Wu, T. Sugiyama, and S. C. Kowalczykowski DNA Annealing Mediated by Rad52 and Rad59 Proteins J. Biol. Chem., June 2, 2006; 281(22): 15441 - 15449. [Abstract] [Full Text] [PDF]

				R. M. Spell and S. Jinks-Robertson Examination of the Roles of Sgs1 and Srs2 Helicases in the Enforcement of Recombination Fidelity in Saccharomyces cerevisiae Genetics, December 1, 2004; 168(4): 1855 - 1865. [Abstract] [Full Text] [PDF]