The Main Role of Srs2 in DNA Repair Depends on Its Helicase Activity, Rather than on Its Interactions with PCNA or Rad51

ABSTRACT Homologous recombination (HR) is a mechanism that repairs a variety of DNA lesions. Under certain circumstances, however, HR can generate intermediates that can interfere with other cellular processes such as DNA transcription or replication. Cells have therefore developed pathways that abolish undesirable HR intermediates. The Saccharomyces cerevisiae yeast Srs2 helicase has a major role in one of these pathways. Srs2 also works during DNA replication and interacts with the clamp PCNA. The relative importance of Srs2’s helicase activity, Rad51 removal function, and PCNA interaction in genome stability remains unclear. We created a new SRS2 allele [srs2(1-850)] that lacks the whole C terminus, containing the interaction site for Rad51 and PCNA and interactions with many other proteins. Thus, the new allele encodes an Srs2 protein bearing only the activity of the DNA helicase. We find that the interactions of Srs2 with Rad51 and PCNA are dispensable for the main role of Srs2 in the repair of DNA damage in vegetative cells and for proper completion of meiosis. On the other hand, it has been shown that in cells impaired for the DNA damage tolerance (DDT) pathways, Srs2 generates toxic intermediates that lead to DNA damage sensitivity; we show that this negative Srs2 activity requires the C terminus of Srs2. Dissection of the genetic interactions of the srs2(1-850) allele suggest a role for Srs2’s helicase activity in sister chromatid cohesion. Our results also indicate that Srs2’s function becomes more central in diploid cells.

H omologous recombination (HR) is important for maintaining the stability of the genome; it helps repair double-strand breaks (DSBs) and participates in the recovery of damaged replication forks. However, HR mechanisms can generate intermediates that may block replication forks, or nucleoprotein complexes that can lead to cell cycle arrest and even cause cell death in certain genetic backgrounds (1). That is why HR must be tightly regulated to prevent untimely events that could interfere with other DNA replication or repair mechanisms.
The yeast Saccharomyces cerevisiae is an excellent model to isolate and study mutants that shed light on the processes that maintain genome stability (2,3). The Srs2 helicase has a major role in HR regulation; it is generally thought that its role is to suppress HR events at an early stage by dismantling the Rad51-presynaptic filament (4,5). This "antirecombinase" role of Srs2 was first inferred from genetic studies: srs2 mutants show a hyperrecombination phenotype believed to be caused by an inappropriate channeling of the lesions into the homologous recombination pathway (6)(7)(8)(9). The Srs2 protein exhibits single-stranded DNA (ssDNA)-dependent ATPase activity that unwinds DNA with 3=¡5= polarity with a k cat of Ն3,000 min Ϫ1 (10,11), and the Walker A motif is absolutely required for both ATPase and helicase activities (12). It can unwind a variety of substrates, including those containing forks, flaps, D-loops, 3= and 5= single-stranded DNA overhangs, blunt-end double-stranded DNA (dsDNA) substrates, and Holliday junctions (11,13). Biochemical and electron microscopy analysis revealed that Srs2 can efficiently dismantle the presynaptic filament formed by Rad51, an early HR intermediate (4,5). It seems that the helicase activity is not responsible for the dissociation, but rather Srs2's ATP hydrolysis fuels a translocase activity: mutants that cannot bind or hydrolyze ATP fail to disrupt Rad51-presynaptic filaments (12). ATPase mutants show the same sensitivities to genotoxic agents, hyperrecombination phenotype, and genetic interactions as the srs2 deletion mutant (12). Some studies suggest that Srs2 is guided to the Rad51 filament through a physical interaction with Rad51 (5). Rad51 that cannot interact with Srs2 is resistant to Srs2 antirecombinase activity (14,15). Other studies suggested that the direct interaction between Srs2 and Rad51 not only targets Srs2 to the HR intermediates but also triggers ATP hydrolysis within the Rad51 filament, causing Rad51 to dissociate from DNA (16). It seems therefore that Srs2 dismantles Rad51 by ATP-driven motor activities of Srs2 that can dissociate both DNA structures and protein-DNA complexes. Recently, it was also shown that Srs2 is able to disrupt extended D-loops created by the activity of polymerase ␦ (17). Moreover, in vitro experiments have shown that Srs2 can unwind structures that resemble D-loops (recombination intermediates) and that this activity is stimulated by Rad51 bound to dsDNA (18). Srs2 is also needed in the restart of collapsed replication forks together with other members of the Rad6 epistasis group in a process called DNA damage tolerance (DDT). In fact, SRS2 was first identified because mutations in the gene could suppress the DNA damage sensitivity of both rad6 and rad18 mutants (suppressor of RAD six mutant 2), and this suppression requires functional HR (19)(20)(21)(22)(23)(24). The main function of the error-free DDT pathway (which includes the Rad6, Rad18, Rad5, Ubc13, and Mms2 proteins) is to ubiquitinate PCNA at its lysine at position 164. If this step is not accomplished, Srs2 is recruited to the replication forks through its binding to SUMOylated PCNA (mediated by adjacent SIM-SUMO-interacting and PIM-PCNA-interacting motifs, which reside at the very end of the protein), and this recruitment seems to prevent HR (23,25,26). Thus, when the DDT pathway is impaired, Srs2 activity prevents a possible alternative rescue, resulting in sensitivity to DNA damage. Mutations in SRS2 seem to open the path for HR and thus suppress the sensitivity of DDT mutants (23,24).
Although initially Srs2 was considered an inhibitor of HR, later work showed additional roles for Srs2 that favor HR. Srs2 was shown to be required for the repair of DSBs. Cells deleted for SRS2 show low survival when a single DSB is created, and it appears to act during HR repair, possibly by unwinding the invading strand from the D-loop to allow reannealing with the other broken chromosomal arm (27). Accordingly, Srs2 acts in the promotion of synthesis-dependent strand annealing (SDSA) and inhibition of crossover events (28)(29)(30), as well as in additional forms of HR, such as single-strand annealing (SSA), break-induced replication (BIR), as well as in nonhomologous end joining (NHEJ) (31)(32)(33)(34)(35).
In recent years, new roles of Srs2 were identified. Srs2 association with SUMOylated PCNA was shown to limit the DNA synthesis by detaching polymerases ␦ and from PCNA; this function is independent of the interaction with Rad51 (36). Moreover, Srs2 helicase activity can unwind triplet repeat hairpins at the replication fork; this activity is also independent of Rad51 and plays a critical role in maintaining normal replication without expansion or contraction of repeats (37). Another role for Srs2 is in preventing mutations as a result of Top1 topoisomerase activity on misincorporated ribonucleotides. Srs2 can process the nick after Top1 activity and promote resection by enhancing Exo1 activity. Again, this role of Srs2 was shown to be Rad51 independent (38). Recently, in vitro experiments showed that Srs2 can remove replication protein A (RPA) and Rad52-RPA complex from ssDNA, although the in vivo significance of these findings is still unclear (39).
Altogether, Srs2 functions as a multifunctional tool that acts in replication, recombination, and DNA repair. By creating a new Srs2 allele that lacks the whole C terminus [srs2(1-850)], we show here that the DNA helicase domain alone is sufficient to deal with various types of DNA damage, to complete efficient DSB repair, and to promote meiosis. The synthetic lethality (SL) of ⌬srs2 with other deletion mutations is also largely dependent on the DNA helicase activity. The analysis of the SL screen suggests that Srs2 might be involved in sister chromatin cohesion (SCC). Last, we show that the ploidy state of the cell dictates the importance of Srs2's activity, and diploids rely more on the helicase's C terminus in order to maintain genome stability.

RESULTS
The helicase domain of Srs2 is the major player in dealing with DNA damage. Our recent study provided evidence that Srs2 has functions that are independent of its role in the eviction of Rad51 and of its interaction with PCNA. Neither the Rad51 interaction region, PIM (PCNA-interacting motif), nor SIM (SUMO-interacting motif), which allow recruitment of Srs2 to SUMOylated PCNA, are required to deal with DNA damage caused by methyl methanesulfonate (MMS) (34).
We were interested in further investigating the importance of the helicase domain for genome stability. To answer this question, we created a new truncation mutation of Srs2 that lacks 324 amino acids (aa) from the C terminus (out of 1,174 aa). This mutation [srs2(1-850)] lacks all the known interaction sites of Srs2, such as those needed to interact with PCNA, Rad51, Nej1, Mre11, Sgs1, Esc2, Ubc9, Siz1, Siz2, Mus81, Rad5, and Rad18 (23,(40)(41)(42)(43)(44). Figure S1 in the supplemental material shows that the srs2(1-850) strain produces protein, which is expressed from its natural promoter at a slightly higher level than the level produced from the wild type (wt). Importantly, this N-terminal region of Srs2 has been shown to lack the ability to bind Rad51 (15).
We first tested how strains carrying this allele handle DNA damage. Figure 1 shows that, surprisingly, the mutants are as proficient as a wt strain when it comes to handling different kinds of DNA damage (MMS [DNA alkylation], hydroxyurea [deoxynucleoside triphosphate {dNTP} depletion], camptothecin [topoisomerase poison], and zeocin [DSBs]). These results imply that Srs2 deals with DNA damage through its helicase region and that the interactions with other proteins are dispensable for its main DNA repair activity. Δsrs2 mutants are more sensitive to DNA damage as diploid cells than as haploid cells. Consistent with a more central role of HR repair in diploids (34,45), ⌬srs2 diploids are more sensitive to the DSB-forming agent zeocin. We show that, in contrast, a diploid strain homozygous for the srs2(1-850) allele is as proficient as the wt parent for growth in the presence of DNA-damaging agents, with the possible exception of MMS, where a barely detected defect can be seen (Fig. 1).
The activity of Srs2 creates toxic intermediates in strains impaired in the DDT pathway and sensitizes them to DNA damage. Deletion of SRS2 suppresses the DNA damage sensitivity (24). This suppression was shown to be caused by mutations in the C terminus of Srs2 (the SIM and PIM motifs) (23,42,46,47). As expected, the srs2(1-850) allele (also lacking these motifs) suppressed the MMS sensitivity of pol30-K164R, ⌬rad18, and ⌬rad5 mutants that are impaired in the DDT pathway (Fig. 1E). Thus, cells with the srs2(1-850) allele act like the wt when the cells are confronted with external insults to their DNA, but when the DDT pathway is inactivated, it behaves like a mutant with the whole SRS2 gene deleted. The helicase part of Srs2 is important for dealing with DNA damage; however, when there are no modifications on lysine 164 of PCNA, Srs2 exerts its negative effects through its C terminus, probably via its interactions with PCNA.
The C terminus of Srs2 is dispensable for Srs2's role in DNA repair during replication and DSB repair. Truncation of the C terminus of Srs2 does not sensitize the cells to DNA damage. As Srs2 is involved in several repair pathways, we characterized the repair capacity of the srs2(1-850) mutant. First, we measured the ability of cells carrying the allele to carry out homologous recombination (48). Strain MK166 allows the measurement of the rates of ectopic gene conversion (GC) and direct-repeat recombination (DRR) during normal cell division (Fig. S2). Relative to the wt, a strain deleted for SRS2 showed elevated rates of DRR and GC of about 1.5-to 2-fold (49). In contrast, an isogenic strain with the new allele showed levels of both GC and DRR similar to those of the wt ( Fig. 2A).
Next, we tested whether the srs2(1-850) mutant is proficient for the repair of a single DSB. Since diploid srs2(1-850) strains showed some minor sensitivity to MMS, we tested the proficiency of the allele in the repair of an inducible DSB in both haploid and diploid strains.
In the strains used, a single, defined DSB break is created by an inducible HO endonuclease; in the haploid strain, the cells can then repair the damage by an ectopic gene conversion and thus survive and form a colony. Two different diploid strains were used: in the first strain (allelic), the DSB can be repaired by a gene conversion event in which the donor sequence originates at the homologous chromosome. In the second strain (ectopic), the two copies of chromosome V undergo DSBs, and they can be repaired only by recombination with the ectopic donor (Fig. S3). By comparing the number of colonies created when cells are plated on galactose-containing media (continuous HO expression) versus glucose-containing media (no DSB creation), it is possible to calculate the efficiency of repair. Figure 2B shows that wt diploids exhibit an efficiency of repair close to 100% in the presence of an allelic donor. Haploids and diploids that can repair the broken chromosome only by ectopic recombination show about 60% survival. SRS2 is essential for DSB repair: ⌬srs2 strains exhibit very low repair efficiency in haploid and diploid strains; they are defective for both allelic and ectopic recombination. In contrast, the srs2(1-850) mutant did not show any significant difference from the wt in any of the three systems tested (Fig. 2B). Since srs2(1-850) does not seem to affect interchromosomal gene conversion, we tested whether it might have an effect on intrachromosomal recombination. Using strain NA3 (50) (Fig. S4), we measured the abilities of the various strains to repair the DSB by intrachromosomal (Rad51 dependent) gene conversion or by single-strand annealing (SSA), which is Rad51 independent. Again, the efficiency of repair is assessed by comparing the number of cells able to form colonies on galactose-versus glucose-based medium. The wt strain showed a repair efficiency of about 85%, whereas the ⌬srs2 mutant had less than 2% repair (Fig. 2C). These results confirm that SRS2 is also required for intrachromosomal recombination initiated by a DSB. In contrast, the srs2(1-850) mutant exhibited a repair efficiency similar to that of the wt; furthermore, the distribution between intrachromosomal GC and SSA was similar to that of the wt (Fig. 2D). Taken together, our results show that only the helicase activity of SRS2 is needed for the repair of DSBs by all types of recombination tested, whereas the C terminus is dispensable.
The helicase domain of Srs2 is sufficient to undergo proficient meiosis. The Srs2 protein has a pivotal role in meiotic progression (51). ⌬srs2 diploid cells are unable to undergo a proper meiosis and form few asci. Moreover, most of these asci give rise to dead spores. These defects are caused by the need of Srs2 for efficient homologous recombination in meiosis, which is essential for proper chromosomal segregation during the meiotic divisions (51).
To test the srs2(1-850) allele for possible meiotic defects, we subjected wt, ⌬srs2/ ⌬srs2, and srs2(1-850)/srs2(1-850) diploids to meiosis. Diploid cells were allowed to  , in contrast to Δsrs2, acts as a wt in haploids and diploids. (C and D) A strain that measures intrachromosomal recombination and DRR following a single DSB shows that, in contrast to ⌬srs2, the strain with the allele has the same repair efficiency as the wt. Error bars represent 95% confidence intervals. Asterisks represent P values below 0.001. The P value between the values for wt and srs2(1-850) strains was above 0.05 and not statistically different. IGC, interchromosomal gene conversion; EGC, ectopic gene conversion.
Role of Srs2 in DNA Repair ® sporulate for 6 days, and the percentage of cells that completed meiosis to form asci was determined. ⌬srs2 homozygotes formed very few asci, and most of the spores from these asci were unable to form colonies (Fig. 3A). Homozygous srs2(1-850) diploids, in contrast, showed no impairment in meiosis. Tetrad dissection showed that, contrary to what is seen in ⌬srs2/⌬srs2 strains, viability of the srs2(1-850)/srs2  spores was as high as that of the wt spores (Fig. 3B). We conclude that srs2(1-850) does not have meiotic defects, again pointing to Srs2's helicase domain as the main need for Srs2's activity during meiosis.
Genetic interactions of the ⌬srs2 and srs2(1-850) alleles. The data we showed so far provide strong evidence for the hypothesis that the main role of Srs2 in genome stability is carried out by the helicase activity present at its N terminus and does not require interactions with additional repair proteins or with PCNA. Systematic screens (52)(53)(54) have shown that there is a large number of genes that, when deleted, show a dependence on Srs2 function for survival. For most of the genes, it is unknown why, if mutated, they are synthetic sick or lethal with ⌬srs2. We saw an opportunity to identify the regions of Srs2 that are responsible for the impaired growth in these genetic backgrounds. We chose genes involved in DNA metabolism that, when deleted, show the most severe negative interactions with the ⌬srs2 allele. Diploid strains heterozygous for various deletion allele and for srs2  or ⌬srs2 were subjected to meiosis, and tetrads were dissected. The viability and growth rate of double mutant spores defective for each of the chosen genes and carrying the srs2(1-850) allele was compared to that of the double mutant with ⌬srs2. The results (Table 1) are grouped according to their phenotypes. We could distinguish three categories.
• Ctf18. Ctf18 is part of a replication factor C (RFC)-like complex that moves with the replication fork and participates in sister chromatid cohesion and checkpoint response (55).
• Csm3. Csm3, together with Tof1, is located at the replication fork, where it contributes to fork stability by inhibiting fork rotation caused by topological stress of unwinding the dsDNA ahead of the polymerases (56).
• Mrc1. Mrc1 is involved in replication checkpoint activation; it mediates phosphorylation of Rad53 by Mec1 and is active through S phase. It is also required for proper DNA replication (57,58). The synthetic lethality between ⌬mrc1 and ⌬srs2 was found to be independent of the checkpoint activity of MRC1: the mrc1-AQ allele, defective in the checkpoint function of MRC1, does not show synthetic sickness or lethality with ⌬srs2 (53). Site-specific mutations of MRC1 were found to impair specifically its role at the replication fork but not in DNA damage signaling (59). We crossed one such MRC1 mutant (mrc1-C14) with ⌬srs2 and observed synthetic lethality (SL). However, when combined with mrc-C14, the srs2(1-850) allele led to normal cell growth (Table 1A). We thus conclude that the DNA replication functions of Mrc1 and the helicase activity of Srs2 are synthetic lethal, and not the other functions of these proteins.
• Ctf4 and Chl1. The Ctf4 and Chl1 proteins are involved in sister chromatid cohesion and genome integrity and interact with Ctf18 (63, 64).
• Rrm3. Rrm3 is another helicase that assists in the replication of regions of the genome with secondary structures or bound proteins (65,66). • Hst3. Hst3 is the histone deacetylase that removes the acetyl groups from H3K56 (69). Surprisingly, both lack of acetylation and too much acetylation have a similar synthetic sick phenotype in the absence of Srs2 helicase activity.
• Slx5. Slx5 forms with Slx8 a SUMO-targeted ubiquitin (Ub) ligase (STUbL) complex that attaches ubiquitin to poly-SUMOylated proteins, supposedly in order to send them for degradation during replication and DNA repair (70).
• Vid22. Vid22 acts as a chromatin remodeler and removes nucleosome from DNA damage repair site. This in turn allows the recruitment of the MRX complex, which initiate the repair (71). 3. A third group of mutants were synthetic lethal with ⌬srs2 and showed synthetic sickness with the srs2(1-850) allele (Table 1C and Fig. 4C). This implies that both Srs2's helicase activity and the C terminus of Srs2 are required for proper cell activity in these genetic backgrounds.
• Rad54. Rad54 is a chromatin remodeling factor that is needed for DSB repair; it participates in D-loop formation, extension, and resolution (77).
• Sgs1, Rmi1, and Top3. The Sgs1, Rmi1, and Top3 proteins form a complex required for many different aspects of genome stability and DNA repair, including DNA resection, the resolution of Holiday junction intermediates, and the relaxation of supercoiled DNA (78).
Taken together, the results point to the fact that in the absence of Srs2 function, histone deposition, checkpoint activation, and sister chromatid cohesion become impaired (see Discussion).
Ploidy dictates Srs2 activity. Previous results have shown that in certain genetic backgrounds (for example, in the absence of the Elg1 RFC-like subunit), deletion of SRS2 has little effect in haploids, but diploids fail to form colonies (72). We therefore tested the double mutants that showed normal growth as haploids for their phenotype as diploids. Double mutants were mated, and the homozygous diploid zygotes (Ͼ24 per strain) were manipulated to predetermined locations on rich medium plates. All mutants that showed synthetic sickness with ⌬srs2 were not viable as diploids and generated no, or only a few, viable colonies. Thus, all the genetic interactions of ⌬srs2 are stronger in diploids than in haploids. In contrast, most of the combinations of the srs2(1-850) allele and various deletions that showed normal growth as haploids were able to form viable diploids (although in some cases only 2/3 of the zygotes grew). The striking exceptions to this rule were diploids homozygous for the srs2(1-850) allele and for ⌬rad54, ⌬pol32, Δsgs1, or ⌬top3. Whereas these double mutant strains grew slowly as haploids, they failed to form diploid colonies, similarly to the double mutants with ⌬srs2. This illustrates that in certain situations, the C terminus of Srs2 becomes important in diploids (see Discussion).
In conclusion, Srs2 has various roles in DNA replication and chromosome maintenance, which depend on the genome state of the cell. In haploids and diploids, the DNA helicase activity of SRS2 is required for supporting proper DNA replication and chromosome segregation. However, in some genetic backgrounds (such as in the absence of Pol32, Rad54, or Sgs1), the C terminus is also important for reliable DNA replication in diploids.

DISCUSSION
The srs2 mutant was originally isolated as a suppressor of the DNA damage sensitivity of mutants with an impaired DDT pathway; genetic evidence suggested that this suppression depends on Rad51 (6,8,24). In vitro experiments showed that Srs2 is able to disrupt Rad51 nucleofilaments (4,5) and inhibit recombination at D-loops and replication forks by binding to PCNA (4,5,23). After these convincing biochemical experiments, it was widely assumed that the role of Srs2 in the maintenance of genomic integrity is to inhibit recombination by removing Rad51 from the DNA. Thus, all phenotypes of ⌬srs2 were interpreted in light of this activity.
Although many times described as an "antirecombinase," Srs2 is essential for DSB repair by HR (27); thus, its activity is both pro-and antirecombinational (44). Moreover, even Srs2 alleles that lack the region required for interactions with Rad51 or PCNA (or with any of the proven Srs2 interactors) are still proficient in promoting synthesisdependent strand annealing (SDSA) over crossover resolution (79) and perfectly complement the sensitivity of ⌬srs2 mutants to DNA damage (34,37,38,80 this work). The new srs2 allele [srs2(1-850)], which lacks the entire C terminus and has only the DNA helicase domain (15), is unable to interact with Rad51 (15) or with any of the known partners of Srs2 (PCNA, Rad51, Nej1, Mre11, Sgs1, Esc2, Ubc9, Siz1, Siz2, Mus81, Rad5, and Rad18 [23,[40][41][42][43][44]). We showed that in vivo, the helicase of Srs2 was enough to fully deal with various DNA-damaging agents and with HO-induced DSB in both haploids and diploids. Mitotic recombination, meiosis progression, and spore survival were also unaffected. Only when the DDT pathway was impaired was the helicase domain insufficient to enable viability. Thus, binding to PCNA through the C-terminal PIM and SIM motifs becomes essential in the absence of the PCNA ubiquitination that allows DDT pathways to work. Our work thus defines two separate domains of the Srs2 protein with different biological relevance.
Srs2 supports sister chromatid cohesion. Srs2 plays an important role during DNA replication and chromosome segregation, as evidenced by strong negative genetic interactions with mutants defective in these processes. Our analysis showed that the srs2(1-850) allele is less affected than the ⌬srs2 allele to inactivation of additional DNA processing functions. This implies that the helicase activity of Srs2 is sufficient for normal growth in most of the mutant backgrounds.
Interestingly, ⌬srs2 and ⌬mre11 synthetic lethality was not dependent on the nuclease activity of the MRX complex (which is necessary for MRX's role in end resection) or on active HR (87,88). These results suggest that Srs2 and Mre11 are required for proper DNA replication, but not in their classical role of repairing the DNA damage during replication. Thus, their alternative role in DNA replication could also be in SCC, as the MRX complex has been shown to affect this process (89).
Asf1, Rtt109, Rtt107, Hst3, and Mms22 also have a role in SCC, as histone acetylation metabolism was found to act in the regulation of SCC (90)(91)(92). It seems that proper regulation of histone H3 acetylation is important for chromosome cohesion and segregation.
Two nonessential pathways were proposed to promote SCC. The first pathway is composed of Tof1, Csm3, Ctf4, and Chl1, and the second pathway is composed of Mrc1, the Ctf18 RFC-like complex, and the Sgs1-Top3-Rmi1 complex (82,93). Srs2 does not seem to belong specifically to one of the nonessential SCC pathways. Rather, it has a supporting role for the two SCC pathways. This is evident also by the supporting role of Srs2 in histone acetylation metabolism during SCC. The role of helicases in SCC is unknown; it was suggested that helicases might prepare the DNA for targeting of new cohesin rings by removing old cohesin units left on the DNA and by stimulating the loading of new cohesins during replication (94,95). These functions could be executed by Srs2's DNA helicase and translocase capabilities.
The C terminus of Srs2 is required for specific functions during DNA replication to promote genome integrity. In contrast to the previous lack of synthetic phenotypes, the srs2(1-850) allele was synthetic sick when combined with ⌬sgs1, ⌬pol32, or ⌬rad54.
Both Pol32 and Srs2 are implicated in BIR (76). BIR is divided into two pathways: a Rad51-dependent branch and a Rad51-independent branch (96). In the absence of POL32, the C terminus of Srs2 is partly required for cell viability. The C terminus contains the Rad51-interacting motif, a fact that may implicate Srs2 in the Rad51-dependent BIR. Alternatively, it may be the interaction of Srs2 to PCNA that is required. When the Srs2 pathway is disrupted, cells become completely dependent on the Pol32-mediated repair pathway (97).
The synthetic lethality between ⌬srs2 and ⌬rad54 is more complex. It has been proposed that Srs2 and Rad54 actually act in the same pathway, and the SL interaction is due to the generation of toxic intermediates that are trapped and making the cells unable to proceed with the repair without Rad54, but the generated intermediates also cannot recede to an alternative repair pathway due to lack of Srs2 antirecombinase activity (98). This is consistent with our results showing that the C terminus (Rad51 interaction region) of Srs2 is required for proper DNA replication in ⌬srs2 ⌬rad54 cells. A similar explanation could also be applied to the synthetic sickness of the srs2(1-850) allele in the absence of a functional Sgs1 helicase. Sgs1's activity affects many stages of the HR process, from resection to resolution, as well as having a role in SCC (99).
Ploidy regulates a wider range of Srs2's activities. All of the double mutants with ⌬srs2 that are synthetic sick as haploid cells become essential in diploid cells. The DNA helicase of Srs2 is, however, sufficient to suppress the SL phenotype in diploids. This implies that the DNA helicase part of Srs2 becomes more central in diploids and is crucial for cell viability in the absence of other factors (72). In certain genetic backgrounds, the C terminus of Srs2 also becomes important in diploids. Mutants that are synthetic sick with srs2(1-850) as haploids (⌬rad54, ⌬pol32, and ⌬sgs1) are inviable as diploids. Ploidy seems to affect the fundamental regulation of the pathways involved in dealing with DNA replication stress. Haploids rely more on the DDT pathways, whereas diploids seem to rely more on HR (34,45). It seems that PCNA and its modifications affect the regulation of DNA repair during replication, depending on the ploidy of the cell. The fact that diploids rely more on HR to deal with DNA damage is consistent with our finding that Srs2 C terminus and probably its antirecombinase activity is important in diploids, when other factors of HR are unavailable.
In conclusion, we show that the helicase activity of Srs2, and not its physical interactions with Rad51 or PCNA, plays a major role in genome maintenance. PCNA interaction becomes important only in the absence of the DDT pathway. We also show that Srs2 plays a role in SCC and that its helicase activity becomes more important in diploid cells.

MATERIALS AND METHODS
Yeast strains. Saccharomyces cerevisiae strains used in this study are listed in Table 2. Unless otherwise stated, strains used were of one of these backgrounds.  Determination of recombination rates. Strain MK166 carries substrates that allow easy scoring of direct-repeat recombination (DRR) (His ϩ colonies) and ectopic gene conversion (GC) (Lys ϩ colonies). Colonies isolated from plates with various concentrations of methyl methanesulfonate (MMS) were subjected to fluctuation tests, and the rates were calculated as described previously (48). The MMS concentrations used were low and did not cause cell death in the wild-type (wt) strain.