TMCnet News

Shared Genetic Pathways Contribute to the Tolerance of Endogenous and Low-Dose Exogenous DNA Damage in Yeast [Genetics]
[October 30, 2014]

Shared Genetic Pathways Contribute to the Tolerance of Endogenous and Low-Dose Exogenous DNA Damage in Yeast [Genetics]


(Genetics Via Acquire Media NewsEdge) ABSTRACT DNA damage that escapes repair and blocks replicative DNA polymerases is tolerated by bypass mechanisms that fall into two general categories: error-free template switching and error-prone translesion synthesis. Prior studies of DNA damage responses in Saccharomyces cerevisiae have demonstrated that repair mechanisms are critical for survival when a single, high dose of DNA damage is delivered, while bypass/tolerance mechanisms are more important for survival when the damage level is low and continuous (acute and chronic damage, respectively). In the current study, epistatic interactions between DNA-damage tolerance genes were examined and compared when haploid yeast cells were exposed to either chronic ultraviolet light or chronic methyl methanesulfonate. Results demonstrate that genes assigned to error-free and error-prone bypass pathways similarly promote survival in the presence of each type of chronic damage. In addition to using defined sources of chronic damage, rates of spontaneous mutations generated by the Pol ? translesion synthesis DNA polymerase (complex insertions in a frameshift-reversion assay) were used to infer epistatic interactions between the same genes. Similar epistatic interactions were observed in analyses of spontaneous mutation rates, suggesting that chronic DNA-damage responses accurately reflect those used to tolerate spontaneous lesions. These results have important implications when considering what constitutes a safe and acceptable level of exogenous DNA damage.



DAMAGE to cellular DNA that results from normal met- abolic processes is classified as spontaneous, while that resulting from exogenous physical or chemical agents is con- sidered induced. Spontaneous damage occurs at low, contin- uous levels and hence is chronic in nature. Though induced damage is typically delivered in a single, acute dose, it can also be chronic, with examples including daily exposure to solar radiation or cigarette smoke. Spontaneous and induced damage are dealt with by two general, evolutionarily con- served mechanisms: one that permanently removes the dam- age and one that temporarily bypasses the damage, allowing it to be tolerated (for a general overview, see Ciccia and Elledge 2010). Most damage removal occurs via conserved excision repair pathways, but a small number of damages can be directly reversed enzymatically (e.g., thymine-thymine dimer reversal by photolyase). Damage removal by excision repair generates a single-strand gap that is filled using the complementary, undamaged strand as a template and hence is a high-fidelity process. Such repair only operates, how- ever, when a relevant lesion is present in double-strand DNA. Damage that escapes repair and is encountered during replication can block the progress of DNA synthesis, leading to the formation of a single-strand gap opposite the lesion, replication fork collapse, cell-cycle arrest, and/or death. DNA-damage bypass/tolerance pathways thus become criti- cally important during S and G2 phases, as these allow completion of replication and timely cell-cycle progression. Following bypass, the damage is once again present in du- plex DNA and has another opportunity to be permanently removed via excision repair.

There are several distinct mechanisms of lesion bypass, and these are collectively referred to as DNA damage tolerance (DDT) pathways (reviewed by Friedberg 2005; Sale 2012). DDT can be subdivided into (1) error-free tem- plate switching that involves copying information from the undamaged sister chromatid and (2) error-prone bypass that is effected by specialized translesion synthesis (TLS) DNA polymerases. Though the exact nature of template switching remains elusive, current data suggest that it in- volves homologous recombination (HR) functions (Vanoli et al. 2010). TLS polymerases directly replicate over a tem- plate lesion, but do so with an elevated risk of introducing mutations; an incorrect nucleotide can be inserted either directly opposite the lesion or during coincident copying of adjacent, undamaged DNA. Whether insertion opposite the lesion is error-free or error-prone depends on the nature of the lesion and the particular TLS polymerase involved (reviewed by Waters et al. 2009).


Saccharomyces cerevisiae has served as a model for elucidat- ing the genetic control and basic mechanisms of DDT in eu- karyotes (reviewed by Boiteux and Jinks-Robertson 2013). Central to regulating DDT are post-translational modifications to PCNA, the sliding clamp that promotes processivity of rep- licative DNA polymerases and serves as a landing pad for nu- merous proteins involved in DNA metabolism (reviewed by Moldovan et al. 2007). Monoubiquitination (mono-Ub) of PCNA by the Rad6-Rad18 complex promotes TLS (Hoege et al. 2002; Stelter and Ulrich 2003) by one of three yeast TLS polymerases: Pol zeta (z), Pol eta (h), or Rev1.Polz works in concert with Rev1 and its activity generates .50% of spon- taneous and 90% of induced mutations (Lawrence 1994). Whereas mono-Ub of PCNA promotes TLS, extension of the single Ub moiety into a polyubiquitin (poly-Ub) chain by Rad5 and Mms2-Ubc13 promotes template switching (Hoege et al. 2002; Stelter and Ulrich 2003). Finally, sumoylation of PCNA by Siz1 promotes interaction with the Srs2 helicase (Pfander et al. 2005), which acts as an antirecombinase and prevents accumulation of inappropriate and potentially toxic HR inter- mediates (Krejci et al. 2003; Veaute et al. 2003). Defining the precise relationship between HR and template switching has been difficult, with members of the RAD52 epistasis group of HR genes being required for template switching (Zhang and Lawrence 2005; Gangavarapu et al. 2007; Vanoli et al. 2010), as well as for bypass via more canonical HR when template switching and/or TLS fails. Additionally, whether DDT nor- mally occurs directly at a stalled replication fork or as a gap- filling process after fork passage, and which mechanism of DDT normally predominates, have been controversial. Though it is possible to limit expression of Rad18 to specificpoints in the cell cycle and thereby uncouple DDT from replication (Daigaku et al. 2010; Karras and Jentsch 2010), recent analyses suggest that template switching and TLS primarily occur during Sand G2,respectively(Huang et al. 2013). The strong mutator phenotype observed when error-free DDT is eliminated sug- gests that TLS is normally the minor bypass/tolerance path- way (Broomfield et al. 1998; Minesinger and Jinks-Robertson 2005).

Ultraviolet (UV) light is a common mutagen used to define DNA-damage responses in yeast. Though acute doses are typically employed, recent studies have shown that cellular and genetic responses to chronic, low levels of UV damage are distinctly different from those observed following acute expo- sure (Hishida et al. 2009). This difference was presaged by early genetic studies demonstrating a decrease in induced, but an increase in spontaneous, mutagenesis in a rad6 mutant (Lawrence 1994). Subsequent work demonstrated that RAD5 is required for a separate pathway of spontaneous mutagenesis that can be detected in the absence of RAD6/RAD18 (Liefshitz et al. 1998; Cejka et al. 2001; Minesinger and Jinks-Robertson 2005). In terms of survival following exposure to acute UV, early studies demonstrated that the nucleotide excision repair (NER) pathway is much more important for survival than is the RAD6/RAD18 DDT pathway (Haynes and Kunz 1981). This hierarchy reverses, however, when cells are continuously ex- posed to a chronic, low dose of UV, with RAD18-mediated DDT being much more important for survival than NER (Hishida et al. 2009). In the absence of RAD18-dependent DDT, chronic damage was associated with the accumulation of single-strand gaps during replication and a subsequent G2 arrest. The critical role of NER for survival following an acute UV dose likely reflects the triggering of a G1 or G2 checkpoint by NER- generatedgaps,whichallowssufficient time to remove what would otherwise be a lethal load of damage during S or M phase, respectively (Novarina et al. 2011). In addition to ge- netic differences in response to acute vs. chronic UV exposure, dose-dependent differences are observed when yeast are con- tinuously exposed to methyl methanesulfonate (MMS) (Huang et al. 2013). Based on genetic analyses, the transition from chronic to acute MMS exposure occurs at a drug concentration of ^0.001%. Finally, NER is more important than HR for surviving an acute dose of formaldehyde, which causes DNA- protein crosslinks, while HR is more important during chronic formaldehyde exposure (De Graaf et al. 2009). A general framework of lesion repair/bypass that incorporates these observations is presented in Figure 1.

In the current study, the survival of yeast DDT mutants during chronic UV or MMS exposure was compared, and we find similar relationships and hierarchies between the genes examined. Importantly, analyses reveal a previously unappre- ciated role of TLS during the bypass of chronic UV. In addition to examining genetic responses to chronic damage, a Pol z- dependent mutation signature was used to examine the genetic control of error-free and error-prone DDT pathways that deal with spontaneous, replication-blocking lesions. These analyses reveal striking parallels between responses to spontaneous and chronic DNA damage in the mutagenesis and survival assays, respectively, indicating that chronic exposure to defined DNA- damaging agents provides an accurate and physiologically relevant reflection of how diverse types of endogenous DNA damage are tolerated.

Materials and Methods Strain construction All yeast strains were derived from strain SJR1467 (MATa lys2DA746,NR ade2-101oc his3D200 ura3DNco (Lehner et al. 2012). Mutant strains were constructed by one-step gene disruption using PCR-generated cassettes containing a select- able marker. A complete list of strains in given in Table 1.

Mutation rates and spectra Mutation rates were determined using at least 16 cultures, which were composed of at least 8 cultures from each of two independent strain isolates. Cultures were grown to satura- tion at 30^ in nonselective YEPGE medium (1% yeast ex- tract, 2% Bacto-peptone, 2% glycerol, 2% ethanol, and 250 mg/liter adenine). Appropriate dilutions were plated onto YEPD medium (YEP plus 2% dextrose) to determine total cell number and onto lysine-deficient synthetic glucose me- dium to select Lys+ revertants. Mutation rates were calcu- lated by the Lea-Coulson method of the median using the FALCOR application (http://www.mitochondria.org/protocols/ FALCOR.html).Themutationrateofaspecificmutationtype wascalculatedbymultiplyingitsproportioninthecorrespond- ing spectrum by the total Lys+ reversion rate. Distributions of mutation frequencies between genotypes were compared using the nonparametric Mann-Whitney-Wilcoxon test in R (func- tion: "pairwise.wilcox.test").

To generate mutation spectra, DNA was extracted from purified Lys+ colonies isolated from independent cultures (http://jinks-robertsonlab.duhs.duke.edu/protocols/yeast_ prep.html). The frameshift reversion window of the LYS2 gene was amplified by PCR and sequenced by the Duke Uni- versity DNA Analysis Facility (Durham, NC) using primer 59- GTAACCGGTGACGATGAT.

Chronic, low-dose mutagen treatments Cells were grown to near saturation for 2 days in liquid YEPGE medium with shaking at 30^. For chronic UV experi- ments, 10-fold serial dilutions were spotted on YEPD plates; plates were covered with thin sheets of vinyl (cut from Sta- ples heavy duty sheet protectors) and exposed to a chronic dose of UV (254-nm germicidal lamp) for 2 days at 30^. The UV dose was attenuated using wire mesh placed between the light source and plates. The experimental UV doses used were below the level of detection of a UV meter (UVX-25, Ultraviolet Products), but the 33 (highest) dose generated results phenotypically similar to those in other studies (Hishida et al. 2009, 2010). Control plates were treated the same, but covered with aluminum foil to prevent UV exposure. For MMS experiments, 10-fold serial dilutions were spotted on YEPD or YEPD + MMS plates and grown for 2 days at 30^. Fresh YEPD + MMS plates were prepared 18-24 hr before use.

Results The relative importance of NER and DDT following acute vs. chronic UV exposure was confirmed in the haploid strain background used here. As shown in Figure 2A, eliminating NER by deleting RAD14 resulted in much greater sensitivity to acute UV (5 J/m2) than did deleting RAD18, while a rad18 D strain was much more sensitive to chronic UV than was a rad14D strain. Interactions between and hierarchies of genes specifically involved in DDT were then examined by growingmutantsinthepresenceofachronic,lowdoseof UV or MMS (CLUV and CLMMS, respectively) that had no effect on growth of the wild-type (WT) parent strain. In the case of CLMMS, serial dilutions of cells were plated on rich medium containing 0.00025-0.001% MMS. For CLUV experi- ments, serial dilutions of cells spottedontorichmediumwere continuously exposed to a low level of UV light, which was arbitrarily defined as a 13-33 dose. The use of variable doses of UV or MMS allowed a dynamic range of growth inhibition to be monitored, allowing a more accurate assessment of possible genetic interactions. Genes relevant to the current study, along with a brief description of encoded protein function, are listed in Table 2.

Both subpathways of Rad18-dependent DDT are important during exposure to chronic damage The DDT pathway regulated by the Rad6-Rad18 complex is composed of two subpathways: an error-prone pathway me- diated by TLS polymerases and an error-free pathway that requires Rad5 and the Ubc13-Mms2 complex (reviewed by Boiteux and Jinks-Robertson 2013). The importance of er- ror-free DDT for survival during CLUV treatment was pre- viously inferred through analyses of rad18D and rad5D single mutants; loss of TLS alone (rev1D rev3D rad30D triple mutant) did not affect survival (Hishida et al. 2009). As observed in acute damage studies, deletion of RAD18 con- ferred greater sensitivity to CLUV than did RAD5 deletion, consistent with the role of Rad5 only in the error-free pathway of Rad18-regulated DDT. Because spontaneous-mutagenesis studies have implicated RAD5 in RAD18-independent DDT as well (Liefshitz et al. 1998; Cejka et al. 2001; Minesinger and Jinks-Robertson 2005), however, we compared the CLUV sensitivity of rad18D and rad5D single mutants to that of a rad18 D rad5 D double mutant. If Rad5 acts only in the Rad18 error-free bypass pathway, then the double mutant should be no more sensitive than the most sensitive single mutant. As expected, a rad18D mutant was much more sen- sitive to 13 CLUV than was a rad5D mutant. The combined effect of the deletions was additive/synergistic rather than epistatic, however, demonstrating that Rad5 has a Rad18- independent function when cells are exposed to CLUV (Figure 2B). A similar relationship was observed when the single- and double-mutant strains were grown in the presence of CLMMS.

In previous analyses, rad5D and rad18D mutants were indistinguishable in terms of triggering a checkpoint- dependent G2 arrest, suggesting that it is predominantly error-free DDT that promotes survival (Hishida et al. 2009). Because our data indicated that using RAD5 to specifically define error-free DDT can be potentially misleading, we reex- amined the importance of the RAD18-dependent subpath- ways for tolerating chronic damage. This was done by deleting MMS2, which, like RAD5, is required during RAD18-dependent error-free bypass, and REV3, which enc- odes the catalytic subunit of Pol z and is required for most TLS. Although growth of neither the mms2D nor rev3D sin- gle mutant was impaired in the presence of CLUV or CLMMS, the mms2D rev3D double mutant was very sensitive to each, exhibiting sensitivity similar to the rad18D mutant (Figure 2C). We conclude that RAD18-dependent error- prone as well as error-free DDT is important for surviving multiple types of chronic DNA damage and that these sub- pathways are functionally redundant.

Homologous recombination is an alternative to error-prone DDT An earlier study reported that deletion of RAD52 alone, which is required for HR, did not confer CLUV sensitivity; in the absence of RAD18 or RAD5, however, RAD52 pro- moted survival (Hishida et al. 2009). Use of RAD5 as a proxy for error-free DDT in these studies prompted us to reex- amine the relationship between HR and the two RAD18- dependent bypass pathways. As shown in Figure 2D, a rad52D rev3D double mutant was very sensitive to CLUV or CLMMS, whereas neither single mutant exhibited sensitivity. By con- trast, the rad52D mms2D double mutant, like the single mutants, had little sensitivity to either CLUV or CLMMS. These results suggest that Rad52-mediated HR primarily provides an alternative to Pol z-mediated damage tolerance and are consistent with a role for Rad52 in Mms2-mediated error-free DDT. With regard to spontaneous lesion bypass, it should be noted that the rad52D strain exhibited a slight growth defect in the absence of exogenous damage, and this impairment was further exacerbated by deletion of REV3, but not by loss of MMS2.

MPH1 and MMS2 regulate separate pathways of chronic DNA damage bypass Prior analyses using CLMMS implicated MPH1 and MMS2 in separate pathways of replication-associated lesion tolerance, each of which involves formation of recombination inter- mediates (Choi et al. 2010). At the levels of chronic MMS and UV damage used here, neither the single mutants nor the mms2D mph1D double exhibited growth inhibition (Fig- ure 3A). In a rev3D background, however, a subtle effect on growth was detected when MPH1 was deleted (Figure 3A), and the strong sensitivity of the mms2D rev3D double mu- tant was further enhanced by loss of MPH1 (Figure 3B). Together, these data indicate that MMS2 and MPH1 regulate separate, Pol z-independent pathways for tolerating CLUV- as well as CLMMS-induced damage. The much greater sensi- tivity of the mms2D rev3D than the mph1D rev3D double mutant to chronic damage indicates that the MMS2-dependent pathway is more important than the MPH1-dependent path- way for tolerating the corresponding lesions.

The precise mechanisms of error-free DDT during replica- tion have yet to be determined, but generally are assumed to involve a Rad52-mediated template switch to the undamaged sister chromatid (Branzei et al. 2008).Consistentwitharole for HR in both MPH1-andMMS2-mediated alternatives to Pol z-dependent bypass, the sensitivity of the rad52D rev3D dou- ble mutant was similar to that of the mms2D mph1D rev3D triple mutant (Figure 3B). In addition, the rad52D mms2D mph1D triple mutant, like the rad52D mms2D double mutant, exhibited no sensitivity to chronic damage (Figure 3C).

RAD5 and RAD18 independently regulate Pol z -dependent bypass of chronic damage Data presented above indicate that RAD5 has a RAD18- independent role in tolerating chronic damage (Figure 2B). To investigate the roles of RAD5 and RAD18 during the error- prone bypass of chronic damage, we made use of a rad52D background, where Pol z plays a critical role in survival. The relationship between rad18D and rad5D observed in the ab- sence of RAD52 was the same as in its presence; deletion of RAD18 conferred greater CLUV and CLMMS sensitivity than loss of RAD5, and the rad52D rad18D rad5D triple mutant was more sensitive than was the rad52D rad18D double mu- tant (Figure 4A). The rad52D rad18D rad5D rev3D quadruple mutant was no more sensitive than the rad52D rad18D rad5D triple mutant, indicating that Pol z-dependent bypass was eliminated upon deletion of RAD18 and RAD5.

Our expectation was that the rad52D rev3D double mu- tant would be similar to the rad52D rad18D rad5D triple mutant in terms of chronic damage sensitivity, but it clearly was less sensitive (Figure 4A). Because RAD18 regulates lesion bypass by Pol h as well as Pol z (McDonald et al. 1997), the gene encoding Pol h (RAD30) was deleted from appropriate mutant backgrounds. A rad52D rev3D rad30D triple mutant exhibited CLUV and CLMMS sensitivity similar to that of the rad52D rad18D rad5D triple mutant and greater than that of the rad52D rev3D double mutant (Fig- ure 4B). In contrast to REV3 deletion, however, loss of only RAD30 did not enhance the sensitivity of a rad52D mutant (Figure 4C), indicating that Pol z is more important than Pol h for tolerating potentially toxic lesions generated by CLUV or CLMMS.

Reversion of the lys2DA746,NR allele as an assay for spontaneous lesion bypass The lys2DA746 allele contains a 1-bp deletion and reverts by compensatory, net +1 frameshift mutations that occur in an ^150-bp reversion window defined by stop codons in the alternative reading frames (Harfe and Jinks-Robertson 1999). In a WT, repair-proficient background, ^80% of revertants contain a simple 1-bp insertion and almost all insertions are within homopolymer runs .3N. In a NER- defective background, however, complex events in which the selected 1-bp insertion is accompanied by one or more nearby base substitutions are prominent (Harfe and Jinks-Robertson 2000). Because complex insertions (cins) are proportionally elevated in the absence of NER and are completely depen- dent on the presence of Pol z, these events are assumed to reflect the Pol z-dependent bypass of spontaneous damage normally removed by NER (Harfe and Jinks-Robertson 2000). We thus have used the error-prone cins signature produced in the lys2DA746 assay to dissect the genetic con- trol of spontaneous lesion bypass, which is similarly regu- lated in the presence or absence of NER (Minesinger and Jinks-Robertson 2005).

The lys2DA746 assay was made a more sensitive detector of events other than simple 1-bp insertions by deleting all runs .3N from the reversion window (Lehner et al. 2012). The resulting "no run" lys2DA746,NR allele was used in the current study and all analyses were done in an NER-proficient background. The spectrum of lys2DA746,NR reversion eventsobtainedinWTisshowninFigure5A.Twospecific types of events are relevant: cins, which are Pol z dependent and account for ^5% of compensatory frameshift mutations (cins are shaded), and large deletions (dels), which have endpoints in 7- or 10-bp direct repeats and comprise ^25% of WT events. In the analyses that follow, the frequency distributions of Lys+ revertants obtained in different genet- ic backgrounds were compared using the Mann-Whitney- Wilcoxon test. To compare rates of cins or large dels between strains, a frequency distribution of the relevant event was obtained by multiplying the Lys+ frequency in individual cul- tures by the cins/dels proportion in the corresponding spec- trum. The resulting frequencies were then compared by the Mann-Whitney-Wilcoxon test.

MMS2 and MPH1 reside in separate, RAD52-dependent pathways that counteract Pol z -dependent mutagenesis The CLUV and CLMMS data presented above demonstrate that Rad52-dependent HR processes, which are considered to be error-free, provide an alternative to Pol z-dependent DDT (Figure 2D). In accordance with these observations, the rate of lys2DA746,NR reversion was elevated ^4-fold in a rad52 D background (Table 3). This increase was accom- panied by a strong enrichment for cins in the corresponding spectrum, where they accounted for ^50% of reversion events (Figure 5B). The rate of cins thus increased ^30-fold upon deletion of RAD52, indicating that .95% of spontane- ous DDT is normally error-free and HR dependent. When REV3 was additionally deleted in the rad52D background, the reversion rate was reduced to that of WT, and there were no cins detected among 78 Lys+ revertants sequenced (Table 3 and Figure 5C).

The chronic MMS and UV survival data are consistent with the interpretation that Mms2 and Mph1 are each rele- vant during HR-mediated template switching (Figure 3). To explore whether these proteins have similar roles during spontaneous damage tolerance, the corresponding genes were deleted singly or together from the WT or rad52D background. Relative to WT, there was a 13-fold elevation in the cins rate in an mph1D or mms2D single mutant (Table 3; P , 0.001), and the cins rate was further increased in the mph1D mms2D double mutant (P , 0.01). Importantly, the cins rate in the mph1D mms2D double mutant was not sig- nificantly different from that in the rad52D single- or rad52- D mph1D mms2D triple-mutant background ( P = 0.25 and P = 0.59, respectively). We conclude that the error-free, spontaneous lesion bypass pathways defined by MPH1 and MMS2 are distinct and that, as during chronic damage ex- posure, RAD52 is relevant in each.

RAD5 and RAD18 define two pathways of Pol z -dependent bypass Separate RAD18- and RAD5-dependent pathways of Pol z-dependent bypass were previously inferred in the CAN1 forward mutation assay (Liefshitz et al. 1998; Cejka et al. 2001) or using the cins signature in the lys2DA746 assay (Minesinger and Jinks-Robertson 2005). Here, a rad52D background was used to reduce error-free bypass and thereby enhance dependence of survival on Pol z-dependent TLS, allowing us to focus on the roles of RAD18 and RAD5 during error-prone DDT in the lys2DA746,NR assay. The large, 32-fold increase in cins observed upon deletion of RAD52 was reduced to 7.5-fold (a 75% reduction) upon additional loss of either RAD18 or RAD5 (Table 3; P , 0.001). The overall Lys+ rates in the rad52D rad18D rad5D triple- and rad52D rev3D double-mutant strains were indis- tinguishable (P = 1); there were only 3/79 and 0/78 cins in the corresponding spectra, respectively. Together, these data confirm that RAD18 and RAD5 define independent pathways of Pol z-dependent bypass. Furthermore, the 75% cins re- duction in each single mutant suggests that Rad18 and Rad5 may sometimes work together to promote Pol z-dependent bypass. We previously observed a similar, dual requirement with regard to UV-induced, NER-dependent mutagenesis that occurs in nongrowing cells (Kozmin and Jinks-Robertson 2013).

RAD52 and REV3 limit the formation of large deletions We previously described large deletions (dels) among reversion events in the lys2DA746 and lys2DA746,NR as- says (Minesinger and Jinks-Robertson 2005; Lehner and Jinks-Robertson 2009). Although most of these events are 95 bp in size and have endpoints in 10-bp direct repeats, 5-10% have endpoints in 7-bp direct repeats and delete 131 bp. Here, we define these as type 1 and type 2 dels, respec- tively. As illustrated in Figure 6A, type 1 and type 2 events have a partially overlapping endpoint that is within the re- version window; the other endpoint is distinct and upstream of the window. Because large deletions are REV3 indepen- dent, are dependent on the direction of replication fork move- ment through LYS2 (Abdulovic et al. 2007), are elevated upon loss of the Pol32 subunit of Pol d (Minesinger and Jinks-Robertson 2005), and are not generated through dou- ble-strand breaks (Lehner et al. 2012), we previously sug- gested that they reflect direct-repeat-mediated realignment of a blocked 39 end during lagging-strand synthesis. Table 4 presents rates for total large dels, as well as those for type 1 and type 2 events individually. Because there were very few (fewer than five) type 2 events in most of the spectra analyzed, changes in this class are emphasized only in those genetic backgrounds where there was a clear proportional increase among total large dels.

Relative to WT, there was a significant, ^1.5-fold in- crease in large dels upon loss of RAD52 (P , 0.001). In contrast to the similar increases in cins observed upon de- letion of MPH1 or MMS2 (Table 3), their loss had different effects on large dels (Table 4). Relative to WT, the rate of large dels increased 2.7-fold in the mms2D background (P , 0.0001), but was unaffected in the mph1D background (P = 0.25). Interestingly, in the mms2D mph1D double- or the rad52D mms2D mph1D triple-mutant background, the rate of large dels was reduced to the level observed in the mph1D single mutant (P = 0.62 and P = 0.82, respectively). These data indicate that RAD52 and MMS2 limit the formation of large dels, and that the additional events detected in their absence depend on the presence of MPH1.

In the rad52D background, additional deletion of RAD5, RAD18, RAD5 and RAD18,orREV3 had little effect on the total rate of large dels (P . 0.05 in each case). The rate of type 2 events, however, was significantly elevated in the rad52D rad5D double, rad52D rev3D double, and rad52D rad5D rad18D triple mutants relative to the rad52D single mutant (P , 0.001, P = 0.010, and P = 0.016, respectively). These data suggest that the type 2 events reflect an alterna- tive to Pol z-dependent bypass in this system, and are con- sistent with the cins data, suggesting that RAD5 and RAD18 promote separate pathways of Pol z-dependent bypass.

Discussion Bypass/tolerance mechanisms fall into two general classes that together ensure the progression of replication in the presence of unrepaired damage as well as the subsequent filling of lesion-containing gaps left behind replication forks. In the work presented here, two complementary approaches were used to genetically dissect lesion-bypass pathways when low levels of damage are present and normal repair mechanisms and cell-cycle checkpoints are unperturbed. First, we assessed growth on solid medium when cells were continuously exposed to CLUV or CLMMS. Even though these agents produce very different lesions, the relationships between the genes examined were indistinguishable. In the second approach, a unique mutation signature (cins in a frameshift-reversion assay) produced by Pol z was used as a readout of bypass-pathway utilization in the absence of exogenous damage. Elimination of a gene whose product promotes Pol z-dependent bypass decreased its mutagenic footprint, while loss of a protein promoting alternate, error- free bypass increased the mutagenic signal. The similar epis- tatic relationships between genes in the reversion and growth assays suggest that low levels of UV or MMS accu- rately mimic spontaneous damage and can be used to assess the corresponding physiological responses, even though nei- ther the nature nor the complexity of spontaneous lesions is known. Importantly, results revealed functional redundancies not evident in prior studies and clarify the hierarchies be- tween the various bypass/tolerance pathways when DNA repair and checkpoint pathways are intact.

The seminal study of Hishida et al. (2009) documented distinct genetic responses to acute UV, where repair is most important for survival, and CLUV, where tolerance path- ways are most important. In the absence of RAD18, single-strand gaps accumulated during replication and triggered a subsequent G2 arrest, allowing time for gaps to be filled by RAD52-mediated HR. RAD5 was used to distinguish RAD18-mediated error-free and error-prone by- pass, leading to the conclusion that TLS contributes little, if at all, to CLUV tolerance. Our use of MMS2 to define error- free bypass revealed clear functional redundancy between the two RAD18-regulated subpathways; neither an mms2D nor rev3D single mutant was sensitive to CLUV, but the double mutant was exquisitely sensitive. Similar synergism with respect to CLMMS has been previously reported (Huang et al. 2013) and was confi rmed here. RAD18- mediated error-free and error-prone pathways were inter- changeable in terms of promoting growth, but epistasis analysis provided no information about the hierarchies or relative importance of these subpathways. This can be inferred, however, from our analysis of Pol z-dependent spontaneous mutagenesis. Deletion of MMS2 strongly ele- vated the Pol z-dependent cins mutation signature, indicat- ing that RAD18-dependent error-free bypass normally predominates. Based on prior studies, we suggest that this reflects temporal separation of bypass mechanisms during the cell cycle, with MMS2- and REV3-dependent bypass operating primarily in S and G2, respectively (Daigaku et al. 2010; Huang et al. 2013). This is consistent with MMS2-regulated template switching at a lesion-stalled rep- lication fork in S phase (Choi et al. 2010) and with the strong accumulation of Rev1 in G2/M (Waters and Walker 2006). It also implies that most damage is dealt with dur- ing S phase, with the small amount that remains being subsequently handled by TLS.

RAD52 is required during RAD18-mediated error-free by- pass, for efficient strand invasion during HR and for strand- annealing reactions (reviewed by San Filippo et al. 2008). MPH1 was originally identified by virtue of the mutator phenotype observed in its absence (Schurer et al. 2004), and recently has been genetically implicated in HR-mediated sister chromatid interactions (Ede et al. 2011). MMS2 and MPH1 independently promote formation of X-shaped recom- bination structures on 2D gels when DNA replicates in the presence of MMS, and it was suggested that MMS2 regulates template switching while MPH1 promotes replication fork regression (Choi et al. 2010). Our observation that the CLMMS or CLUV sensitivity of a rev3D single or mms2D rev3D double mutant was further enhanced by deletion of MPH1 is consistent with the separate roles of these proteins in promoting error-free bypass. The epistasis with rad52D likewise is consistent with a role for HR in both MPH1- and MMS2-regulated bypass mechanisms.

The CLUV or CLMMS sensitivity of double mutants indicated that MMS2-regulated bypass is a more robust alternative to Pol z-mediated TLS than MPH1-regulated bypass. In terms of the mutagenic Pol z-dependent signa- ture in our frameshift reversion assay, however, loss of Mph1 or Mms2 had equivalent effects. Although one might consider these data to be contradictory, it must be empha- sized that our mutagenesis assay is very selective because it only detects the subset of events that restore the correct LYS2 reading frame. One intriguing aspect of the mutagen- esis data concerns the roles of MPH1 and MMS2 in large dels, which are Pol z-independent events. Large dels were elevated in the absence of MMS2, and the elevation re- quired MPH1. Mph1 has helicase activity that has been implicated in the dismantling of recombination intermedi- ates as well as in replication fork regression (reviewed by Whitby 2010). In the case of large dels, one possibility is that the helicase activity of Mph1 is required to unwind a39 end that is stalled within the first copy of the relevant direct repeat. Once unwound, the 39 end could then anneal tothesecond,complementarycopyoftherepeatonthe same or different DNA strand (Figure 6B). Resumption of DNA synthesis would then delete the region between the repeats.

Efficient TLS requires Rad18-dependent mono-Ub of PCNA, with subsequent extension into a poly-Ub chain by Rad5 and Mms2-Ubc13 promoting error-free bypass; Rad5 in this context interacts with both Rad18 and Ubc13 (UlrichandJentsch2000).WhetherornotRad5 addition- ally has a role in TLS, and whether this role is indepen- dent of Rad18 has been unclear. In early studies of Pol z-dependent induced mutagenesis, there was an absolute re- quirement for RAD6,butRAD5 was only necessary in specific reversion assays (Lawrence 1994). By contrast, in genetic studies of spontaneous Pol z-dependent mutagenesis, RAD5 and RAD18 were shown to regulate distinct TLS pathways (Liefshitz et al. 1998; Cejka et al. 2001; Minesinger and Jinks-Robertson 2005). More recently, a requirement of RAD5 for the Pol z-dependent bypass of a defined lesion engineered into a single- or double-stranded plasmid was reported (Pages et al. 2008; Kuang et al. 2013). Although the helicase and Ub-ligase activities of Rad5, which are required for error-free bypass, were not necessary for Pol z- dependent TLS in this context, a physical interaction with Rev1 was required. A specificmodelwasproposedinwhichRad5 interacts with Rad18, and then recruits either Ubc13 or Rev1 to promote error-free or error-prone bypass, respec- tively (Kuang et al. 2013). Consistent with a dual require- ment for RAD18 and RAD5 during TLS, most (75%) of the Pol z-dependent cins in our frameshift reversion assay re- quired both genes; the remaining events required only RAD18 or RAD5, consistent with each also independently promoting TLS. Separate contributions of RAD18 and RAD5 during TLS are further supported by the sensitivity of mutant strains to CLUV and CLMMS. In a rad52D back- ground where error-free bypass is disabled, RAD18 and RAD5 contributed independently to survival. With regard to the role of Rad5 in Pol z-dependent TLS, a consistent requirement for RAD5 has only been evident when damage levels are very low. The involvement of Rad5 in TLS might thus depend on whether damage is extensive enough to trigger a checkpoint response, as would be the case with acute doses. Finally, the clear increase in type 2 dels in rad52D rad5D rad18D and rad52D rev3D backgrounds sug- gests that this class may reflect a specific alternative to Pol z-dependent bypass of the relevant lesion. Given that Pol z activity is largely confined to G2, type 2 dels might reflect a gap- fi lling process that occurs after the completion of bulk DNA synthesis. This distinctive signature could potentially be exploited to identify small molecules that inhibit Pol z activity.

The bypass pathways deployed by yeast to deal with persistent damage are diverse and functionally redundant, and together ensure successful genome duplication. The current study has revealed clear parallels between the genetic responses to endogenous DNA damage vs. that gen- erated by chronic exposure to low doses of UV or MMS. Whether similar parallels will extend to additional agents likely depends on the specific types of lesions generated. Nevertheless, results reinforce the view that there likely is no "safe" dose of external mutagens in terms of an ability to exert physiological effects.

Acknowedgements This work was supported by a grant from the National Institutes of Health (S.J.-R.) (GM038464).

Literature Cited Abdulovic, A. L., B. K. Minesinger, and S. Jinks-Robertson, 2007 Identification of a strand-related bias in the PCNA- mediated bypass of spontaneous lesions by yeast Polh. DNA Repair (Amst.) 6: 1307-1318.

Boiteux, S., and S. Jinks-Robertson, 2013 DNA repair mechanisms and the bypass of DNA damage in Saccharomyces cerevisiae. Genetics 193: 1025-1064.

Branzei, D., F. Vanoli, and M. Foiani, 2008 SUMOylation regu- lates Rad18-mediated template switch. Nature 456: 915-920.

Broomfield, S., B. L. Chow, and W. Xiao, 1998 MMS2, encoding a ubiquitin-conjugating-enzyme-like protein, is a member of the yeast error-free postreplication repair pathway. Proc. Natl. Acad. Sci. USA 95: 5678-5683.

Cejka, P., V. Vondrejs, and Z. Storchova, 2001 Dissection of the func- tions of the Saccharomyces cerevisiae RAD6 postreplicative repair group in mutagenesis and UV sensitivity. Genetics 159: 953-963.

Choi, K., B. Szakal, Y. H. Chen, D. Branzei, and X. Zhao, 2010 The Smc5/6 complex and Esc2 influence multiple replication- associated recombination processes in Saccharomyces cerevisiae. Mol. Biol. Cell 21: 2306-2314.

Ciccia, A., and S. J. Elledge, 2010 The DNA damage response: making it safe to play with knives. Mol. Cell 40: 179-204.

Daigaku,Y.,A.A.Davies,andH.D.Ulrich,2010 Ubiquitin- dependent DNA damage bypass is separable from genome repli- cation. Nature 465: 951-955.

De Graaf, B., A. Clore, and A. K. Mccullough, 2009 Cellular path- ways for DNA repair and damage tolerance of formaldehyde- induced DNA-protein crosslinks. DNA Repair (Amst.) 8: 1207-1214.

Ede, C., C. J. Rudolph, S. Lehmann, K. A. Schurer, and W. Kramer, 2011 Budding yeast Mph1 promotes sister chromatid interac- tions by a mechanism involving strand invasion. DNA Repair (Amst.) 10: 45-55.

Freedman, J. A., and S. Jinks-Robertson, 2002 Genetic require- ments for spontaneous and transcription-stimulated mitotic re- combination in Saccharomyces cerevisiae. Genetics 162: 15-27.

Friedberg, E. C., 2005 Suffering in silence: the tolerance of DNA damage. Nat. Rev. Mol. Cell Biol. 6: 943-953.

Gangavarapu, V., S. Prakash, and L. Prakash, 2007 Requirement of RAD52 group genes for postreplication repair of UV-damaged DNA in Saccharomyces cerevisiae. Mol. Cell. Biol. 27: 7758-7764.

Goldstein, A. L., and J. H. Mccusker, 1999 Three new dominant drug resistance cassettes for gene disruption in Saccharomyces cerevisiae. Yeast 15: 1541-1553.

Gueldener, U., J. Heinisch, G. J. Koehler, D. Voss, and J. H. Hegemann, 2002 A second set of loxP marker cassettes for Cre-mediated multiple gene knockouts in budding yeast. Nucleic Acids Res. 30: e23.

Harfe, B. D., and S. Jinks-Robertson, 1999 Removal of frameshift intermediates by mismatch repair proteins in Saccharomyces cer- evisiae. Mol. Cell. Biol. 19: 4766-4773.

Harfe, B. D., and S. Jinks-Robertson, 2000 DNA polymerase z in- troduces multiple mutations when bypassing spontaneous DNA damage in Saccharomyces cerevisiae.Mol.Cell6:1491-1499.

Haynes, R. H., and B. A. Kunz, 1981 DNA repair and mutagenesis in yeast, pp. 371-414 in The Molecular Biology of theYeast Sac- charomyces: Life Cycle and Inheritance, edited by J. N. Strathern, E. W. Jones, and J. R. Broach. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

Hishida, T., Y. Kubota, A. M. Carr, and H. Iwasaki, 2009 RAD6- RAD18-RAD5-pathway-dependent tolerance to chronic low- dose ultraviolet light. Nature 457: 612-615.

Hishida, T., Y. Hirade, N. Haruta, Y. Kubota, and H. Iwasaki, 2010 Srs2 plays a critical role in reversible G2 arrest upon chronic and low doses of UV irradiation via two distinct homol- ogous recombination-dependent mechanisms in postreplication repair-deficient cells. Mol. Cell. Biol. 30: 4840-4850.

Hoege, C., B. Pfander, G. L. Moldovan, G. Pyrowolakis, and S. Jentsch, 2002 RAD6-dependent DNA repair is linked to mod- ification of PCNA by ubiquitin and SUMO. Nature 419: 135- 141.

Huang, D., B. D. Piening, and A. G. Paulovich, 2013 The prefer- ence for error-free or error-prone postreplication repair in Saccha- romyces cerevisiae exposed to low-dose methyl methanesulfonate is cell cycle dependent. Mol. Cell. Biol. 33: 1515-1527.

Karras, G. I., and S. Jentsch, 2010 The RAD6 DNA damage toler- ance pathway operates uncoupled from the replication fork and is functional beyond S phase. Cell 141: 255-267.

Kozmin, S. G., and S. Jinks-Robertson, 2013 The mechanism of nucleotide excision repair-mediated UV-induced mutagenesis in nonproliferating cells. Genetics 193: 803-817.

Krejci, L., S. Van Komen, Y. Li, J. Villemain, M. S. Reddy et al., 2003 DNA helicase Srs2 disrupts the Rad51 presynaptic fila- ment. Nature 423: 305-309.

Kuang, L., H. Kou, Z. Xie, Y. Zhou, X. Feng et al., 2013 A non- catalytic function of Rev1 in translesion DNA synthesis and mu- tagenesis is mediated by its stable interaction with Rad5. DNA Repair (Amst.) 12: 27-37.

Lawrence, C., 1994 The RAD6 DNA repair pathway in Saccharo- myces cerevisiae: What does it do, and how does it do it? Bio- Essays 16: 253-258.

Lehner, K., and S. Jinks-Robertson, 2009 The mismatch repair system promotes Polz-dependent translesion synthesis in yeast. Proc. Natl. Acad. Sci. USA 106: 5749-5754.

Lehner, K., S. V. Mudrak, B. K. Minesinger, and S. Jinks-Robertson, 2012 Frameshift mutagenesis: the roles of primer-template misalignment and the nonhomologous end-joining pathway in Saccharomyces cerevisiae. Genetics 190: 501-510.

Liefshitz, B., R. Steinlauf, A. Friedl, F. Eckardt-Schupp, and M. Kupiec, 1998 Genetic interactions between mutants of the 'error-prone' repair group of Saccharomyces cerevisiae and their effect on recombination and mutagenesis. Mutat. Res. 407: 135-145.

Longtine, M. S., A. Mckenzie Iii, D. J. Demarini, N. G. Shah, A. Wach et al., 1998 Additional modules for versatile and eco- nomical PCR-based gene deletion and modification in Saccharo- myces cerevisiae. Yeast 14: 953-961.

McDonald, J. P., A. S. Levine, and R. Woodgate, 1997 The Sac- charomyces cerevisiae RAD30 gene, a homologue of Escherichia coli dinB and umuC, is DNA damage inducible and functions in a novel error-free postreplication repair mechanism. Genetics 147: 1557-1568.

Minesinger, B. K., and S. Jinks-Robertson, 2005 Roles of RAD6 epistasis group members in spontaneous Polz-dependent trans- lesion synthesis in Saccharomyces cerevisiae. Genetics 169: 1939-1955.

Moldovan, G. L., B. Pfander, and S. Jentsch, 2007 PCNA, the maestro of the replication fork. Cell 129: 665-679.

Novarina, D., F. Amara, F. Lazzaro, P. Plevani, and M. Muzi-Falconi, 2011 Mind the gap: keeping UV lesions in check. DNA Repair (Amst.) 10: 751-759.

Pages, V., A. Bresson, N. Acharya, S. Prakash, R. P. Fuchs et al., 2008 Requirement of Rad5 for DNA polymerase z-dependent translesion synthesis in Saccharomyces cerevisiae. Genetics 180: 73-82.

Pfander, B., G. L. Moldovan, M. Sacher, C. Hoege, and S. Jentsch, 2005 SUMO-modified PCNA recruits Srs2 to prevent recombi- nation during S phase. Nature 436: 428-433.

Sale, J. E., 2012 Competition, collaboration and coordination- determining how cells bypass DNA damage. J. Cell Sci. 125: 1633-1643.

San Filippo, J., P. Sung, and H. Klein, 2008 Mechanism of eukaryotic homologous recombination. Annu. Rev. Biochem. 77: 229-257.

Schurer, K. A., C. Rudolph, H. D. Ulrich, and W. Kramer, 2004 Yeast MPH1 gene functions in an error-free DNA damage bypass pathway that requires genes from homologous recombi- nation, but not from postreplicative repair. Genetics 166: 1673- 1686.

Stelter, P., and H. D. Ulrich, 2003 Control of spontaneous and damage-induced mutagenesis by SUMO and ubiquitin conjuga- tion. Nature 425: 188-191.

Ulrich, H. D., and S. Jentsch, 2000 Two RING finger proteins mediate cooperation between ubiquitin-conjugating enzymes in DNA repair. EMBO J. 19: 3388-3397.

Vanoli, F., M. Fumasoni, B. Szakal, L. Maloisel, and D. Branzei, 2010 Replication and recombination factors contributing to recombination-dependent bypass of DNA lesions by template switch. PLoS Genet. 6: e1001205.

Veaute, X., J. Jeusset, C. Soustelle, S. C. Kowalczykowski, E. Le Cam et al., 2003 The Srs2 helicase prevents recombination by disrupting Rad51 nucleoprotein filaments. Nature 423: 309-312.

Waters, L. S., and G. C. Walker, 2006 The critical mutagenic translesion DNA polymerase Rev1 is highly expressed during G(2)/M phase rather than S phase. Proc. Natl. Acad. Sci. USA 103: 8971-8976.

Waters, L. S., B. K. Minesinger, M. E. Wiltrout, S. D'Souza, R. V. Woodruff et al., 2009 Eukaryotic translesion polymerases and their roles and regulation in DNA damage tolerance. Microbiol. Mol. Biol. Rev. 73: 134-154.

Whitby, M. C., 2010 The FANCM family of DNA helicases/ translocases. DNA Repair (Amst.) 9: 224-236.

Zhang, H., and C. W. Lawrence, 2005 The error-free component of the RAD6/RAD18 DNA damage tolerance pathway of bud- ding yeast employs sister-strand recombination. Proc. Natl. Acad. Sci. USA 102: 15954-15959.

Communicating editor: N. Hollingsworth Kevin Lehner and Sue Jinks-Robertson1 Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, North Carolina 27710 Copyright © 2014 by the Genetics Society of America doi: 10.1534/genetics.114.168617 Manuscript received June 22, 2014; accepted for publication July 21, 2014; published Early Online July 24, 2014.

Supporting information is available online at http://www.genetics.org/lookup/suppl/ doi:10.1534/genetics.114.168617/-/DC1.

1Corresponding author: Department of Molecular Genetics and Microbiology, DUMC 3020, 213 Research Dr., 384 CARL Bldg., Duke University Medical Center, Durham, NC 27710. E-mail: [email protected] (c) 2014 Genetics Society of America

[ Back To TMCnet.com's Homepage ]