Post-replication repair suppresses duplication-mediated genome instability

Abstract

RAD6 is known to suppress duplication-mediated gross chromosomal rearrangements (GCRs) but not single-copy sequence mediated GCRs. Here, we found that the RAD6- and RAD18-dependent post-replication repair (PRR) and the RAD5-, MMS2-, UBC13-dependent error-free PRR branch acted in concert with the replication stress checkpoint to suppress duplication-mediated GCRs formed by homologous recombination (HR). The Rad5 helicase activity, but not its RING finger, was required to prevent duplication-mediated GCRs, although the function of Rad5 remained dependent upon modification of PCNA at Lys164. The SRS2, SGS1, and HCS1 encoded helicases appeared to interact with Rad5, and epistasis analysis suggested that Srs2 and Hcs1 act upstream of Rad5. In contrast, Sgs1 likely functions downstream of Rad5, potentially by resolving DNA structures formed by Rad5. Our analysis is consistent with models in which PRR prevents replication damage from becoming double strand breaks (DSBs) and/or regulates the activity of HR on DSBs.

Figure 1. PRR defects result in increased rates of duplication-mediated translocations.

A. The pre-duplication (yel068c::CAN1/URA3) and post-duplication (yel072w::CAN1/URA3) assays differ by whether or not they include the DSF1-HXT13 homology in the breakpoint region (the left arm of chromosome V between the CAN1/URA3 cassette and the most telomeric essential gene, PCM1). The hygromycin resistance marker is indicated by hph. Grey boxes indicate regions of homologies between the chromosomes. B. The rates of the total Can^R 5FOA^R product and the rates of t(V;XIV) and t(V;IV or X) translocations, and non-duplication-mediated GCR products in the yel072w::CAN1/URA3 assay are depicted in a bar graph. Error bars indicate 95% confidence intervals and the fold increase for each rate is displayed in parentheses, (<) indicates that no isolates of that class were identified. The number of isolates analyzed is shown in parentheses after the genotype. The numerical GCR rates are presented in Tables 1, 2, 4 and 5.

Figure 2. Models for the suppression of duplication-mediated GCRs by PRR.

A. The most important RAD6-dependent pathway that suppresses duplication-mediated GCRs (thick lines) corresponds to the “error-free” PRR branch, which is downstream of Srs2. Other Rad6- and Rad18-dependent branches are less important (thin lines). The presence of specific PCNA and DNA states are inferred based on the genes involved in the pathway. Sgs1 appears to act downstream of the Rad5-dependent branches. The existence of Rad5 branches that are independent of Ubc13 and Rev3 that could be dependent upon Rad6 and Rad18 or independent of Rad6 is inferred by the observation of synergistic interactions between mutations in RAD5 and mutations in RAD6, UBC13 and REV3. Our data do not directly address the previously identified Rad5- and Rev3-dependent branch . B. PRR could potentially suppress duplication-mediated GCRs by preventing replication damage from being converted into DSBs and other HR substrates. Suppression of single-copy GCRs also requires that PRR promotes other GCR forming pathways (such as NHEJ and de novo telomere addition) or requires PRR-dependent suppression of HR. C. PRR could potentially suppress duplication-mediated GCRs by functioning as an alternative to HR. Suppression of single-copy GCRs also requires that PRR promotes other GCR forming pathways (such as NHEJ and de novo telomere addition) or requires PRR-dependent suppression of HR. The red arrows and Xs in B and C indicate the consequences of PRR defects.