DAMAGE-INDUCED LOCALIZED HYPERMUTABILITY (LHM). Mutations are important in evolution as well as many diseases. While mutations are generally considered to accumulate independently, most single base substitutions in coding sequences fail to significantly alter the activity of the corresponding protein. Multiple mutations may be needed to produce dramatic genetic consequences such as gene inactivation or generation of alleles with novel function. While somatic hyper-mutation is an example of programmed localized hyper-mutability confined to a small region in immunoglobulin genes, non-programmed pathways for generating mutation clusters without creating overall high mutation loads has been elusive. We proposed that lesions in transient single-strand DNA (ssDNA) are especially threatening to genome stability. To test this, we designed systems in budding yeast that could generate several kb of persistent ssDNA next to double-strand breaks (DSBs) or uncapped telomeres. The systems allowed restoration to the double-strand state after applying DNA damage. We found that lesions induced by UV-light and methyl methanesulfonate (MMS) can be tolerated in long single-strand regions and are hypermutagenic. Both mutagens caused multiple, strand-biased premutagenic lesions in the transient ssDNA resulting in mutation clusters. The MMS induced ssDNA during DSB repair can result in over 0.2 mutations/kb. This remarkably high level is 20,000 fold higher than the mutation density without a DSB. The MMS-induced mutations associated with DSB-repair were primarily due to substitutions via translesion DNA synthesis at damaged cytosines, even though there are nearly 10 times more MMS-induced lesions at other bases. The promutagenic lesion dominating the LHM is likely the single-strand specific lesion 3-methylcytosine. Thus, the dramatic increase in mutagenesis at a DSB is concluded to result primarily from generation of nonrepairable, highly mutagenic ssDNA-specific lesions in ssDNA associated with DSB-repair. We also found that mutation clusters can occur in yeast grown in the presence of MMS, which causes random DNA damage. A novel mutation reporter was developed for selecting forward mutations in URA3 and CAN1. Chronic exposure to MMS caused joint inactivation of URA3 and CAN1 when they were close (separated by 1 kb) but not when they were separated by 85 kb, indicating that the double mutations occurred primarily via a single localized event. To identify mutation clusters and underlying mechanisms, we improved whole genome sequencing to a level where single base substitutions in 85% of the yeast genome could be detected (collaboration with Dr. Piotr Mieczkowsk). Surprisingly, inactivation of URA3 and CAN1 is often accompanied by additional mutations (up to 30) in clusters that span up to 250 kb. The cluster densities were as high as 1/kb. Unlike mutations in the rest of the genome, clusters were predominantly composed of mutations of G:C pairs and contained a strand bias consistent with the mutation spectra of error-prone TLS occurring during restoration of MMS-damaged ssDNA. This base specificity and strand bias indicates DSB associated strand-resection as a major mechanism for the LHM. Thus, we identify a new pathway of damage-induced mutagenesis in which the combination of localized inability to repair DNA damage along with error-prone translesion synthesis leads to localized severe genetic alteration within a single generation. This scenario could result in rapid diversification and selective advantage in adaptive evolution. It also identifies a possible new source of genetic disease and cancer. DSBS RESULTING FROM DNA BASE DAMAGE. Our studies on MMS have also been extended to mechanisms by which MMS can cause derived DSBs. Little is known about derived DSBs, which might arise during normal or defective base excision repair (BER) of single-strand lesions. We employed our recent approaches for detecting MMS-induced DSBs (MCB 2009) and end-processing (resection) of ionizing radiation-induced DSBs (PLoS Genetics, 2009) to address the role of BER in generation of derived DSBs and mechanisms of repair. Previously, we found that MMS damage, which does not cause single strand breaks directly, led to accumulation of unrepairable DSBs in G1 haploid yeast that lack AP endonucleases (Apn1/2). We now show that the MMS-derived DSBs in G2 cells could be repaired via the RAD52 pathway utilizing homologous recombination between sister chromatids. Although the ends are dirty, we find that they are efficiently resected as part of the recombinational repair process. Interestingly, using pulsed-field gel electrophoresis (PFGE) of chromosomal DNA we identified a very slow-moving DNA intermediate (SMD). The SMD was quickly formed after short MMS exposure and was present up to 4 hr later. It is unlikely due to recombination since RAD50 or RAD51 mutants still present SMD. To address the origin and resection of the derived DSBs, we blocked either the first step in BER, formation of AP sites, by deletion of the MAG1 glycosylase or the following step, cleavage of AP sites, by methoxyamine treatment of cells. Since the MMS derived DSBs as well as SMD were completely blocked, they must arise through the processing of AP sites. However, additional deletion of NTG1 and NTG2 lyases (i.e., apn1/2 ntg1/2) which are involved in the alternative pathway for processing AP sites, reduced the amount of DSBs but not SMD, suggesting that SMD does not arise during the generation or processing of DSBs. Collectively, these results reveal a complex repair profile of base lesions in the absence of AP endonucleases. Our results highlight the pathway for generation, processing and repair of DSBs caused by base alkylation. GENE DOSAGE OF GENOME STABILITY GENES. The sister chromatid cohesion (SCC) complex known as cohesin assures close association between sister chromatids and faithful transmission of chromosomes in mitosis. Cohesin is also involved in chromosome structure, maintenance, transcription DNA repair, and cohesin mutations are associated with cancer and developmental defects. Using a gene dosage approach developed in tetraploid yeast we reported (PLoS Genetics, 2010) that cohesin is present in limiting amounts and that reduced levels greatly enhance the risk of DNA damage-induced genome instability in G2 cells. This is because cohesin restricts repair to sister chromatids so that reduced cohesin levels open the genome to recombinational interactions between homologous chromosomes which can then lead to loss-of-heterozygosity. Given the importance of cohesin levels in genome stability, we have extended these findings to address the role of regulators and the mode of establishment of SCC in genome stability. SCC is established during replication;however, it is positively regulated (i.e., more cohesin) in response to DNA damage. Our recent results with a conditional cohesin mutant reveal a major role for the S-phase cohesin in preventing gamma-induced recombination between homologues. Furthermore, this cohesin is also highly protective against inter-homologue recombination induced by UV-lesions, which do not induce DSBs. We have also examined Rad61 since this protein counteracts SSC establishment and is a negative regulator of SCC during S-phase by preventing over-cohesion. We found that in the absence of Rad61 the levels of spontaneous and damage-induced inter-homologue and heterologous recombination were increased, consistent with Rad61 playing a positive, rather than a negative role, in SCC-mediated genome stability. Given that the sister chromatid cohesion process is sensitive to environmental perturbations we suggest that defects in S phase SCC in combination with DNA damage are risk factors in human health.
Showing the most recent 10 out of 20 publications
