Rad52-mediated homologous recombination is required for the resolution of double strand break intermediates generated by interstrand crosslink repair

Abstract

1566

DNA Interstrand crosslinks (ICLs) are potentially lethal, premutagenic, and DNA polymerase arresting lesions (PALs). Hexavalent chromium [Cr(VI)] is an occupational carcinogen that can be reduced intracellularly to Cr(III), which can interact with DNA to form ICLs. Additionally, cisplatin (CDDP), a chemotherapeutic agent, has also been shown to produce ICLs. The mechanisms involved in the repair of ICLs are not well understood. It has previously been suggested that homologous recombinational repair (HRR) plays a role in the efficient repair of ICLs. To determine the role of HRR in the removal and/or repair of Cr-PALs, we monitored these lesions using quantitative PCR (QPCR) in WT and rad52-deficient S. cerevisiae at various time points following treatment with Cr(VI). QPCR is a highly sensitive method for the detection of DNA damage associated PALs at the single gene level. In this procedure, the enzymatic amplification of target DNA can be inhibited by several lesions in template DNA including DNA strand breaks and ICLs. We have previously shown that rad52 yeast are hypersensitive to Cr(VI) lethality which is due to an inability to exit Cr(VI)-induced G2 arrest. Because HRR has been implicated in the repair of ICLs, we hypothesized that rad52 yeast would exhibit more PALs as a result of the incomplete repair of Cr-ICLs and CDDP-ICLs. Our results demonstrate that rad52 yeast display a significantly higher level of PALs (3.9 Cr-PALs per 100 Kb) compared to wildtype (wt) yeast at 120 min following treatment with 10 mM Cr(VI). Similarly, rad52 yeast treated with equitoxic doses of cisplatin (CDDP) (0.5 mM), exhibited a significant increase in PALs (5.8 CDDP-PALs per 100Kb) at 120 min recovery. Additionally, no difference in the removal of CDDP-ICLs was observed between the wt and rad52 strains during this recovery period, suggesting that the ICLs were not responsible for the difference in PALs. Because double strand breaks (DSBs) may be formed at stalled replication forks during replication and are postulated to be intermediates of HRR mediated ICL repair, we next investigated whether unrepaired DSBs were generated in rad52 yeast 120 min post-treatment. No difference existed between DSBs in wt and rad52 yeast immediately following treatment with Cr(VI) or CDDP. However, in contrast to wt yeast, which efficiently removed DSBs, rad52 yeast accumulated DSBs at 120 min recovery. These results collectively suggest that rad52 yeast are deficient in the resolution of DSBs generated as a consequence of ICL repair and help explain the hypersensitivity of this strain to ICL agents. (Supported by NIH grants ES-05304 and ES-09661 to SRP).