Search

Drugs that produce DNA interstrand crosslinks (ICLs), between the two complementary strands of the double helix, have an important role in chemotherapy regimens for cancer. Novel crosslinking agents, and targeting strategies involving DNA crosslinking agents, continue to be developed. The ability of cells to repair DNA ICLs is a critical determinant of sensitivity, and recent dinical studies indicate that DNA repair capacity is strongly implicated in both inherent tumour sensitivity and acquired drug resistance. A detailed understanding of the cellular mechanisms that act to eliminate these critical DNA lesions is clearly important. DNA ICLs present a complex challenge to DNA repair mechanisms because of the involvement of both DNA strands. It is now clear that cells from bacteria and yeast to mammals eliminate interstrand ICLs through the coordinated action of several DNA repair pathways. Recently, a model of ICL repair has been proposed, in which mammalian cells use novel excision repair reactions (requiring the XPF and ERCC1 proteins) to uncouple the crosslink. This is followed by a homologous recombination step to provide the genetic information needed to complete repair. This new knowledge may permit the development of screens for tumour response to crosslinking agents, and should also aid the design of more effective crosslinking agents that evade DNA repair. In addition, the proteins mediating the repair reactions represent potential targets for therapeutic intervention.

Abstract Background DNA interstrand cross-links (ICLs) are critical lesions produced by several cancer chemotherapy agents including platinum drugs and nitrogen mustards. We have previously shown in haematological (multiple myeloma) and solid tumours (ovarian cancer) that clinical sensitivity to such agents can result from a defect in DNA ICL processing leading to their persistence. Conversely, enhanced repair can result in clinical acquired resistance following chemotherapy. The repair of ICLs is complex but it is assumed that the ‘unhooking’ step is common to all ICLs. Methods Using a modification of the single cell gel electrophoresis (Comet) assay we measured the formation and unhooking of melphalan and cisplatin-induced ICLs in cell lines and clinical samples. DNA damage response in the form of γ-H2AX foci formation and the formation of RAD51 foci as a marker of homologous recombination were also determined. Real-time PCR of 84 genes involved in DNA damage signalling pathways was also examined pre- and post-treatment. Results Plasma cells from multiple myeloma patients known to be clinically resistant to melphalan showed significant unhooking of melphalan-induced ICLs at 48 hours, but did not unhook cisplatin-induced ICLs. In ovarian cancer cells obtained from patients following platinum-based chemotherapy, unhooking of cisplatin-induced ICLs was observed at 48 hours, but no unhooking of melphalan-induced ICLs. In vitro, A549 cells were proficient at unhooking both melphalan and cisplatin-induced ICLs. γ-H2AX foci formation closely followed the formation of ICLs for both drugs, and rapidly declined following the peak of formation. RPMI8226 cells unhooked melphalan, but not cisplatin-induced ICLs. In these cells, although cross-links form with cisplatin, the γ-H2AX response is weak. In A549 cells, addition of 3nM gemcitabine resulted in complete inhibition of cisplatin-induced ICL unhooking but no effect on repair of melphalan ICLs. The RAD51 foci response was both drug and cell line specific. Real time PCR studies highlighted differences in the damage response to melphalan and cisplatin following equi-ICL forming doses. Conclusions These data suggest that the mechanisms by which melphalan and cisplatin-induced ICLs are ‘unhooked’ in vitro are distinct, and the mechanisms of clinical acquired resistance involving repair of ICLs, are drug specific.