DNA double-strand breaks (DSBs) are primarily repaired in mammalian cells by non-homologous end joining (NHEJ). Initially, the Ku 70/80 heterodimer and DNA-PKcs, the catalytic subunit of DNA-PK, were identified as critical components of the NHEJ pathway. It has been demonstrated that Ku binds to DSBs ends and recruits DNA-PKcs to form DNA-dependent protein kinase (DNA-PK), which has been proposed to be the first step in initiating NHEJ repair. This model is supported by experiments demonstrating that DNA end joining can be suppressed by wortmannin, an inhibitor of DNA-PK. Recent studies suggest that DNA-PKcs independent NHEJ pathways exist. Cells lacking DNA-PKcs, such as M059J, still have the capacity to repair DSBs. In addition, in vitro DNA end joining has been observed with human cell extracts after DNA-PKcs depletion. In this study, we provide further evidence of a DNA-PK-independent NHEJ pathway with activity that predominates dependent only upon the in vitro reaction conditions. We have assayed the DNA end joining activity of whole cell extracts (WCEs) from three cell lines (HeLa, M059J, and M059K) at 17°C with and without polyethylene glycol (PEG). DNA end joining activity was observed both in the presence and absence of PEG. However, Ku dependent end joining reactions in the presence of PEG were insensitive to inhibition by wortmannin. In contrast, end joining in the absence of PEG was strongly inhibited by wortmannin. The wortmannin insensitive end joining observed in the presence of PEG suggests a DNA-PK independent pathway, which is further supported by our observation of a loss of kinase activity in the presence of PEG. These data suggest that there are biochemically distinct end joining mechanisms that are a function of the repair complex architecture, which in turn is dependent upon the reaction conditions. This is supported by the observation that unique protein-DNA complexes form in the presence and absence of PEG, as detected by electrophoretic mobility shift assays. Supershift assays with antibodies to Ku and DNA-PKcs also indicated structural differences between the complexes formed in the presence and absence of PEG. Also, a comparison of DNA end joining rates with M059J WCE (lacking DNA-PKcs) and M059K WCE (with DNA-PKcs), showed that even in the presence of PEG, M059K WCE was far more efficient, suggesting that although the kinase is not active, DNA-PKcs may still have a role in DNA end joining. In conclusion, we have established conditions for observing both DNA-PK-dependent and DNA-PK-independent DNA end joining in vitro without removing DNA-PKcs. Furthermore, although the kinase activity of DNA-PK is inhibited in the presence of PEG at 17°C, and different protein complexes are formed on the DNA ends, DNA-PKcs still appears to play role in DNA end joining under these conditions.

[Proc Amer Assoc Cancer Res, Volume 46, 2005]