DNA in eukaryotes is packaged into nucleosomes, which present steric impediments to many of the factors and enzymes that act on DNA, including DNA repair enzymes. Within the nucleosome, DNA remains vulnerable to oxidative damage that can result from normal cellular metabolism, ionizing radiation, and various chemical agents. Oxidatively damaged DNA is repaired in a stepwise fashion via the base excision repair (BER) pathway. Other DNA repair pathways, including Nucleotide Excision Repair (NER), Mismatch Repair (MMR), Homologous Recombination (HR), and Non-homologous End-Joining (NHEJ) are all thought to require nucleosome remodeling or disruption. In contrast, it was reported that the first step of BER does not require or induce nucleosome disruption. For example, the human DNA glycosylase hNTH1 (human Endonuclease III) was discovered to excise thymine glycol lesions from nucleosomes without nucleosome disruption, and could excise optimally oriented lesions with an efficiency approaching that seen for naked DNA (Prasad, Wallace, and Pederson 2007). To determine if the properties of hNTH1 are shared by other human DNA glycosylases, we compared hNTH1 with NEIL1, a human DNA glycoylase that also excises thymine glycol from DNA, with respect to their activities on nucleosome substrates. We found that the cellular concentrations and apparent kcat/KM ratios for hNTH1 and NEIL1 are similar. However, NEIL1 and hNTH1 differ in that NEIL1 binds undamaged DNA far more avidly than hNTH1. After adjustment for non-specific DNA binding, hNTH1 and NEIL1 proved to have similar intrinsic activities towards nucleosome substrates. We next wanted to examine the effects of nucleosomes on enzymes that catalyze the remaining steps in BER. We therefore assembled the entire four-step BER reaction with model, lesion-containing nucleosomes. The rates of substrate processing during the first three steps in BER, catalyzed by a DNA glycosylase, AP endonuclease, and DNA Polymerase Pol), varied with the helical orientation of the substrate relative to the underlying histone octamer. In contrast, the rate of action by DNA Ligase III- (in association with XRCC1) was independent of lesion orientation. These results are consistent with structural studies of BER enzymes and the previously proposed DNA unwrapping model for how BER enzymes gain access to lesions in nucleosomes (Prasad, Wallace, and Pederson 2007). During these investigations, we also discovered a synergistic interaction between Pol and Ligase III- complexed with XRCC1 that enhances the repair of lesions in nucleosomes. Together, our results support the hypothesis that DNA glycosylases have evolved to function in specific cellular environments (e.g. NEIL1 may function exclusively during DNA replication), but also possess DNA binding motifs and mechanisms of substrate recognition that impart a similar intrinsic activity on nucleosomes. In addition to hNTH1 and NEIL1, we have discovered that lesion orientation is also an important factor to the activities of APE and Pol and that the complete BER reaction can occur without requiring or inducing nucleosome disruption. Finally, protein-protein interactions between XRCC1 and Pol may be important for the efficient in vivo repair of lesions in nucleosomes.

Clustered damage to DNA is a signature mark of radiation-induced damage, which involves damage to the nucleobases and/or DNA backbone. Double-strand breaks created by damaging agents are detrimental to cell survival leading to chromosomal translocations. Normal cells employ Non-homologous end-joining because of its faster kinetics, to suppress chromosomal translocations. However, the presence of complex DNA ends constitutes a significant challenge to NHEJ. Location of Thymine glycol (Tg) at DSB ends was a potential hindrance to end joining. The substrate with Tg at the third position (Tg3) from the DSB joined better than when present at the fifth position (Tg5). However, hNTH1 assay showed Tg5 to be a better substrate than Tg3 for BER, potentially explaining the increased Tg removal and decreased end joining of Tg5 in extracts. Nonetheless, there appeared to be no preference in the susceptibility of 5’-Tg substrates with Tg at the second and third positions from DSB ends. Polynucleotide kinase phosphatase is crucial in restoring the 3′ hydroxyl, and 5′ phosphate ends at strand breaks. No other enzyme is known to possess PNKP’s activity in mammalian cells at DSBs. Experiments done with PNKP knockout cells have shown some activity similar to PNKP, which appeared to be a part of NHEJ and was not pharmacologically inhibited by PNKP inhibitor. Additionally, core NHEJ factors XRCC4 and XLF influenced the activities of PNKP. Overall, these experiments suggest that Tg repair is dependent on the position from DSB and an alternative enzyme processes 3′- PO, and 5′-OH ends.