Resolution of complex ends by non-homologous end joining - better to be lucky than good?

DNA double strand breaks (DSBs) arise after replication, aberrant repair of spontaneous damage, and exposure to exogenous damaging agents, especially those used in cancer therapies. DSBs are also intermediates in several developmentally-programmed recombinations. Failed DSB repair is typically lethal, while aberrant DSB repair can lead to developmental defects, progeria, and cancer. Repair pathways include Homologous recombination (HR), non-homologous end joining (NHEJ), and Alternate end joining (Alt-EJ) (reviewed in e.g. [1]). Importantly, HR is dependent on extensive (100s to 1000s of nucleotides) DNA synthesis, a sister chromatid template to direct this synthesis, and a homology search step needed to find the template in a sister chromatid. In contrast, NHEJ is primarily a ligation reaction and can act independently of S-phase restricted sister chromatids, dNTP generation [2], and other requirements for extensive synthesis. Finally, a fraction of ligation-mediated repair is independent of factors required for classically defined NHEJ, and is thus termed “Alt-EJ”.The primary disadvantage to resolving ends by ligation is that biological sources of DSBs often produce “dirty” or complex end structures that can interfere with this step (Figure 1). DNA flanking the break may possess nucleotide damage, most frequently oxidized bases, various classes of abasic sites, and 3′ phosphate or 5′ hydroxyl termini [3]. Such damage is especially likely in the case of ionizing radiation-induced breaks, which are associated with damage clusters [3,4]. DSB ends may be further occluded by proteins, both non-covalently associated (e.g. chromatin) and covalently adducted type II topoisomerases [5]. Ends can also possess secondary structures including hairpins or, after a pair of ends are aligned together, gaps, mismatches, or flaps.NHEJ resolves complex ends by employing a sophisticated machine engineered to facilitate ligation despite ligation-blocking lesions. A series of core facto