The pendulum of the Ku-Ku clock.

Canonical DNA non-homologous end-joining (c-NHEJ) and homologous recombination (HR), the two major DNA double-strand break (DSB) repair pathways, have long been depicted as competitors, fighting a race to rejoin DSBs. In human cells, Ku, an upstream component of NHEJ, is highly abundant and has exquisite end-binding capacity. Emerging evidence has suggested that Ku is the first protein binding most, if not all, DSBs, and creates a block to resection. Although most c-NHEJ proceeds without resection, recent studies have provided strong evidence for a process of resection-dependent c-NHEJ, that repairs a subset of DSBs. HR also repairs a subset of two-ended DSBs in G2 phase and processes one-ended DSBs that arise following replication fork stalling or collapse to promote replication restart. HR also necessitates end-resection. This raises the question of how end-resection takes place despite Ku's avid end-binding capacity. Insight into this enigma has been gained from the analysis of DSBs generated by Spo11 or TOP2, which create protein-bridged DSBs. The progression of repair by HR or NHEJ requires removal of the end-blocking lesions. The MRE11-RAD50-NBS1 (MRN) complex, CtIP and EXO1 play critical roles in this process. Here, we review our current understanding of how resection arises at lesions blocked by covalently bound Spo11 or TOP2 or following Ku binding, which effectively creates a distinct resection-blocking lesion due to its avid end-binding activity and abundance. Our review reveals that Ku plays an active role in determining pathway choice and exposes similarities yet distinctions in the progression of resection that is suited to the optimal repair pathway choice.