Non-reparable DNA strand breaks and cell killing studied in CHO cells after X-irradiation at different passage numbers

The absence of Ku but not defects in classical non-homologous end-joining is required to trigger PARP1-dependent end-joining

Classical-non-homologous end-joining (C-NHEJ) is considered the main pathway for repairing DNA double strand breaks (DSB) in mammalian cells. When C-NHEJ is defective, cells may switch DSB repair to an alternative-end-joining, which depends on PARP1 and is more erroneous. This PARP1-EJ is suggested to be active especially in tumor cells contributing to their genomic instability. Here, we define conditions under which cells would switch the repair to PARP1-EJ. Using the end jining repair substrate pEJ, we revealed that PARP1-EJ is solely used when Ku is deficient but not when either DNA-PKcs or Xrcc4 is lacking. In the latter case, DSB repair, however, could be shuttled to PARP1-EJ after additional Ku80 down-regulation, which partly rescued the DSB repair in these mutants. We demonstrate here that PARP-EJ may work on DSB ends at high fidelity manner, as evident from the unchanged efficiency upon blocking end resection by either roscovitin or mirin. Furthermore, we demonstrate for that PARP-EJ is likewise involved in the repair of multiple DSBs (I-PpoI- and IR-induced). Importantly, we identified a chromatin signature associated with the switch to PARP1-EJ which is characterized by a strong enrichment of both PARP1 and LigIII at damaged chromatin. Together, these data indicate that Ku is the main regulator for the hierarchal organization between C-NHEJ and PARP1-EJ.

Stochastic simulation of DNA double-strand break repair by non-homologous end joining based on track structure calculations

A Monte Carlo simulation model for DNA repair via the non-homologous end-joining pathway has been developed. Initial DNA damage calculated by the Monte Carlo track structure code PARTRAC provides starting conditions concerning spatial distribution of double-strand breaks (DSBs) and characterization of lesion complexity. DNA termini undergo attachment and dissociation of repair enzymes described in stochastic first-order kinetics as well as step-by-step diffusive motion considering nuclear attachment sites. Pairs of DNA termini with attached DNA-PK enter synapsis under spatial proximity conditions. After synapsis, a single rate-limiting step is assumed for clean DNA ends, and step-by-step removal of nearby base lesions and strand breaks is considered for dirty DNA ends. Four simple model scenarios reflecting different hypotheses on the origin of the slow phase of DSB repair have been set up. Parameters for the presynaptic phase have been derived from experimental data for Ku70/Ku80 and DNA-PK association and dissociation kinetics. Time constants for the post-synaptic phase have been adapted to experimental DSB rejoining kinetics for human fibroblasts after (137)Cs gamma irradiation. In addition to DSB rejoining kinetics, the yields of residual DSBs, incorrectly rejoined DSBs, and chromosomal aberrations have been determined as a function of dose and compared with experimental data. Three of the model scenarios obviously overestimate residual DSBs after long-term repair after low-dose irradiation, whereas misrejoined DSBs and chromosomal aberrations are in surprisingly good agreement with measurements.