RepairSwitch: simultaneous functional assessment of homologous recombination and end joining DNA repair pathways in living cells

DNA repair pathways are frequently defective in human cancers. DNA double strand breaks (DSBs) are most often repaired by either homologous recombination (HR) or non-homologous end joining (NHEJ). Alterations in repair pathways can indicate sensitivity to therapeutic agents such as PARP inhibitors, cisplatin, and immunotherapy. Thus, functional assays to measure rates of HR and NHEJ are of significant interest. Several methods have been developed to measure rates of HR or NHEJ; however, there is a need for functional cell-based assays that can measure rates by both major DNA DSB pathways simultaneously. Here, we describe the RepairSwitch assay, a flow cytometry assay to assess rates of HR and NHEJ mediated repair of Cas9 programmed DSB simultaneously using a novel fluorescence switching reporter system. The assay exhibits low background signal and is capable of detecting rare repair events in the 1 in 10,000 range. We demonstrate the utility of RepairSwitch by measuring the potency of inhibitors of ATM (KU-60019, KU-55933), DNA-PK (NU7441), and PARP (Olaparib) on modulating DSB repair rates in HEK293FT cells. The selective ATM inhibitor KU-60019 inhibited HR rates with IC50 of 915 nM. Interestingly, KU-60019 exposure led to a dose responsive increase in rates of NHEJ. In contrast, the less selective ATM inhibitor KU-55933, which also has activity on DNA-PK, showed inhibition of both HR and NHEJ. The selective DNA-PK inhibitor NU7441 inhibited NHEJ efficiency with an IC50 of 299 nM, and showed a dose responsive increase in HR. The PARP inhibitor Olaparib showed lower potency in modulating HR and NHEJ. We next used the RepairSwitch assay to assess how pharmacological and genetic inhibition of DNA methyltransferases (DNMT) impacted rates of HR and NHEJ. The DNMT inhibitor decitabine reduced HR, but increased rates of NHEJ, both in a dose responsive manner, in both HEK293FT and HCT116 cells (IC50 for HR of 187 nM and 1.4 uM respectively). Knockout of DNMT1 and DNMT3B increased NHEJ, while knockout of DNMT3B, but not DNMT1, reduced HR. These results illustrate the utility of RepairSwitch as a functional assay for measuring changes in rates of DSB repair induced by pharmacological or genetic perturbation. Furthermore, the findings illustrate the potential for one DNA repair mechanism to compensate in part for loss of another. Finally, we showed that inhibition of DNMT can lead to reduction of HR and increase in NHEJ, providing some additional insight into recently observed synergy of DNMT inhibitors with PARP inhibitors for cancer treatment.