Polymerase theta (Polθ) acts in DNA replication and repair, and its inhibition is synthetic lethal in BRCA1 and BRCA2-deficient tumor cells. Novobiocin (NVB) is a first-in-class inhibitor of the Polθ ATPase activity, and it is currently being tested in clinical trials as an anti-cancer drug. Here, we investigated the molecular mechanism of NVB-mediated Polθ inhibition. Using hydrogen deuterium exchange-mass spectrometry (HX-MS), biophysical, biochemical, computational and cellular assays, we found NVB is a non-competitive inhibitor of ATP hydrolysis. NVB sugar group deletion resulted in decreased potency and reduced HX-MS interactions, supporting a specific NVB binding orientation. Collective results revealed that NVB binds to an allosteric site to block DNA binding, both in vitro and in cells. Comparisons of The Cancer Genome Atlas (TCGA) tumors and matched controls implied that POLQ upregulation in tumors stems from its role in replication stress responses to increased cell proliferation: this can now be tested in fifteen tumor types by NVB blocking ssDNA-stimulation of ATPase activity, required for Polθ function at replication forks and DNA damage sites. Structural and functional insights provided in this study suggest a path for developing NVB derivatives with improved potency for Polθ inhibition by targeting ssDNA binding with entropically constrained small molecules.

Orderly chromosome segregation is enabled by crossovers between homologous chromosomes in the first meiotic division. Crossovers arise from recombination-mediated repair of programmed DNA double-strand breaks (DSBs). Multiple DSBs initiate recombination, and most are repaired without crossover formation, although one or more generate crossovers on each chromosome. Although the underlying mechanisms are ill-defined, the differentiation and maturation of crossover-specific recombination intermediates requires the cyclin-like CNTD1. Here, we identify PRR19 as a partner of CNTD1. We find that, like CNTD1, PRR19 is required for timely DSB repair and the formation of crossover-specific recombination complexes. PRR19 and CNTD1 co-localise at crossover sites, physically interact, and are interdependent for accumulation, indicating a PRR19-CNTD1 partnership in crossing over. Further, we show that CNTD1 interacts with a cyclin-dependent kinase, CDK2, which also accumulates in crossover-specific recombination complexes. Thus, the PRR19-CNTD1 complex may enable crossover differentiation by regulating CDK2.