Background: PARP inhibitors (PARPi) are under clinical investigation as anticancer therapies in several indications with a focus on monotherapy in tumors with homologous recombination deficits and on combination regimens with DNA damaging chemotherapy. The combination activity of PARPi with DNA alkylating agents is due in part to trapping of PARP1 onto base excision repair intermediates. Recent studies in mice indicate that potent trapping activity limits the tolerability of PARPi in this setting. While trapping has not been directly demonstrated with PARP inhibitors alone, genetic and correlative data suggest that it may be important in the monotherapy setting as well. The objective of this study was to further delineate the role of trapping in single-agent cytotoxicity of PARPi in cancer cells and healthy bone marrow.
Methods: Cytotoxic potency was determined for cancer cells using CellTiter-Glo and for erythroid and myeloid progenitors using colony forming unit (CFU) assays. PARP1 trapping was evaluated by cellular fractionation, TR-FRET and BLI. PARPi binding affinities and off-rates were determined via SPR. Cellular PAR levels were determined by ELISA.
Results: Talazoparib, rucaparib and the novel PARPi A-934935 bound to PARP1 with similar affinities and off-rates and inhibited cellular PAR synthesis with identical potencies. Despite these similarities, talazoparib was a more potent trapping agent. Consistent with the increased trapping activity, talazoparib was >10-fold more potent than rucaparib and A-934935 at inhibiting proliferation of erythroid progenitors and >50-fold more potent in myeloid progenitors. Similar differences were observed in a panel of human cancer cell lines including BRCA−/− breast cancer and small cell lung cancer.
Conclusions: All NAD+-competitive PARPi elicit PARP1 trapping, but different inhibitors display varying degrees of resolution between trapping and cellular PAR depletion. Understanding the relative importance of these mechanisms to the tolerability and efficacy of PARP inhibitors is of critical importance. Myelosuppression is a commonly observed toxicity in PARPi clinical trials. In the present study we used ex vivo bone marrow CFU assays to model myelosuppression. The IC50s observed were within the range of exposures observed at recommended phase II doses in monotherapy clinical trials, highlighting the clinical relevance of this experimental approach. Our data reveal that compounds inhibiting cellular PAR synthesis with equal potency span a broad range of potency in BM-CFU assays in a manner predicted by PARP1 trapping activity. These results suggest that PARP1 trapping is a primary driver of myelosuppression by PARP inhibitors. Our observation that trapping drives cytotoxicity to a similar extent in bone marrow and cancer cells sensitive to single-agent PARP inhibition suggests that PARP1 trapping activity is not advantageous with respect to the therapeutic index of PARPi in a monotherapy setting. This is consistent with observations related to the therapeutic index of PARP inhibitors in the clinic. These results also have important implications for selection of a PARPi for combination regimens where trapping is not required for efficacy and in which myelosuppression may be a concern, such as those containing platinum drugs or topoisomerase I inhibitors.
Citation Format: Todd A. Hopkins, Julie L. Wilsbacher, Enrico L. DiGiammarino, Sanjay C. Panchal, Gui-Dong Zhu, Thomas D. Penning, Eric F. Johnson, David Maag. PARP1 trapping activity of PARP inhibitors is associated with cytotoxicity in both cancer cells and healthy bone marrow. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr C52.
- ©2015 American Association for Cancer Research.