Deletions at 11q are common in human malignancies and associate with poor clinical outcomes. In this context, deletions at 11q22, containing the Ataxia-Telangiectasia Mutated (ATM) locus, are frequently observed in B cell lymphomas (BCL) and, combined with mutation of the second allele, render these tumors ATM null. In contrast, deletions at 11q23, containing the histone H2AX locus, result in monoallelic H2AX expression in a subset of neuroblastomas (NB). Given the prominent roles for the ATM kinase and its substrate histone H2AX in the signaling and repair of DNA double-strand breaks (DSB), alterations in their copy number may result in differential responses to DNA damaging agents in cancer cells. In this context, PARP inhibitors (PARPi) are a novel class of DNA damaging chemotherapeutic agents that preferentially eliminate cells with defects in replication-associated DSBs. Based on these previous observations, we propose here to test the hypothesis that monoallelic expression of H2AX in 11q-deleted malignancies sensitizes them to PARPi, by conferring a defect in Homologous Recombination (HR)-mediated repair of PARPi-induced DSBs in human cancer cells. In support of our hypothesis, we provide strong genetic evidence that H2AX deficiency is synthetic lethal with PARP inhibition. First, we find that combined loss of H2AX and either of the two main PARPi targets, PARP1 or PARP2, results in embryonic lethality in the mouse. Secondly, we demonstrate a defect in the repair of replication-associated DSBs after treatment with PARPi, in an H2AX gene dose- dependent manner. Lastly, we demonstrate that H2AX is also limiting for DSB repair in NB cells with 11q23 deletion and monoallelic H2AX expression. To test our hypotheses, we will employ a combination of biochemical, molecular and cytogenetic assays on mouse primary and transformed cells and in human NB cells. Specifically, experiments in Aim 1 will employ murine cells to test the hypothesis that H2AX becomes limiting for HR-mediated repair of PARPi-induced lesions, characterize the underlying genetic pathway and define the relative contribution of PARP1 and PARP2 to these phenotypes.
In Aim 2, we will assess PARPi sensitivity in 11q23-deleted NB cells as a function of H2AX gene dose and define roles for PARP1 and PARP2 in DSB repair in human cancer cells with reduced H2AX expression.
In Aim 3, we will employ a novel murine model to examine nonoverlapping functions for ATM and H2AX in the repair of PARPi-induced lesions, modeling the subset of human cancers that co-delete the two factors. In the longer term, knowledge gained from these exploratory studies will increase our understanding of how PARPi interfere with the mechanisms that normally protect the genome during replication and facilitate the development of future clinical trials for biomarker development in a variety of human malignancies with 11q abnormalities.
Some human cancers carry a deletion in the long arm of chromosome 11 that results in the loss of the DNA repair factors ATM and/or H2AX. We find that loss of either ATM or H2AX makes cell more sensitive to therapy with a novel class of anticancer agents, PARP inhibitors. Our proposed studies will investigate the molecular mechanisms underlying this sensitivity and also determine whether tumor cells that have lost the two factors are more sensitive to PARP inhibitors relative to tumor cells that have lost ATM alone.
|Ghosh, Rajib; Roy, Sanchita; Kamyab, Johan et al. (2016) Common and unique genetic interactions of the poly(ADP-ribose) polymerases PARP1 and PARP2 with DNA double-strand break repair pathways. DNA Repair (Amst) 45:56-62|