Inhibition of Histone Deacetylase Activity Compromises Homologous Recombination Repair and Increases Sensitivity of Chemo-Resistant Chronic Lymphocytic Leukemia Cells to Olaparib

Abstract 3891

Loss of ATM or TP53 gene is associated with chemo-resistant form of B cell chronic lymphocytic leukemia due to inability of tumour cells to activate DNA damage induced apoptosis. For ATM deficient tumours an alternative treatment strategy is to impose requirement for double strand break (DSB) repair by inhibition of single strand break (SSB) DNA repair. DNA DSBs can be repaired using either the error-free homologous recombination repair (HRR) pathway or the error-prone non-homologous end-joining (NHEJ) pathway, both of which are under the ATM control. Consequently, ATM-deficient B cell chronic lymphocytic (CLL) leukemia cells are selectively sensitive to Olaparib, the inhibitor that targets SSB repair protein PARP. We have recently observed that this effect can be enhanced by addition of (HDAC) inhibitor sodium valproate (VPA).

Given the relatively low toxicity of both HDAC and PARP inhibitors, in this study we have focused on the combination of Olaparib and the range of clinically applicable HDAC inhibitors. We tested the hypothesis that HDAC inhibition can both alter DNA repair and induce apoptosis in CLL cells treated with Olaparib irrespective of ATM and TP53 status through its ability to modify chromatin structure. We investigated the cytotoxic effect and mode of cellular death induced by combined of HDAC/PARP inhibition in 18 proliferating primary CLL tumour cells and CLL cell lines with different ATM and TP53 status.

First, we have analysed three class I/II HDAC inhibitors (sodium valproate, Vorinostat and Belinostat) each of which has shown promise in the treatment of haematological malignancies in phase I and II clinical trials. We interrogated isogenic CLL cell lines (with and without ATM knock down) containing either functional p53 (PGA, CII) or mutant p53 (Mec1). We observed that all cell lines could be sensitised to all three HDAC inhibitors by 48hrs pre-treatment with Olaparib, and that this sensitisation was independent of TP53 and ATM status. This observation was recapitulated in primary CLL tumour cells induced to proliferate in the presence of CD40L expressing mouse fibroblasts and IL21.

To elucidate the mechanism of killing we addressed the impact of HDAC inhibition on HRR. Quantification of gamma H2AX and HRR associated RAD51 intranuclear DNA foci following induction of DSBs with ionising radiation showed that HRR was significantly reduced by HDAC inhibitor treatment in ATM-wt but not ATM-deficient CLL cells which were already HRR deficient. Furthermore, using western blotting, we observed that Caspase 3 and PARP cleavage was induced by all three HDAC inhibitors in primary CLL tumour cells irrespective of their ATM and TP53 status. Finally, the absence of increased p53 levels in HDAC treated cells suggested a Caspase-dependent but p53-independent mechanism of apoptosis.

Together, these findings suggest that HDAC inhibitors can both compromise HRR and induce apoptosis in CLL cells independent of ATM and TP53 status. As PARP inhibitors are already in clinical trials for ATM deficient tumours, the addition of HDAC inhibitors to Olaparib-based schedules should be considered as a treatment option for a wider range of chemo-resistant CLL tumours.