The Novel PARP Inhibitor, AZD2281, Targets Apoptotic Resistant ATM Mutant CLL Tumours for Cellular Killing.

B-CLL is incurable and the development of resistance to standard chemotherapeutics remains an important problem. ATM mutations lead to chemo-resistance in a significant proportion of CLL patients and this is due to impairment in the activation of the DNA-damage induced ATM/p53 apoptotic pathway. Therefore, there is a requirement for novel treatments for ATM mutant CLL tumours, which induce cell death by mechanisms that are independent of this pathway.

Breast cancer cells with homozygous BRCA1/2 mutations are deficient in the repair of DNA double strand break (DSB) by homologous recombination (HR). It has recently been shown that inhibition of PARP activity, which is required for repair of DNA single strand breaks (SSB), can lead to selective sensitisation of tumours harbouring BRCA1/2 mutations in replicating cells. The mechanism involves the continual progression of DNA SSBs into DSBs, which in the presence of defective HR, results in accumulation of DSBs and activation of cell death via mitotic catastrophe. ATM regulates the balance between the repair of DNA DSBs and the induction of the DNA DSB apoptotic pathway. Therefore, the phenotype of ATM mutant CLL cells includes a repair defect as well as an apoptotic defect. By analogy with the BRCA study, we investigated whether PARP inhibition can sensitise ATM mutant CLL tumours.

We addressed the in vitro cytotoxicity of a similar PARP inhibitor (PARPi), AZD2281 produced by KuDOS, in 20 CLL tumours, including 10 ATM mutant and 10 ATM wild type. We analysed each of the CLL tumours in non-cycling and cycling states in culture. The aim was to mimic the CLL tumour populations in vivo, which are believed to consist of non-cycling peripheral blood tumour cells and cycling lymphoid tissue tumour cells. Given the mechanism of PARP activity, we predicted that its inhibition would preferentially be toxic in the cycling CLL population. In our experiments, CLL cycling was induced using a CD40L/IL4 support system and verified by incorporation of tritiated thymidine or BrdU. Consistent with our expectations, non-cycling CLL cells did not show significant cytotoxicity to increasing doses of AZD2281 (0.5–10μM). By comparison, in cycling CLLs there was increased sensitivity to AZD2281 at concentrations of 1.5μM or greater in ATM mutant compared to ATM wild type tumours. Furthermore, protein analysis revealed that treatment with AZD2281 did not induce the up regulation of p53 or the cleavage of caspases and that the killing in ATM mutant cells did not require induction of Atm/p53 dependent apoptosis. In non-cycling ATM mutant CLL cells, we also investigated whether pre-incubation with the AZD2281 could sensitize cells to DNA damaging chemotherapeutics. Interestingly, we found that 24 hours pre-treatment with AZD2281 rendered ATM mutant non-cycling cells sensitive to Fludarabine.

We conclude that the PARP inhibitor AZD2281 is capable of targeting CLL cells with defective ATM function for cellular killing. Furthermore, our results indicate that inhibition of PARP is particularly important in the induction of cell death within proliferating ATM mutant CLL cells, which play a major role in tumour progression. Also, the addition of AZD2281 can sensitise non-cycling ATM mutant tumour cells to killing by Fludarabine. Our results suggest that this compound could be effective in the treatment of CLL patients with apoptotic resistant ATM mutant tumours.