Abstract
Abstract 3865
Chronic Lymphocytic Leukaemia (CLL) is characterised by marked clinical heterogeneity and tumours with deletion or mutation of the TP53 and ATM genes on chromosomes 17p and 11q are associated with poor outcome. ATM is a protein kinase that, following induction of DNA double strand breaks (DSBs), phosphorylates a number of downstream targets and synchronises a network of cellular responses including p53 activation that leads to induction of pro-apoptotic genes and activation of apoptosis. Consequently, loss of integrity of ATM/p53 pathway results in apoptotic resistance, retention of cells with genomic damage, and tumour progression.
Cellular p53 levels are regulated through a p53-Mdm2 regulatory loop whereby Mdm2, a ubiquitin ligase, facilitates p53 polyubiquitination and targeting of p53 for proteasome degradation. It was recently shown that stabilisation of p53 can be achieved by manipulation of this regulatory loop through use of small molecule inhibitors of this p53 degradation pathway, termed nutlins, which prevent p53 ubiquitination. More recently, a class of deubiquitinating enzymes (DUBs) highlighted an additional level of p53-Mdm2 regulation. In particular, a specific DUB, (USP7/HAUSP), has a high affinity for Mdm2 and functions by antagonizing Mdm2 ubiquitination. Unlike nutlins, USP7 also has been implicated in the regulation of cell cycle, mitosis and DNA damage response and we reasoned that USP7 inhibition may sensitise CLL tumours with ATM and TP53 defects.
Our analysis of 25 primary CLL tumours with different ATM and TP53 status indicated that USP7 was robustly expressed in all CLL tumour cells tested. Through collaboration with Hybrigenics we obtained a specific USP7 inhibitor, HBX19818. To determine if cell killing could be induced in ATM or TP53 deficient tumours by inhibiting the USP7-Mdm2 pathway, we analysed the induction of cell death over a range of HBX19818 concentrations using both isogenic CLL cell lines with and without ATM and/or p53, as well as 18 primary CLL tumours. We observed a significant cytotoxic effect of HBX19818 in isogenic CLL lines, at concentrations between 1–10μM, irrespective of their ATM and TP53 status. Strikingly, the majority of primary CLL tumours were sensitive to HBX19818 concentrations between 8μM and 16μM to which non-tumour PBMCs were resistant.
Western blotting was used to monitor the induction of p53, apoptosis (associated with caspase 7 and PARP1 cleavage), and also whether DNA damage was induced (as measured by H2AX phosphorylation) in response to HBX19818. Our analysis revealed that pharmacological inhibition of USP7 led to p53 upregulation in the p53 proficient CLL cells associated with a robust induction of p21 indicating that the stabilised p53 was active. This response was absent in Mec-1 cell line with non-functional p53. Interestingly, caspase 7 and PARP1 were only cleaved in the ATM wild type CLL cell lines suggesting that activated p53 was capable of inducing cell death. In contrast, the ATM and p53 deficient CLL cell lines did not exhibit any markers suggesting an apoptotic mode of cell death. Rather, these cell lines displayed elevated levels of phospho-H2AX, suggesting the induction of DNA damage, possibly caused by an underlying DNA repair defect. Consistent with our hypothesis, both p53 proficient and p53 non-functional CLL cell lines failed to induce the recruitment of the HR protein Rad51 to sites of IR-induced DNA double strand breaks.
Taken together, our data implies that in addition to p53 activation, pharmacological inhibition of USP7 can exert a cytotoxic effect by further mechanisms, possibly by modulating DNA double strand break repair. This is consistent with previous reports suggesting that USP7 regulates monoubiquitination of transcription factor FOXO4 and is involved in the regulation of DNA repair and mitotic progression via its interactions with Claspin and Chfr respectively. We suggest that pharmacological inhibition of USP7 represents a promising target for the treatment of tumours with defective ATM and p53 signalling.
Disclosures:
No relevant conflicts of interest to declare.
Stochastic multicellular modeling of x-ray irradiation, DNA damage induction, DNA free-end misrejoining and cell death
