Targeting an RNaseH2 Defect in Chronic Lymphocytic Leukaemia with PARP Inhibitors

The therapeutic exploitation of molecular defects within the DNA damage response (DDR) in tumour cells has become an important treatment paradigm. 'Synthetic lethality' relies on pharmacological inhibition of pathways upon which DDR-deficient tumour cells have become dependent for their survival. This induces an intolerable level of unrepaired DNA damage in the tumour cells resulting in cell death, whilst sparing DDR-proficient normal cells

Deletion of 13q14 is a frequent, early event in the pathogenesis of CLL. Alongside well-described tumour suppressor genes this genomic region also encompasses the DDR gene, RNASEH2B, which encodes a subunit of the heterotrimeric enzymatic complex, RNaseH2. This complex is a principal component of ribonucleotide excision repair (RER), a DDR pathway that removes ribonucleotides incorporated into DNA by error prone DNA repair polymerases. If unremoved, these DNA-incorporated ribonucleotides lead to DNA damage, chromosome instability and mutagenesis (Reijns et al, Cell. 2012;149:1008).

We recently reported a synthetically lethal interaction between the functional loss of RNaseH2 enzymatic complex and PARP inhibition (Zimmerman et al, Nature 2018, 559:285). We observed that inactivation of any of the three RNAseH2 subunits (A,B,C) leads to loss of enzymatic activity of this complex and also that primary CLL tumours with 13q14 deletion involving the RNASEH2B locus are sensitive to PARP inhibitors (PARPi) in vitro.

In light of these preliminary observations, we addressed the following questions: a) Do monoalleic and biallelic RNASEH2B deletions have equal consequences for RNAseH2 enzymatic activity and sensitivity to PARP inhibition in CLL? d) Can loss of RNAseH2 activity be caused by an alternative mechanism, such as mutations in RNASEH2B? c) Can the PARPi sensitivity of RNaseH2-deficient CLLs be demonstrated in vivo, in patient-derived xenografts? d) Is PARP inhibition an option for RNAseH2 deficient tumours with limited response to other treatments?

Analysis of 100 primary CLL tumours through a combination of multiplex ligation-dependent probe amplification (MLPA), CGH microarrays and Sanger sequencing identified 29 tumours with monoallelic and 14 with biallelic RNASEH2B deletions. None of the analysed tumours had mutations in RNASEH2B. Increased levels of genomic ribonucleotides were confirmed in all RNASEH2B deleted tumours by two complementary methods: alkaline gel electrophoresis and DNA nick translation. We found that the RNaseH2 enzymatic defect and sensitivity to PARP inhibition were evident in all RNASEH2B deleted tumours, but were more profound in those harbouring biallelic deletion compared to tumours that have lost only one RNASEH2B allele. Furthermore, sensitivity to PARP inhibitors was dependent on PARP-trapping capacity and therefore cytotoxicity was most prominent in response to PARP-inhibitors with a potent PARP trapping capacity such as talazoparib. In vivo experiments revealed similar trends, with CLL xenografts derived from tumours with biallelic RNASEH2B deletion being differentially sensitive to Talazoparib. Notably, the PARP inhibition sensitivity of RNAseH2-deficient primary CLLs was independent of patients' response to different treatments.

In summary, we conclude that the RNASEH2B loss associated with 13q14 deletion represents a frequent cause of RNaseH2 enzymatic defect that renders primary CLL tumours sensitive to PARP-trapping inhibitors. Our findings expand the range of molecular defects in CLL that are amenable to treatment with clinically applicable PARP inhibitors and may have implications for the management of patients with limited response to other treatments.