What is the nature of the DNA replication vulnerability that belies PARG inhibitor sensitivity,
and how do we exploit this insight to develop predictive biomarkers?


ADP-ribosylation is a reversible post-translational modification that influences a plethora of biological functions including transcription, translation, stress responses, mitosis, microbial pathogenicity and DNA damage repair. In humans, there are at least 17 ADP-ribosyl transferases, with the poly(ADP-ribose) (PAR) polymerases PARP1, PARP2 and the Tankyrases assembling branched PAR chains on themselves and various other targets. PAR chains are turned over rapidly with the activity of Poly (ADP-ribose) glycohydrolase (PARG) mainly responsible for dismantling long PAR chains.

Pioneering studies in 2005 showed that BRCA-mutant cells are exquisitely sensitive to PARP1/2 inhibition, as a result of their deficiency in homologous recombination (HR), leading to PARP inhibitors now being used to treat various BRCA-mutant cancers. While ~20% of high-grade serous ovarian cancer cases have BRCA1/BRCA2 mutation, a further ~30% are HR-deficient due to other oncogenic lesions and thus might also benefit from treatment with PARP inhibitors. This leads to our central question: how can we improve outcomes for women with HR-proficient disease? One way by which we aim to target HR-proficient ovarian cancers is by exploiting PAR dynamics through inhibition of PARG.

We initially exposed a panel of six ovarian cancer cell lines to a PARG inhibitor (PDD00017273) and the PARP1/2 inhibitor olaparib (PARPi). While four lines proliferated in both inhibitors, Kuramochi and OVCAR3 displayed differential sensitivities; Kuramochi proliferation was suppressed by PARGi, OVCAR3 proliferation was suppressed by PARPi. This differential sensitivity indicates that PARG inhibitors present a new opportunity to treat HR-proficient ovarian cancers not expected to respond to PARP inhibitors.

More recent characterisation of 10 ovarian cancer cell lines, including four PARGi-sensitive lines, finds that intrinsic PARGi sensitivity correlates with both DNA replication catastrophe and reduced expression of DNA replication genes compared with PARGi resistant lines. Indeed, our initial “synthetic lethal” siRNA screen identified DNA replication genes, in particular TIMELESS (Figure), supporting the role of replication catastrophe. CHK1 also emerged from the synthetic lethality siRNA screen, presenting the opportunity to broaden the therapeutic potential of PARGi; combination matrices with PARGi and the CHK1 inhibitor AZD7762 indicate synergistic effects. Importantly, we have confirmed these cell-line-derived observations using our patient-derived ex vivo cultures.

We now aim to address several questions to support development of PARG inhibitors towards first-in-human trials. In particular, what is the nature of the DNA replication vulnerability that belies PARG inhibitor sensitivity, and how do we exploit this insight to develop predictive biomarkers? As PARG inhibitor-sensitive cell lines exhibit reduced levels of DNA replication genes, a ‘‘replication stress” signature may have merit as a predictive biomarker. Therefore, an important next step will be to also evaluate these genes in a wider collection of our patient-derived HGSOC samples with defined PARGi sensitivity. Indeed, we are using the transcriptomics and proteomics datasets that we are generating from our living biobank for this purpose.

Key references:


©2020 by The Taylor Lab. Proudly created with