PARG deficiency is neither synthetic lethal with BRCA1 nor PTEN deficiency

Abstract

Background: Poly(ADP-ribose) polymerase (PARP) inhibitors have entered the clinics for their promising anticancer effect as adjuvant in chemo- and radiotherapy and as single agent on BRCA-mutated tumours. Poly(ADP-ribose) glycohydrolase (PARG) deficiency was also shown to potentiate the cytotoxicity of genotoxic agents and irradiation. The aim of this study is to investigate the effect of PARG deficiency on BRCA1- and/or PTEN-deficient tumour cells.

Methods: Since no PARG inhibitors are available for in vivo studies, PARG was depleted by siRNA in several cancer cell lines, proficient or deficient for BRCA1 and/or PTEN. The impact on cell survival was evaluated by colony formation assay and short-term viability assays. The effect of simultaneous PARG and BRCA1 depletion on homologous recombination (HR) efficacy was evaluated by immunodetection of RAD51 foci and using an in vivo HR assay.

Results: The BRCA1-deficient cell lines MDA-MB-436, HCC1937 and UWB1.289 showed mild sensitivity to PARG depletion, whereas no sensitivity was observed for the BRCA1-proficient MDA-MB-231, MDA-MB-468, MCF10A and U2OS cell lines. However, the BRCA1-reconstituted UWB1.289 cell lines was similarly sensitive to PARG depletion than the BRCA1-deficient UWB1.289, and the simultaneous depletion of PARG and BRCA1 and/or PTEN in MDA-MB-231 or U2OS cells was not more cytotoxic than depletion of BRCA1 or PTEN only.

Conclusions: Some tumour cells displayed slight sensitivity to PARG deficiency, but this sensitivity could not be correlated to BRCA1- or PTEN-deficiency. Therefore, PARG depletion cannot be considered as a strategy to kill tumours cells mutated in BRCA1 or PTEN.

Keywords: BRCA; Cancer; Homologous recombination; Poly(ADP-ribose); Synthetic lethality.

Fig. 1

The BRCA1-deficient MDA-MB-436 and HCC1937 cells but not the BRCA-proficient MDA-MB-231 and MCF10A cells are slightly sensitive to PARG depletion. a The MDA-MB-436 cell line shows increased sensitivity to PARP inhibition compared to the MDA-MB-231 cell line. Colony formation assays were performed with indicated concentrations of PARP inhibitor KU0058948. Mean values of triplicates (± SD) from one representative of 5 independent experiments are shown. b The BRCA1-proficient MDA-MB-231 cell line is not sensitive to PARG depletion. Left panels Clonogenic survival of MDA-MB-231 cells transfected with siCTL, siPARG, AllNeg or siPARG5. Results are from 6 (siCTL and siPARG) and 3 (AllNeg and siPARG5) independent experiments. Middle panels Percentage of viable cells relative to non-targeting siRNA transfected cells 72 h post-transfection, time point when cells are re-plated for clonogenic or short-term MTS assay. Results show mean values ± SD of 7 (siCTL and siPARG) and 5 (AllNeg and siPARG5) independent experiments. Right panels Cell viability measured by MTS assays 144 h post-transfection. Results show the percentage of viability relative to cells transfected with non-targeting siRNA from 3 independent experiments. Lower panels PARG depletion was verified by western blot at the time post-siRNA transfection indicated. #: non-specific band. c Clonogenic survival (left panels), cell counting at 72 h post-siRNA transfection (middle panels) and short-term MTS assay at 144 h post-siRNA transfection (right panels) were performed in MDA-MB-436 cell line as described in b. Number of experiments was 7 (siCTL and siPARG) and 3 (AllNeg and siPARG5) for clonogenic assays, 8 (siCTL and siPARG) and 4 (AllNeg and siPARG5) for cell counting at 72 h and 3 for MTS assay at 144 h post-siRNA transfection. Lower panels PARG depletion was verified by western blot at the time post-siRNA transfection indicated. d Clonogenic survival of MCF10A (left panel) and HCC1937 (right panel) cells transfected with siCTL and siPARG. Results are depicted as in a from 6 independent experiments. p: *** < 0.001; 0.001 < ** < 0.01; 0.01 < * < 0.05; 0.05 < ns, not significant

Fig. 2

The BRCA1-mutated UWB1.289 cell line is not more sensitive to PARG depletion than the BRCA1-reconstituted UWB1.289 + BRCA1 cell line. a Left panels Clonogenic survival of UWB1.289 cells transfected with siCTL, siPARG, AllNeg or siPARG5. Results are depicted as box plots showing distribution of data from 7 (siCTL and siPARG) and 4 (AllNeg and siPARG5) independent experiments. Middle panels Percentage of viable cells relative to non-targeting siRNA transfected cells 72 h post-siRNA transfection, time point when cells are re-plated for clonogenic or short-term MTS assay. Results show mean values ± SD of 11 (siCTL and siPARG) and 4 (AllNeg and siPARG5) independent experiments. Right panels Cell viability measured by MTS viability assays 144 h post-transfection. Results show the percentage of viability relative to cells transfected with non-targeting siRNA from 3 independent experiments. Lower panels PARG depletion was verified by western blot at the time post-siRNA transfection indicated. b Clonogenic survival (left panels), relative cell number at 72 h post-siRNA transfection (middle panels) and short-term MTS assay at 144 h post-siRNA transfection (right panels) were performed in UWB1.289 + BRCA1 cell line exactly as described in a. Results are depicted as box plots showing distribution of data from 7 (siCTL and siPARG) and 4 (AllNeg and siPARG5) for clonogenic assays. Number of experiments was 11 (siCTL and siPARG) and 4 (AllNeg and siPARG5) for cell counting at 72 h and 3 for MTS assay at 144 h post-siRNA transfection. Lower panels PARG depletion was verified by western blot at the time post-siRNA transfection indicated. c. Spontaneous PAR accumulation is a consequence of efficient PARG depletion in UWB1.289 (UWB) and UWB1.289 + BRCA1 (UWB + BRCA1) cells. PAR, BRCA1, PARG and actin levels were analysed by western blot using the indicated antibodies. BRCA1 specific signal is indicated by arrowheads. p: *** < 0.001; 0.001 < ** < 0.01; 0.01 < * < 0.05; 0.05 < ns, not significant

Fig. 3

Co-depletion of PARG does not potentiate cytotoxicity by BRCA1 depletion. a Clonogenic survival (left panel) of MDA-MB-231 cells after single or combined siRNA-mediated depletion of BRCA1 and PARG. Results are depicted as box plots showing distribution of data from 4 individual experiments. Middle panel Percentage of viable cells relative to siCTL-transfected cells 72 h post-transfection, time point when cells are re-plated for clonogenic or short-term MTS assay. Results show mean values ± SD of 7 independent experiments. Right panel Cell viability measured by MTS viability assays 144 h post-transfection. Results show the percentage of viability relative to cells transfected with siCTL from 3 independent experiments. Lower panel PARG and BRCA1 depletions were verified by western blot at the times indicated. b Clonogenic survival (left panel) of U2OS cells after single or combined siRNA-mediated depletion of BRCA1 and PARG. Results are depicted as box plots showing distribution of data from 4 individual experiments. Percentage of viable cells relative to siCTL-transfected cells 72 h post-transfection (right panel). Results show mean values ± SD of 5 independent experiments. PARG and BRCA1 depletions at 72 h were verified by western blot (middle panel). BRCA1 specific signal is indicated by arrowhead. p: *** < 0.001; 0.001 < ** < 0.01; 0.01 < * < 0.05; 0.05 < ns, not significant

Fig. 4

The HR defect caused by BRCA1-depletion is not further affected by PARG depletion. a and b The frequency of HR-mediated repair events was analysed by flow cytometry in U2OS-DR-GFP-mCherry-I-SceI-GR cells after transfection with the indicated siRNA and induction of DSB formation by incubation with TA for 48 h. The percentage of Cherry- and GFP-positive cells is indicated. In b, the values correspond to the ratio of GFP-positive cells relative to cells transfected with scrambled siRNA (siCTL), as illustrated in a, and represent the mean ± SD of 7 independent experiments. c and d Defect of etoposide-induced RAD51 foci formation in BRCA1-depleted cells is not further affected by PARG depletion. Cells transfected with the indicated siRNA were incubated with 10 µM etoposide for 1 h, released into fresh medium for 2 h and processed for immunofluorescence using anti-RAD51 and anti-γH2AX antibodies. Representative immunofluorescence image are shown in c. In d, the box plot graph depicts the percentage of cells displaying RAD51 foci (more than 10 RAD51 foci per cell) from 5 independent experiments scoring >200 nuclei for each condition. p: 0.01 < * < 0.05; 0.05 < ns, not significant

Fig. 5

PARG deficiency is not synthetic lethal with PTEN deficiency. a The PTEN-mutated MDA-MB-468 cell line is not sensitive to PARG depletion. Clonogenic survival (left panel) of MDA-MB-468 cells after siCTL and siPARG transfection. Results are depicted as box plots showing distribution of data from 5 individual experiments. Relative cell number 72 h post-siRNA transfection (right panel) is counted from 5 individual experiments. PARG depletion was verified by western blot (middle panel). b Clonogenic survival (left panel) of MDA-MB-231 cells after single or combined siRNA-mediated depletion of BRCA1, PARG and PTEN. Results are depicted as box plots showing distribution of data from 5 independent experiments. Percentage of viable cells relative to siCTL-transfected cells 72 h post-transfection (right panel) is counted from 5 individual experiments. BRCA1, PARG and PTEN siRNA-mediated depletions compared to untransfected or siCTL transfected cells were verified by western blot (middle panel). The arrow points to BRCA1 signal above a non-specific band (#). p: 0.01 < * < 0.05; 0.05 < ns, not significant