PARG loss sensitizes cells to PARGi.

a, Ranking of co-essential genes with PARGi treatment on the basis of a DrugZ analysis of the results of CRISPR/Cas9 screening with a DDR library in HEK293A cells. The NormZ score was used to determine a possible synthetic lethality gene under PARGi treatment. Drug-sensitive genes were marked in red; drug-resistant genes were marked in blue on the basis of the false discovery rate (FDR, 0.05 cut-off). b, HEK293A WT cells and HEK293A PARG KO cells were treated with different doses of PARGi for 72 hours. Cell viability was determined by the CellTiter-Glo assay. c, HEK293A PARG KO cells, re-constituted with either full-length PARG or catalytic domain mutation of PARG, were treated with different doses of PARGi for 72 hours. Cell viability was determined by the CellTiter-Glo assay.

PARGi treatment induces NAD+- and PARP-dependent cell death in PARG KO cells.

a, HEK293A WT, PARG KO, PARP1/2 DKO, and PARG/PARP1/2 TKO cells were treated with PARGi (1µM) for 72 hours. The total cell lysates were immunoblotted with the indicated antibodies. b, HEK293A WT, PARG KO, PARP1/2 DKO, and PARG/PARP1/2 TKO cells were treated with DMSO or 10 µM PARGi for 4 hours and then fixed and stained with anti-pADPr antibody and propidium iodide (PI). c, HEK293A WT, PARG KO, PARP1/2 DKO clls, and PARG/PARP1/2 TKO cells were treated with different doses of PARGi for 72 hours. Cell viability was determined by the CellTiter-Glo assay. d, Relative NAD+ level in HEK293A WT and PARG KO cells with the indicated treatment for 48 hours. PARGi, 10 µM; PARPi, 10 µM; NAM, 100 µM; NMN, 1 mM; FK866, 10 nM. e, PARG KO cells were treated with PARGi (10 µM) or PARGi and NAM (100 µM) or NMN (1mM) for 48 hours. The total cell lysates were immunoblotted with the indicated antibodies. f, Results of clonogenic assays conducted using HEK293A PARG KO cells treated with PARGi (500 nM) or PARGi and NAM (100 µM) or NMN (1 mM) for 7 days.

PARGi treatment induces S phase-specific pADPr signaling in PARG KO cells.

a, HEK293A WT and PARG KO cells were treated with DMSO or 10 µM PARGi for 4 hours or 0.01% MMS for 30 min and then fixed and stained with anti-pADPr antibody or anti-γH2A.X antibody and PI. b, HEK293A WT and HEK293A PARG KO cells were mock-treated or pre-treated with 2 µM emetine for 90 min and then treated with PARGi for an additional 4 hours. Cells were fixed and stained with anti-pADPr antibody and PI. c, HEK293A PARG KO cells were synchronized with double thymidine block (DTB). Cells remained with DTB or were released from DTB, treated with 10 µM PARGi for 4 hours, and then fixed and stained with anti-pADPr antibody and PI. d, Results of clonogenic assays conducted using control cells and DTB synchronized or released HEK293A PARG KO cells treated with the indicated doses of PARGi for 7 days.

Prolonged PARGi treatment induces pADPr throughout the cell cycle and DDR in PARG KO cells.

a, Immunoblots of chromatin-bound PARP1 and PARylated proteins in HEK293A WT and PARG KO cells treated with PARGi (10 µM) for 4 hours. PARylated proteins were enriched by Af1521 beads. b, Sensitivity of HEK 293A PARG/PARP1/2 TKO and PARP1 reconstitution cells to PARGi. Cells were treated with different doses of PARGi for 72 hours, and cell viability was determined by the CellTiter-Glo assay. c, Prolonged PARGi treatment induces pADPr throughout the cell cycle in PARG KO cells. HEK293A WT and PARG KO cells were treated with PARGi (10 µM) for the indicated time and then fixed and stained with anti-pADPr antibody and violet. d, Immunoblotting of γH2A.X signals and other indicated proteins and modifications induced by prolonged PARGi+/-MMS treatment.

CRISPR screening identifies regulators of pADPr and cell viability.

a, Workflow of whole-genome CRISPR screens. For FACS-based CRISPR screening, 5 days after puromycin selection, cells were treated with PARGi (10uM) for 4 hours and then stained with anti-pADPr antibody and sorted with flow cytometry. Cells with strong signals (top 25%, TOP) and weak signals (bottom 25%, BOT) were selected. The sgRNAs from these cells were then sequenced and analyzed. For cell viability screening, cells were treated with or without PARGi for 21 days before collection. b, Scatter plot of DrugZ scores of PARG KO cells treated with or without PARGi treatment. The genes in the same pathway were marked with specific colors. A positive score indicates an enhanced pADPr signal, while a minus score indicates a decreased pADPr signal. c and d, Ranking of PARGi co-essential genes on the basis of a DrugZ analysis of the results of CRISPR/Cas9 screens performed with Toronto Knock Out Library (version 3) in HEK293A PARG KO cells and HEK239A cells. e, Analysis of biological processes of PARGi co-essential genes identified in HEK293A PARG KO cells and HEK239A cells. f, HEK293A WT cells, PARG KO, POLB KO, LIG1 KO, XRCC1 KO, and LIG3 knockdown cells were treated with different doses of PARGi for 72 hours. Cell viability was determined by the CellTiter-Glo assay. g, The cell Viability of HEK293A WT and PARG cells under POLB knockdown to PARGi. Cells were treated with different doses of PARGi for 72 hours.

PARG expression is a potential marker for PARGi sensitivity.

a, PARG mRNA level comparison between sensitive (RMGI, KURAMOCHI, and OVMANA) and resistant cells (COV362, COV318, OV56, OVISE, OVSAHO, CAOV3, and OVCAR3) from Pillay’s work on the basis of CCLE data. b, Clonogenic assay results of control and PARG knockdown HeLa cells with PARGi (2 µM) treatment for 7 days. shRNA knockdown efficiency was confirmed by an immunoblot of PARG. c, Ranked PARG expression level in ovarian cancer cell lines based on the CCLE database. The sensitive and resistant cells from Pillay’s work were labeled. d, Clonogenic assay results of OVCAR3 and RMUGS treated with or without PARGi (2 µM). e, Clonogenic assay results of RMUGS treated with PARGi (2 µM), PARPi (2 µM), or both. f, Left: Representative images of PARG IHC staining in breast and ovarian tissues and tumor samples to determine the PARG expression level. The summary is listed at the bottom. Scale bar, 200 µm. Right: The scatter plot of mean immunostaining intensity of PARG in each sample. The mean of each group was plotted.

PARG is essential for cell survival.

a, Diagram of full-length PARG was presented with the indicated gRNAs which target different regions in the C-terminal catalytic domain. The boundary of the catalytic domain was depicted based on Uniprot annotation. b, Immunoblotting was conducted to confirm the loss of PARG in PARG complete/conditional knockout (cKO) cells derived from HEK293A and HeLa cells, which were cultured in the presence of 100 nM olaparib. c, Clonogenic assay results of WT and PARG cKO cells treated with or without PARPi (100nM) for 7 days. d, Left: The immunoblots to confirm reconstitution with WT PARG or catalytic inactivation PARG in HEK293A PARG cKO cells. Right: Results of clonogenic survival assay with HEK293A PARG cKO cells reconstituted with WT or catalytic inactivation mutant of PARG for 7 days. e, Representative clonogenic results conducted in HEK293A PARG cKO cells treated with NAM (100 µM) or NMN (1 mM) for 7 days. f, HEK293A PARG cKO cells were synchronized with double thymidine block (DTB). Cells remained with DTB or were released from DTB for 4 hours, and then fixed and stained with anti-pADPr antibody and FxCycle Violet dye. g, Immunoblots of soluble and chromatin-bound PARP1 and pADPr levels in HEK293A WT and PARG cKO cells treated with DMSO or olaparib (10 µM) for 2 hours.

PARG KO cells are sensitive to active PARG inhibitors.

a, Chemical structures of active and inactive PARG inhibitors (PDD 00031705). b, Another two PARG inhibitors (PDD00017272 and PDD 00017238) were used to determine cell viability in HEK293A WT and PARG KO cells. c, HEK293A WT cells and PARG KO cells were treated with different doses of inactive PARGi (PDD 00031705) for 72 hours. Cell viability was determined by the CellTiter-Glo assay. d, Top: HeLa WT cells and PARG KO cells were treated with different doses of PARGi for 72 hours. Cell viability was determined by the CellTiter-Glo assay. Bottom: full-length PARG was presented with the indicated gRNA targeting regions, the C-terminal catalytic domain, the catalytic inactive mutation, and the PCNA-interaction motif based on Uniprot database annotation. e, HEK293A WT, PARG KO, and PARG KO cells reconstituted with WT PARG or catalytic inactivation PARG cells were treated with 0.01% MMS for 30 min. Blotting with anti-PARG and anti-pADPr antibodies was conducted to confirm the expression of PARG and PARG activity.

PARP-dependent pADPr and reduced NAD+ may contribute to cell death induced by PARGi treatment in PARG KO cells.

a, HEK293A WT and PARG KO cells were treated with DMSO or PARGi (10 µM) for 4 hours or 0.01% MMS for 30 min or the combination of PARGi (10 µM) for 4 hours and an additional 30 min with 0.01% MMS. The total cell lysates were immunoblotted with the indicated antibodies. b, The total cell lysates in Figure 2a were blotted with anti-pADPr antibody. c, Results of clonogenic assays conducted using HEK293A WT, PARG KO, PARP1/2 DKO, and PARG/PARP1/2 TKO cells treated with PARGi (1 µM) or PARPi (2 µM) or the combination of PARGi and PARPi for 7 days. d, Relative NAD+ level in HEK293A WT and PARG KO cells with the indicated treatment for different times. PARGi, 10 µM; FK866, 10 nM. e, Results of clonogenic assays conducted using HEK293A WT and PARG KO cells with the indicated treatment for 48 hours. f, Results of clonogenic assays were conducted using HeLa WT and PARG KO cells treated with PARGi (1 µM) or the combination of PARGi and NAM (100 µM) or NMN (1 mM) for 7 days.

Treatment with DNA damaging agents did not induce S phase-specific pADPr signaling.

HEK293A WT and PARG KO cells were treated with DMSO, 10 µM PARGi for 4 hours, or other DNA damaging agents and then fixed and stained with PI and anti-pADPr antibody (upper panel) or anti-γH2A.X antibody (bottom panel).

Uncontrolled S phase pADPr accumulation eventually leads to DNA damage and cell death.

a, Immunoblots of soluble and chromatin-bound PARP1/2 levels in different knockout cells treated with MMS, PARPi, or both. b, PARG/PARP1/2 TKO cells, reconstituted with PARP1 WT or trapping-deficient mutant (del.p119K120S), were treated with PARGi (10 µM) for 4 hours. The total cell lysates and PARylated proteins enriched by Af1521 beads were immunoblotted with the indicated antibodies. c, Alkaline comet assay results of HEK293A WT and PARG KO with PARGi (10 µM) treatment for different times. The comet-tail moments from 100 cells in each condition were measured and are shown in the box plot. The center line indicates the median, the box bounds indicate the first and third quartiles, and the whiskers indicate the maximum and minimum. d, Flow cytometry analysis of pADPr signaling in HeLa WT and PARG KO cells treated with PARGi (10 µM) for different times. e, PARG and PARP1 expression levels in HEK293A and HeLa cells. f, HeLa WT and PARG KO cells were treated with PARGi (10 µM) for different times or for an additional 30 min with 0.01% MMS. The total cell lysates were immunoblotted with the indicated antibodies.

Proteins involved in pADPr regulation contribute to PARGi sensitivity.

a, NormZ scores of sgRNA abundance between top and bottom groups in FACS-based anti-pADPr screens with HEK293A treated with PARGi (10 µM) for 4 hours. b, ARH3 loss and c, HPF1 loss show modestly increased sensitivity to PARGi in PARG KO cells. ARH3 or HPF knockout was confirmed by Western blotting and DNA sequencing. Cell viability was measured with the CellTiter-Glo assay after 3 days. d, XRCC1, LIG1, and POLB knockout in HEK293A was confirmed by Western blotting and DNA sequencing. LIG3 knockdown was validated by Western blotting, in which XRCC1 KO cells were included as controls. e, Flow cytometry analysis of pADPr signaling in cells treated with PARGi (10 µM) for 4 hours. PARG KO, LIG1 KO, and LIG3 KD cells treated with PARGi showed S phase-specific pAPDr signaling, while XRCC1 KO and POLB KO cells showed pAPDr signaling throughout the cell cycle. f, POLB was knocked down by shRNA in both HEK293A WT and PARG KO cells.

HR deficiency renders cells sensitive to PARGi.

a, Results of clonogenic assays conducted with RPE1 FLAG-Cas9 P53/BRCA1 DKO and PRE FLAG-Cas9 P53 KO cells with the indicated treatment for 14 days. b, Results of clonogenic assays conducted with inducible BRCA1 depletion of HeLa cells and corresponding 53BP1 KO cells, with or without BRCA1 depletion, which were co-treated with the indicated compounds for 7 days. BRCA1 was inducibly depleted by indole-3-acetic acid (IAA) and doxycycline (Dox) co-treatment; this was confirmed by Western blotting.

PARG loss is a robust marker of PARGi sensitivity.

a, Immunoblot of PARG in cells treated with PARG shRNA to knock down PARG. b, Results of clonogenic assays conducted in PARG knockdown cells. PARG was knocked down in the inducible BRCA1 depletion HeLa cells and corresponding 53BP1 KO cells. BRCA1 was inducibly depleted by IAA and Dox co-treatment. Cells were cultured for 7 days with the indicated treatment.

KO cells were validated by DNA sequencing.

gDNA sequences are also indicated for each KO cells.

DePARylation activity of PARG is essential for cell survival.

a, HEK293A cKO cells were validated by DNA sequencing. gDNA sequences are indicated at the top. b, The dePARylation activity was assessed in whole cell lysates (WCLs) prepared from control wild-type (WT), PARG KO, and PARG cKO HEK293A cells. The dePARylation activity was measured with remnant pADPr recognized by anti-pADPr antibody, and normalized with the Control and the Blank samples as indicated in Materials and Methods. n = 4 independent measurements. c, The PARG dePARylation activity in WCL prepared from control WT and PARG KO HEK293A was inhibited in a dose-dependent manner in the presence of PARGi. WCL prepared from PARG cKO cells was included as a control. n = 3 independent measurements. d, The PARG dePARylation activity shown in c is presented with PARGi concentrations.