Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Jul;559(7713):285-289.
doi: 10.1038/s41586-018-0291-z. Epub 2018 Jul 4.

CRISPR Screens Identify Genomic Ribonucleotides as a Source of PARP-trapping Lesions

Affiliations
Free PMC article

CRISPR Screens Identify Genomic Ribonucleotides as a Source of PARP-trapping Lesions

Michal Zimmermann et al. Nature. .
Free PMC article

Abstract

The observation that BRCA1- and BRCA2-deficient cells are sensitive to inhibitors of poly(ADP-ribose) polymerase (PARP) has spurred the development of cancer therapies that use these inhibitors to target deficiencies in homologous recombination1. The cytotoxicity of PARP inhibitors depends on PARP trapping, the formation of non-covalent protein-DNA adducts composed of inhibited PARP1 bound to DNA lesions of unclear origins1-4. To address the nature of such lesions and the cellular consequences of PARP trapping, we undertook three CRISPR (clustered regularly interspersed palindromic repeats) screens to identify genes and pathways that mediate cellular resistance to olaparib, a clinically approved PARP inhibitor1. Here we present a high-confidence set of 73 genes, which when mutated cause increased sensitivity to PARP inhibitors. In addition to an expected enrichment for genes related to homologous recombination, we discovered that mutations in all three genes encoding ribonuclease H2 sensitized cells to PARP inhibition. We establish that the underlying cause of the PARP-inhibitor hypersensitivity of cells deficient in ribonuclease H2 is impaired ribonucleotide excision repair5. Embedded ribonucleotides, which are abundant in the genome of cells deficient in ribonucleotide excision repair, are substrates for cleavage by topoisomerase 1, resulting in PARP-trapping lesions that impede DNA replication and endanger genome integrity. We conclude that genomic ribonucleotides are a hitherto unappreciated source of PARP-trapping DNA lesions, and that the frequent deletion of RNASEH2B in metastatic prostate cancer and chronic lymphocytic leukaemia could provide an opportunity to exploit these findings therapeutically.

Conflict of interest statement

Competing Financial Interests DD and TH are advisors to Repare Therapeutics.

Figures

ED Figure 1
ED Figure 1
Related to Figure 1. a, Cas9 immunoblot of whole cell extracts (WCEs) from parental HeLa, RPE1-hTERT and SUM149PT cells and clones stably transduced with a lentiviral FLAG-Cas9-2A-Blast construct (representative of n ≥ 2 biologically independent experiments). Tubulin, loading control. b, Validation of CRISPR/Cas9 gene editing efficiency in Cas9-expressing HeLa, RPE1-hTERT and SUM149PT clones. Cell proliferation was monitored after transduction with a control sgRNA construct (sgLacZ) or sgRNAs targeting essential genes PSMD1, PSMB2 and EIF3D. Solid circles, individual values. Bars, mean ±SD (normalized to sgLacZ, n = 3 technical replicates), c, Gene ontology (GO) terms significantly (P < 0.05, binomial test with Bonferroni correction) enriched among hits from olaparib screens common to at least two cell lines. Enrichment was analyzed using PANTHER. d, esyN network analysis of interactions between hits common to at least two cell lines. Node size corresponds to mean DrugZ score across cell lines. 77/155 genes are mapped on the network.
ED Figure 2
ED Figure 2
Related to Figure 2a,b. a, CRISPR-mediated inactivation of RNASEH2A or RNASEH2B in the cell lines used in this manuscript. WCEs of indicated cell lines and genotypes were processed for immunoblotting using antibodies against RNASEH2A, RNASEH2B or RNASEH2C. Vinculin, tubulin and GAPDH, loading controls. Representative immunoblots (of n ≥ 2 biologically independent experiments). b-d. Abolished RNase H2 enzymatic activity and increased levels of genome-embedded ribonucleotides in RNASEH2A-KO cells. b. Representative (n = 3 biologically independent experiments) analysis of total nucleic acids from WT and RNASEH2A-KO HeLa cells treated with recombinant RNase H2 and separated by alkaline agarose gel electrophoresis. Ribonucleotide-containing genomic DNA from RNASEH2A-KO HeLa cells is nicked and therefore has increased electrophoretic mobility. c, Densitometric quantification of the alkaline gel shown in b. d, Cleavage of an RNase H2-specific double-stranded DNA oligonucleotide with a single incorporated ribonucleotide (DRD:DNA; ribonucleotide position is shown in red) by WT and RNASEH2A-KO WCEs of the indicated cell types was measured using a fluorescence quenching-based assay. Individual values (open circles) with mean (red lines, n = 3 biologically independent experiments). e-l, RNase H2 deficiency leads to PARPi sensitivity in multiple cell types. e-h, Clonogenic survival assays of the indicated cell lines treated with the indicated PARPi. Mean ±SD, normalized to untreated cells (n = 3 biologically independent experiments). Solid lines, nonlinear least squares fit of the data to a three-parameter dose response model. h. EC50 values for olaparib (left) and talazoparib (right) in the indicated cell lines as determined by nonlinear least squares fitting of data in e, f, g and Fig 2a,b. Bars, EC50 value ± 95% confidence interval. i-l, Increased apoptosis in HeLa RNASEH2A-KO, SUM149PT Cas9 RNASEH2B-KO and HCT116 RNASEH2A-KO cells following PARPi treatment. i, Representative (n = 3 biologically independent experiments) cleaved caspase-3 immunofluorescence / flow cytometry (IF/FACS) profiles of untreated and talazoparib-treated HeLa WT and RNASEH2A-KO cells. FSC = forward scatter. j-l, Percentages of cleaved caspase-3-positive (caspase-3+) cells of the indicated genotypes treated with the indicated PARPi. Individual values (coloured symbols) with mean (solid lines, n = 3 biologically independent experiments). Inset: Levels of cleaved caspase-3+ cells without PARPi treatment. Red lines, mean (n = 3 biologically independent experiments). P values, unpaired two-tailed t-test. In a, d, g and l, HCT116 RNASEH2A-KO cells were transduced either with an empty vector (+EV) or a full-length RNASEH2A expression construct (+WT), where indicated.
ED Figure 3
ED Figure 3
Related to Figure 2. a-d, HR is not affected by inactivation of RNase H2. a, Representative micrographs (n = 3 biologically independent experiments) of RPE1-hTERT Cas9 TP53-KO (WT) and RNASEH2A-KO cells exposed to 3 Gy of X-rays (IR) and processed for γ-H2AX and RAD51 immunofluorescence (IF) 4 h later. b. Quantification of the experiment in a at the indicated time points after IR, plotted as percentage of cells with >5 γ-H2AX and RAD51 colocalizing foci. Individual values (open circles) with mean (red lines, n = 3 biologically independent experiments). P values, unpaired two-tailed t-test. c, Representative (n = 3 biologically independent experiments) quantitative image-based cytometry (QIBC) plots of DR-GFP experiments in Fig 2e. Each point shows the mean GFP and RNASEH2A IF intensities per nucleus of mock- or I-SceI-transfected HeLa DR-GFP cells transduced with indicated Cas9/sgRNA constructs (EV = empty vector). Dashed lines separate RNASEH2A+/- and GFP+/- cell populations. d, Quantification of RNASEH2A+ cells in DR-GFP experiments shown in c and Fig 2e as determined by QIBC. Individual values (open circles) with mean (red lines; n = 3 biologically independent experiments). e-h, Replication-dependent endogenous DNA damage in RNase H2-deficient cells. e, Representative (n = 3 biologically independent experiments) micrographs for experiments quantified in Fig 2g. γ-H2AX immunofluorescence (IF) in EdU positive (EdU+) and negative (EdU-) WT and RNASEH2A-KO HeLa cells. Scale bars, 5 µm. f, Quantification of γ-H2AX foci per nucleus in experiments shown in e and Fig 2g. Dots, foci number in individual nuclei. Red lines, mean (n = 3 biologically independent experiments). g,h. HeLa WT and RNASEH2A-KO cells were treated with aphidicolin and EdU as indicated in the schematic (top), and immunostained with antibodies to γ-H2AX. Mean number of foci per EdU-positive (EdU+) nucleus in each experiment (g, open circles) or the number of foci in individual EdU+ nuclei (h, dots). Red lines, mean (n = 3 biologically independent experiments, ≥100 cells / sample / experiment analyzed). P value, unpaired two-tailed t-test. i, j, Increased poly(ADP-ribosylation) of PARP1 in G1 as well as in S/G2/M phases in RNASEH2A-KO cells. i, Representative (n = 2 biologically independent experiments) FACS plots of HeLa WT and RNASEH2A KO cells expressing the FUCCI cell cycle reporters mKO2-Cdt1 and mAG-Geminin. j, PARP1 immunoprecipitates from WCEs of FUCCI-sorted G1 or S/G2/M HeLa WT and RNASEH2A-KO cells, probed with the indicated antibodies in immunoblotting (representative of n = 2 biologically independent experiments). Tubulin, loading control. Densitometric quantification of PAR signals normalized to immunoprecipitated PARP1 is shown as fold changes from WT to RNASEH2A-KO cells. k-o, Inactivation of RNase H2 in BRCA1- or BRCA2-deficient backgrounds results in synthetic lethality. k, BRCA1 and BRCA2 expression, respectively, in RPE1-hTERT TP53-KO WT and BRCA1-KO (top) or DLD-1 WT and BRCA2-KO (bottom) cells. WCEs were processed for immunoblotting with the indicated antibodies. Tubulin and KAP1, loading controls. Representative of n ≥ 2 biologically independent experiments. l, RNase H2 levels in cells used in m, n, o (bottom) and Fig 2i. Cells were transduced with the indicated sgRNA- (top) or Cas9/sgRNA vectors (bottom; EV = empty vector) and processed for RNASEH2A IF. Each point represents mean RNASEH2A intensity per nucleus as measured by QIBC (n = 1 experiment). ≥2000 cells analyzed per sample. Percentages of RNASEH2A+ cells in individual samples are shown above each plot. m, Representative images (n = 3 biologically independent experiments) of clonogenic survival assays quantified in Fig 2i. n, o, Synthetic lethality after inactivation of RNASEH2A or RNASEH2B in BRCA2-deficient cells. Clonogenic survival of DLD-1 WT and BRCA2-KO cells was assessed after transduction with indicated Cas9/sgRNA vectors. n, Representative images of n = 3 biologically independent experiments. o, Quantification of the experiment in n. Individual values (open circles) with mean (red lines; n = 3 biologically independent experiments). P values, unpaired two-tailed t-test.
ED Figure 4
ED Figure 4
Related to Figure 2j. RNASEH2A P40D/Y210A is a separation-of-function mutant that cannot excise single DNA-embedded ribonucleotides, but cleaves RNA:DNA heteroduplexes (similar to the yeast rnh201-P45D-Y219A mutant16). a, Schematic depicting enzymatic activity against two different RNase H2 substrates (DRD:DNA, dsDNA with embedded ribonucleotide, or RNA:DNA hybrids) in cell lines used in b-d and Fig 2j. WT and RNASEH2A-KO cells were transduced with either an empty vector (EV) or the indicated RNASEH2A constructs. b, Complementation of HeLa RNASEH2A-KO cells with FLAG-tagged RNASEH2A variants restores RNase H2 complex protein levels. WCEs from HeLa WT and RNASEH2A-KO cells stably expressing indicated lentiviral constructs were processed for immunoblotting with the indicated antibodies. Vinculin, loading control. Asterisk indicates a non-specific band. Representative of n = 3 biologically independent experiments. c,d, Complementation of HeLa RNASEH2A-KO cells with WT RNASEH2A, but not with the D34A/D169A (catalytic-dead) or P40D/Y210A (separation-of-function) mutants, rescues increased levels of genome-embedded ribonucleotides. c, Total nucleic acids from the cell lines shown in a, b were treated with recombinant RNase H2 and separated by alkaline agarose gel electrophoresis (representative of n = 4 experiments). d, Densitometric quantification of alkaline gel shown in c. e, Purified human RNase H2 complexes consisting of RNASEH2B, RNASEH2C and either RNASEH2A WT, P40D/Y210A or D34A/D169A subunits separated by SDS-PAGE and stained with Coomassie Blue (n = 1). f-k, RNase H2 activity assays with fluorescein-labeled RNA:DNA substrate (f) or double-stranded DNA with a single incorporated ribonucleotide (DRD:DNA) (g) and increasing amounts of recombinant WT, P40D/Y210A or D34A/D169A RNase H2. Products were separated by polyacrylamide gel electrophoresis and detected by fluorescence imaging. Representative of n = 3 biologically independent experiments. h,k, Quantification of f, g. Product signal plotted relative to substrate signal per lane. Mean ±SD (n = 3 biologically independent experiments).
ED Figure 5
ED Figure 5
Related to Fig 3a-c. PARP1 trapping is the underlying cause of PARPi sensitivity in RNase H2-deficient cells. a, Schematic representation of CRISPR screens for suppressors of talazoparib sensitivity in RNase H2-deficient cells. Cas9-expressing cells were transduced with the TKOv1 library, talazoparib was added on day 6 (t6; HeLa: 20 nM, RPE1-hTERT: 50 nM) and cells were cultured in its presence until day 18 (t18). Cells were subcultured once at day 12 (RPE1) or 13 (HeLa). sgRNA representations in the initial (t6) and final (t18) populations were quantified by next-generation sequencing. Gene knockouts that were enriched at t18 over t6 were identified by MAGeCK. b, CRISPR-mediated inactivation of RNASEH2A and/or PARP1 in cell lines used in c-e and Fig 3b. WCEs were processed for immunoblotting with the indicated antibodies. KAP1, loading control. Representative of n = 2 biologically independent experiments. c-e, Loss of PARP1 restores PARPi-resistance in RNASEH2A-KO cells. c, Percentage of cleaved caspase-3+ HeLa cells of indicated genotypes with or without olaparib treatment measured by flow cytometry (FACS). Individual values (open circles) with mean (red lines, n = 3 biologically independent experiments; P-value, unpaired two-tailed t-test). d,e. Clonogenic survival assays with HeLa (d) and RPE1-hTERT (e) cells of the indicated genotypes treated with olaparib (left) or talazoparib (right). Mean ±SD (n = 3 biologically independent experiments). Solid lines, nonlinear least squares fit to a three-parameter dose response model. f. Trapping activity of PARPi correlates with the ability to induce apoptosis in RNASEH2A-KO cells. Quantification of cleaved caspase-3-positive HeLa WT and RNASEH2B-KO cells without treatment or treated with the indicated PARPi. Individual values with mean (black lines, n = 3 biologically independent experiments). Note that PARP-trapping activity decreases as follows: talazoparib > olaparib > veliparib ,. g, PARPi-induced S-phase arrest in RNASEH2A-KO cells is alleviated in the absence of PARP1. Top, schematic of talazoparib and EdU treatment. Bottom, representative (n = 3 biologically independent experiments) EdU (pseudocolor plots) and DNA content (histograms) FACS profiles of untreated and talazoparib-treated HeLa WT, PARP1-KO, RNASEH2A-KO and PARP1-KO/RNASEH2A-KO cells. DNA content was determined by propidium iodide (PI) staining. h Quantification of mean γ-H2AX intensities in experiments shown in Fig 3c. Individual values (open circles) with mean (red lines, n = 3 biologically independent experiments, ≥10,000 cells / sample / experiment analyzed).
ED Figure 6
ED Figure 6
Related to Figure 3d-g. TOP1-mediated cleavage at genome-embedded ribonucleotides leads to PARPi sensitivity in RER-deficient cells. a, Reduced endogenous DNA damage in TOP1-depleted RNASEH2A-KO cells. Quantification of γ-H2AX foci per nucleus in experiments shown in Fig 3e,f. Dots, focus number in individual nuclei. Red lines, mean (n = 5 biologically independent experiments). b-i, TOP1 depletion alleviates PARPi-induced apoptosis and S-phase arrest in HeLa RNASEH2A-KO cells (b-e) and in RNASEH2A P40D/Y210A separation-of-function mutant cells (f-h). b, Representative (n = 3 biologically independent experiments) cleaved caspase-3 FACS plots for experiments quantified in Fig 3g. FSC, forward scatter. c, HeLa WT and RNASEH2A-KO cells were transfected with non-targeting (siCTRL-SP) or TOP1-targeting (siTOP1-SP) SMARTpool siRNAs. WCEs analyzed by immunoblotting with antibodies to TOP1 and actin (loading control). Representative of n = 3 biologically independent experiments. d, Representative (n = 3 biologically independent experiments) FACS plots of cleaved caspase-3 in siCTRL-SP or siTOP1-SP-transfected WT and RNASEH2A-KO HeLa cells after talazoparib treatment. FSC, forward scatter. e, Quantification of the experiment shown in d. f, HeLa RNASEH2A-KO cells stably expressing the indicated FLAG-tagged constructs were transfected with non-targeting (siCTRL) or TOP1-targeting (siTOP1) siRNAs. WCEs were analyzed by immunoblotting with antibodies to TOP1, FLAG and actin (loading control). Representative of n = 3 biologically independent experiments. g, Representative (n = 3 biologically independent experiments) FACS plots of cleaved caspase-3 in siCTRL- or siTOP1-transfected HeLa RNASEH2A-KO cells expressing RNASEH2A-WT or P40D/Y210A mutant. h, Quantification of the experiment shown in g. Data in e,h, mean ±SD normalized to untreated cells (n = 3 biologically independent experiments, ≥10,000 cells / sample / experiment analyzed; P values, unpaired two-tailed t-test). i, Representative (n = 3 biologically independent experiments) cell cycle profiles, prior and post talazoparib treatment, of HeLa WT and RNASEH2A-KO cells transfected with the indicated siRNAs. DNA content was assessed by PI staining and FACS.
ED Figure 7
ED Figure 7
Related to Figure 4a-c. Collateral loss of RNASEH2B in CLL and metastatic castration-resistant prostate cancer (CRPC). a, b, Multiplex ligation-dependent probe amplification (MLPA) analysis (a) and comparative genomic hybridization (CGH) array profiles for chromosome 13q (b) of representative CLL samples carrying two wild-type (WT) RNASEH2B alleles (top), a monoallelic RNASEH2B deletion (middle) or biallelic deletion (bottom). a, For MLPA analysis, genomic DNA from reference and experimental samples was analyzed using probes targeting control loci and individual RNASEH2B exons (Exon 1-11). MLPA ratio calculated per probe and normalised to control probes and reference samples. Error bars indicate SD of the mean from 8 control probes for each sample. Dashed lines indicate the threshold set for diploid copy number. b, For each CGH array profile the y-axes of the top and bottom plots indicate copy number probe intensity (log R ratio) and the x axes mark the position on chromosome 13 represented by the ideogram (middle). An enlargement of the frequently deleted 13q14.2-14.3 region, including the miRNA-15A/16-1 gene cluster and the RNASEH2B gene, is shown in the bottom plot. n = 1 experiment. c, RNASEH2B is frequently co-deleted with RB1 in CRPC. Copy number alterations (CNA) in the RB1-RNASEH2B region in CRPC (n = 226 cases) are shown. Horizontal lines represent the CNA profile for individual CRPC samples (dark blue, homozygous loss; light blue, heterozygous loss; grey, no change; pink, copy number gain (CNA 3-4); red, copy number amplification (CNA > 4); white, insufficient data to determine CNA). Samples are clustered based on RNASEH2B gene status. CNA frequencies for RNASEH2B and the RB1-RNASEH2B region without a copy number breakpoint are shown on the right.
ED Figure 8
ED Figure 8
Related to Figure 4. a,b, Proliferating cells, and not quiescent cells, are the major population of viable cells in ex-vivo cultured primary CLL patient samples irrespective of treatment group. Quantification of absolute (a) and relative (b) quiescent and proliferating cell numbers as determined by FACS analysis of the primary CLL samples used in Fig 4b,c. (RNASEH2B WT, n = 8 individual samples; monoallelic deletion, n = 4 individual samples; biallelic deletion, n = 9 individual samples). Mean ± SD (n = 3 technical replicate). FACS gating strategy for stimulated peripheral blood lymphocytes (PBLs) from CLL patients is shown in Supplementary Fig 2. c, RNase H2-deficient primary CLL cells have reduced survival when cultured with olaparib. Mean of individual samples ± s.e.m. (n = 3 biologically independent CLL samples / group, each analyzed in technical triplicates). P-value, two-way ANOVA. d, Talazoparib selectively inhibits the growth of RNASEH2A-KO xenograft tumours. RNASEH2A-KO cells complemented either with empty vector (EV) or RNASEH2A-WT were injected subcutaneously into bilateral flanks of CD-1 nude mice. Mice were randomized to either vehicle or talazoparib (0.333 mg/kg) treatment groups (n = 8 animals / group) and tumour volumes were measured twice-weekly. Data plotted as mean ± s.e.m. P-values. two-way ANOVA under the null hypothesis that talazoparib does not supress the tumour growth.
ED Figure 9
ED Figure 9
RNase H2-deficient cells are more sensitive to PARPi than DNA polymerase β mutants. a, Schematic representation of the POLBΔ188-190 CRISPR mutation. The Mg2+-coordinating aspartate residues (D190, D192 and D256, red triangles) are highlighted in the domain structure of the human Polβ protein. The sgRNA target site and antibody epitope are indicated by black lines. b, WCEs from parental RPE1-hTERT Cas9 TP53-KO cells and two POLBΔ188-190 clones were processed for immunoblotting with antibodies to Polβ and tubulin (loading control). Representative of n = 2 biologically independent experiments. c, The POLBΔ188-190 mutation impairs base excision repair. RPE1-hTERT Cas9 TP53-KO WT or POLBΔ188-190 cells were exposed to different concentrations of methyl-methanesulfonate (MMS) for 24 h, followed by growth in drug-free media for an additional 48 h. Cell viability was determined by the Cell Titer Glo assay. d, Sensitivity of RPE1-hTERT Cas9 TP53-KO WT, RNASEH2A-KO and POLBΔ188-190 cells to indicated talazoparib concentrations in clonogenic survival assays. Data in c and d represent mean ±SD, normalized to untreated cells (n = 3 biologically independent experiments). Solid lines denote a nonlinear least-squares fit to a three-parameter dose response model.
Figure 1
Figure 1. CRISPR screens identify determinants of PARP inhibitor (PARPi) sensitivity.
a, Schematic of screening pipeline. b, Venn diagram of all high-confidence hits (FDR ≤ 0.01 + FDR ≤ 0.1 in ≥ 2 cell lines) in individual cell lines. c, Gene ontology (GO) terms significantly (P < 0.05, binomial test with Bonferroni correction) enriched among hits common to ≥ 2 cell lines. d, esyN network analysis of interactions between hits common to ≥ 2 cell lines. Node size represents the mean DrugZ score across cell lines. 31/73 genes are mapped on the network. See also ED Fig 1.
Figure 2
Figure 2. Defective ribonucleotide excision repair causes PARPi sensitivity, DNA damage and synthetic lethality with BRCA1 deficiency.
a,b, Reduced survival of HeLa RNASEH2A-KO cells after treatment with indicated PARPi. Mean ±SD, normalized to untreated cells. Solid lines, nonlinear least-squares fit to a three-parameter dose-response model. c-f, RNASEH2A-KO cells are HR-proficient. c,d, Normal RAD51 focus formation in RNASEH2A-KO HeLa cells after X-ray exposure. c, Representative micrographs of HeLa WT and RNASEH2A-KO cells stained with indicated antibodies (n = 3 biologically independent experiments). Scale bar, 10 μm. d, Quantification. Percentage of cells with >5 RAD51/γ-H2AX colocalizing foci at indicated time points. e, HR is not impaired in RNase H2-null cells. Quantification of gene conversion in DR-GFP reporter cells transduced with Cas9 + sgRNASEH2A/B or empty vector (EV) ± I-SceI transfection. Values normalized to transfection efficiency of control GFP vector. f, Increased sister chromatid exchanges (SCEs) in RNASEH2A-KO cells. Representative micrographs of SCEs in WT and RNASEH2A-KO metaphases. Below, numbers of SCEs / chromosome (mean ±SD, n = 3 biologically independent experiments). Scale bars, 10 μm. g,h, Spontaneous replication-associated damage and increased PARP1 activation in RNASEH2A-KO cells. g, Quantification of mean γ-H2AX immunofluorescent foci number / nucleus in EdU positive (+) and -negative (-) WT and RNASEH2A-KO cells. h, Representative poly(ADP-ribose) (PAR) immunoblot of PARP1 immunoprecipitates (IP) from whole cell extracts (WCE). Mean fold-increase in PARylation between WT and RNASEH2A-KO indicated (n = 3 biologically independent experiments, normalized to immunoprecipitated PARP1 levels). Tubulin and IgG heavy chain, loading controls. i, Synthetic lethality in combined absence of RNase H2 and BRCA1. Quantification of colony formation of BRCA1-proficient (WT) and BRCA1-KO RPE1-hTERT Cas9 TP53-KO cells transduced with sgLacZ or sgRNASEH2B constructs. Open circles, individual values normalized to sgLacZ; red lines, mean (n = 3 biologically independent experiments). j, PARPi sensitivity is associated with ribonuclease excision repair (RER) deficiency. Survival of olaparib-treated HeLa WT and RNASEH2A-KO cells transduced with indicated FLAG-tagged constructs. Mean ±SD, normalized to untreated cells (n = 3 biologically independent experiments). Solid lines, nonlinear least squares fit to a three-parameter dose response model. For d, e, and g: open circles, individual values; red lines, mean [n = 3 biologically independent experiments; ≥100 (d, g) and ≥1000 (e) cells / sample / experiment analyzed]. P values in d-g and i, unpaired two-tailed t-test. See also ED Fig 2-4.
Figure 3
Figure 3. PARPi-induced PARP1 trapping occurs in RER-deficient cells as a result of TOP1-mediated processing of genomic ribonucleotides.
a,b, PARP1 is required for PARPi-induced toxicity in RNASEH2A-KO cells. CRISPR screens for talazoparib sensitivity suppressors in RNASEH2A-deficient HeLa Cas9 and RPE1 Cas9 TP53-KO cell lines. MAGeCK positive scores for each gene plotted. Colors indicate gene density in each hexagonal bin. b, Percentage of cleaved caspase-3+ cells of indicated genotypes with or without talazoparib treatment measured by flow cytometry (FACS). Open circles, individual experiments; red lines, mean (n = 3 biologically independent experiments). c. DNA damage persists on withdrawal of PARPi in RNASEH2A-KO cells. HeLa WT and RNASEH2A-KO cells were treated with talazoparib and released into fresh medium for the indicated times before being processed for γ-H2AX immunofluorescence and propidium iodide (PI) staining. Representative (n = 3 biologically independent experiments) γ-H2AX (pseudocolor plots) and cell cycle (histograms) IF/FACS profiles shown. d-f, Increased γ-H2AX foci formation in RNASEH2A-KO cells depends on TOP1 (images representative of n = 5 biologically independent experiments). d, HeLa WT and RNASEH2A-KO cells were transfected with non-targeting (siCTRL) or TOP1-targeting (siTOP1) siRNAs. Immunoblot of WCEs, probed for TOP1. Actin, loading control. e, Representative micrographs of HeLa WT and RNASEH2A-KO cells transfected with siCTRL or siTOP1 immunostained for γ-H2AX. Scale bars, 10 µm. f, Quantification of experiments shown in e. Mean number of foci / nucleus / experiment (open circles) with mean of n = 5 biologically independent experiments (red lines). ≥100 cells / sample / experiment analyzed. g, TOP1 depletion alleviates PARPi-induced apoptosis in RNASEH2A-KO cells. Quantification of cleaved caspase-3+ WT and RNASEH2A-KO cells transfected with indicated siRNAs, with or without talazoparib treatment. Mean ± SD normalized to untreated cells (n = 3 biologically independent experiments). ≥10,000 cells / sample / experiment. P values in b, f, g, unpaired two-tailed t-test. See also ED Fig 5 and 6.
Figure 4
Figure 4. Talazoparib selectively suppresses growth of RNase H2 deficient tumours.
a-c, PARP inhibitors selectively kill RNASEH2B-deficient chronic lymphocytic leukemia (CLL) primary cancer cells. a, RNASEH2B deletion frequency in a panel of 100 primary CLL samples, determined by multiplex ligation-dependent probe amplification (MLPA). b, Reduced RNase H2 activity in lysates from CLL samples with monoallelic and biallelic RNASEH2B deletions. Top, substrate schematic. Individual data points, mean of technical duplicates for each sample. Red lines, mean of individual genotypes (n = 8 WT, 4 monoallelic and 9 biallelic deleted biologically independent primary CLL samples). Data normalized to mean of RNASEH2B-WT samples. c, Reduced survival of CLL cells with monoallelic and biallelic RNASEH2B loss following treatment with talazoparib. Individual points, mean ± s.e.m. (n = 8, 4 and 9 CLL samples as in b), each analysed in technical triplicates. P-values, unpaired two-tailed t-test (b) and two-way ANOVA (c). d, Selective inhibition of RNASEH2A-KO xenograft tumour growth. HCT116 TP53-KO RNASEH2A-WT or -KO cells were injected subcutaneously into bilateral flanks of CD-1 nude mice. Mice were randomized to either vehicle or talazoparib (0.333 mg/kg) treatment groups (n = 8 animals / group) and tumour volumes measured twice-weekly. Mean ± s.e.m. P-value, two-way ANOVA. e, Model. Genome-embedded ribonucleotides (R) can be processed by TOP1 as an alternative to RNase H2-dependent RER. DNA lesions that engage PARP1 (black circles) are formed as a result, and PARP inhibitors induce PARP1 trapping on these TOP1-dependent lesions, causing replication arrest, persistent DNA damage and cell death. See also ED Fig 7, 8, ED Table 1 and Supplementary Table 3.

Comment in

  • A ribonucleotide trap.
    Song Y. Song Y. Nat Chem Biol. 2018 Sep;14(9):831. doi: 10.1038/s41589-018-0127-z. Nat Chem Biol. 2018. PMID: 30120354 No abstract available.

Similar articles

See all similar articles

Cited by 48 articles

See all "Cited by" articles

References

    1. Lord CJ, Ashworth A. PARP inhibitors: Synthetic lethality in the clinic. Science. 2017;355:1152–1158. doi: 10.1126/science.aam7344. - DOI - PMC - PubMed
    1. Pommier Y, O'Connor MJ, de Bono J. Laying a trap to kill cancer cells: PARP inhibitors and their mechanisms of action. Sci Transl Med. 2016;8:362ps317. doi: 10.1126/scitranslmed.aaf9246. - DOI - PubMed
    1. Hopkins TA, et al. Mechanistic Dissection of PARP1 Trapping and the Impact on In Vivo Tolerability and Efficacy of PARP Inhibitors. Molecular cancer research : MCR. 2015;13:1465–1477. doi: 10.1158/1541-7786.MCR-15-0191-T. - DOI - PubMed
    1. Murai J, et al. Trapping of PARP1 and PARP2 by Clinical PARP Inhibitors. Cancer Res. 2012;72:5588–5599. doi: 10.1158/0008-5472.CAN-12-2753. - DOI - PMC - PubMed
    1. Cerritelli SM, Crouch RJ. The Balancing Act of Ribonucleotides in DNA. Trends Biochem Sci. 2016;41:434–445. doi: 10.1016/j.tibs.2016.02.005. - DOI - PMC - PubMed

Publication types

MeSH terms

Feedback