Diffuse intrinsic pontine glioma (DIPG) is an invariably fatal brain tumor occurring predominantly in children. Up to 90% of pediatric DIPGs harbor a somatic heterozygous mutation resulting in the replacement of lysine 27 with methionine (K27M) in genes encoding histone H3.3 (H3F3A, 65%) or H3.1 (HIST1H3B, 25%). Several studies have also identified recurrent truncating mutations in the gene encoding protein phosphatase 1D, PPM1D, in 9%–23% of DIPGs. Here, we sought to investigate the therapeutic potential of targeting PPM1D, alone or in combination with inhibitors targeting specific components of DNA damage response pathways in patient-derived DIPG cell lines. We found that GSK2830371, an allosteric PPM1D inhibitor, suppressed the proliferation of PPM1D-mutant, but not PPM1D wild-type DIPG cells. We further observed that PPM1D inhibition sensitized PPM1D-mutant DIPG cells to PARP inhibitor (PARPi) treatment. Mechanistically, combined PPM1D and PARP inhibition show synergistic effects on suppressing a p53-dependent RAD51 expression and the formation of RAD51 nuclear foci, possibly leading to impaired homologous recombination (HR)-mediated DNA repair in PPM1D-mutant DIPG cells. Collectively, our findings reveal the potential role of the PPM1D–p53 signaling axis in the regulation of HR-mediated DNA repair and provide preclinical evidence demonstrating that combined inhibition of PPM1D and PARP1/2 may be a promising therapeutic combination for targeting PPM1D-mutant DIPG tumors.

Implications:

The findings support the use of PARPi in combination with PPM1D inhibition against PPM1D-mutant DIPGs.

Diffuse intrinsic pontine glioma (DIPG) is an invariably fatal pediatric brain tumor for which median survival is <1 year. Patients with DIPG have very limited treatment options given the diffuse, invasive nature of the disease in the critical structures of the brainstem. DIPG tumors are not amenable to surgical resection due to the anatomic localization within the brainstem and infiltrative nature. Radiotherapy is the current standard of care for DIPG but confers only a 3- to 4-month survival benefit (1). Multiple trials with different chemotherapies have proven ineffective in treating DIPGs (2). Integrated genetic profiling of DIPGs has identified Histone H3 mutations (H3K27M) in genes encoding histone H3.3 (H3F3A, 65%) or H3.1 (HIST1H3B, 25%) and several other recurrent alterations that may serve as promising therapeutic targets for DIPGs (3–6). Our laboratory and others identified recurrent mutations in the gene encoding protein phosphatase 1D, PPM1D, in 9%–23% of DIPGs (4,7–9). PPM1D mutations have been shown to cooccur with Histone H3 mutations (H3K27M) and are generally found to be mutually exclusive to tumor suppressor protein 53 (TP53)-inactivating mutations, but two cases displayed cooccurrence of a PPM1D mutation and TP53 mutation (likely subclonal) have also been reported in DIPGs (4, 7, 8, 10, 11).

PPM1D, also known as wild-type p53-induced phosphatase 1 (Wip1), is defined as an oncogenic phosphatase because it functions to inactivate the tumor suppressor p53 (12). Increased PPM1D activity, through amplification or overexpression, has been found in multiple tumor types, including breast cancer, ovarian cancer, neuroblastoma, and medulloblastoma (12). In addition to its role as an oncogene, PPM1D also functions as a homeostatic regulator of the DNA damage response (DDR) by dephosphorylating checkpoint kinases, including ATM, ATR, and Chk1/2 (12). Moreover, PPM1D plays multiple roles in the regulation of DNA repair. It has been observed that PPM1D overexpression negatively regulates both base excision repair (BER) and nucleotide excision repair (NER; refs. 13, 14). PPM1D has also been implicated in the negative regulation of DNA double-strand break (DSB) repair by directly dephosphorylating γH2AX and subsequently prevents recruitment of other repair factors to the DNA damage sites (15).

In DIPGs, PPM1D mutations invariably occur as nonsense or frameshift alterations that cluster in exon 6, the terminal exon of PPM1D, thus generating a truncated PPM1D protein. Truncated PPM1D mutants retain phosphatase activity and appear to be stabilized because of the loss of COOH-regulatory elements (7, 16, 17). It has been demonstrated that PPM1D-truncating mutations result in the accumulation of PPM1D-truncated protein, which causes increased dephosphorylation of PPM1D downstream substrates (e.g., ATM, p53, Chk2, and γH2AX), suppresses the DDR, and impairs the p53-dependent G1-phase checkpoint (16, 18, 19). Truncating PPM1D mutations also confer a chemoresistance phenotype and lead to the selective outgrowth of PPM1D-mutant hematopoietic cells after exposure to certain cytotoxic DNA-damaging agents (20, 21). Collectively, these observations suggest that inhibition of PPM1D may be a viable therapeutic strategy for augmenting the DNA damage–induced toxicity when combined with inhibitors targeting DDR pathways.

In this study, we sought to investigate the therapeutic potential of pharmacologic PPM1D inhibition and the extent to which PPM1D inhibition could be combined with inhibitors targeting the specific components of DDR pathways for DIPG therapy. We observed that GSK2830371, a PPM1D allosteric inhibitor (22), suppressed the proliferation of PPM1D-mutant DIPG cells. Using isogenic PPM1D−/− cell lines derived from the SF7761 line, a patient-derived DIPG cell line harboring native PPM1D truncating mutation (p.E540X), we tested a panel of small molecules targeting various nodes of the DDR pathway for their potential to augment or synergize with PPM1D inhibition. We found that genetic deletion of PPM1D or pharmacologic inhibition of PPM1D sensitized SF7761 cells to concurrent treatment with PARP inhibitors (PARPi). Notably, we observed that combined inhibition of PARP1/2 and PPM1D reduced the growth of three-dimensional (3D) tumor spheroids and promoted apoptosis in all three PPM1D-mutant DIPG cell lines. Mechanistically, we demonstrated that PPM1D inhibition in combination with PARPi treatment dramatically reduced the formation of Rad51 nuclear foci and caused a p53-dependent suppression of Rad51 expression in PPM1D-mutant DIPG cells, suggesting that PPM1D inhibition confers PARPi sensitivity by comprising the homologous recombination (HR)-mediated DNA repair pathway in these cells.

DIPG neurosphere-forming cultures and inhibitors

SF7761 modified with hTERT (human telomerase ribonucleoprotein reverse transcriptase) and with a luciferase reporter was a gift from Dr. C. David James (Northwestern University, Chicago, IL) and cultured in NeuroCult NS-A Basal Medium (Human) supplemented with NeuroCult NS-A Proliferation Kit (Human), heparin, human recombinant bFGF, and human recombinant EGF (23). GBM-002, HSJD-DIPG-012, HSJD-DIPG-013, and HSJD-DIPG-007 were gifts from Dr. Angel Montero Carcaboso (Hospital Sant Joan de Déu, Esplugues de Llobregat, Barcelona), and were cultured as described previously (24). TT10714 and TT10728 cell lines were established by Liwei Zhang at Beijing Tiantan Hospital (Beijing, China) and were cultured as described previously (25). All genetically modified clones were cultured with the same medium as the parental cells. All cells were maintained in a humidified atmosphere at 37°C and 5% CO2. Cell line authentication was conducted by short tandem repeat analyses (January 2019) and experiments were performed within the first 15 passages of thawing cells. SF7761, HSJD-DIPG-007, TT10714, and TT10728 were tested for Mycoplasma (August 2018) at IDEXX BioResearch and all cells tested negative. GBM-002, HSJD-DIPG-012, and HSJD-DIPG-013 were not tested for Mycoplasma by the authors.

The inhibitors used in this study (GSK2830371, olaparib, rucaparib, veliparib, KU60019, MK-1775, NU7441, TH287, RI-1, VE-821, and NU7026) were purchased from Selleckchem.

Western blot analysis

Cell pellets were lysed in ice-cold lysis buffer (8 mol/L urea, 2 mmol/L EDTA, 10 mmol/L dithiothreitol, and 20 mmol/L Tris-HCl, pH 8.5) supplemented with Halt Protease Inhibitor Cocktail (Thermo Fisher Scientific). After incubation for 30 minutes on the wet-ice, the cell lysates were further sonicated in an ice-cold water bath for 5 minutes using Bioruptor UCD-200TM sonicator at high energy setting and cleared by centrifugation at 13,000 rpm for 10 minutes at 4°C. The supernatants were transferred, and the protein concentration of the lysates was measured using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Approximately 20 μg of total proteins containing lysates were resolved using (4%–12%) NuPAGE Bis-Tris gradient gels. Subsequently, the proteins were transferred to nitrocellulose membrane (0.2-μm pore size) using an XCell II Blot Module. After transfer, the membranes were rinsed briefly in TBST (150 mmol/L NaCl, 50 mmol/L Tris-HCl, pH 7.5, 0.1% Tween-20) and then blocked for 1–2 hours in Pierce TBST Protein-Free Blocking Buffer (catalog number 37571) or 5% BSA at room temperature. After blocking, primary antibodies were diluted (1:1,000) in TBST protein-free blocking buffer or 5% BSA blocking buffer and incubated overnight at 4°C. Membranes were washed and then incubated with mouse or rabbit horseradish peroxidase (HRP)-conjugated secondary antibody (1:3,300) in TBST buffer for 1 hour at room temperature, and HRP signals were detected by chemiluminescence and the images were captured with a Bio-Rad ChemiDoc MP System.

Antibodies

The antibodies used in study included: rabbit GAPDH antibody (FL-335) (sc-25778), mouse monoclonal p53 (DO-1) antibody (sc-126), rabbit RAD51 antibody (H-92) (sc-8349), mouse monoclonal PPM1D antibody (sc-376257), and rabbit polyclonal 53BP1 antibody (sc-22760) were purchased from Santa Cruz Biotechnology, Inc.; rabbit PPM1D antibody (CST#11901), mouse monoclonal β-actin (CST #3700), rabbit p21 antibody (CST #2947), rabbit PARP antibody (CST#9532), rabbit Tri-Methyl-Histone H3 (Lys27) (C36B11) antibody (CST#9733), rabbit polyclonal anti-Histone H3.3K27M mutant, mouse monoclonal phospho-Histone H2A.X (Ser139) (D7T2V) antibody (CST#80312), and rabbit anti-Rad51 (D4B10) rabbit mAb (CST#8875) were purchased from Cell Signaling Technology.

Retrovirus vector generation and transgene expression

Human full-length PPM1D or truncated PPM1DE472X cDNA were cloned from pCMV-Neo-Bam-PPM1D into a retroviral MIGR1 vector expressing a GFP reporter (an efficient way to monitor retroviral transduction efficiency and transgene expression) under control of an MSCV promoter.

Retroviral particles for transgene delivery were generated as described previously (26) with slight modifications at the virus-harvesting steps. The virus-containing supernatants were collected and spun at 3,000 rpm for 15 minutes at 4°C to remove debris. Retro-X Concentrator (Clontech) was used to further concentrate the viruses and remove the serum-containing medium. All the virus particles were resuspended with 500 μL of PBS and the aliquots were stored at −80°C. For retroviral infection, a total of 80,000 cells were seeded in a 24-well dish in 450 μL of growth media. The next day, 50 μL of viral particle–containing aliquot was added to the well. After incubation for 48 hours, the infected cells were washed to remove the extra viral particles. The transduction efficiency was examined by fluorescence microscopy and successfully transduced cells were further collected by flow cytometry sorting performed at Duke Flow Cytometry Shared Resource (FCSR).

Generation of lentiviral particles and transduction of cells

The lentiviral particles were generated as reported previously (27). SF7761 and HSJD-DIPG-007 cells were transduced with lentiviral particles at 1 multiplicity of infection (MOI) in medium containing polybrene at 4 μg/mL. After transduction for 48 hours, the puromycin selection (2 μg/mL) was started. After 2 weeks of puromycin selection, the knockout efficiency of nontargeting control (NTC) and TP53 in the drug-resistant cells was examined by Western blotting.

Immunofluorescence staining

For RAD51 and 53BP1 foci assays, cells were grown on laminin-coated 4-well ibidi μ-Slide (chambered coverslip). The next day, cells were treated with olaparib (5 μmol/L) for 48 hours. Immunofluorescence staining was performed as described previously, using the primary antibodies against mouse anti-γ-H2AX (1:200, Cell Signaling Technology #2577), rabbit anti-RAD51(H-92) (1:200, sc-8349), or rabbit anti-53BP1 (1:200, sc-22760) in blocking solution (1 mg/mL BSA, 3% goat serum, 0.1% Triton X-100, 1 mmol/L EDTA) overnight at 4°C. Cells were then washed and stained with 1:500 Alexa Fluor 488 goat anti-mouse IgG (Thermo Fisher Scientific) and Alexa Fluor 594 goat anti-rabbit IgG for 1 hour at room temperature, protected from light, and then stained with DAPI and sealed with mounting medium (ibidi#50001). Images were taken by Zeiss 880 Airyscan Inverted Confocal Microscope and analyzed with ImageJ-Fiji (28).

RNA extraction and cDNA conversion

Total RNA was extracted from cell pellets using Maxwell RSC SimplyRNA Cells Kit (Promega) according to the manufacturer's instructions. After quantification with NanoDrop (Thermo Fisher Scientific), 1 μg of total RNA from each sample was used as a template and converted to cDNA using the RNA to cDNA EcoDry Premix (Double Primed), cDNA Synthesis Kit (Takara).

qRT-PCR

RAD51 expression levels were determined by qRT-PCR using KAPA SYBR FAST qPCR Master Mix (2X) Kit (Kapa Biosystems). Results were expressed as fold change calculated by the ΔΔCt method relative to the control sample. The GAPDH was used as an internal normalization control. The primers used for RAD51 and GAPDH qRT-PCR are listed in Supplementary Table S1.

Sanger sequencing

PCR purification and sequencing reactions were performed by Eton Biosciences or Genewiz. The primers used for PCR reactions are listed in Supplementary Table S1. Mutation status was determined using Mutation Surveyor (Soft Genetics).

Cell viability assay

DIPG cells were plated at a density of 3,000 cells/well in 96-well white and clear bottom microplates (Greiner Bio-One). After 24-hour incubation, cells were treated for 7 days with different inhibitors against the DNA repair pathway at the indicated concentrations. The effects of all the inhibitors on cell viability was measured by CellTiter-Glo Luminescent Cell Viability Assay (Promega) per the manufacturer's instruction. The luminescence signals were recorded using a Tecan Infinite M200 PRO microplate reader. Relative viabilities were calculated by normalizing luminescence values for each treatment condition to DMSO-treated wells. Dose–response curves were fit using GraphPad Prism software (version 8.02).

Annexin V and propidium iodide apoptosis assay

Cells were plated at a density of 3 × 105 cells per well in 6-well plates. After 24 hours of incubation, cells were treated for 72 hours with DMSO or indicated drug/drug combinations. Then, the cells were harvested and further stained with the Alexa Fluor 488 Annexin V/Dead Cell Apoptosis Kit (Thermo Fisher Scientific), according to the manufacturer's instructions. The percentage of apoptotic cells was determined by flow cytometry (BD FACSCanto IIs) at Duke FCSR and analyzed by FlowJo (Version 9.9.6).

X-ray irradiation

X-ray irradiation was performed using the X-RAD 320 Irradiator (Precision X-Ray) operating at 320 kVp with a dose rate of 1.97 Gy/minute. Cells were plated at a density of 3,000 cells/well in 96-well white and clear bottom microplates (Greiner Bio-One). The cells were then treated with DMSO or indicated drug/drug combinations for 12 hours before irradiation. Cells were then exposed to X-ray irradiation (0, 2, and 4 Gy) at room temperature. At 3 days postirradiation, the cell viabilities were measured by CellTiter-Glo Luminescent Cell Viability Assay (Promega) as described above.

Analysis of synergy/antagonism from drug combination studies

To evaluate possible additive and synergistic effects when using combinations of GSK2830371 and olaparib, the data from viability assays were analyzed using Combenefit software (www.cruk.cam.ac.uk/research-groups/jodrell-group/combenefit; ref. 29). The synergy/antagonism score for each combination was assessed by the classical Loewe synergy and antagonism method and was presented in a matrix format, where a positive score indicates synergy, a score of 0 is additive, and a negative score indicates antagonism. GSK2830371 and olaparib were diluted orthogonally using a three-fold serial dilution and then combined to produce a 6 × 10 combination dose–response matrix. After 7 days of treatments, cell viability was measured by CellTiter-Glo Luminescent Cell Viability Assay.

CRISPR/Cas9-mediated PPM1D genetic targeting

CRISPR guides were designed for minimal off-targets for the coding region of PPM1D utilizing the MIT-Broad design tool (http://crispr.mit.edu/). Two guides (see Supplementary Table S1) targeting exons 2 and 6 of PPM1D were selected to increase the probability of achieving a potent gene deletion and easy screening. Complementary oligonucleotides encoding the guides were annealed and cloned into pSpCas9(BB)-2A-GFP (PX458), which was a gift from Feng Zhang, Broad Institute of MIT and Harvard, Cambridge, MA (Addgene plasmid #48138). Two sgRNA plasmids (5 μg of each) were cotransfected into SF7761 (2 × 106 cells) using the Neon Transfection System 100 μL Kit (Thermo Fisher Scientific) with the program (1,300 V, 30 ms, 2 pulses), and GFP-positive cells were FACS sorted (Astrios, Beckman Coulter, Duke FCSR) at day 6. The GFP-positive cells were further cultured for 2 weeks to let the cells recover from sorting. Limiting dilution was performed to plate cells into 96-well plates. Clones were expanded over 3 to 6 months and the PPM1D−/−clones were screened as reported previously (27). Briefly, DNA was isolated by the addition of DirectPCR Lysis Reagent (ViaGen) with proteinase K (Sigma-Aldrich) and incubation of plates at 55°C for 30 minutes, followed by 95°C for 45 minutes. Crude lysate (1 μL) was used as a template for junction-spanning qPCR (to detect dual-sgRNA–induced deletion products) with KAPA SYBR FAST (Kapa Biosystems). The junction-spanning amplicon was detected by qPCR signal, using the parental (not transfected) line as a negative control. The targeted exons and junction products were sequenced to validate the presence of indels. Clones were then expanded further and screened by Western blot analysis to ensure the absence of PPM1D protein expression. All primers used for screening are listed in Supplementary Table S1.

CRISPR/Cas9-mediated TP53 genetic targeting

To knockout TP53 in a pool population of cells, we applied the high fidelity CRISPR system pLentiCRISPR-E (pLentiCRISPR-E was a gift from Dr. Phillip Abbosh (Fox Chase Cancer Center, Philadelphia, PA), Addgene plasmid # 78852). Two sgRNAs targeting exons 5 and 8 of TP53 (listed in Supplementary Table S1) were designed and cloned into the pLentiCRISPR-E system. Lentivirus was packaged and tittered by the Functional Genomics Shared Resource at Duke University (Durham, NC). The cells were infected with the pLenti-CRISPR-E virus at 1 MOI. The efficiency of TP53 knockout (KO) was validated by Western blot analysis.

3D tumor spheroid invasion assay

The 3D tumor spheroid invasion assay was performed as reported previously (30). Briefly, on day 0, the cells were plated at a density of 1,000 cells/well in a 96-well round-bottom plate in 100 μL of seeding medium (DIPG neurosphere-forming culture media). Then the cell-containing plates were spun down at 1,000 rpm for 5 minutes and cultured for 1 to 2 days (depending on the cell line) until spheroids formed. A total of 100 μL of fresh thawed Corning Matrigel matrix was added into each well using cold peptide tips. After the Matrigel matrix solidified (incubation at 37°C for 1 hour), 100 μL of the drug-containing medium was added on top of the solidified Matrigel. The growth and invasion of the DIPG tumor spheroids were monitored daily. The images were taken by Nikon Elclipse TE2000-E Microscope.

Statistical analyses

Statistical analyses were performed with GraphPad Prism software (version 8.02) or Excel. The detailed statistical analysis methods for each comparison or test are described in the figure legends. A P value of less than 0.05 was considered statistically significant. The adjusted P value is listed in each figure.

Identification and characterization of patient-derived DIPG cell lines with PPM1D mutations

To assess the therapeutic potential of PPM1D mutations in DIPG, we first collected a number of patient-derived DIPG cell lines and analyzed them for PPM1D mutations. Among six collected DIPG cell lines, three cell lines (SF7761, HSJD-DIPG-007, and TT10714) harbored PPM1D-truncating mutations (Fig. 1A), while the remaining three DIPG cell lines (HSJD-DIPG-012, HSJD-DIPG-013, and TT10728) did not harbor truncating mutations in PPM1D, and instead have TP53 mutations. All six DIPG cell lines harbor H3F3A mutations resulting in the H3K27M histone mutant, characteristic of this tumor type. The expressions of truncated PPM1D, H3.3K27M, and p53 were confirmed by Western blotting (Fig. 1B). We confirmed that all six DIPG cell lines exhibited the loss of H3K27 trimethylation, consistent with their expression of H3.3K27M expression, as compared with the H3F3A wild-type cell line GBM-002 (Fig. 1B). Moreover, the abundance of truncated PPM1D was markedly higher than full-length PPM1D in all six DIPG cell lines examined (Fig. 1B), consistent with previous reports demonstrating that truncated PPM1D is more stable than full-length PPM1D (16, 17, 20). Detailed information regarding TP53, PPM1D, and H3F3A mutational status for all six DIPG cell lines and GBM-002 are listed in Fig. 1A.

Figure 1.

Characterizations of PPM1D mutations in DIPG cell lines. A, Chromatograms of three truncating mutations identified by Sanger sequencing of PPM1D exon 6 in three PPM1D-mutant DIPG cell lines (SF7761, TT10714, and HSJD-DIPG-007, top). Red arrows indicate the mutation sites. The detailed information of TP53, PPM1D, and H3F3A mutation status in six DIPG cell lines and GBM-002 (bottom). B, Western blot analysis of the expressions of both full-length (as indicated by red arrows) and truncated PPM1D mutants, H3.3K27M, p53, and H3K27me3 levels in the whole-cell lysates from indicated cell lines. GBM-002 was included as a negative control for H3.3K27M and positive control for H3K27me3. β-actin was served as a loading control. PPM1D-SC (Santa Cruz Biotechnology, Inc.) is a mouse mAb raised against amino acid residues 306–605 of PPM1D and it recognizes full-length PPM1D and the truncated PPM1D mutants in SF7761 (PPM1DE540X) and TT10714 (PPM1DC478X) but not the shortest truncated PPM1D mutant (PPM1DP428Qfs*2) in HSJD-DIPG-007. PPM1D-CST (Cell Signaling Technology) is a rabbit antibody raised against a synthetic peptide corresponding to residues surrounding Gly240 of PPM1D and it detects all three truncated PPM1D mutants but it is not sensitive enough to detect the full length PPM1D in all DIPG cell lines. WT, wild-type.

Figure 1.

Characterizations of PPM1D mutations in DIPG cell lines. A, Chromatograms of three truncating mutations identified by Sanger sequencing of PPM1D exon 6 in three PPM1D-mutant DIPG cell lines (SF7761, TT10714, and HSJD-DIPG-007, top). Red arrows indicate the mutation sites. The detailed information of TP53, PPM1D, and H3F3A mutation status in six DIPG cell lines and GBM-002 (bottom). B, Western blot analysis of the expressions of both full-length (as indicated by red arrows) and truncated PPM1D mutants, H3.3K27M, p53, and H3K27me3 levels in the whole-cell lysates from indicated cell lines. GBM-002 was included as a negative control for H3.3K27M and positive control for H3K27me3. β-actin was served as a loading control. PPM1D-SC (Santa Cruz Biotechnology, Inc.) is a mouse mAb raised against amino acid residues 306–605 of PPM1D and it recognizes full-length PPM1D and the truncated PPM1D mutants in SF7761 (PPM1DE540X) and TT10714 (PPM1DC478X) but not the shortest truncated PPM1D mutant (PPM1DP428Qfs*2) in HSJD-DIPG-007. PPM1D-CST (Cell Signaling Technology) is a rabbit antibody raised against a synthetic peptide corresponding to residues surrounding Gly240 of PPM1D and it detects all three truncated PPM1D mutants but it is not sensitive enough to detect the full length PPM1D in all DIPG cell lines. WT, wild-type.

Close modal

PPM1D inhibitor GSK2830371 specifically suppresses the proliferation of PPM1D-mutant DIPG cells

We sought to evaluate the effect of treatment with GSK2830371, an allosteric PPM1D inhibitor that binds to a site that depends on a structural flap subdomain unique to PPM1D, in the three PPM1D-mutant DIPG cell lines. Similar to the previous report (22), treatment with GSK2830371 for 48 hours decreased the abundance of both full-length and truncated PPM1D proteins in all three PPM1D-mutant DIPG cell lines (Fig. 2A). We also observed that expression of p21, a downstream target of p53, was increased in a dose-dependent manner (Fig. 2A), indicating that the activity of the p53 pathway was restored in response to GSK2830371 treatment. We further evaluated the effect of GSK2830371 on both PPM1D-mutant and TP53-mutant DIPG cell lines. As shown in Fig. 2B, we observed that treatment with GSK2830371 specifically increased expression of p21 in PPM1D-mutant, but not TP53-mutant cell lines. Analysis of GSK2830371 treatment on DIPG cell viability revealed a dose-dependent inhibition of cell viability in all three PPM1D-mutant DIPG cells, but not the TP53-mutant cells in a 6-day treatment (Fig. 2C). Within a concentration range of 0–5 μmol/L, GSK2830371 exhibited specific activities against the PPM1D-mutant but not the TP53-mutant DIPG cells. Among them, TT10714, a treatment-naïve cell line, is the most sensitive line to GSK2830371 treatment with the IC50 of approximately 0.05 μmol/L (Fig. 2C). Furthermore, we observed that GSK2830371 specifically inhibited the proliferation of PPM1D-mutant cells (SF7761, TT10714, and HSJD-DIPG-007) but not TP53-mutant cells (HSJD-DIPG-012, HSJD-DIPG-013, and TT10728; Fig. 2D).

Figure 2.

PPM1D inhibitor GSK2830371 activates the p53 pathway and inhibits the proliferation of PPM1D-mutant DIPG cell lines. A, Three PPM1D-mutant DIPG cell lines (SF7761, TT10714, and HSJD-DIPG-007) were treated with vehicle (DMSO) or the indicated concentrations of GSK2830371 for 48 hours. Cell lysates were subjected to Western blot analysis to measure protein expression of PPM1D, p21, p53, and GAPDH (loading control). B, Western blot analysis showing changes in p21 and p53 expression in response to treatment with DMSO or GSK2830371 (2 μmol/L) for 48 hours in six DIPG cell lines. GAPDH was used as a loading control. C, Cell viability assay of six DIPG cell lines treated with DMSO or increasing concentrations (three-fold serial dilution) of GSK2830371 for 6 days. Luminescence units were normalized to DMSO-treated controls and are represented as relative cell viability. Data shown are mean ± SD of three independent biological replicates (the red broken line indicates 50% viability). The listed P values were determined by unpaired t tests to evaluate the differences between three PPM1D-mutant and three TP53-mutant DIPG cell lines in response to indicated doses of GSK2830371 treatments.D, Effect of GSK2830371 on cell growth of DIPG cells. Six DIPG cell lines were treated with DMSO or indicated concentrations (0.19 and 1.67 μmol/L) of GSK2830371 and relative cell growth rates were determined by CellTiter-Glo Assay reagent at days 1, 4, and 7. Results shown are mean ± SEM of three independent biological replicates. The adjusted P values were determined by two-way ANOVA followed by Dunnett multiple comparisons test.

Figure 2.

PPM1D inhibitor GSK2830371 activates the p53 pathway and inhibits the proliferation of PPM1D-mutant DIPG cell lines. A, Three PPM1D-mutant DIPG cell lines (SF7761, TT10714, and HSJD-DIPG-007) were treated with vehicle (DMSO) or the indicated concentrations of GSK2830371 for 48 hours. Cell lysates were subjected to Western blot analysis to measure protein expression of PPM1D, p21, p53, and GAPDH (loading control). B, Western blot analysis showing changes in p21 and p53 expression in response to treatment with DMSO or GSK2830371 (2 μmol/L) for 48 hours in six DIPG cell lines. GAPDH was used as a loading control. C, Cell viability assay of six DIPG cell lines treated with DMSO or increasing concentrations (three-fold serial dilution) of GSK2830371 for 6 days. Luminescence units were normalized to DMSO-treated controls and are represented as relative cell viability. Data shown are mean ± SD of three independent biological replicates (the red broken line indicates 50% viability). The listed P values were determined by unpaired t tests to evaluate the differences between three PPM1D-mutant and three TP53-mutant DIPG cell lines in response to indicated doses of GSK2830371 treatments.D, Effect of GSK2830371 on cell growth of DIPG cells. Six DIPG cell lines were treated with DMSO or indicated concentrations (0.19 and 1.67 μmol/L) of GSK2830371 and relative cell growth rates were determined by CellTiter-Glo Assay reagent at days 1, 4, and 7. Results shown are mean ± SEM of three independent biological replicates. The adjusted P values were determined by two-way ANOVA followed by Dunnett multiple comparisons test.

Close modal

Genetic inhibition of PPM1D increases PARPi sensitivity in SF7761 cells

We observed that GSK2830371 specifically inhibited the proliferation of PPM1D-mutant DIPG cells, but only a moderate decrease in viability was observed after treatment with GSK2832731 in both SF7761 and HSJD-DIPG-007 cells lines (Fig. 2C). Similarly, a previous study reported that GSK2830371 treatment slows, but does not completely inhibit, the growth of a lymphoma xenograft (22). PPM1D is best known for its role in dephosphorylating the key components (e.g., p53, γH2AX, ATM/ATR, and CHK1/2) involved in the DDR pathway (13–15). Combination therapies targeting key components of the DDR have been shown to induce synthetic lethality and overcome acquired resistance to single-agent DDR inhibitors (31–33). We, therefore, hypothesized that the therapeutic effects of PPM1D inhibition against PPM1D-mutant DIPG cell lines could be further enhanced by combining PPM1D inhibition with inhibitors targeting specific components of DDR pathways. To test this hypothesis, we first generated PPM1D−/− isogenic derivative cells in SF7761 using CRISPR/Cas9 gene editing to facilitate the screening (34). As shown in Fig. 3A, two sgRNAs targeting exons 2 and 6 of PPM1D were used to create a large deletion and increase the efficiency of PPM1D KO (Fig. 3A; Supplementary Table S1). We identified four single-cell–derived isogenic PPM1D−/− clones (c10, c41, c49, and c57) as confirmed by Sanger sequencing (Supplementary Fig. S1) and Western blotting (Fig. 3B). As expected, we observed an increase of p21 expression in all four PPM1D−/− clones as compared with the parental line (Fig. 3B), indicating that the activity of wild-type p53 was restored in response to PPM1D genetic inhibition in SF7761 cells. Moreover, we found that rescuing PPM1D expression by retroviral delivery of either full-length PPM1D (MIGR1-PPM1DWT) or truncated PPM1D (MIGR1-PPM1DE472X) in PPM1D−/− c10 could suppress p21 expression (Fig. 3C), supporting that the functional response of wild-type p53 pathway is intact in those PPM1D−/− clones.

Figure 3.

Loss of PPM1D sensitizes SF7761 cells to the treatments of PARPi. A, The schematic strategy of CRISPR/Cas9-mediated PPM1D KO. Two sgRNAs (red triangles) targeting exons 2 and 6 of PPM1D, respectively, were designed to knockout PPM1D gene in SF7761 cells. Western blot analyses of whole-cell lysates from SF7761 parental, PPM1D−/− clones (c10, c41, c49, and c57; B), and rescued cells (c10 MIGR1, c10 MIGR1-PPM1DWT, and c10 PPM1DE472X) analyzed for p21, PPM1D, and β-actin (loading control; C). D and E, Cell viability assays of SF7761 parental, PPM1D−/− clones (c10, c41, c49, and c57), and rescued cells (c10 MIGR1, c10 MIGR1-PPM1DWT, and c10 PPM1DE472X) treated with DMSO or indicated concentrations of PARPi (olaparib, rucaparib, or veliparib) for 7 days. Results shown are mean ± SD of triplicate wells and representative of two independent biological replicates. The comparisons of the dose–response curve fit between parental and PPM1D−/− clones (D) or c10 with parental/rescued clones were performed using extra sum-of-square F test. The adjusted maximum P values are listed in the figure.

Figure 3.

Loss of PPM1D sensitizes SF7761 cells to the treatments of PARPi. A, The schematic strategy of CRISPR/Cas9-mediated PPM1D KO. Two sgRNAs (red triangles) targeting exons 2 and 6 of PPM1D, respectively, were designed to knockout PPM1D gene in SF7761 cells. Western blot analyses of whole-cell lysates from SF7761 parental, PPM1D−/− clones (c10, c41, c49, and c57; B), and rescued cells (c10 MIGR1, c10 MIGR1-PPM1DWT, and c10 PPM1DE472X) analyzed for p21, PPM1D, and β-actin (loading control; C). D and E, Cell viability assays of SF7761 parental, PPM1D−/− clones (c10, c41, c49, and c57), and rescued cells (c10 MIGR1, c10 MIGR1-PPM1DWT, and c10 PPM1DE472X) treated with DMSO or indicated concentrations of PARPi (olaparib, rucaparib, or veliparib) for 7 days. Results shown are mean ± SD of triplicate wells and representative of two independent biological replicates. The comparisons of the dose–response curve fit between parental and PPM1D−/− clones (D) or c10 with parental/rescued clones were performed using extra sum-of-square F test. The adjusted maximum P values are listed in the figure.

Close modal

Using the parental cells and the above isogenic derivatives of SF7761, we performed a focused screen for inhibitors targeting ATM, ATR, PARP1/2, MTH1, Wee1, DNA-PK, or RAD51 (Supplementary Fig. S2A), which might display synergistic effects with PPM1D genetic inhibition. As shown in Supplementary Fig. S2B, no differential responses to treatments with the inhibitors targeting ATM, ATR, MTH1, Wee1, DNA-PK, or RAD51 between parental and PPM1D−/− clones were observed. However, we found that compared with the parental cells, the four PPM1D−/− clones were invariably more sensitive to the treatments with the inhibitors targeting PARP1/2, including two FDA-approved PARPi, olaparib and rucaparib, as well as veliparib (Fig. 3D). Moreover, rescuing PPM1D expression in the PPM1D−/− clone c10 decreased PARPi sensitivity to a level similar to the parental cells but no obvious differential responses between overexpressing either PPM1DWT or PPM1DE472X were observed (Fig. 3E).

Pharmacologic inhibition of PPM1D sensitizes PPM1D-mutant but not TP53-mutant DIPG cells to PARPi treatments

On the basis of the observation that genetic inhibition of PPM1D sensitizes the PPM1D-mutant (SF7761) cell line to PARPi treatment, we sought to investigate whether pharmacologic inhibition of PPM1D would sensitize PPM1D-mutant DIPG cell lines to PARPi treatment, indicating a possible benefit of combined pharmacologic therapy. Using Combenefit software and a classical Loewe synergy model (29), we analyzed drug combination effects in terms of synergy versus antagonism of the combined treatment of GSK2830371 and olaparib in six DIPG cell lines. As shown in Supplementary Fig. S3 (highlighted in red squares), we observed a substantial synergistic interaction between GSK2830371 and olaparib in the two PPM1D-mutant cell lines (SF7761 and HSJD-DIPG-007), which is consistent with the observed PARPi sensitivity conferred by PPM1D genetic inhibition in SF7761 cells (Fig. 3D). In contrast, in TT10714 cells, which are more sensitive to individual GSK2830371 treatment, we observed a much milder synergistic effect in response to combination treatment (Supplementary Fig. S3, bottom left). We speculated that the increased sensitivity of TT10714 cells to GSK2830371 treatment might mask the synergistic effect of the combinational treatments. Therefore, we adjusted the GSK2830371 concentration to a lower dose range to further evaluate the synergistic effect of GSK2830371 and olaparib. We observed a similarly mild synergistic effect as compared with the high-dose range of GSK2830371 gradient (Supplementary Fig. S4), with maximal, albeit mild, synergy occurring at the same dose combinations as (Supplementary Figs. S3 and S4). No obvious synergy was observed in the three PPM1D wild-type (TP53-mutant) cell lines (HSJD-DIPG-012, HSJD-DIPG-013, and TT10728; Supplementary Fig. S3, right). The effects of the combined treatment of GSK2830371 and olaparib on cell viabilities are shown in Supplementary Fig. S5. We further observed that the combination treatments with GSK2830371 and olaparib (at the concentrations associated with high synergy scores) significantly increased the percentage of Annexin V–positive cells as compared with GSK2830371 or olaparib treatment alone (Fig. 4A), indicating that the combination treatments significantly increased apoptotic cell death in PPM1D-mutant DIPG cells. In addition, we assessed the effect of combination GSK2830371 and olaparib treatment using a 3D tumor spheroid invasion assay. As shown in Supplementary Figs. S6–S8, we found that the combination treatments caused a further reduction in the size of the tumor spheroids grown in Matrigel in all three PPM1D-mutant cell lines relative to the single-agent treatments.

Figure 4.

Analysis of the cytotoxic effects of single agent, or combination treatments on PPM1D-mutant DIPG cells. A, Flow cytometry analysis of Annexin V/propidium iodide (PI) staining of cells following single-agent or combination treatments. HSJD-DIPG-007 cells, TT10714 cells, or SF7761 cells were treated with the indicated dose of GSK2830371 and/or olaparib for 3 days and stained with Annexin V/PI. Representative flow cytometry dot plot graphs (top). Results shown are mean ± SD % of apoptotic cells (Annexin V–positive + Annexin V and PI double-positive cells) from three biological distinct samples (bottom). The adjusted P values were determined by one-way ANOVA followed by Tukey multiple comparisons tests (**, P < 0.01; ***, P < 0.001). B, Cell viability analysis of PPM1D-mutant DIPG cells in response to single or combination treatments with radiotherapy. All three PPM1D-mutant DIPG cell lines were treated with vehicle (DMSO), GSK2830371, or GSK2830371 plus olaparib for 12 hours before exposure to radiation (2 Gy or 4 Gy). Cells were assayed for viability by CellTiter-Glo after 3 days post-irradiation incubation. Data shown are mean ± SD (n = 5) and representative of two independent experiments. The adjusted P values were determined by two-way ANOVA followed by Tukey multiple comparisons tests (*, P < 0.05; ***, P < 0.001; ****, P < 0.0001; n.s., not significant).

Figure 4.

Analysis of the cytotoxic effects of single agent, or combination treatments on PPM1D-mutant DIPG cells. A, Flow cytometry analysis of Annexin V/propidium iodide (PI) staining of cells following single-agent or combination treatments. HSJD-DIPG-007 cells, TT10714 cells, or SF7761 cells were treated with the indicated dose of GSK2830371 and/or olaparib for 3 days and stained with Annexin V/PI. Representative flow cytometry dot plot graphs (top). Results shown are mean ± SD % of apoptotic cells (Annexin V–positive + Annexin V and PI double-positive cells) from three biological distinct samples (bottom). The adjusted P values were determined by one-way ANOVA followed by Tukey multiple comparisons tests (**, P < 0.01; ***, P < 0.001). B, Cell viability analysis of PPM1D-mutant DIPG cells in response to single or combination treatments with radiotherapy. All three PPM1D-mutant DIPG cell lines were treated with vehicle (DMSO), GSK2830371, or GSK2830371 plus olaparib for 12 hours before exposure to radiation (2 Gy or 4 Gy). Cells were assayed for viability by CellTiter-Glo after 3 days post-irradiation incubation. Data shown are mean ± SD (n = 5) and representative of two independent experiments. The adjusted P values were determined by two-way ANOVA followed by Tukey multiple comparisons tests (*, P < 0.05; ***, P < 0.001; ****, P < 0.0001; n.s., not significant).

Close modal

Radiotherapy is the standard of care and the only proven beneficial treatment for patients with DIPG (1). We, therefore, determined the combinatory effect of the dual GSK2830371 and olaparib treatment plus ionizing radiation (IR). As shown in Fig. 4B, we found that the presence of either or both agents together with IR displayed potent cytotoxicity. Notably, we observed a stronger (in HSJD-DIPG-007 and TT10714 lines) or similar (in SF7761 line) cytotoxic effect of a lower dose of IR (2 Gy) in the presence of both agents when compared with the cytotoxic effect of a higher dose of IR (4 Gy) without either agent, indicating the triple-combination therapy provides even more compelling results in vitro in terms of killing PPM1D-mutant cells.

Given that the synergistic effects of GSK2830371 and olaparib were only observed in PPM1D-mutant but not in PPM1D wild-type (TP53-mutant) DIPG cell lines, we investigated whether this synergism requires a functional p53 pathway. We knocked out TP53 in both SF7761 and HSJD-DIPG-007 cells using lentiviral delivery of Cas9 and two guide RNAs targeting exons 5 and 8 of TP53 and a nontargeting sgRNA (NTC sgRNA) was used as a control (27). The knockout efficiency of TP53 in both cell lines was confirmed by loss of p53 protein expression and decreased expression of p21 in response to GSK2830371 treatment (Fig. 5A). We then evaluated the synergistic effects of combined treatments with GSK2830371 and olaparib in both NTC control and TP53-KO cells. As shown in Fig. 5B, we detected a synergistic effect of the combined treatments in the NTC condition, but not in TP53-KO cells (highlighted in red squares), suggesting that the synergistic interaction between PPM1D inhibition and PARPi requires p53 pathway function. To further explore how the combination treatments affect the p53 pathway, we examined the effect of combination treatments (at the doses of two drugs with high synergy scores) on the induction of p21 by p53 activation in HSJD-DIPG-007 NTC and HSJD-DIPG-007 TP53-KO cells. As shown in Fig. 5C, we detected an upregulation (∼3-fold) of p21 expression in response to GSK2830371 treatment alone in HSJD-DIPG-007 NTC cells but not in HSJD-DIPG-007 TP53-KO cells. However, no obvious induction of p21 in olaparib treatment alone was observed. Consistent with the synergistic effect of these two drugs on cell viability in NTC control cells, a strong induction (∼6-fold) of p21 expression in the combination treatment group in HSJD-DIPG-007 NTC cells but not in HSJD-DIPG-007 TP53-KO cells was observed (Figs. 5C and D).

Figure 5.

Synergistic interactions between olaparib and GSK2830371 are p53 dependent. A, Western blot analysis of p53 and p21 confirming the knockout of p53 protein and the loss of p53 response in SF7761 and HSJD-DIPG-007 cells. GAPDH was used as the loading control. B, Matrix synergy plots showing the synergy/antagonism score for each combination of GSK2830371 and olaparib in SF7761 and HSJD-DIPG-007 cells infected with sgRNAs against NTC or TP53. Values are calculated from cell viabilities after treatment with indicated drug concentrations for 7 days, as measured by CellTiter-Glo Assay. Loewe synergy scores were calculated by Combenefit software. The highlighted red squares indicate that the synergistic interaction between olaparib and GSK2830371 is lost in TP53 KO SF7761 and HSJD-DIPG-007 cells. Asterisks indicate the level of significance (*, P < 0.05; **, P < 0.001; ***, P < 0.0001). C, Western blot analysis of the relative expressions of p21 and p53 in HSJD-DIPG-007 isogenic cells treated with GSK2830371 (0.19 μmol/L) or/and olaparib (0.63 μmol/L) for 48 hours. β-actin was used as a loading control. D, Densitometry quantification of the relative expression of p21 from three independent biological replicates. The adjusted P values were determined by ordinary one-way ANOVA followed by Tukey multiple comparisons test. Error bars represent SD. The 95% confidence intervals are: NTC GSK2830371, (1.38–5.08); NTC olaparib, (0.83–1.65); and NTC combination (1.81–8.75).

Figure 5.

Synergistic interactions between olaparib and GSK2830371 are p53 dependent. A, Western blot analysis of p53 and p21 confirming the knockout of p53 protein and the loss of p53 response in SF7761 and HSJD-DIPG-007 cells. GAPDH was used as the loading control. B, Matrix synergy plots showing the synergy/antagonism score for each combination of GSK2830371 and olaparib in SF7761 and HSJD-DIPG-007 cells infected with sgRNAs against NTC or TP53. Values are calculated from cell viabilities after treatment with indicated drug concentrations for 7 days, as measured by CellTiter-Glo Assay. Loewe synergy scores were calculated by Combenefit software. The highlighted red squares indicate that the synergistic interaction between olaparib and GSK2830371 is lost in TP53 KO SF7761 and HSJD-DIPG-007 cells. Asterisks indicate the level of significance (*, P < 0.05; **, P < 0.001; ***, P < 0.0001). C, Western blot analysis of the relative expressions of p21 and p53 in HSJD-DIPG-007 isogenic cells treated with GSK2830371 (0.19 μmol/L) or/and olaparib (0.63 μmol/L) for 48 hours. β-actin was used as a loading control. D, Densitometry quantification of the relative expression of p21 from three independent biological replicates. The adjusted P values were determined by ordinary one-way ANOVA followed by Tukey multiple comparisons test. Error bars represent SD. The 95% confidence intervals are: NTC GSK2830371, (1.38–5.08); NTC olaparib, (0.83–1.65); and NTC combination (1.81–8.75).

Close modal

Genetic inhibition of PPM1D impairs the formation of RAD51 nuclear foci in SF7761 cells induced by olaparib treatment

Previous studies have demonstrated that deficiencies of BRCA1/2, PALB2, ATM, ATR, XRCC2/3, PTEN1, or the FANC gene family confer sensitivities to PARPi treatments, a phenotype most likely caused by defects in RAD51-mediated DSB repair by HR (35). We evaluated whether PPM1D inhibition would affect the formation of RAD51 nuclear foci, which is a commonly used biomarker to indicate HR activity and PARPi resistance/sensitivity in preclinical models and clinical samples (36, 37). As shown in Fig. 6A and B, in vehicle-treated cells, over 80% of (SF7761 parental, c10, c10 PPM1DWT, or c10 PPM1DE472X) cells were negatively (<5 foci) stained for RAD51 nuclear foci, and we observed no statistically significant difference of RAD51 nuclear foci staining between the PPM1D−/− clone c10 and SF7761 parental or rescued cells (Fig. 6B). In olaparib-treated cells, we observed a dramatic increase in the percentage (∼65%) of RAD51 nuclear foci–positive cells (≥5 foci) in parental and PPM1D-rescued cells but not in PPM1D−/− clone c10 (∼20%; Fig. 6A and B). In addition, we observed the immunofluorescence staining signals (Fig. 6A) of γH2AX, a marker for DSBs and a direct substrate of PPM1D (15), was increased in PPM1D−/− clone c10 relative to parental cells or rescued cells (c10 PPM1DWT and c10 PPM1DE472X), in both vehicle and olaparib treatment conditions. Next, we assessed the non-homologous end joining (NHEJ) repair pathway for changes in response to olaparib treatment, using 53BP1 nuclear foci as a marker for DSB for repair by NHEJ (38). As shown in Supplementary Fig. S9A and S9B, 53BP1 nuclear foci were dramatically increased in response to olaparib treatment in both SF7761 parental and PPM1D−/− clone c10, suggesting that the NHEJ pathway remains functional in response to PPM1D genetic inhibition.

Figure 6.

Genetic deletion of PPM1D reduces the formation of nuclear RAD51 foci in response to olaparib treatment. Representative stitched confocal images of RAD51 (red) and γH2AX (green) foci. DAPI nuclear staining is shown in blue. The SF7761 parental and isogenic derivatives treated with DMSO or olaparib (5 μmol/L) for 48 hours were fixed and stained with antibodies against γH2AX and RAD51. A, Representative stitched confocal images of RAD51 (red) and γH2AX (green) foci in SF7761 isogenic cells treated with DMSO (left) or olaparib (right). B, Quantitative analysis of RAD51 nuclear foci in DMSO (left) or olaparib (right) treated cells. The number of RAD51 nuclear foci from three different stitched confocal images was counted by ImageJ_Fiji software. Frequency of RAD51 foci per nucleus categorized in 0–4, ≥5 foci was presented as a percentage of cells in the bar graph. Data are mean ± SD of three randomly selected pictures and were analyzed by two-way ANOVA followed by Tukey multiple comparisons tests (****, P < 0.0001).

Figure 6.

Genetic deletion of PPM1D reduces the formation of nuclear RAD51 foci in response to olaparib treatment. Representative stitched confocal images of RAD51 (red) and γH2AX (green) foci. DAPI nuclear staining is shown in blue. The SF7761 parental and isogenic derivatives treated with DMSO or olaparib (5 μmol/L) for 48 hours were fixed and stained with antibodies against γH2AX and RAD51. A, Representative stitched confocal images of RAD51 (red) and γH2AX (green) foci in SF7761 isogenic cells treated with DMSO (left) or olaparib (right). B, Quantitative analysis of RAD51 nuclear foci in DMSO (left) or olaparib (right) treated cells. The number of RAD51 nuclear foci from three different stitched confocal images was counted by ImageJ_Fiji software. Frequency of RAD51 foci per nucleus categorized in 0–4, ≥5 foci was presented as a percentage of cells in the bar graph. Data are mean ± SD of three randomly selected pictures and were analyzed by two-way ANOVA followed by Tukey multiple comparisons tests (****, P < 0.0001).

Close modal

PPM1D inhibition suppresses RAD51 expression in a p53-dependent manner

Previous studies have shown that decreased RAD51 expression and loss of RAD51 foci are correlated with a decrease in HR efficiency (39–41). To further define how PPM1D inhibition affects the formation of RAD51 nuclear foci in response to DNA damage breaks and confers PARPi sensitivity, we examined the protein expression changes of RAD51 in response to olaparib treatments in SF7761 and its isogenic derivatives. We observed a decrease of RAD51 expression in all three PPM1D−/− clones (c10, c49, and c57) as compared with SF7761 parental and rescued isogenic derivatives, and a further substantial decrease of RAD51 expression in three PPM1D−/− clones after olaparib treatment (Fig. 7A). Consistent with the results observed in Fig. 3B, the expression of p21 in PPM1D−/− clones was increased in SF7761 parental and rescued isogenic derivatives (Fig. 7A).

Figure 7.

PPM1D inhibition suppresses RAD51 expression in a p53-dependent manner. A, SF7761 PPM1D−/− clones treated with DMSO or olaparib (5 μmol/L) for 48 hours were harvested and Western blot analysis was performed for RAD51, PPM1D, p53, p21, and GAPDH (loading control). B, Western blot analysis of RAD51 expression in SF7761, TT10714, and HSJD-DIPG-007 cells treated with increasing concentration of GSK2830371 for 48 hours. GAPDH was used as a loading control. Please note that the GAPDH panel for SF7761 is the same as used in Fig. 2B because the RAD51 data were derived from the same experiment. C, qRT-PCR analysis of RAD51 expression in six DIPG cell lines treated with DMSO or GSK2830371 (2 μmol/L) for 48 hours. All values were normalized to GAPDH gene (the dashed line indicates the normalized expression of RAD51 in DMSO control). Data are mean ± SD of three independent biological replicates. Each dot represents one biological replicate. P values were determined by one sample t test. The 95% confidence intervals are: SF7761, (0.79–0.92); HSJD-DIPG-007, (0.26–0.87); TT10714, (0.41–0.64); TT10728, (0.63–1.39); HSJD-DIPG-012, (0.56–1.72); and HSJD-DIPG-013, (0.64–1.36). D, HSJD-DIPG-007 NTC and HSJD-DIPG-007 TP53 KO cells treated with GSK2830371 (0.19 μmol/L) or/and olaparib (0.63 μmol/L) for 48 hours were harvested and subjected to Western blot analysis for RAD51, p53, and GAPDH. This experiment has been repeated three times and showed the same trends. E, Densitometry quantification of the relative expression of RAD51 from three independent biological replicates. The adjusted P values were determined by ordinary one-way ANOVA followed by Tukey multiple comparisons test. Error bars represent SD. The 95% confidence intervals are: NTC GSK2830371, (0.21–1.31); NTC olaparib, (0.21–1.82); and NTC combination (−0.05–0.84).

Figure 7.

PPM1D inhibition suppresses RAD51 expression in a p53-dependent manner. A, SF7761 PPM1D−/− clones treated with DMSO or olaparib (5 μmol/L) for 48 hours were harvested and Western blot analysis was performed for RAD51, PPM1D, p53, p21, and GAPDH (loading control). B, Western blot analysis of RAD51 expression in SF7761, TT10714, and HSJD-DIPG-007 cells treated with increasing concentration of GSK2830371 for 48 hours. GAPDH was used as a loading control. Please note that the GAPDH panel for SF7761 is the same as used in Fig. 2B because the RAD51 data were derived from the same experiment. C, qRT-PCR analysis of RAD51 expression in six DIPG cell lines treated with DMSO or GSK2830371 (2 μmol/L) for 48 hours. All values were normalized to GAPDH gene (the dashed line indicates the normalized expression of RAD51 in DMSO control). Data are mean ± SD of three independent biological replicates. Each dot represents one biological replicate. P values were determined by one sample t test. The 95% confidence intervals are: SF7761, (0.79–0.92); HSJD-DIPG-007, (0.26–0.87); TT10714, (0.41–0.64); TT10728, (0.63–1.39); HSJD-DIPG-012, (0.56–1.72); and HSJD-DIPG-013, (0.64–1.36). D, HSJD-DIPG-007 NTC and HSJD-DIPG-007 TP53 KO cells treated with GSK2830371 (0.19 μmol/L) or/and olaparib (0.63 μmol/L) for 48 hours were harvested and subjected to Western blot analysis for RAD51, p53, and GAPDH. This experiment has been repeated three times and showed the same trends. E, Densitometry quantification of the relative expression of RAD51 from three independent biological replicates. The adjusted P values were determined by ordinary one-way ANOVA followed by Tukey multiple comparisons test. Error bars represent SD. The 95% confidence intervals are: NTC GSK2830371, (0.21–1.31); NTC olaparib, (0.21–1.82); and NTC combination (−0.05–0.84).

Close modal

The above results reveal that genetic PPM1D inhibition suppresses RAD51 expression, as well as the formation of RAD51 nuclear foci in SF7761 cells induced by olaparib treatment. We then evaluated the effect of pharmacologic PPM1D inhibition on the RAD51 expression on all three PPM1D-mutant DIPG cell lines. As shown in Fig. 7B, GSK2830371 treatment caused a concentration-dependent decrease of RAD51 protein expression in all three PPM1D-mutant DIPG cell lines. Moreover, we also observed that GSK2830371 treatment suppressed the RAD51 mRNA expression in all three PPM1D-mutant cells but not in all three TP53-mutant cells (Fig. 7C), Together, these results suggest that both genetic and pharmacologic PPM1D inhibition suppresses RAD51 expression in PPM1D-mutant DIPG cell lines.

As the synergistic effects of GSK2830371 and olaparib observed in PPM1D-mutant cell lines are dependent on the function of p53 (Supplementary Figs. S3 and S4B), we then assessed whether the effect of PPM1D inhibition on RAD51 expression is also mediated via the p53 pathway. To answer this question, we first evaluated the RAD51 expression in HSJD-DIPG-007 NTC control cells and TP53-KO isogenic cells. RAD51 expression was slightly decreased in response to GSK2830371 treatment but a profound decrease of RAD51 expression was observed in the combination treatment with GSK2830371 and olaparib in NTC cells (Fig. 7D and E). However, no statistically significant changes of RAD51 expression were observed in TP53-KO isogenic cells in response to either single treatment or combination treatment (Fig. 7D and E), which further supports that PPM1D regulates RAD51 expression through the functional p53 pathway.

Recent genomic studies profiling DIPGs have identified the major underlying genetic alterations in these tumors, including alterations of H3.3/H3.1, PDGFRA, ACVR1, TP53, and PPM1D (42). Somatic mutations of PPM1D are found in 9% to 29% of DIPGs and generally occur in a mutually exclusive fashion with mutations in TP53 (4, 7, 8). Stabilized truncated PPM1D increases its ability to dephosphorylate checkpoint kinases and several other key components involved in DDR. We evaluated the therapeutic efficacy of PPM1D inhibition alone or in combination with inhibitors targeting the DDR pathways with the aim to develop a more effective combination therapy for PPM1D-mutant DIPGs. In this study, we found that inhibition of PPM1D sensitizes PPM1D-mutant DIPG cell lines to PARPi treatments. We demonstrated that in PPM1D-mutant, but not TP53-mutant DIPG cells, PPM1D inhibition sensitizes the PPM1D-mutant DIPG cells to PARPi treatment, as evidenced by dramatically reduced cell viability and the growth of 3D tumor spheroids, as well as induced cell apoptosis. This effect may be due to PPM1D inhibition impairing the formation of RAD51 nuclear foci at the transcriptional level, downregulating the expression of RAD51, and impairing the formation of RAD51 nuclear foci, which would possibly impair RAD51-mediated HR and increase the DSBs induced by PARPi.

GSK2830371 is an orally active, allosteric PPM1D inhibitor and has been shown to induce p53-dependent growth inhibition in vitro and in vivo (22). A very recent study in DIPG cells reveals that GSK2830371 significantly reduces the proliferation of HSJD-DIPG-007 (IC50 4.6 μmol/L), but not the TP53-mutant DIPG VI (IC50 226 μmol/L) in vitro. The investigators reported that treatment with GSK2830371 augments IR-induced apoptosis in HSJD-DIPG-007 cells, similar to our results showing a synergistic decrease of cell viability in response to combined PPM1D and PARP inhibition (43). Our results further show that one of the mechanisms underlying the synergy of these combinations is a functional reconstitution of the p53 pathway, p53-mediated suppression of Rad51 expression, and possibly a decrease in HR-mediated DNA repair.

Previous studies have shown that ectopic expression of PPM1D suppresses the DNA repair mediated by BER and NER as well as the DNA DSB repair by disrupting the recruitment of critical DNA repair factors to damaged sites (13–15). Here, we discovered an unanticipated role of PPM1D in the regulation of RAD51, a key component involved in DNA repair through HR. It has been shown that hypoxia, HDAC inhibition, and PPP2R2A loss could inhibit HR activity by downregulating RAD51 expression (41, 44, 45). Our study identifies PPM1D as another regulator of RAD51 expression. Therefore, we propose that the synergistic effect of PPM1D inhibition and PARPi treatment is mediated at least partially by impairing the RAD51-mediated HR pathway. In agreement with our findings, a recent study in non-DIPG cell models also reveals that PPM1D modulates sensitivity to PARPi via regulating the recruitment of BRCA1 (46). Collectively, these studies reveal the critical role of PPM1D in regulating HR via multiple mechanisms.

We show that the synergistic effects of PPM1D inhibition and PARPi require an intact p53 pathway function. A previous study reported that wild-type p53 modulates HR by transcriptional regulation of the RAD51 gene and inhibits the formation of RAD51 nuclear foci (40). In addition to its role in the regulation of RAD51, wild-type p53 has also been reported to inhibit expression and control nuclear/cytoplasmic localization of BRCA1, a key factor required for HR-mediated DNA repair (47). A recent study in breast and glioma cells further revealed that the synthetic lethality of PARPi and IR is p53-dependent (48). Similarly, our results here demonstrate that PPM1D inhibition suppresses RAD51 expression in a p53-dependent manner.

Tumors with defects in HR-mediated DNA repair are dependent on PARP-mediated DNA repair and have a profound sensitivity to PARPi. Four PARPi, olaparib, rucaparib, niraparib, and talazoparib, have been approved by FDA to treat certain types of ovarian, breast, and prostate cancers with germline or somatic mutations in BRCA1/2 (49, 50). Recent discoveries of HR deficiency caused by genetic alterations other than BRCA1/2 mutations have led to the design of clinical trials to test the PARPi for cancers with BRCAness phenotype (50). Moreover, clinical trials investigating combination regimens of PARPi with cytotoxic agents, molecular-targeted agents or immunotherapies are underway with the hope of expanding the patient populations that may benefit from PARPi (50). In this study, we found that PPM1D inhibition decreases indicators of HR activity and ultimately leads to PARPi sensitivity in PPM1D-mutant DIPG. Therefore, our findings provide a proof-of-concept that combination treatment with a PPM1D inhibitor and PARPi may represent a therapeutic strategy for patients with PPM1D-mutant DIPG. Future studies to test this combination treatment in an expanded cohort of patient-derived DIPG cell lines and preclinical PPM1D-mutant DIPG mouse models are warranted to explore the efficacy of dual inhibition of PPM1D and PARP for treating the patients with PPM1D-mutant DIPGs.

H. Yan is CSO of Genetron Holdings and has ownership interest in Genetron Holdings. H. Yan receives royalties from Genetron Holdings, Agios, and PGDX. No potential conflicts of interest were disclosed by the other authors.

Conception and design: Z. Wang, L. Zhang, H. Yan

Development of methodology: Z. Wang, C. Xu, B.H. Diplas

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Z. Wang, C. Xu, C.J. Moure, C.-P.J. Chen, H. Zhu, P.K. Greer

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Z. Wang, B.H. Diplas, L.H. Chen, M.S. Waitkus

Writing, review, and/or revision of the manuscript: Z. Wang, C. Xu, B.H. Diplas, C.J. Moure, H. Zhu, Y. He, M.S. Waitkus, H. Yan

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C. Du, P.K. Greer, M.S. Waitkus

Study supervision: Y. He, M.S. Waitkus, H. Yan

Z. Wang received a fellowship (BRF1600002) grant from The American Brain Tumor Association in honor of the Bradley Benton Davis Memorial Foundation. Z. Wang and H. Yan received a research grant (SBTF-2015) from the Southeastern Brain Tumor Foundation. The authors would like to thank Drs. C. David James and Angel Montero Carcaboso for the generous gifts of DIPG cell lines. We also thank Dr. Jiaoti Huang for the generous gift of the RAD51 antibody. The authors would like to thank the core facilities used in this study, including the Duke Cancer Institute Flow Cytometry Shared Resource (Nancy Martin, Lynn Martinek, and Michael Cook), the Light Microscopy Core Facility (Lisa Cameron, Benjamin Carlson, and Yasheng Gao), and the Functional Genomics Core Facility (Sufeng Li and So Young Kim).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Hoffman
LM
,
Van Zanten
SEMV
,
Colditz
N
,
Baugh
J
,
Chaney
B
,
Hoffmann
M
, et al
Clinical, radiologic, pathologic, and molecular characteristics of long-term survivors of diffuse intrinsic pontine glioma (DIPG): a collaborative report from the International and European Society for Pediatric Oncology DIPG registries
.
J Clin Oncol
2018
;
36
:
1963
72
.
2.
Clymer
J
,
Kieran
MW
. 
The integration of biology into the treatment of diffuse intrinsic pontine glioma: a review of the North American clinical trial perspective
.
Front Oncol
2018
;
8
:
169
.
3.
Buczkowicz
P
,
Hoeman
C
,
Rakopoulos
P
,
Pajovic
S
,
Letourneau
L
,
Dzamba
M
, et al
Genomic analysis of diffuse intrinsic pontine gliomas identifies three molecular subgroups and recurrent activating ACVR1 mutations
.
Nat Genet
2014
;
46
:
451
6
.
4.
Taylor
KR
,
Mackay
A
,
Truffaux
N
,
Butterfield
YS
,
Morozova
O
,
Philippe
C
, et al
Recurrent activating ACVR1 mutations in diffuse intrinsic pontine glioma
.
Nat Genet
2014
;
46
:
457
61
.
5.
Fontebasso
AM
,
Papillon-Cavanagh
S
,
Schwartzentruber
J
,
Nikbakht
H
,
Gerges
N
,
Fiset
PO
, et al
Recurrent somatic mutations in ACVR1 in pediatric midline high-grade astrocytoma
.
Nat Genet
2014
;
46
:
462
6
.
6.
Wu
G
,
Broniscer
A
,
McEachron
TA
,
Lu
C
,
Paugh
BS
,
Becksfort
J
, et al
Somatic histone H3 alterations in pediatric diffuse intrinsic pontine gliomas and non-brainstem glioblastomas
.
Nat Genet
2012
;
44
:
251
3
.
7.
Zhang
L
,
Chen
LH
,
Wan
H
,
Yang
R
,
Wang
Z
,
Feng
J
, et al
Exome sequencing identifies somatic gain-of-function PPM1D mutations in brainstem gliomas
.
Nat Genet
2014
;
46
:
726
30
.
8.
Wu
G
,
Diaz
AK
,
Paugh
BS
,
Rankin
SL
,
Ju
B
,
Li
Y
, et al
The genomic landscape of diffuse intrinsic pontine glioma and pediatric non-brainstem high-grade glioma
.
Nat Genet
2014
;
46
:
444
50
.
9.
El Ayoubi
R
,
Boisselier
B
,
Rousseau
A
. 
Molecular landscape of pediatric diffuse intrinsic pontine gliomas: about 22 cases
.
J Neurooncol
2017
;
134
:
465
7
.
10.
Mueller
S
,
Jain
P
,
Liang
WS
,
Kilburn
L
,
Kline
C
,
Gupta
N
, et al
A pilot precision medicine trial for children with diffuse intrinsic pontine glioma—PNOC003: a report from the Pacific Pediatric Neuro-Oncology Consortium
.
Int J Cancer
2019
;
145
:
1889
1901
.
11.
Nikbakht
H
,
Panditharatna
E
,
Mikael
LG
,
Li
R
,
Gayden
T
,
Osmond
M
, et al
Spatial and temporal homogeneity of driver mutations in diffuse intrinsic pontine glioma
.
Nat Commun
2016
;
7
:
11185
.
12.
Pecháčková
S
,
Burdová
K
,
Macurek
L
. 
WIP1 phosphatase as pharmacological target in cancer therapy
.
J Mol Med
2017
;
95
:
589
99
.
13.
Nguyen
TA
,
Slattery
SD
,
Moon
SH
,
Darlington
YF
,
Lu
X
,
Donehower
LA
. 
The oncogenic phosphatase WIP1 negatively regulates nucleotide excision repair
.
DNA Repair
2010
;
9
:
813
23
.
14.
Lu
X
,
Bocangel
D
,
Nannenga
B
,
Yamaguchi
H
,
Appella
E
,
Donehower
LA
. 
The p53-induced oncogenic phosphatase PPM1D interacts with uracil DNA glycosylase and suppresses base excision repair
.
Mol Cell
2004
;
15
:
621
34
.
15.
Moon
SH
,
Lin
L
,
Zhang
X
,
Nguyen
TA
,
Darlington
Y
,
Waldman
AS
, et al
Wild-type p53-induced phosphatase 1 dephosphorylates histone variant γ-H2AX and suppresses DNA double strand break repair
.
J Biol Chem
2010
;
285
:
12935
47
.
16.
Kleiblova
P
,
Shaltiel
IA
,
Benada
J
,
Ševčík
J
,
Pecháčková
S
,
Pohlreich
P
, et al
Gain-of-function mutations of PPM1D/Wip1 impair the p53-dependent G1 checkpoint
.
J Cell Biol
2013
;
201
:
511
21
.
17.
Fons
NR
,
Sundaram
RK
,
Breuer
GA
,
Peng
S
,
McLean
RL
,
Kalathil
AN
, et al
PPM1D mutations silence NAPRT gene expression and confer NAMPT inhibitor sensitivity in glioma
.
Nat Commun
2019
;
10
:
1
10
.
18.
Dudgeon
C
,
Shreeram
S
,
Tanoue
K
,
Mazur
SJ
,
Sayadi
A
,
Robinson
RC
, et al
Genetic variants and mutations of PPM1D control the response to DNA damage
.
Cell Cycle
2013
;
12
:
2656
64
.
19.
Zhu
Y-H
,
Bulavin D
V
. 
Wip1-dependent signaling pathways in health and diseases
.
Prog Mol Biol Transl Sci
2012
;
106
:
307
25
.
20.
Gibson
C
,
Mcconkey
M
,
Adams
D
,
Mar
B
,
Mertins
P
. 
PPM1D truncating mutations confer resistance to chemotherapy and sensitivity to PPM1D inhibition in hematopoietic cells
.
Blood
2018
;
132
:
1095
105
.
21.
Hsu
JI
,
Dayaram
T
,
Tovy
A
,
De Braekeleer
E
,
Jeong
M
,
Wang
F
, et al
PPM1D mutations drive clonal hematopoiesis in response to cytotoxic chemotherapy
.
Cell Stem Cell
2018
;
23
:
700
13
.
22.
Gilmartin
AG
,
Faitg
TH
,
Richter
M
,
Groy
A
,
Seefeld
MA
,
Darcy
MG
, et al
Allosteric Wip1 phosphatase inhibition through flap-subdomain interaction
.
Nat Chem Biol
2014
;
10
:
181
7
.
23.
Hashizume
R
,
Smirnov
I
,
Liu
S
,
Phillips
JJ
,
Hyer
J
,
McKnight
TR
, et al
Characterization of a diffuse intrinsic pontine glioma cell line: Implications for future investigations and treatment
.
J Neurooncol
2012
;
110
:
305
13
.
24.
Tsoli
M
,
Liu
J
,
Franshaw
L
,
Shen
H
,
Cheng
C
,
Jung
M
, et al
Dual targeting of mitochondrial function and mTOR pathway as a therapeutic strategy for diffuse intrinsic pontine glioma
.
Oncotarget
2018
;
9
:
7541
56
.
25.
Xu
C
,
Liu
X
,
Geng
Y
,
Bai
Q
,
Pan
C
,
Sun
Y
, et al
Patient-derived DIPG cells preserve stem-like characteristics and generate orthotopic tumors
.
Oncotarget
2017
;
8
:
76644
55
.
26.
Waitkus
MS
,
Pirozzi
CJ
,
Moure
CJ
,
Diplas
BH
,
Hansen
LJ
,
Carpenter
AB
, et al
Adaptive evolution of the GDH2 allosteric domain promotes gliomagenesis by resolving IDH1R132H-induced metabolic liabilities
.
Cancer Res
2018
;
78
:
36
50
.
27.
Diplas
BH
,
He
X
,
Brosnan-Cashman
JA
,
Liu
H
,
Chen
LH
,
Wang
Z
, et al
The genomic landscape of TERT promoter wildtype-IDH wildtype glioblastoma
.
Nat Commun
2018
;
9
:
1
11
.
28.
Schindelin
J
,
Arganda-Carreras
I
,
Frise
E
,
Kaynig
V
,
Longair
M
,
Pietzsch
T
, et al
Fiji: an open-source platform for biological-image analysis
.
Nat Methods
2012
;
9
:
676
82
.
29.
Di Veroli
GY
,
Fornari
C
,
Wang
D
,
Mollard
S
,
Bramhall
JL
,
Richards
FM
, et al
Combenefit: an interactive platform for the analysis and visualization of drug combinations
.
Bioinformatics
2016
;
32
:
2866
8
.
30.
Vinci
M
,
Box
C
,
Eccles
SA
. 
Three-dimensional (3D) tumor spheroid invasion assay
.
J Vis Exp
2015
;
2015
:
1
9
.
31.
Pilié
PG
,
Tang
C
,
Mills
GB
,
Yap
TA
. 
State-of-the-art strategies for targeting the DNA damage response in cancer
.
Nat Rev Clin Oncol
2019
;
16
:
81
104
.
32.
Kim
H
,
George
E
,
Ragland
RL
,
Rafail
S
,
Zhang
R
,
Krepler
C
, et al
Targeting the ATR/CHK1 axis with PARP inhibition results in tumor regression in BRCA-mutant ovarian cancer models
.
Clin Cancer Res
2017
;
23
:
3097
108
.
33.
Brandsma
I
,
Fleuren
EDG
,
Williamson
CT
,
Lord
CJ
. 
Directing the use of DDR kinase inhibitors in cancer treatment
.
Expert Opin Investig Drugs
2017
;
26
:
1341
55
.
34.
Ran
FA
,
Hsu
PD
,
Wright
J
,
Agarwala
V
,
Scott
DA
,
Zhang
F
. 
Genome engineering using the CRIPR-Cas9 system
.
Nature
2013
;
8
:
2281
308
.
35.
Lord
CJ
,
Ashworth
A
. 
BRCAness revisited
.
Nat Rev Cancer
2016
;
16
:
110
20
.
36.
Castroviejo Bermejo
M
,
Cruz
C
,
Llop Guevara
A
,
Gutiérrez Enríquez
S
,
Ducy
M
,
Ibrahim
YH
, et al
A RAD51 assay feasible in routine tumor samples calls PARP inhibitor response beyond BRCA mutation
.
EMBO Mol Med
2018
;
10
:
e9172
.
37.
Cruz
C
,
Castroviejo-Bermejo
M
,
Gutiérrez-Enríquez
S
,
Llop-Guevara
A
,
Ibrahim
YH
,
Gris-Oliver
A
, et al
RAD51 foci as a functional biomarker of homologous recombination repair and PARP inhibitor resistance in germline BRCA-mutated breast cancer
.
Ann Oncol
2018
;
29
:
1203
10
.
38.
Gupta
A
,
Hunt
CR
,
Chakraborty
S
,
Pandita
RK
,
Yordy
J
,
Ramnarain
DB
, et al
Role of 53BP1 in the regulation of DNA double-strand break repair pathway choice
.
Radiat Res
2014
;
181
:
1
8
.
39.
Mendes-Pereira
AM
,
Martin
SA
,
Brough
R
,
McCarthy
A
,
Taylor
JR
,
Kim
JS
, et al
Synthetic lethal targeting of PTEN mutant cells with PARP inhibitors
.
EMBO Mol Med
2009
;
1
:
315
22
.
40.
Arias-Lopez
C
,
Lazaro-Trueba
I
,
Kerr
P
,
Lord
CJ
,
Dexter
T
,
Iravani
M
, et al
p53 modulates homologous recombination by transcriptional regulation of the RAD51 gene
.
EMBO Rep
2006
;
7
:
219
24
.
41.
Kalev
P
,
Simicek
M
,
Vazquez
I
,
Munck
S
,
Chen
L
,
Soin
T
, et al
Loss of PPP2R2A inhibits homologous recombination DNA repair and predicts tumor sensitivity to PARP inhibition
.
Cancer Res
2012
;
72
:
6414
24
.
42.
Buczkowicz
P
,
Hawkins
C
. 
Pathology, molecular genetics, and epigenetics of diffuse intrinsic pontine glioma
.
Front Oncol
2015
;
5
:
1
9
.
43.
Akamandisa
MP
,
Nie
K
,
Nahta
R
,
Hambardzumyan
D
,
Castellino
RC
. 
Inhibition of mutant PPM1D enhances DNA damage response and growth suppressive effects of ionizing radiation in diffuse intrinsic pontine glioma
.
Neuro Oncol
2019
;
21
:
786
99
.
44.
Maranto
C
,
Udhane
V
,
Hoang
DT
,
Gu
L
,
Alexeev
V
,
Malas
K
, et al
STAT5A/B blockade sensitizes prostate cancer to radiation through inhibition of RAD51 and DNA repair
.
Clin Cancer Res
2018
;
24
:
1917
31
.
45.
Adimoolam
S
,
Sirisawad
M
,
Chen
J
,
Thiemann
P
,
Ford
JM
,
Buggy
JJ
. 
HDAC inhibitor PCI-24781 decreases RAD51 expression and inhibits homologous recombination
.
Proc Natl Acad Sci U S A
2007
;
104
:
19482
7
.
46.
Burdova
K
,
Storchova
R
,
Palek
M
,
Macurek
L
. 
WIP1 Promotes homologous recombination and modulates sensitivity to PARP inhibitors
.
Cells
2019
;
8
:
1258
.
47.
Jiang
J
,
Yang
ES
,
Jiang
G
,
Nowsheen
S
,
Wang
H
,
Wang
T
, et al
p53-dependent BRCA1 nuclear export controls cellular susceptibility to DNA damage
.
Cancer Res
2011
;
71
:
5546
57
.
48.
Sizemore
ST
,
Mohammad
R
,
Sizemore
GM
,
Nowsheen
S
,
Yu
H
,
Ostrowski
MC
, et al
Synthetic lethality of PARP inhibition and ionizing radiation is p53-dependent
.
Mol Cancer Res
2018
;
16
:
1092
102
.
49.
Litton
JK
,
Rugo
HS
,
Ettl
J
,
Hurvitz
SA
,
Gonçalves
A
,
Lee
KH
, et al
Talazoparib in patients with advanced breast cancer and a germline BRCA mutation
.
N Engl J Med
2018
;379:753–63.
50.
Lord
CJ
,
Ashworth
A
. 
PARP inhibitors: synthetic lethality in the clinic
.
Science
2017
;
355
:
1152
8
.