The DNA damage response is critical for maintaining genome integrity and is commonly disrupted in the development of cancer. PPM1D (protein phosphatase, Mg2+/Mn2+ dependent 1D) is a master negative regulator of the response; gain-of-function mutations and amplifications of PPM1D are found across several human cancers making it a relevant pharmacologic target. Here, we used CRISPR/Cas9 screening to identify synthetic-lethal dependencies of PPM1D, uncovering superoxide dismutase-1 (SOD1) as a potential target for PPM1D-mutant cells. We revealed a dysregulated redox landscape characterized by elevated levels of reactive oxygen species and a compromised response to oxidative stress in PPM1D-mutant cells. Moreover, we observed marked genomic instability in mutant cells, which is exacerbated upon inhibition of SOD1. Altogether, our results demonstrate the protective role of SOD1 against oxidative stress and DNA damage in PPM1D-mutant leukemia cells and highlight a new potential therapeutic strategy against PPM1D-mutant cancers.
Gain-of-function mutations and amplifications of PPM1D are found across several human cancers and are associated with advanced tumor stage, worse prognosis, and increased lymph node metastasis. Unfortunately, clinical translation has so far not been possible due to the lack of PPM1D inhibitors with favorable pharmacokinetic properties. This useful study leveraged CRISPR/Cas9 screening to identify that SOD1 inhibition is synthetic lethal with PPM1D mutation in leukemia, although the mechanistic analyses were incomplete.
Cellular DNA is frequently damaged by both endogenous and exogenous factors (Hoeijmakers, 2009). Unresolved DNA damage can lead to genomic instability, which is a hallmark of aging and cancer (Hanahan and Weinberg, 2011). Cells have evolved intricate mechanisms to detect and repair DNA lesions. The DNA damage response (DDR) is a complex network of signaling pathways that coordinate various cellular processes initiated by p53, such as DNA repair (Ciccia and Elledge, 2010), cell cycle checkpoint activation (Harper et al., 1993), and apoptosis (Yonish-Rouach et al., 1991). However, upon resolution of DNA damage, the cell must terminate the DDR to avoid prolonged cell cycle arrest and apoptosis. One critical mechanism for DDR termination is the expression of Protein Phosphatase Mg2+/Mn2+–Dependent 1D (PPM1D) (Fiscella et al., 1997), which is induced by p53 and plays a key role in attenuating the response. PPM1D is a member of the PP2C family of serine/threonine protein phosphatases and has been shown to dephosphorylate a wide range of DDR signaling molecules including p53, p38 MAPK, CHK1, CHK2, and H2AX (Bulavin et al., 2002; Cha et al., 2010; Lu et al., 2005; Oliva-Trastoy et al., 2007; Takekawa et al., 2000). These dephosphorylation events generally lead to reduced activity of the targets, ultimately resulting in deactivation of the DDR.
Dysregulation of PPM1D has been associated with the development of diverse cancers, including breast, ovarian, esophagus, brain, and others (Khadka et al., 2022; Li et al., 2002; Li et al., 2020b; Ruark et al., 2013; Zhang et al., 2014). PPM1D is located on chromosome 17q and therefore frequently amplified in breast and ovarian cancers exhibiting 17q23 amplifications (Li et al., 2002; Ruark et al., 2013). These amplifications result in overexpression of the wildtype PPM1D protein and consequent suppression of p53 and other PPM1D targets in the DDR (Bulavin et al., 2002; Lambros et al., 2010). In addition, PPM1D can also become dysregulated through mutations in its terminal exon. These mutations produce a truncated protein that is stabilized, evading proteasome-mediated degradation (Tokheim et al., 2021). The resulting mutant protein maintains its phosphatase activity and is found at high levels even in the absence of DNA damage. Excessive PPM1D activity leads to constitutive dephosphorylation and downregulation of PPM1D targets including multiple members of the DDR (Hsu et al., 2018). These gain-of-function PPM1D mutations are observed in diverse solid cancers including osteosarcoma (He et al., 2021), colorectal carcinoma (Peng et al., 2014; Yin et al., 2013), diffuse midline gliomas (Wang et al., 2011; Zhang et al., 2014) and others. Moreover, PPM1D mutations and overexpression are associated with advanced tumor stage, worse prognosis, and increased lymph node metastasis (Fu et al., 2014; Jiao et al., 2014; Li et al., 2020a; Li et al., 2020b; Peng et al., 2014; Zhang et al., 2014).
More recently, PPM1D-mutations have been shown to drive expansion of hematopoietic stem cells (Bolton et al., 2020; Hsu et al., 2018; Kahn et al., 2018) in association with clonal hematopoiesis (CH), a pre-malignant state associated with an increased risk of hematologic malignancies and elevated all-cause mortality (Genovese et al., 2014; Jaiswal et al., 2014). PPM1D-mutations are particularly enriched in patients with prior exposure to cytotoxic therapies, who have a high risk of therapy-related myeloid neoplasms (t-MN) (Hsu et al., 2018; Lindsley et al., 2017). Given the prevalence of PPM1D aberrations in cancer, PPM1D is an attractive therapeutic target. Ongoing efforts are focused on elucidating the structure of PPM1D to improve drug design and development (Miller et al., 2022). While several inhibitors thus far have shown efficacy in vitro, few have been studied in vivo and none have progressed to clinical trials due to poor bioavailability. Therefore, identifying targetable, synthetic-lethal partners to exploit the genetic defects of PPM1D-altered cells can offer an alternative therapeutic approach.
In this study, we performed an unbiased, whole-genome CRISPR screen to investigate genes essential for cell survival in PPM1D-mutated leukemia cell lines. We identified superoxide dismutase-1 (SOD1) as a novel synthetic-lethal dependency of PPM1D which was validated by genetic and pharmacologic approaches. We showed that the mutant cells display compromised responses to oxidative stress and DNA damage, leading to increased reactive oxygen species and genomic instability. These results provide valuable insights into the biological processes corrupted by mutant PPM1D and underscore the potential of SOD1 as a targetable vulnerability in this context.
SOD1 is a synthetic lethal vulnerability of PPM1D-mutant leukemia cells
CRISPR dropout screens have emerged as a powerful tool to assess the functional importance of individual genes within a particular pathway by measuring the impact of their depletion on cell viability or fitness. To identify genes essential for PPM1D-mutant cell survival, we first created isogenic wild-type (WT) and PPM1D-mutant Cas9-expressing OCI-AML2 leukemia cell lines and selected two PPM1D-mutant clones for CRISPR screening (Figure 1–figure supplement 1A). We transduced the cells with a whole-genome lentiviral library containing 90,709 guide RNAs (gRNAs) targeting 18,010 human genes (Tzelepis et al., 2016). At day ten post-transduction, the cells were harvested for the first timepoint and then subsequently passaged for an additional 18 days to allow for negatively selected gene-knockout cells to “drop out”. The remaining pool of cells were collected for deep sequencing analysis of gRNA abundance (Figure 1A). We analyzed genes that were specifically depleted in the mutant but not WT cells using the MaGECK-VISPR pipeline (Li et al., 2014). Differentially depleted genes are those for which the knockout or depletion of the gene results in a significant impact on the viability or growth of PPM1D-mutant cells compared to WT control cells. Through this analysis, we identified 409 differentially depleted genes in one of the PPM1D-mutant clones and 92 differentially depleted genes in the other clone while adhering to the maximum false discovery rate (FDR) cutoff of 25%. Among these genes, we found 37 common candidates that were depleted in both PPM1D-mutant biological replicates that were not depleted in the WT control cells (Figure 1–figure supplement 1B, Figure 1–source data 1).
Gene ontology analysis of these top essential genes demonstrated an enrichment in pathways related to DNA repair, interstrand crosslink (ICL) repair, and cellular responses to stress (Figure 1B). Pathway analyses with the KEGG and REAC databases revealed a significant enrichment of the Fanconi anemia (FA) repair pathway, with notable genes such as BRIP1 (FANCJ), FANCI, FANCA, SLX4 (FANCP), UBE2T (FANCT), and C19orf40 (FAAP24) (Figure 1–figure supplement 1C). Interestingly, our dropout screen revealed that superoxide dismutase 1, or SOD1, was the top essential gene based on fitness score (Figure 1C). SOD1 is a crucial enzyme involved in scavenging superoxide (O2–) radicals, which are harmful byproducts of mitochondrial cellular metabolism. Excessive reactive oxygen species (ROS) causes oxidative stress, which can damage cellular structures including DNA, proteins, and lipids. SOD1 is an attractive therapeutic target due to the availability of SOD1 small molecule inhibitors that are being tested in clinical trials (Lin et al., 2013; Lowndes et al., 2008). Therefore, we decided to further investigate the role of SOD1 in promoting PPM1D-mutant cell survival.
To validate the essentiality of SOD1 in PPM1D-mutant cells, we performed in vitro competitive proliferation assays in two different acute myeloid leukemia (AML) cell lines, OCI-AML2 and OCI-AML3. We transduced isogenic WT and PPM1D-mutant Cas9-expressing cells with either empty vector (EV) or two different SOD1-gRNA-expressing lentiviral vectors containing a blue-fluorescent protein (BFP) reporter. While loss of SOD1 had minimal effects on the fitness of WT cells, PPM1D-mutant cells with SOD1 deletion had significant reduction in cellular growth in both OCI-AML2 and OCI-AML3 cells (Figure 1D). Collectively, these results show that loss of SOD1 confers a significant disadvantage specific to PPM1D-mutant leukemia cells.
PPM1D-mutant cells are sensitive to SOD1 inhibition and have increased oxidative stress
We next wanted to test if pharmacologic inhibition of SOD1 could mimic the genetic deletion of SOD1. We used two different SOD1 inhibitors, Lung Cancer Screen-1 (LCS-1) and Bis-choline tetrathiomolybdate (ATN-224), which work by different mechanisms. LCS-1 is a small molecule that binds to SOD1 and disrupts its activity (Somwar et al., 2011), while ATN-224 is a copper chelator that reduces SOD1 activity by decreasing the availability of copper ions, which are an essential SOD1 cofactor (Juarez et al., 2006). To study the sensitivity of PPM1D-mutant cells to SOD1 inhibition, we engineered three patient-derived AML cell lines, MOLM-13, OCI-AML2, and OCI-AML3 which harbor distinct genetic backgrounds and AML driver mutations. Upon treatment with either ATN-224 or LCS-1, we found that SOD1 inhibition induced a significantly greater proportion of apoptotic PPM1D-mutant than PPM1D-WT cells (Figure 2–figure supplement 1A). PPM1D-truncating mutations conferred significant sensitivity to SOD1 inhibition compared to their WT counterparts in all three AML cell lines (Figure 2A, Figure 2–figure supplement 1B). To determine if this cytotoxicity was dependent on oxidative stress, we treated the cells with SOD1 inhibitors in combination with an antioxidant, N-acetylcysteine (NAC). Importantly, NAC supplementation was able to completely rescue the sensitivity of mutant cells to both LCS-1 and ATN-224 treatment (Figure 2B, Figure 2–figure supplement 1C), suggesting that ROS generation contributes to the sensitivity of mutant cells to SOD1 inhibition.
Activating mutations in oncogenes often lead to increased ROS generation by altering cellular metabolism, inducing replication stress, or dysregulating redox homeostasis (Maya-Mendoza et al., 2015; Park et al., 2014). We therefore hypothesized that PPM1D-mutant cells have increased oxidative stress, leading to reliance on SOD1 for protection. SOD1 catalyzes the breakdown of superoxide into hydrogen peroxide and water. Therefore, we first assessed mitochondrial superoxide levels using the fluorogenic dye, Mitosox Green. This mitochondrial-targeted dye is rapidly oxidized by superoxide, but not other types of ROS, to produce green fluorescence. We observed that in the absence of exogenous stressors, PPM1D-mutant cells had a moderate increase in superoxide radicals (Fig. 2C). Free radicals can be detrimental to cells due to their ability to oxidize proteins, lipids, and DNA. Therefore, we also measured levels of lipid peroxidation as an additional measure of oxidative stress. Consistent with the increase in superoxide radicals, we observed a concurrent increase in lipid peroxidation in the PPM1D-mutant cells (Figure 2D). Using 2’7’-dichlorofluorescein diacetate (DCFDA) staining to measure total ROS levels, we observed that PPM1D-mutant cells harbored more total ROS compared to WT cells and that upon inhibition of SOD1, there was an increase in ROS that was specific to the mutant cells (Figure 2E).
To investigate whether the observed elevated ROS was a characteristic of other PPM1D-mutant cell lines, we measured ROS levels in two different germline models. Humans with germline mutations in PPM1D were first described by Jansen et al. in 2017 in patients with intellectual disability. This neurodevelopmental condition is named Jansen-de Vries syndrome (JdVS, OMIM #617450) and is characterized by frameshift or nonsense mutations in the last or second-to-last exons of the PPM1D gene. These mutations result in functionally active, truncated mutant proteins like those exhibited in human cancers and clonal hematopoiesis. Lymphoblastoid cell lines (LCLs) were generated from these JdVS patients by Jansen et al (Jansen et al., 2017; Wojcik et al., 2023). In addition to human PPM1D-mutant LCLs, we also generated mouse embryonic fibroblasts (MEFs) from a germline mouse model harboring a heterozygous truncating mutation in the terminal exon of Ppm1d (Hsu et al., 2018). When we measured total ROS from both the JdVS LCLs and the Ppm1d-mutant MEFs compared to their WT counterparts, both mutant models exhibited greater levels of total ROS (Figure 2–figure supplement 1D, E). Additionally, PPM1D-mutant LCLs were also more sensitive to pharmacologic SOD1 inhibition compared to the WT LCL line, GM12878 (Figure 2–figure supplement 1F). These results demonstrate that PPM1D mutations not only increase ROS in the context of cancer, where cellular metabolism is often altered, but can also alter redox homeostasis in non-transformed cells.
To determine if mutant PPM1D was associated with ROS generation, we treated isogenic OCI-AML2 WT and PPM1D-mutant cells with a PPM1D inhibitor, GSK2830371, for 24 hours. We found that pharmacologic inhibition of PPM1D mildly decreased ROS levels in both WT and PPM1D mutant cells (Figure 2F), suggesting a link between PPM1D and ROS production. Altogether, these data demonstrate that SOD1 inhibition leads to preferential cytotoxicity in the mutant cells due to increased oxidative stress induced by mutant PPM1D.
PPM1D-mutant leukemia cells have altered mitochondrial function
Mitochondria are the primary source of ROS within the cell, as the electron transport chain is a major site of ROS production during oxidative phosphorylation. We next asked whether the observed increase in ROS in PPM1D-mutant cells was due to differences in mitochondrial abundance. We used two independent methods to measure mitochondrial mass, including MitoTracker Green flow cytometry (Figure 3A) and western blot analysis of mitochondrial complex proteins (Figure 3B). However, we did not observe a difference in mitochondrial mass with either method. This finding suggests that mechanisms other than a change in mitochondrial abundance are responsible for the increase in ROS levels in mutant cells, such as alterations in mitochondrial metabolism or changes in ROS scavenging systems.
To assess mitochondrial function, we performed seahorse assays in WT and PPM1D-mutant OCI-AML2 cells. Our seahorse assays revealed that the mutant cells have decreased mitochondrial respiration, as indicated by decreased basal, maximal, and ATP-linked respiration (Figure 3C). While PPM1D-mutant MOLM-13 and OCI-AML3 cells also had decreased basal respiration, there were variable differences in maximal and ATP-linked respiration compared to WT, suggesting possible cell line differences affecting mitochondrial respiration (Figure 3–figure supplement 1A, B). In addition to analyzing respiratory capacity, we also examined mitochondrial membrane potential (MMP) using the fluorescent dye Mitotracker CMXRos, which accumulates in the mitochondria in an MMP-dependent manner. We stained both WT and mutant cells with Mitotracker CMXRos and observed a decrease in MMP in the mutant cells (Figure 3D). Tracking cell numbers between the WT and mutant cell lines over time established this decrease in MMP was not due to altered cellular growth rates (Figure 3–figure supplement 1C). Altogether, these findings, along with decreased respiratory capacity and increased mitochondrial ROS, indicate a mitochondrial defect in PPM1D-mutant cells.
PPM1D-mutant cells have a reduced oxidative stress response
Mitochondrial dysfunction and increased ROS production are closely intertwined. On one hand, mitochondrial dysfunction leads to increased ROS production as a result of impaired oxidative phosphorylation and increased electron leakage (Turrens, 2003). On the other hand, sustained oxidative stress can directly damage mitochondrial components and mtDNA and compromise their function (Wallace, 2005). To better understand the molecular basis for the observed mitochondrial dysfunction and dependency on SOD1, we performed bulk RNA-sequencing (RNA-seq) on Cas9-expressing WT and PPM1D-mutant OCI-AML2 cells transduced with SOD1-sgRNA to induce SOD1 deletion or the empty vector (EV) control (Figure 4–figure supplement 1A). Both EV and SOD1-sgRNA vectors were tagged with a BFP reporter to identify transduced cells. The cells were collected ten-days post-transduction, the timepoint at which we observed 50% reduction of the SOD1-deletion cells during the in vitro proliferation assays, reasoning this would capture the effects of SOD1-deletion on cellular and metabolic processes while avoiding excessive cell death.
Analysis of the RNA-seq data revealed 3530 differentially expressed genes, with 2028 downregulated genes and 1502 upregulated genes in the mutant cells compared to WT cells at baseline (Figure 4–source data 1). Gene set enrichment analysis (GSEA) of the differentially expressed genes showed an upregulation in genes related to cell cycle (GO: 0007049), cell division (GO: 0051301), DNA replication (GO: 005513), and mitophagy (GO: 0000423) in the PPM1D-mutant cells (Figure 4A). Interestingly, there was a significant downregulation of pathways related to the regulation of the oxidative stress response (GO: 1902882, Figure 4– figure supplement 1B), ROS metabolic processes (GO: 0072593), and oxidation reduction (GO: 0055114).
Following SOD1 deletion, the WT cells displayed notable upregulation of pathways associated with cell cycling, chromosome organization, cell division, and DNA repair. In contrast, the mutant cells showed significant downregulation of these same pathways (Figure 4–figure supplement 1C). Intriguingly, upon SOD1 deletion, the mutant cells exhibited an upregulation in the response to oxidative stress (GO:0006979, Figure 4–figure supplement 1D). This finding suggests a reactive transcriptional response to the heightened ROS levels resulting from the loss of SOD1.
As PPM1D is a phosphatase that can directly modulate the activation state of proteins, we examined whether there were alterations in protein and phosphoprotein levels in PPM1D-mutant cells using reverse phase protein array (RPPA) analysis, mirroring the experimental design used for bulk RNA-seq (Figure 4–figure supplement 1A). By focusing on differential protein expression between wild-type (WT) and PPM1D-mutant cells, we aimed to capture the post-translational regulatory events that could contribute to the mitochondrial dysfunction observed in the mutants. The RPPA analysis of over 200 (phospho-)proteins covering major signaling pathways identified 128 differentially expressed proteins between PPM1D-mutant and control WT OCI-AML2 cells (a panel of 264 proteins), with 67 downregulated proteins and 61 upregulated proteins (Figure 4–figure supplement 2A, source data 2). Notably, over-representation analysis showed that among the differentially expressed proteins, there was a significant enrichment in the “Response to Oxidative Stress” pathway in the mutant cells (-log10(pValue) = 24.164) compared to WT, with a particular emphasis on the downregulated proteins of this pathway (-log10(pValue 15.457, Figure 4B, Figure 4–source data 3). While the RNA-seq suggested a transcriptional upregulation of the response to oxidative stress in the mutant cells after SOD1 deletion, the RPPA data revealed that the mutant cells continued to exhibit decreased expression in proteins associated with the oxidative stress response (Figure 4–figure supplement 2B). Taken together, these findings suggest that PPM1D-mutant cells have an inherent impairment in their baseline response to oxidative stress.
To further explore the diminished oxidative stress response in the mutant cells, we assessed their total-and small-molecule-antioxidant capacity. Total antioxidant capacity refers to the overall ability of the cells to counteract free radicals and reduce oxidative damage. This includes enzymatic antioxidants such as catalase, SODs, and peroxidases. Small molecule antioxidant capacity measures the capacity of low molecular weight antioxidants, such as glutathione and vitamin E, to neutralize ROS (Hawash et al., 2022). Our results showed that PPM1D-mutant cells have significantly reduced total and small-molecule antioxidant capacity compared to WT cells (Figure 4C). We also measured the protein levels of key antioxidant enzymes by western blot. While we saw similar protein levels of SOD1 in both WT and mutant cells, we observed a mild reduction in the thioredoxin and catalase levels (Figure 4D). These results provide evidence to support the RNA-seq and RPPA findings that PPM1D-mutant cells have impaired antioxidant defense mechanisms, leading to an elevation in ROS levels and reliance on SOD1 for protection.
PPM1D mutations increase genomic instability and impair non-homologous end-joining repair
In addition to a decreased response to oxidative stress, the RNA-seq GSEA analysis also revealed differential responses to DNA repair. Upon SOD1 deletion, WT cells significantly upregulated the regulation of DNA repair (GO:0006281), double-stranded break repair (GO:0006302), homologous recombination (GO:0035825), and more. However, there was a striking downregulation of DNA repair pathways after deletion of SOD1 in the mutant cells (Figure 4–figure supplement 1C). PPM1D plays a key role in suppressing the DNA damage response (DDR) by dephosphorylating, thereby inactivating, p53 and other key upstream and downstream effectors of the pathway. Truncating mutations and amplifications in PPM1D that lead to increased PPM1D activity may therefore inhibit DNA damage repair and increase genomic instability. Oxidative stress and ROS also pose endogenous challenges to genomic integrity. Therefore, we hypothesized that due to the increase in ROS within the mutant cells, loss of SOD1 may lead to unsustainable accumulation of DNA damage and overwhelm the mutant cell’s DNA repair capacity.
To test this hypothesis, we first sought to establish the baseline levels of DNA damage in PPM1D-altered cells. We performed alkaline comet assays in mouse embryonic fibroblasts and found a significant increase in single- and double-stranded DNA breaks in mutant cells compared to WT (Figure 5A). we also performed metaphase spreads in mouse primary B-cells to investigate chromosomal aberrations, which are consequences of abnormal double-stranded break repair. WT and Ppm1d-mutant mouse primary resting CD43+ B-cells were purified from spleens and stimulated with LPS, IL-4, and CD180 to induce proliferation. The cells were then treated with either low- or high-dose cisplatin for 16-hours. Consistent with our comet assay findings, we observed that Ppm1d-mutant cells harbored approximately two-fold more chromosomal breaks per metaphase after exposure to cisplatin (Figure 5B). When we classified the chromosomal aberrations into subtypes, we observed that the mutant cells had increased numbers of each type of aberration. These results demonstrate that mutations in PPM1D increase genomic instability.
To further assess the DNA repair efficiency of PPM1D-mutant cells, we utilized U2OS DNA repair reporter cell lines which express a green fluorescent protein (GFP) cassette when specific DNA repair pathways are active after stimulation when the I-SceI restriction enzyme is induced to stimulate a double-stranded break (DSB). To test for homologous recombination (HR), tandem defective GFP genes can undergo HR to generate GFP+ cells. Non-homologous end joining (NHEJ) repairs a defective GFP in a distinct cassette (Weinstock et al., 2006). Because the U2OS parental line harbors an endogenous heterozygous PPM1D truncating mutation (R458X), we corrected the lines to generate the isogenic PPM1D WT control (Figure 5–figure supplement 1A).
With two isogenic clones for each reporter cell line, we transfected the PPM1D-WT and - mutant U2OS clones with I-SceI and measured GFP expression by flow cytometry after 48 hours. Our results showed similar levels of repair via HR in both WT and mutant clones (Figure 5C). This finding is consistent with prior studies showing that PPM1D promotes HR by forming a stable complex with BRCA1-BARD1, thereby enhancing their recruitment to DSB sites (Burdova et al., 2019). In contrast, we saw significantly decreased NHEJ repair in the PPM1D-mutant clones (Figure 5D). Prior studies have shown that PPM1D modulates lysine-specific demethylase 1 (LSD1) activity, which is important for facilitating the recruitment of 53BP1 to DNA damage sites through RNF168-dependent ubiquitination (Peng et al., 2015). PPM1D mutations may therefore lead to impairment of NHEJ through dysregulation of 53BP1 recruitment. To confirm this, we performed immunofluorescence imaging of Rad51 and 53BP1 foci. The recruitment of Rad51 and 53BP1 to the sites of DNA damage are important for the activation of HR and NHEJ, respectively. We analyzed mouse embryonic fibroblasts at baseline and after irradiation (10 Gy) and observed similar numbers of Rad51 foci in Ppm1d-mutant and WT cells (Figure 5–figure supplement 1B). In contrast, Ppm1d-mutant MEFs had fewer 53BP1 foci, indicating decreased NHEJ repair capacity that was consistent with our U2OS reporter line findings (Figure 5–figure supplement 1C). Comet assays were performed in parallel with the immunofluorescence experiments to show that the mutant cells had increased DNA damage (Figure 5–figure supplement 1D). Therefore, the decrease in foci was not due to resolution of DNA damage, but rather due to inefficient DNA repair.
Lastly, to investigate the impact of SOD1 inhibition on DNA damage, we treated cells with a SOD1 inhibitor and performed comet assays. Consistent with our earlier observations, PPM1D-mutant cells exhibited a higher burden of DNA breaks. Upon treatment with ATN-224, the mutant cells displayed a significant increase in DNA damage compared to their baseline levels, while the WT cells maintained relatively stable levels of DNA damage (Figure 5E). As ROS are known to contribute to oxidative DNA damage, we further assessed the levels of 8-oxo-2′-deoxyguanosine (8-Oxo-dG), a well-established marker of oxidative DNA damage. Strikingly, the mutant cells demonstrated elevated levels of oxidative DNA damage both at baseline and following SOD1 inhibition with ATN-224, whereas the WT cells once again did not exhibit a substantial increase in oxidative DNA damage upon treatment with ATN-224 (Figure 5F). Collectively, these findings demonstrate that inhibiting SOD1 in PPM1D-mutant cells exacerbates DNA damage, ultimately contributing to death of the PPM1D-mutant cells.
The search for synthetic-lethal strategies for cancer therapy has gained significant attention in recent years due to the potential to identify new therapeutic targets that exploit tumor-specific vulnerabilities. In this study, we utilized whole-genome CRISPR/Cas9 screening to uncover synthetic-lethal vulnerabilities of PPM1D-mutant leukemia cells. Our screen revealed that the top essential gene in PPM1D-mutant cells was SOD1. We investigated the mechanisms underlying this dependency and characterized the redox landscape of PPM1D-mutant cells, which revealed significant oxidative stress and mitochondrial dysfunction. Recent studies have suggested that PPM1D is indirectly associated with energy metabolism via dephosphorylation of the ataxia telangiectasia mutated (ATM) protein. ATM promotes mitochondrial homeostasis, and therefore sustained inactivation of ATM could lead to potential mitochondrial dysfunction (Bar et al., 2023; Guleria and Chandna, 2016). However, oxidative stress and mitochondrial dysfunction are closely related, and it is difficult to dissect the driving factor. We therefore performed RNA-sequencing and RPPA analysis to better understand the underlying processes contributing to the heightened oxidative stress observed in the mutant cells. Our analyses indicated a diminished response to oxidative stress in the mutant cells. These findings may suggest a self-amplifying cycle whereby dysregulation of ROS scavenging systems increases ROS levels, which in turn leads to mitochondrial dysfunction, which further exacerbates oxidative stress. Hence, the additional impairment of ROS detoxification mechanisms within the cell, such as the loss of SOD1, has detrimental consequences for the viability of mutant cells.
Multiple mechanisms may underlie the suppressed oxidative stress response observed in PPM1D-mutant cells. One possible explanation is through PPM1D-mediated inhibition of p53. p53 exhibits complex and context-dependent roles in cellular responses to oxidative stress, and its functions can vary depending on the severity of stress encountered by the cell (Kang et al., 2013; Liang et al., 2013; Sablina et al., 2005). Under mild or moderate oxidative stress conditions, p53 may protect the cell from ROS by inducing the transcription of genes such as superoxide dismutase (SOD), glutathione peroxidase (GPx), and others (Dhar et al., 2011; Peuget et al., 2014; Sablina et al., 2005; Tan et al., 1999). However, under severe or prolonged oxidative stress, the pro-apoptotic functions of p53 may promote ROS production to eliminate cells that have accumulated excessive DNA damage or irreparable cellular alterations. The duality of these anti- and pro-oxidant functions of p53 highlight its intricate role in modulating responses to oxidative stress. How PPM1D affects the switch between these functions of p53 is not understood. Furthermore, the extent to which the dependency on SOD1 observed in PPM1D-mutant cells is mediated through p53 remains unclear and requires deeper exploration to better understand the context in which SOD1 inhibitors can be used in cancer therapy.
Oxidative stress and DNA damage are intimately linked processes that frequently co-occur. Our study also investigated the interplay between PPM1D, DNA damage, and oxidative stress. We demonstrated significant genomic instability of PPM1D-mutant cells at baseline and further characterized the effects of mutant PPM1D on specific DNA repair pathways. While previous studies have suggested a role for PPM1D in modulating HR and NHEJ (Burdova et al., 2019; Peng et al., 2015), our study is the first to demonstrate impaired NHEJ in PPM1D-mutant cells. Additionally, our study corroborated previous research demonstrating the synthetic-lethal relationship of SOD1 and other DNA damage genes such as RAD54B, BLM, and CHEK2 (Sajesh et al., 2013; Sajesh and McManus, 2015). In the context of PPM1D mutations, we showed that SOD1 inhibition amplifies oxidative stress and DNA damage, ultimately culminating in mutant cell death. These findings align with previous studies demonstrating that the inhibition of SOD1 by LCS-1 leads to the degradation of key DNA repair proteins, PARP1 and BRCA1, resulting in ROS-dependent death of glioma cells (Ling et al., 2022). Similar observations were made in colorectal cancer models, where LCS-1 treatment enhanced the effectiveness of PARP1 inhibition. These results highlight the potential of combinatorial therapies to achieve therapeutic synergism and underscores the intricate relationship between ROS and DNA damage.
Interestingly, our screen also uncovered sensitivity of PPM1D-mutant cells to dropout of genes in the Fanconi Anemia (FA) DNA repair pathway including BRIP1 (FANCJ), FANCI, FANCA, SLX4 (FANCP), UBE2T (FANCT), and C19orf40 (FAAP24). The FA pathway plays a crucial role in facilitating the repair of interstrand crosslinks by coordinating the recruitment of various DNA repair pathways, such nucleotide excision repair, translesion synthesis, and homologous recombination (Ceccaldi et al., 2016; Kottemann and Smogorzewska, 2013). Furthermore, the FA genes are also essential for safeguarding genomic integrity during replication stress by protecting replication forks (Schlacher et al., 2012). Previous studies have highlighted the essentiality of the FA pathway in p53-deficient cells (Ardeljan et al., 2020). Another study also showed that several FA pathway genes are shared synthetic-lethal partners in cells with various loss of tumor suppressors including PBRM1, LKB1, TP53, TP53BP1, and CHD1 (Feng et al., 2022). In the context of cancer, cells with defects in DNA repair mechanisms heavily rely on compensatory pathways for survival. A prominent example is the successful treatment of BRCA1/2-deficient breast cancers with PARP1 inhibitors (50, 51). Given the observed suppression of p53 and increased genomic instability, it is reasonable that the FA pathway is also a dependency in PPM1D-mutant cells. However, the lack of small molecule inhibitors specifically targeting the FA pathway hinders our ability to directly assess its therapeutic potential in targeting PPM1D-mutant cells in our study.
In summary, our investigation sheds light on the role of mutant PPM1D in modulating cellular responses to oxidative stress and DNA repair in leukemia cells, offering valuable insights into the underlying molecular mechanisms. This research not only enhances our understanding of PPM1D-mediated cellular responses, but also identifies potential therapeutic targets against PPM1D-mutant leukemia cells. However, it is important to acknowledge the limitations of our study. We recognize that while PPM1D mutations are frequently observed in patients with t-MN, they are rare in de novo AML (Hsu et al., 2018). While there is ample evidence that PPM1D is an oncogenic driver in many types of cancers (Ali et al., 2012; Khadka et al., 2022; Li et al., 2002; Nguyen et al., 2010; Wu et al., 2016), the clinical importance of targeting pre-malignant PPM1D-associated clonal expansion in the hematopoietic system is not clear. However, the prevalence of PPM1D somatic mutations in other tissues such as the esophagus, suggests the need for further investigation (Yokoyama et al., 2019).
Materials and Methods
Cell lines and Reagents
Cas9-expressing OCI-AML2 cells were generated by lentiviral transduction using pKLV2-EF1aBsd2ACas9-W plasmid obtained from Dr. Kosuke Yusa from the Sanger Institute (Addgene #67978). Four days post-transduction, cells underwent blasticidin selection. Single clones were obtained by fluorescence-activated cell sorting and functionally tested for Cas9 activity using a lentiviral reporter pKLV2-U6gRNA5(gGFP)-PGKBFP2AGFP-W (Addgene #67980). PPM1D-mutant cell lines were generated using the RNP-based CRISPR/Cas9 delivery method using a single sgRNA (GCTAAAGCCCTGACTTTA). Two PPM1D-mutant clones were validated and selected for the CRISPR dropout screen.
CRISPR Dropout Screen and Analyses
For large-scale production of lentivirus, 15 cm plates of 80-90% confluent 293T cells were transfected using Lipofectamine 2000 (Invitrogen) with 7.5 ug of the Human Improved Whole-Genome Knockout CRISPR library V1 (by Kosuke Yuya, Addgene #67989), 18.5 ug of psPax2, and 4 ug of pMD2.G. For the CRISPR dropout screen, one WT and two independent PPM1D mutant Cas9-expressing OCI-AML2 cell lines were used as biological replicates, with three technical replicates per line. 3 x 107 cells were transduced with the lentivirus library supernatant. Three days post-transduction, the cells were selected with puromycin for three days. Cells were collected on day 28 for genomic DNA isolation using isopropanol precipitation. Illumina adapters and barcodes were added to samples by PCR as previously described (Tzelepis et al., 2016). Single-end sequencing was performed on the HiSeq 2000 V4 platform and cell-essential genes were identified using the MaGECK-VISPR (Li et al., 2014).
Competitive Proliferation Assay
Gene-specific sgRNAs were cloned into the pKLV2-U6gRNA5(BbsI)-PGKpuro2ABFP (Addgene #67974) lentiviral backbone. 293T cells (0.4 x 106 cells per well) were seeded in a six-well plate the day prior and transfected using lipofectamine 3000 with pMD2G (0.8 ug), pAX2 (1.6 ug), and the sgRNA-BFP (1.6 ug) plasmids. Cas9-expressing cells were then seeded in 12-well plates (200k cells per well, in triplicates) in media supplemented with 8ug/ml polybrene and 5 ug/mL blasticidin, and lentivirally transduced at a titer that yields 50% infection efficiency. Cells were assayed using flow cytometry for BFP expression between 4- and 16-days post-transduction and normalized to the BFP percentage at day 4.
Drug and Proliferation Assays
Drug and proliferation assays were done using the Cell Proliferation MTT Kit (Sigma) as per manufacturer’s protocol. Briefly, 1 x 104 cells were plated in 96-well, flat bottom plates and treated with vehicle or drugs in a total volume of 100 uL. Plates were incubated at 37°C for at least 24 hours. 10 uL of MTT labeling reagent was added to each well and incubated for 4 hours. 100 uL of solubilization buffer was added to each well and incubated overnight. Plates were analyzed using a fluorometric microplate reader at 550 nm.
Alkaline Comet Assay
Comet assays were conducted as previously described (Greve et al., 2012; Schmezer et al., 2001). Cells were resuspended to 1 x 105 cells/mL and mixed with 1% low-melting agarose (R&D Systems) at a 1:10 ratio and plated on 2-well comet slides (R&D Systems). Cells were then lysed overnight and immersed in alkaline unwinding solution as per manufacturer’s protocol (Trevigen). Fluorescence microscopy was performed at 10X magnification using the Keyence BZ-X800 microscope and analyses of comet tails were performed using the Comet Assay IV software (Instem). At least 150 comet tails were measured per sample.
Chromosome aberration analysis mitotic chromosome spreads
Primary resting mouse splenic B-cells were isolated using anti-CD43 microbeads (Miltenyi Biotec) and activated with 25 ug/mL LPS (Sigma), 5 ng/mL IL-4 (Sigma), and 0.5 ug/mL anti-CD180 (BD Pharmingen) for 30 hours. The cells were then treated with cisplatin for 16 hours at two concentrations - 0.5 uM and 5 uM cisplatin. Metaphases were prepared as previously described (Zong et al., 2019). Briefly, cells were arrested at mitosis with colcemid (0.1 ug/mL, ThermoFisher) for 1 hour. Cells were then incubated in a prewarmed, hypotonic solution of potassium chloride (75 mM) for 20 minutes to induce swelling and fixed in methanol/glacial acetic acid (3:1). Droplets were spread onto glass slides inside a cytogenetic drying chamber. Fluorescence in situ hybridization was performed using a Cy3-labeled peptide nucleic acid probe to stain telomeres and DNA was counterstained by DAPI. At least 50 metaphases were scored for chromosome aberrations for each experimental group.
To measure superoxide, total cellular reactive oxygen species (ROS), and lipid peroxidation, 1 x 106 cells were collected after the indicated treatments and washed with PBS. The cells were stained with 1 uM Mitosox Green (Thermofisher), 20 uM 2’7’-dichlorofluorescin diacetate (DCFDA, Abcam), or 2.5 uM BODIPY 581/591 (Thermofisher) in FBS-free Hanks’ buffered saline solution (HBSS, Thermofisher), and incubated at 37°C for 30 minutes. The staining was quenched with complete HBSS (2% FBS, 1% HEPES) and washed twice before being resuspended in DAPI-containing PBS to assess ROS in viable cells. The data was acquired using an LSRII (BD Biosciences) and analyzed on Flowjo.
Reverse-Phase Protein Array
Reverse phase protein array assays for antibodies to proteins or phosphorylated proteins in different functional pathways were carried out as described previously (Coarfa et al., 2021; Lu H.-Y., 2021; Wang et al., 2022). Specifically, protein lysates were prepared from cultured cells with modified Tissue Protein Extraction Reagent (TPER) (Life Technologies Corporation, Carlsbad, CA) and a cocktail of protease and phosphatase inhibitors (Roche, Pleasanton, CA) (Lu et. al. 2021). The lysates were diluted into 0.5 mg/ml in SDS sample buffer and denatured on the same day. The Quanterix 2470 Arrayer (Quanterix, Billerica, MA) with a 40 pin (185 µm) configuration was used to spot samples and control lysates onto nitrocellulose-coated slides (Grace Bio-labs, Bend, OR) using an array format of 960 lysates/slide (2880 spots/slide). The slides were processed as described and probed with a set of 264 antibodies against total proteins and phosphoproteins using an automated slide stainer Autolink 48 (Dako, Santa Clara, CA). Each slide was incubated with one specific primary antibody and a negative control slide was incubated with antibody diluent without any primary antibody. Primary antibody binding was detected using a biotinylated secondary antibody followed by streptavidin-conjugated IRDye680 fluorophore (LI-COR Biosciences, Lincoln, NE). Total protein content of each spotted lysate was assessed by fluorescent staining with Sypro Ruby Protein Blot Stain according to the manufacturer’s instructions (Molecular Probes, Eugene, OR).
Fluorescence-labeled slides were scanned on a GenePix 4400 AL scanner, along with accompanying negative control slides, at an appropriate PMT to obtain optimal signal for this specific set of samples. The images were analyzed with GenePix Pro 7.0 (Molecular Devices, Silicon Valley, CA). Total fluorescence signal intensities of each spot were obtained after subtraction of the local background signal for each slide and were then normalized for variation in total protein, background and non-specific labeling using a group-based normalization method as described (Lu H.-Y., 2021). For each spot on the array, the-background-subtracted foreground signal intensity was subtracted by the corresponding signal intensity of the negative control slide (omission of primary antibody) and then normalized to the corresponding signal intensity of total protein for that spot. Each image, along with its normalized data, was evaluated for quality through manual inspection and control samples. Antibody slides that failed the quality inspection were either repeated at the end of the staining runs or removed before data reporting. A total of 261 antibodies remained in the list. Multiple t-tests with Benajimini Hochberg correction were performed for statistical analysis and filtering was based on an FDR <0.2 and linear fold change of >1.25.
Bulk RNA-sequencing was performed on WT and PPM1D-mutant OCI-AML2 cells after lentiviral SOD1 CRISPR knockout. Cells were transduced with pKLV2-U6-sgRNA-BFP lentivirus (either empty vector or with SOD1-sgRNA). Transduced cells were then cultured for 10 days and BFP+ cells were sorted directly into Buffer RLT Plus with ß-mercaptoethanol. RNA was isolated using the Allprep DNA/RNA Micro Kit (Qiagen) per manufacturer’s protocols. RNA-sequencing library preparation was done using the True-Seq Stranded mRNA kit (Illumina) per manufacturer’s protocol. Quality control of libraries was performed using a TapeStation D1000 ScreenTape (Agilent, 5067-5584). Libraries were then sequenced using an Illumina Nextseq 2000 sequencer, aiming for >20 million reads per biological replicate. Paired-end RNA-sequencing reads were obtained. obtained and trimmed using trimGalore (https://github.com/FelixKrueger/TrimGalore). Mapping was performed using the STAR package (Dobin et al., 2013) against the human genome build UCSC hg38 and counts were quantified with featureCounts (Liao et al., 2014). Differential expression analysis was performed using the DESeq2 R package (1.28.1) (Love et al., 2014). P-values were adjusted with Benjamini and Hochberg’s approach for controlling the false discovery rate (FDR). Significant differentially expressed genes between the indicated comparisons were filtered based on an FDR< 0.05 and absolute fold change exceeding 1.5.
Mitochondrial bioenergetics in AML cell lines were performed using the Seahorse XFp Cell Mito Stress Kit (Agilent Technologies) on the Seahorse XFe96 Analyzer. Cells were resuspended in XF RPMI base media supplemented with 1 mM pyruvate, 2 mM L-glutamine, 10 mM glucose. 1 x 105 cells/well were seeded in poly-D-lysine (Thermofisher) coated XFe96 plates. The plate was incubated in a non-CO2 incubator at 37°C for 1 hour to equilibrate. OCR and ECAR measurements were taken at baseline and every 8 minutes after sequential addition of oligomycin (2 uM), FCCP (0.5 uM), and rotenone/ antimycin A (0.75 uM). All measurements were normalized to the number of viable cells.
Generation of PPM1D WT U2OS cells using CRISPR editing
U2OS cells containing the DR-GFP (for homologous recombination) or EJ5-GFP (for non-homologous end-joining) DNA repair reporter cassettes were kindly provided by the Bertuch Lab at Baylor College of Medicine. To establish PPM1D-WT isogenic lines, knock-in CRISPR editing was performed with a single-stranded oligodeoxynucleotide (ssODN) template: TGCCCTGGTTC GTAGCAATGCCTTCTCAGAGAATTTTCTAGAGGTTTCAGCTGAGATAGCTCGTGAGAATGT ACAAGGTGTAGTCATACCCTAAAAGATCCAGAACCACTTGAAGAAAATGCGCTAAAGCCCT GACTTTAAGGATACA. The PPM1D sgRNA sequence used was: ATAGCTCGAGA GAATGTCCA. 1.3 ug of Cas9 (IDT) was incubated with 1 ug of sgRNA for 15 minutes at room temperature. 1 ug of the ssODN template was then added to the Cas9-sgRNA complexes and mixed with 20,000 U2OS cells and resuspended in 10 uL of Buffer R, immediately prior to electroporation. The neon electroporation system was used with the following conditions: 1400v, 15 ms, 4 pulses. Single cell-derived clones were genotyped by Sanger sequencing and PPM1D protein expression was validated by western blot.
GFP reporter-based DNA repair assays
For the DNA repair reporter assay, 100,000 U2OS cells were seeded in a 12-well plate in antibiotic-free Dulbecco’s Modified Eagle Medium (DMEM, Thermofisher) supplemented with 10% FBS. Cells were transfected with 3.6 uL of Lipofectamine 2000 (Invitrogen) in 200 uL of OptiMEM with 0.8 ug of the I-SceI expression plasmid (pCBASce, Addgene #60960). The media was replaced the next morning and the cells were trypsinized 48-hours post-transfection for analysis of GFP expression by flow cytometry (BD Biosciences).
12 mm glass coverslips were coated with 50 ug/mL poly-D-lysine (Thermofisher) for 30 minutes at room temperature and washed with sterile PBS. 0.5 x 106 suspension cells/well were seeded on coverslips and incubated for 1 hour at 37°C to allow for adherence. Samples were then fixed with 4% paraformaldehyde for 10 minutes at 37°C and washed three times with 0.01% Triton-X PBS (PBS-T). Fixed cells were permeabilized with 0.5% PBS-T for 20 minutes, washed three times, and incubated with 5% goat serum (Thermofisher) for 1 hour at room temperature. Afterwards, samples were incubated overnight at 4°C with the following primary antibodies: rabbit anti-Rad51 (Cell Signaling #8875S 1:100) or rabbit anti-53BP1 (ThermoFisher #PA1-16565, 1:500). The following day, samples were washed and incubated at room temperature for 1 hour with Alexafluor 488-conjugated goat anti-rabbit IgG (#111-545-144, Jackson ImmunoResearch, 1:500). After secondary antibody incubation, the coverslips were washed three times with PBS and mounted with fluoromount-G mounting medium with DAPI (Thermofisher) on glass microscope slides and sealed with nail polish. Imaging was done on the Keyence BZ-X800 microscope and foci analysis was performed using CellProfiler.
Cells were lysed with 1x RIPA buffer supplemented with Halt Protease and Phosphatase inhibitor cocktail (Thermofischer) for 1 hour at 4°C. Protein concentration was quantified using the Pierce BCA protein assay kit (Thermofischer) and boiled at 95°C in 1x Laemmli (Biorad) for 7 minutes. The samples in which mitochondrial proteins were probed were not boiled, as boiling can cause signal reduction. Instead, samples were warmed to 37°C for 30 minutes prior to loading. The proteins were separated by SDS-PAGE on 4-15% gradient gels (Biorad) and transferred on to PVDF membranes using the iBlot Dry Blotting system (Thermofisher). Membranes were incubated for 1 hour at room temperature in 5% milk in Tris-buffered saline solution with Tween-20 (TBST). After washing, the membranes were incubated overnight at 4°C with the following primary antibodies: mouse anti-PPM1D (F-10, Santa Cruz, 1:1000), mouse anti-GAPDH (MAB374, Millipore, 1:200), mouse total OXPHOS Human antibody cocktail (ab110411, Abcam, 1:1000), mouse anti-Vinculin (V9131, Sigma Aldrich, 1:2000). The following day, membranes were washed twice with TBST and incubated for 1 hour with HRP-linked anti-rabbit IgG or anti-mouse IgG (Cell Signaling, 1:5000 – 1:10,000) at room temperature. Blots were imaged on the Bio-Rad ChemiDoc platform.
Statistical analysis incorporated in the MaGECK-VISPR algorithm includes p-value and FDR calculations. GraphPad Prism 6.0 was used for other statistical analyses. The sample size (n) specified in the Figure Legends was used for statistical analysis and denotes the number of independent biological replicates. The main conclusions were supported by data obtained from at least two biological replicates. The graphs presented in the figures are shown with error bars indicating either mean ± SEM or mean ± SD, as mentioned in the Figure Legends. Two-tailed t-tests were performed to calculate statistics, assuming unequal standard deviations, unless mentioned otherwise. Significance levels are indicated in the figures and were determined using GraphPad PRISM. Results were considered statistically significant at *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.
This work was supported by R01CA237291 and P01CA265748. This work was also supported by the NCI Cancer Center Support Grant P30CA125123 which partly supports the Cytometry Core, the Proteomics & Metabolomics Core, and the Antibody-based Proteomics Core. Support for the cores was also provided by the Cancer Prevention and Research Institute of Texas (CPRIT) from grants: RP180672, RR024574, and RP210227 and NIH S10OD028648. LZ was supported by the Baylor Research Advocates for Student Scientists (BRASS) Foundation, and the Janice McNair Medical Foundation.
The authors declare they have no competing interests.
Data and materials availability
All data are available in the main text or in the supplementary materials. All raw and processed data generated in this work will be publicly available at GEO data repositor pending scientific review. The human cell lines generated by the Goodell laboratory for this study are available upon request and will require a standard Materials Transfer Agreement (MTA). Any additional information required to analyze the data reported in this paper is available from the lead contact upon request.
- 1.The oncogenic phosphatase PPM1D confers cisplatin resistance in ovarian carcinoma cells by attenuating checkpoint kinase 1 and p53 activationOncogene 31:2175–2186
- 2.Cell fitness screens reveal a conflict between LINE-1 retrotransposition and DNA replicationNat Struct Mol Biol 27:168–178
- 3.Whole exome/genome sequencing in cyclic vomiting syndrome reveals multiple candidate genes, suggesting a model of elevated intracellular cations and mitochondrial dysfunctionFront Neurol 14
- 4.Cancer therapy shapes the fitness landscape of clonal hematopoiesisNat Genet 52:1219–1226
- 5.Amplification of PPM1D in human tumors abrogates p53 tumor-suppressor activityNat Genet 31:210–215
- 6.WIP1 Promotes Homologous Recombination and Modulates Sensitivity to PARP InhibitorsCells 8
- 7.The Fanconi anaemia pathway: new players and new functionsNat Rev Mol Cell Biol 17:337–349
- 8.Wip1 directly dephosphorylates gamma-H2AX and attenuates the DNA damage responseCancer Res 70:4112–4122
- 9.The DNA damage response: making it safe to play with knivesMol Cell 40:179–204
- 10.Reverse-Phase Protein Array: Technology, Application, Data Processing, and IntegrationJ Biomol Tech
- 11.Manganese superoxide dismutase is a p53-regulated gene that switches cancers between early and advanced stagesCancer Res 71:6684–6695
- 12.STAR: ultrafast universal RNA-seq alignerBioinformatics 29:15–21
- 13.Genome-wide CRISPR screens using isogenic cells reveal vulnerabilities conferred by loss of tumor suppressorsSci Adv 8
- 14.Wip1, a novel human protein phosphatase that is induced in response to ionizing radiation in a p53-dependent mannerProc Natl Acad Sci U S A 94:6048–6053
- 15.Proto-oncogene Wip1, a member of a new family of proliferative genes in NSCLC and its clinical significanceTumour Biol 35:2975–2981
- 16.Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequenceN Engl J Med 371:2477–2487
- 17.Evaluation of different biomarkers to predict individual radiosensitivity in an inter-laboratory comparison--lessons for future studiesPLoS One 7
- 18.ATM kinase: Much more than a DNA damage responsive proteinDNA Repair (Amst 39:1–20
- 19.Hallmarks of cancer: the next generationCell 144:646–674
- 20.The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin-dependent kinasesCell 75:805–816
- 21.In vitro and in vivo assessment of the antioxidant potential of isoxazole derivativesSci Rep 12
- 22.PPM1D accelerates proliferation and metastasis of osteosarcoma by activating PKP2Eur Rev Med Pharmacol Sci 25:78–85
- 23.DNA damage, aging, and cancerN Engl J Med 361:1475–1485
- 24.PPM1D Mutations Drive Clonal Hematopoiesis in Response to Cytotoxic ChemotherapyCell Stem Cell 23:700–713
- 25.Age-related clonal hematopoiesis associated with adverse outcomesN Engl J Med 371:2488–2498
- 26.De Novo Truncating Mutations in the Last and Penultimate Exons of PPM1D Cause an Intellectual Disability SyndromeAm J Hum Genet 100:650–658
- 27.PPM1D as a novel biomarker for prostate cancer after radical prostatectomyAnticancer Res 34:2919–2925
- 28.Copper binding by tetrathiomolybdate attenuates angiogenesis and tumor cell proliferation through the inhibition of superoxide dismutase 1Clin Cancer Res 12:4974–4982
- 29.PPM1D-truncating mutations confer resistance to chemotherapy and sensitivity to PPM1D inhibition in hematopoietic cellsBlood 132:1095–1105
- 30.The critical role of catalase in prooxidant and antioxidant function of p53Cell Death Differ 20:117–129
- 31.PPM1D mutations are oncogenic drivers of de novo diffuse midline glioma formationNat Commun 13
- 32.Fanconi anaemia and the repair of Watson and Crick DNA crosslinksNature 493:356–363
- 33.PPM1D gene amplification and overexpression in breast cancer: a qRT-PCR and chromogenic in situ hybridization studyMod Pathol 23:1334–1345
- 34.PPM1D Knockdown Suppresses Cell Proliferation, Promotes Cell Apoptosis, and Activates p38 MAPK/p53 Signaling Pathway in Acute Myeloid LeukemiaTechnol Cancer Res Treat 19
- 35.Oncogenic properties of PPM1D located within a breast cancer amplification epicenter at 17q23Nat Genet 31:133–134
- 36.PPM1D Functions as Oncogene and is Associated with Poor Prognosis in Esophageal Squamous Cell CarcinomaPathol Oncol Res 26:387–395
- 37.MAGeCK enables robust identification of essential genes from genome-scale CRISPR/Cas9 knockout screensGenome Biol 15
- 38.The regulation of cellular metabolism by tumor suppressor p53Cell Biosci 3
- 39.featureCounts: an efficient general purpose program for assigning sequence reads to genomic featuresBioinformatics 30:923–930
- 40.A non-comparative randomized phase II study of 2 doses of ATN-224, a copper/zinc superoxide dismutase inhibitor, in patients with biochemically recurrent hormone-naive prostate cancerUrol Oncol 31:581–588
- 41.Prognostic Mutations in Myelodysplastic Syndrome after Stem-Cell TransplantationN Engl J Med 376:536–547
- 42.LCS-1 inhibition of superoxide dismutase 1 induces ROS-dependent death of glioma cells and degradates PARP and BRCA1Front Oncol 12
- 43.Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2Genome Biol 15
- 44.Phase I study of copper-binding agent ATN-224 in patients with advanced solid tumorsClin Cancer Res 14:7526–7534
- 45.High-Throughput Evaluation of Metabolic Activities Using Reverse Phase Protein Array TechnologyNeuron Signaling in Metabolic Regulation
- 46.PPM1D dephosphorylates Chk1 and p53 and abrogates cell cycle checkpointsGenes Dev 19:1162–1174
- 47.Myc and Ras oncogenes engage different energy metabolism programs and evoke distinct patterns of oxidative and DNA replication stressMol Oncol 9:601–616
- 48.Allosteric inhibition of PPM1D serine/threonine phosphatase via an altered conformational stateNat Commun 13
- 49.The oncogenic phosphatase WIP1 negatively regulates nucleotide excision repairDNA Repair (Amst 9:813–823
- 50.The Wip1 phosphatase (PPM1D) antagonizes activation of the Chk2 tumour suppressor kinaseOncogene 26:1449–1458
- 51.Novel signaling axis for ROS generation during K-Ras-induced cellular transformationCell Death Differ 21:1185–1197
- 52.Modulation of LSD1 phosphorylation by CK2/WIP1 regulates RNF168-dependent 53BP1 recruitment in response to DNA damageNucleic Acids Res 43:5936–5947
- 53.PPM1D is a prognostic marker and therapeutic target in colorectal cancerExp Ther Med 8:430–434
- 54.Oxidative stress-induced p53 activity is enhanced by a redox-sensitive TP53INP1 SUMOylationCell Death Differ 21:1107–1118
- 55.Mosaic PPM1D mutations are associated with predisposition to breast and ovarian cancerNature 493:406–410
- 56.The antioxidant function of the p53 tumor suppressorNat Med 11:1306–1313
- 57.Synthetic lethal targeting of superoxide dismutase 1 selectively kills RAD54B-deficient colorectal cancer cellsGenetics 195:757–767
- 58.Targeting SOD1 induces synthetic lethal killing in BLM- and CHEK2-deficient colorectal cancer cellsOncotarget 6:27907–27922
- 59.A distinct replication fork protection pathway connects Fanconi anemia tumor suppressors to RAD51-BRCA1/2Cancer Cell 22:106–116
- 60.Rapid screening assay for mutagen sensitivity and DNA repair capacity in human peripheral blood lymphocytesMutagenesis 16:25–30
- 61.Superoxide dismutase 1 (SOD1) is a target for a small molecule identified in a screen for inhibitors of the growth of lung adenocarcinoma cell linesProc Natl Acad Sci U S A 108:16375–16380
- 62.p53-inducible wip1 phosphatase mediates a negative feedback regulation of p38 MAPK-p53 signaling in response to UV radiationEMBO J 19:6517–6526
- 63.Transcriptional activation of the human glutathione peroxidase promoter by p53J Biol Chem 274:12061–12066
- 64.Systematic characterization of mutations altering protein degradation in human cancersMol Cell 81:1292–1308
- 65.Mitochondrial formation of reactive oxygen speciesJournal of Physiology 552:335–344
- 66.A CRISPR Dropout Screen Identifies Genetic Vulnerabilities and Therapeutic Targets in Acute Myeloid LeukemiaCell Rep 17:1193–1205
- 67.A mitochondrial paradigm of metabolic and degenerative diseases, aging, and cancer: a dawn for evolutionary medicineAnnu Rev Genet 39:359–407
- 68.PPM1D silencing by lentiviral-mediated RNA interference inhibits proliferation and invasion of human glioma cellsJ Huazhong Univ Sci Technolog Med Sci 31:94–99
- 69.High-throughput profiling of histone post-translational modifications and chromatin modifying proteins by reverse phase protein arrayJ Proteomics 262
- 70.Assaying double-strand break repair pathway choice in mammalian cells using a targeted endonuclease or the RAG recombinaseMethods Enzymol 409:524–540
- 71.Jansen-de Vries syndrome: Expansion of the PPM1D clinical and phenotypic spectrum in 34 familiesAm J Med Genet A
- 72.PPM1D exerts its oncogenic properties in human pancreatic cancer through multiple mechanismsApoptosis 21:365–378
- 73.Knockdown of protein phosphatase magnesium-dependent 1 (PPM1D) through lentivirus-mediated RNA silencing inhibits colorectal carcinoma cell proliferationTechnol Cancer Res Treat 12:537–543
- 74.Age-related remodelling of oesophageal epithelia by mutated cancer driversNature 565:312–317
- 75.Wild-type p53 induces apoptosis of myeloid leukaemic cells that is inhibited by interleukin-6Nature 352:345–347
- 76.Exome sequencing identifies somatic gain-of-function PPM1D mutations in brainstem gliomasNat Genet 46:726–730
- 77.BRCA1 Haploinsufficiency Is Masked by RNF168-Mediated Chromatin UbiquitylationMol Cell 73:1267–1281