Histone methyltransferases EHMT1 and EHMT2 (GLP/G9A) maintain PARP inhibitor resistance in high-grade serous ovarian carcinoma
- 325 Downloads
Euchromatic histone-lysine-N-methyltransferases 1 and 2 (EHMT1/2, aka GLP/G9A) catalyze dimethylation of histone H3 lysine 9 (H3K9me2) and have roles in epigenetic silencing of gene expression. EHMT1/2 also have direct roles in DNA repair and are implicated in chemoresistance in several cancers. Resistance to chemotherapy and PARP inhibitors (PARPi) is a major cause of mortality in high-grade serous ovarian carcinoma (HGSOC), but the contribution of the epigenetic landscape is unknown.
To identify epigenetic mechanisms of PARPi resistance in HGSOC, we utilized unbiased exploratory techniques, including RNA-Seq and mass spectrometry profiling of histone modifications. Compared to sensitive cells, PARPi-resistant HGSOC cells display a global increase of H3K9me2 accompanied by overexpression of EHMT1/2. EHMT1/2 overexpression was also observed in a PARPi-resistant in vivo patient-derived xenograft (PDX) model. Genetic or pharmacologic disruption of EHMT1/2 sensitizes HGSOC cells to PARPi. Cell death assays demonstrate that EHMT1/2 disruption does not increase PARPi-induced apoptosis. Functional DNA repair assays show that disruption of EHMT1/2 ablates homologous recombination (HR) and non-homologous end joining (NHEJ), while immunofluorescent staining of phosphorylated histone H2AX shows large increases in DNA damage. Propidium iodide staining and flow cytometry analysis of cell cycle show that PARPi treatment increases the proportion of PARPi-resistant cells in S and G2 phases, while cells treated with an EHMT1/2 inhibitor remain in G1. Co-treatment with PARPi and EHMT1/2 inhibitor produces an intermediate phenotype. Immunoblot of cell cycle regulators shows that combined EHMT1/2 and PARP inhibition reduces expression of specific cyclins and phosphorylation of mitotic markers. These data suggest DNA damage and altered cell cycle regulation as mechanisms of sensitization. RNA-Seq of PARPi-resistant cells treated with EHMT1/2 inhibitor showed significant gene expression changes enriched in pro-survival pathways that remain unexplored in the context of PARPi resistance, including PI3K, AKT, and mTOR.
This study demonstrates that disrupting EHMT1/2 sensitizes HGSOC cells to PARPi, and suggests a potential mechanism through DNA damage and cell cycle dysregulation. RNA-Seq identifies several unexplored pathways that may alter PARPi resistance. Further study of EHMT1/2 and regulated genes will facilitate development of novel therapeutic strategies to successfully treat HGSOC.
KeywordsHGSOC Ovarian cancer PARP inhibitor Resistance H3K9me2 EHMT1 EHMT2 DNA repair Cell cycle
Euchromatic histone-lysine-N-methyltransferases 1 and 2
Dimethylated histone H3 lysine 9
High-grade serous ovarian carcinoma
Non-homologous end joining
Poly ADP ribose polymerase
Poly ADP ribose polymerase inhibitor
High-grade serous ovarian carcinoma (HGSOC) is the most common epithelial ovarian cancer histotype and has one of the highest death-to-incidence ratios of all cancers . High mortality is due primarily to frequent late stage diagnosis, high recurrence rates, and the development of therapy resistance. Over 80% of cases recur and ultimately present as chemoresistant disease.
Poly ADP ribose polymerase inhibitors (PARPi) were initially developed to treat homologous recombination (HR) DNA repair-deficient tumors (e.g., BRCA1/2-mutated). Mutation of BRCA1/2 often leads to a defective HR repair pathway and significantly increases the risk of developing HGSOC . Analysis of HGSOC in The Cancer Genome Atlas (TCGA) predicts that close to 50% of HGSOC have some deficiencies in the HR pathway. The three approved PARPi (olaparib, rucaparib, and niraparib) were initially used for recurrent HGSOC with BRCA1/2 mutations as a third- or fourth-line therapy, but the SOLO1 clinical trial showed that first-line maintenance olaparib reduced risk of disease progression or death by 70% in newly diagnosed cases . Further trials showed that BRCA-wildtype tumors also significantly benefit from PARPi [4, 5]. In the future, nearly all patients with HGSOC could receive olaparib. Unfortunately, acquired resistance is an emerging clinical problem that limits PARPi efficacy. Known mechanisms of PARPi resistance, such as BRCA1/2 reversion mutations, restore HR but are found in only a small proportion of resistant cancers [6, 7, 8, 9], suggesting that PARPi resistance has other causes that have yet to be explored .
Epigenetic regulation of transcriptional programming has been associated with chemo- and targeted-therapy resistance [11, 12]. Euchromatic histone-lysine N-methyltransferase 1 and 2 (EHMT1/2, also known as GLP and G9A, respectively) are epigenetic enzymes that promote transcriptional repression through histone modifications [13, 14, 15]. Functionally, EHMT1 and EHMT2 heterodimerize and interact with multi-zinc finger protein ZNF644 to form a complex which catalyzes s-adenosylmethionine (SAM) molecules to H3K9 to form mono- and di-methylated K9 (H3K9me1/H3K9me2). EHMT1/2 both contain acidic regions, ankryin protein-protein interaction domains, and share 80% homology in their methyltransferase SET-domains [13, 16, 17]. Beyond canonical roles in transcriptional repression, the EHMT1/2 complex also directly promotes DNA damage repair via recruitment of BRCA1, 53BP1, and other factors involved in HR and non-homologous end joining (NHEJ) [18, 19]. In the context of ovarian cancer, EHMT2 is frequently amplified and overexpressed [20, 21] and high expression correlates with aggressive peritoneal metastasis and poorer overall survival . Prior to this study, the roles of EHMT1/2 in chemotherapy and PARPi resistance have not been examined in HGSOC.
In this report, we employed an unbiased proteomic approach to profile the epigenetic landscape of PARPi-resistant HGSOC cells. We found an enrichment of H3K9me2 compared to matched PARPi-sensitive cells. Analysis of a tissue microarray (TMA) shows that high H3K9me2 is associated with poorer overall survival. Parallel transcriptome analysis of H3K9-associated epigenetic enzymes revealed a significant increase in EHMT1/2 expression. Tumors from a patient-derived xenograft (PDX) model of PARPi-treated HGSOC also exhibited upregulation of EHMT1/2. Using both genetic and pharmacologic approaches to disrupt EHMT1/2 in PARPi-resistant cells, we resensitized cells to PARPi. Mechanistically, EHMT1/2 disruption did not increase cell death or apoptosis, but did promote increased DNA damage, ablated both HR- and NHEJ-mediated DNA damage repair, and altered cell cycle and transcriptional regulation.
H3K9me2 is globally increased in PARPi-resistant HGSOC cells and correlates with poorer overall survival
To correlate the in vitro H3K9me2 findings to clinically relevant specimens, we performed immunohistochemical staining for H3K9me2 using a TMA of serous tumors (tumor and patient details in Additional file 3). Slides were blinded and H3K9me2 staining was scored from 0 to 3, including half units. Scores < 2 were considered “Low” while scores ≥ 2 were considered “High.” We generated a Kaplan-Meier (K-M) survival curve by correlating scores to overall patient survival, and we observed that high H3K9me2 staining correlated with poorer overall survival (Fig. 1e). Examples of low and high staining are shown in Fig. 1f. H3K9me2 staining within stromal regions was consistent across samples, indicating that changes in H3K9me2 staining intensity were specific to tumor regions (Additional file 2: Figure S2). Although the TMA contains additional samples, to avoid confounding factors in our analysis, we used only the 92 primary, chemonaïve tumors for the K-M curve.
EHMT1 and EHMT2 are overexpressed in PARPi-resistant HGSOC cell lines and patient-derived ascites
Recurrent, PARPi-insensitive HGSOC is difficult to evaluate because in current clinical practice, ascites, and tumors from such patients are rarely collected. Therefore, we utilized a patient-derived xenograft (PDX) model of HGSOC to establish an olaparib-insensitive model. Following intraperitoneal injection of primary HGSOC ascites samples into immunocompromised NOD SCID gamma (NSG) mice, tumor-bearing mice were treated daily with olaparib or vehicle control for 21 days and the mice were monitored for 2 months (design schematic in Additional file 2: Figure S4). Olaparib-treated cells were subsequently shown to be highly olaparib-resistant . Ascites cells were isolated from the control and olaparib-treated mice and examined for EHMT1/2 mRNA and protein expression. EHMT1/2 mRNA and protein expression were significantly upregulated in the olaparib-treated ascites cells compared to vehicle control (Fig. 2f–h).
EHMT2 correlates with HGSOC progression and chemoresistance, and combined EHMT1/2 correlates with poor patient outcomes
Knockdown or inhibition of EHMT1/2 restores PARPi sensitivity
We next determined if pharmacologic inhibition of EHMT1/2 methyltransferase activity recapitulates results from knockdown experiments and restores olaparib sensitivity. We pre-treated PEO1-OR cells with highly-specific EHMT1/2 inhibitor UNC0642  for 48 h and evaluated H3K9me2 abundance via immunoblot and immunofluorescence. UNC0642 treatment reduced H3K9me2 in a dose-dependent fashion, but unrelated H3K27me3 was unaffected (Fig. 4f). Dose-response colony formation assays show that co-treatment with UNC0642 resensitized PEO1-OR cells to olaparib (Fig. 4g), compared to cells treated only with olaparib. Differential response was particularly notable at 600 and 3000 nM olaparib. To determine if EHMT1/2 also convey a broad drug-resistant phenotype to DNA damaging agents, we performed combined UNC0642/cisplatin dose response assays. PEO1-OR cells were relatively sensitive to cisplatin, which was not affected by UNC0642 treatment, suggesting a specific resistance to olaparib in these cells (Additional file 2: Figure S7A). We also performed combined UNC0642/olaparib dose response assays in PARPi-sensitive PEO1 parental HGSOC cells. We noted no difference in olaparib sensitivity between olaparib alone or in combination with 1 μM UNC0642 (Additional file 2: Figure S7B), suggesting that PEO1-OR cells may have developed a dependence on EHMT1/2 that is not present in the parental, sensitive line. To determine if EHMT1/2 inhibition could sensitize BRCA1/2-wildtype cells, we performed combination UNC0642/olaparib dose response assays in PARPi-resistant OVCA433-OR cells  which are TP53-mutant, BRCA1/2-wildtype. Despite a lack of elevated EHMT1/2, these cells show a marked elevation of H3K9me2 relative to parental OVCA433 cells (Additional file 2: Figure S7C-D). One micromolar UNC0642 in combination with olaparib reduced resistance by 1.9× compared to olaparib alone (Additional file 2: Figure S7E), suggesting that EHMT1/2 inhibition may be effective in some BRCA1/2-wildtype tumors, particularly those with elevated H3K9me2, but is likely most effective in the context of HR-deficient tumors.
Inhibition of EHMT1/2 alters cell cycle regulation but does not induce apoptosis or senescence
We next assayed senescence using an assay of beta-galactosidase (β-gal) activity. Senescent cells have an increase in β-gal activity  that we measured using the substrate C12FDG. Cleavage of C12FDG by β-gal generates a fluorescent product that can be detected through flow cytometry. To determine if EHMT1/2 inhibition induced senescence, PEO1-OR cells were treated with UNC0642 and/or olaparib for 72 h and then incubated with C12FDG. β-gal activity was not significantly different following the inhibition of EHMT1/2 in combination with olaparib suggesting that senescence was not induced (Fig. 5c). To determine if EHMT1/2 inhibition altered cell cycle regulation, we pre-treated PEO1-OR cells for 72 h with vehicle control or UNC0642, then treated with olaparib, UNC0642, or combination for an additional 72 h. We then fixed cells and stained with propidium iodide (PI) for DNA content. Flow cytometry analysis allowed for examination of G1, S, and G2 phases (Fig. 5d). We observed that olaparib treatment alone significantly decreased G1%, while increasing S% and G2%. These changes were not observed with UNC0642 alone. Relative to olaparib alone, combined treatment with UNC0642 slightly decreased G2% and slightly increased G1% and S%, suggesting that a proportion of combination-treated cells are halting in G1 or are not completing DNA synthesis.
To further analyze cell cycle, we used the same conditions as for PI staining and then examined protein expression of several cell cycle regulators by immunoblot (Fig. 5e), including Cyclin A (present in S and G2, degraded in M), Cyclin B1 (G2/M-specific), phospho-CDC25C and total CDC25C (a tyrosine phosphatase that directs de-phosphorylation of Cyclin B-bound CDC2 and triggers entry into mitosis), and mitotic markers phospho-MPM2 (a protein motif that is phosphorylated in over 50 proteins) and phospho-H3(Ser28). We observed that olaparib alone increased Cyclin A and Cyclin B1, which is consistent with our observed increases in G2%. Conversely, UNC0642 alone reduced Cyclin A and Cyclin B1. When combined with olaparib, UNC0642 partially prevented the observed increases in Cyclin A and Cyclin B1 due to olaparib. Pretreatment with UNC0642 reduced total and phosphorylated CDC25C, and combined treatment with olaparib/UNC0642 showed the greatest reduction in both, indicating reduced entry into mitosis. Consistent with these data, combined treatment also reduced levels of p-MPM2 and p-H3(Ser28). Densitometry analyses of immunoblots are shown in Additional file 2: Figure S8. Taken together, these data suggest that combined PARP and EHMT1/2 inhibition may sensitize PARPi-resistant cells in a cytostatic manner by preventing entry into mitosis and reducing proliferation.
Disruption of EHMT1/2 induces DNA damage, ablates DNA repair, and causes large transcriptional changes in survival pathways
Gene set enrichment analysis of significantly changed gene expression in UNC0642-treated PEO1-OR HGSOC cells
Gene set name
# Genes in set (K)
# Genes in overlap (k)
FDR q value
Genes upregulated through activation of mTORC1 complex
A subgroup of genes regulated by MYC
Unfolded Protein Response
Genes upregulated during unfolded protein response in the endoplasmic reticulum
Genes upregulated in response to hypoxia
PI3K/AKT/ mTOR Signaling
Genes upregulated by activation of the PI3K/AKT/mTOR pathway
Genes involved in cholesterol homeostasis
Genes involved in oxidative phosphorylation
Genes involved in processing of drugs and other xenobiotics
Genes involved in DNA repair
Genes upregulated by STAT5 in response to IL2 stimulation
PEO1-OR cells were treated with 500 nM UNC0642 for 72 h. RNA was isolated and analyzed by RNA-Seq. The transcriptome of treated cells was compared to the transcriptomes of four untreated PEO1-OR clonal populations to identify significantly changed gene expression. Gene set enrichment analysis of significantly changed genes was performed. The top 10 overlapping pathways are shown.
Recent trials show that PARPi have significant clinical benefit to both newly diagnosed and recurrent HGSOC cases, regardless of BRCA status [3, 4, 5]. Nearly all HGSOC patients are now eligible to receive PARPi, and there is a critical need to identify, understand, and specifically target mechanisms of PARPi resistance. We have previously reported on Wnt signaling as a mechanism of PARPi-resistance in HGSOC . Our RNA-Seq data of UNC0642-treated PEO1 did not show Wnt signaling as a top hit in gene set enrichment analysis. It is unknown if Wnt signaling and the mechanisms in this report are linked, regulate one another, or are entirely independent. Ovarian cancer, especially at later and metastatic stages is a heterogeneous disease, and it must be noted that more than one mechanism of resistance may exist within a patient, a population of cells, or even a single tumor. Therefore, it is of key importance to identify as many mechanisms as possible to develop the best possible treatment options. Going forward, combinatorial approaches will certainly be more effective than single agents.
We observed that there was a significant shift in the epigenetic landscape of BRCA2-mutant PARPi-resistant HGSOC cells, with a prominent enrichment of H3K9me2. Transcriptome analysis of PARPi-resistant cells narrowed the focus to two specific histone methyltransferases, EHMT1/2, which were significantly upregulated in both a cell line model and an in vivo PDX model of PARPi resistance. EHMT1/2 are methyltransferases whose canonical function is to catalyze H3K9me2, a modification associated with transcriptional repression [13, 14, 15]. Both genetic and pharmacologic approaches of disrupting EHMT1/2 activity reduced H3K9me2 levels and resensitized cells to olaparib, suggesting that these enzymes and the H3K9me2 epigenetic modification are playing a role in promoting or maintaining resistance. Single knockdown of EHMT1 or EHMT2 moderately restored olaparib sensitivity, but double knockdown of EHMT1/2 promoted a stronger resensitization phenotype. This is consistent with the known overlapping methyltransferase activities of EHMT1 and EHMT2 . Mechanistically, in the context of PARPi-resistant HGSOC cells, EHMT1/2 disruption significantly reduced functional HR and NHEJ DNA repair pathways, and promoted DNA damage. Disruption only moderately altered cell cycle regulation and did not cause increased apoptosis or induce senescence. However, EHMT1/2 inhibition caused significant transcriptional changes in PARPi-resistant HGSOC cells, particularly in growth and survival pathways. Whether changes in these pathways are responsible for reduced mitotic markers remains to be determined.
Restoration of DNA repair capacity through secondary BRCA mutations have been described as mechanisms of PARPi resistance [6, 7, 8, 9], and our findings strongly indicate that EHMT1/2 overexpression and/or hyperactivity also augment DNA repair in PARPi-resistant HGSOC. Our RNA-Seq analysis of UNC0642-treated PEO1-OR cells indicates that “DNA repair” genes were significantly changed by treatment. EHMT1/2 may therefore regulate DNA repair through epigenetic control of transcription. In addition to canonical roles in epigenetic regulation, several reports have demonstrated direct promotion of DNA damage repair by EHMT1/2, and subsequent knockdown increases sensitivity to chemotherapeutic agents [18, 19]. Our observation that disruption of EHMT1/2 promoted increased γH2AX and prevented repair are consistent with these results, and may be due to inhibition EHMT1/2 in recruiting repair factors to double-strand breaks. Further analysis of repair factor recruitment and repair kinetics will be required to ascertain the importance of these direct roles in PARPi-resistant HGSOC. In addition to roles in DNA repair, the complex of EHMT1/2 and ZNF644 have also been directly implicated in replication fork stability . By disrupting EHMT1/2 in PARPi-resistant cells, we may be causing replication fork stalling or instability, thus slowing or preventing completion of DNA synthesis and subsequent entry into mitosis. DNA fiber or combing analyses may reveal if replication forks are stalled in the context of EHMT1/2 disruption.
In our models, EHMT1/2 disruption did not induce senescence or increase apoptotic response, but we did observe moderate differences in cell cycle when PARPi-resistant cells were co-treated with PARPi and EHMT1/2 inhibitor. Specifically, we noted a decrease in the percentage of cells in G2 and a reduction of inducers and markers of mitosis. It remains unknown if these changes are due to DNA damage or other effects. However, our gene set enrichment analysis of UNC0642-treated PEO1-OR showed an overlap with several pro-survival signaling pathways including mTOR, AKT, PI3K, and MYC. Rather than causing cell death, which we did not observe, disrupted EHMT1/2 interaction with these pathways may potentially sensitize resistant HGSOC cells by preventing growth and survival signaling. It is unknown if EHMT1/2 are acting only as epigenetic regulators of these pathways, or has direct interactions. However, EHMT1/2 often functions in a multi-subunit complex and so it will also be important to identify additional interacting factors. Notably, a recent study by Tu et al. showed that EHMT2 interacts with c-MYC to drive transcriptional repression and tumorigenesis in breast cancer, and that EHMT2 inhibition was a potent suppressor of MYC-dependent tumor growth . c-MYC is often amplified in ovarian cancers and has previously been proposed as a therapeutic target in platinum-resistant cases . Further study is required to determine if EHMT1/2 interacts with MYC, or other pro-survival signaling pathways, in the context of PARPi-resistant ovarian cancer.
Our mass spectrometry profiling of histone modifications revealed increased H3K9me2 in PARPi-resistant cells relative to sensitive cells. This may be due to a broad, but moderate, increase in H3K9me2 across the entire genome. However, given the significant changes in transcriptional programming observed in our RNA-Seq data, we surmise that it is more likely that H3K9me2 is highly enriched at multiple specific gene loci. Further investigation, including ChIP-Seq, is required to reveal the specific genomic loci enriched for H3K9me2 and EHMT1/2 in PARPi-resistant cells. Combined with our RNA-Seq analyses, these data will indicate epigenetically regulated genes that are potential effectors of PARPi resistance. Notably, BRCA1/2-wildtype OVCA433-OR cells did not upregulate EHMT1/2, but did have elevated H3K9me2, suggesting that the direct roles of EHMT1/2 in DNA repair may not be required for resistance in cells with functional BRCA. However, upregulation of H3K9me2 and subsequent transcriptional reprogramming may be a broader resistance mechanism.
Targeting EHMT1/2 to overcome PARPi resistance is a novel approach. The EHMT1/2 inhibitor UNC0642 has suitable pharmacodynamic properties for in vivo studies , which will allow for testing whether targeting EHMT1/2-dependent PARPi resistance is a viable therapeutic strategy in animal models. A recent report found that EGFR-tyrosine kinase inhibitor (erlotinib) resistant non-small cell lung cancer (NSCLC) has increased EHMT2 expression. Moreover, the authors found that combining erlotinib with UNC0642 significantly reduced NSCLC tumor burden in a PDX mouse model . Looking beyond roles in chemoresistance and toward a more broad therapeutic strategy, EHMT1/2 inhibition may become a key target for immunotherapies. Liu et al. showed that EHMT2 inhibition synergized with DNA methyltransferase (DNMT) inhibition in A2780 and CAOV3 ovarian cancer cell lines to induce “viral mimicry,” a state in which expression of endogenous retroviruses and innate antiviral response genes are activated . Studies in appropriate immunocompetent animal models will be essential to determine if such activity can induce the host immune system to kill ovarian cancer cells in vivo.
Our reported findings further highlight the potential of EHMT1/2 inhibition to overcome targeted therapy resistance and prevent cancer progression. In conclusion, the epigenetic landscape is contributing to PARPi resistance through upregulation of EHMT1/2 and targeting EHMT1/2 is a potential approach to managing PARPi resistant HGSOC.
Cell culture, shRNA, and lentivirus
Cell lines were obtained from the Gynecologic Tumor and Fluid Bank (GTFB) at the University of Colorado, and were authenticated at the University of Arizona Genomics Core using short tandem repeat DNA profiling. Regular Mycoplasma testing was performed using MycoLookOut PCR (Sigma). HGSOC lines were cultured in RPMI 1640 supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. 293FT lentiviral packaging cells were cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin. All cells were grown at 37 °C supplied with 5% CO2. shRNA in pLKO.1 lentiviral vector plasmids were purchased from the University of Colorado Functional Genomics Facility. Sequences and The RNAi Consortium numbers are listed in Additional file 4: Table S1. A scrambled non-targeting shRNA was used as control (Sigma-Aldrich #SHC016). Lentivirus was packaged as previously described  in 293FT using third-generation packaging plasmids (Virapower, Invitrogen) with polyethyleneimene (PEI) transfection in a 1:3 DNA:PEI ratio. Culture supernatant was harvested at 48–72 h post-transfection and processed through 0.45 μM filters. Viruses encoded a puromycin resistance gene. Transduced HGSOC cells were selected in 1 μg/mL puromycin. Functional DNA repair plasmids (described below) encode a puromycin resistance gene and thus preclude the use of pLKO.1 vectors, which also encode puromycin resistance. Knockdowns for these experiments were thus performed using pLKO.1-blast (Addgene #26655), which encodes a blasticidin resistance gene in place of puromycin. shControl, shEHMT1#1, and shEHMT2#1 were cloned into pLKO.1-blast between AgeI and EcoRI restriction sites. Virus was produced and cells were transduced and selected as described, except that selection occurred in 1 μg/mL blasticidin.
Colony formation assay
Cell lines were seeded in 24-well plates and treated with increasing doses of olaparib. Media and olaparib were changed every three days for 12 days or until control wells were confluent, whichever occurred first. Colonies were washed twice with PBS, then incubated in fixative (10% methanol and 10% acetic acid in PBS). Fixed colonies were stained with 0.4% crystal violet in PBS. After imaging, crystal violet was dissolved in fixative and absorbance was measured at 570 nm using a Molecular Devices SpectraMax M2e plate reader.
Histone modification profiling
Profiling of histone modifications in olaparib-sensitive and -resistant cells was performed by the Northwestern University Proteomics Core. Briefly, we provided frozen pellets of 5 × 106 PEO1 and PEO1-OR cells. Histone extracts were trypsin digested and histone residues were assayed as previously reported [35, 36] by liquid chromatography coupled to mass spectrometry using a TSQ Quantiva Ultra Triple Quadrupole Mass Spectrometer.
A previously constructed tissue microarray comprised of serous tumors from ovarian cancer patients treated at the University of Colorado was provided by the GTFB (COMIRB #17-7788). Slides were immunohistochemically stained for H3K9me2. Slides were blinded and staining was manually scored from 0 to 3, including half units. Scores < 2 were considered “Low” while scores ≥ 2 were considered “High.” A Kaplan-Meier survival curve was generated by correlating scores to overall patient survival. Only primary, chemonaïve tumors were used for generating the K-M curve.
PDX mouse model of PARPi-resistant HGSOC
All animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals and were approved by the University of Colorado IACUC. Primary ovarian cancer sample GTFB1009 (BRCA1/2-wildtype) was provided by the University of Colorado GTFB. Six to 8-week-old NOD SCID gamma (NSG) mice (Jackson Labs) were given intraperitoneal injections of 5 million GTFB1009 ascites cells each. Following a 7-day incubation period, mice were given once daily intraperitoneal injections of 50 mg/kg olaparib or vehicle control (10% 2-hydroxypropyl-β-cyclodextrin, Sigma-Aldrich #C0926) for 21 days. After treatment, tumors were allowed to recur, and then mice were euthanized and ascites were collected for analysis.
RNA-sequencing of olaparib-sensitive PEO1 and olaparib-resistant PEO1-OR
RNA was isolated from PEO1 olaparib-sensitive (n = 2) and four PEO1 olaparib-resistant clones using RNeasy columns with on-column DNase digest (Qiagen). RNA quality was confirmed using an Agilent Tapestation and all RNA used for library preparation had a RIN > 9. Libraries were created using Illumina TruSEQ stranded mRNA library prep (#RS-122-2102). Strand-specific pair-ended libraries were pooled and run on HiSeq4000 (Illumina). Library creation and sequencing were performed at the University of Colorado Genomics Core. HISAT  was used for alignment against GRCh37 version of the human genome. Samples were normalized using transcripts per kilobase million (TPM) measurement and gene expression using the GRCh37 gene annotation was calculated using home-made scripts. Analysis was performed by the Division of Translational Bioinformatics and Cancer Systems Biology at the University of Colorado School of Medicine. Data have been deposited to NCBI GSE117765.
RNA-sequencing of UNC0642-treated olaparib-resistant PEO1-OR
PEO1-OR cells were treated for 72 h with 500 nM UNC0642. RNA was isolated from cells using the RNeasy Plus Mini Kit (Qiagen). RNA quality was confirmed using an Agilent Tapestation and all RNA used for library preparation had a RIN > 9. Library preparation and sequencing were performed by Novogene Co, Ltd. Using Illumina reagents and the HiSeq platform. Analysis was performed as above. Data have been deposited to NCBI GSE135864.
Reverse-transcriptase quantitative PCR
RNA was isolated from cells using the RNeasy Plus Mini Kit (Qiagen). mRNA expression was determined using SYBR green Luna One Step reverse-transcriptase quantitative PCR (RT-qPCR) Kit (New England BioLabs) on a C1000 Touch (Bio-Rad) or QuantStudio 6 (Applied Biosystems) thermocycler. Expression was quantified by the ΔΔCt method using target-specific and control primers. β-2-microglobulin (B2M) and Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were used as internal controls. mRNA-specific primers were designed to span exon-exon junctions to avoid detection of genomic DNA. Primer sequences are shown in Additional file 4: Table S2.
Ovarian cancer dataset analysis
Publicly available ovarian cancer databases (GSE9899, GSE13813, and GSE1926) were examined for correlations between disease recurrence and chemoresistance and EHMT2 expression (Oncomine, ThermoFisher).
Inhibitors and antibodies
Olaparib (#S1060), UNC0638 (#S8071), and UNC0642 (#S7230) were obtained from SelleckChem. Full details of antibodies and usage for immunoblotting and immunofluorescence are given in Additional file 4: Table S3. Alexa Fluor 488-conjugated donkey anti-mouse secondary antibody (Invitrogen #A21202, 1:1000) was used for immunofluorescence detection of γH2AX and H3K9me2.
For histone blots, extracts were made using the Histone Extraction Kit (Abcam #ab113476). For total protein, cells were lysed and briefly sonicated in RIPA buffer (150 mM NaCl, 1% TritonX-100, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris pH 8.0) supplemented with complete EDTA-free protease inhibitors (Roche #11873580001) and phosphatase inhibitors NaF and NaV. Protein was separated by SDS-PAGE and transferred to PVDF membrane using the TransBlot Turbo (BioRad). Membranes were blocked for 1 h at room temperature. Primary antibody incubation was performed overnight at 4 °C. Membranes were washed three times for 5 min each in TBST (50 mM Tris pH 7.5, 150 mM NaCl, 0.1% Tween-20), then secondary antibodies were applied for one hour at room temperature. Membranes were washed again three times for 5 min each in TBST. For fluorescent detection, bands were visualized using the LI-COR Odyssey Imaging System. For HRP detection, chemiluminescent signal was detected with SuperSignal West Femto (Thermo Scientific #34095) and visualized using a G:Box (SYNGENE). Details of blocking buffers and detection methods are provided in Additional file 4: Table S3.
For immunoblot images captured using the LI-COR Odyssey, band fluorescence intensity was analyzed using LI-COR ImageStudio 4. For all others, G:Box images of immunoblots were analyzed using ImageJ. Immunoblots of histone extracts were normalized to band intensity of total H3. Immunoblots of total protein lysates were normalized to intensity of β-actin.
Cells were seeded on glass coverslips and treated as described in the figure legends. After treatment, cells were washed three times in PBS, then fixed in 4% paraformaldehyde for 10 min at room temperature, followed by three additional PBS washes. Cells were permeabilized for 3 min at room temperature using 0.2% Triton X-100. Primary and secondary antibodies were diluted in 3% BSA/PBS. Primary antibody was applied for 2 h at room temperature, followed by three washes in 1% Triton X-100/PBS and one wash in PBS. Secondary antibody was applied for 1 h at room temperature in the dark. Cells were washed three times with PBS, then mounted on glass slides using SlowFade Gold antifade reagent with DAPI (Invitrogen #S36938). Slides were imaged using an Olympus FV-1000 microscope (University of Colorado Advanced Light Microscopy Core).
Two-plasmid functional DNA repair assay
Two-plasmid functional assays were performed to assess distal non-homologous end joining and homology directed repair. Cells were stably transfected with pimEJ5GFP (NHEJ) or pDRGFP (HR) by maintenance in 0.5 μg/mL puromycin and subsequently transfected with I-SceI restriction enzyme. After 72 h, transfected cells were collected and examined using a Beckman Coulter Gallios 561 flow cytometer (Flow Cytometry Shared Resource, University of Colorado) to quantify GFP positive cells. pimEJ5GFP (Addgene #44026) was a gift from Jeremy Stark. pDRGFP (Addgene #26475) and pCBASceI (Addgene #26477) were gifts from Maria Jasin. To control for I-SceI transfection efficiency, DNA was isolated from cells remaining after flow analysis and qPCR was performed using primers specific for transfected I-SceI DNA. Primers for Claudin 4 (CLDN4) gDNA were used as a DNA loading control. Primer sequences are listed in Additional file 4: Table S2.
Annexin V/propidium iodide assay
Phosphatidylserine externalization was detected using an Annexin V/propidium iodide (PI) staining kit (Life Technologies) following the manufacturer’s instructions. Annexin V/PI positive cells were detected using a Beckman Coulter Gallios 561 Flow Cytometer.
Β-galactosidase staining for senescence
Following drug treatment, PEO1-OR cells were incubated with 33 μM C12FDG (5-dodecanoylaminofluorescein d-β-d-galactopyranoside, Cayman 25583) for 1 h at 37 °C. Fluorescent cleaved substrate was detected using a Beckman Coulter Gallios 561 Flow Cytometer.
Cell cycle staining
PEO1-OR cells were incubated with the indicated concentrations of olaparib and/or UNC0642 for 72 h, then collected by trypsinization and washed once with PBS. Cells were then resuspended and fixed in ice cold 70% ethanol for 1 h at − 20 °C. Cells were washed once with PBS, then resuspended in 200 μL PI staining solution [1× PBS, 50 μg/mL RNase A (Thermo Scientific #EN0531), 50 μg/mL propidium iodide (Thermo Scientific #P3566)] for 30 min at 37 °C, then analyzed by flow cytometry using a Beckman Coulter Gallios 561 Flow Cytometer.
Software and statistical analysis
Flow cytometry analysis was performed using FlowJo 10. Statistical analysis and calculation of P value was performed using GraphPad Prism 7. Quantitative data are expressed as mean ± SD unless otherwise stated. Two-tailed t test was used for single comparisons. Analysis of variance (ANOVA) with Fisher’s least significant difference (LSD) was used in multiple comparisons. For all statistical analyses, the level of significance was set at 0.05.
The authors thank Dr. Philip Owens for microscope use for IF experiments, Christina Looby for assistance with scoring the TMA, and Dr. Peggy Neville and Dr. Heidi K. Baumgartner-Wilson for helpful comments.
ZLW designed the study, performed experiments, analyzed and interpreted data, and prepared and revised the manuscript. TMY, AM, and CJH performed additional experiments and edited the manuscript. HK analyzed RNA-Seq data. LJW, MDP, and KB provided the TMA. BGB aided in study design, data analysis and interpretation, and manuscript preparation and revision. All authors read and approved the final manuscript.
BGB is supported by NIH/NCI grant R00CA194318 and Cancer League of Colorado grant 183478-BB. ZLW is supported by Cancer League of Colorado grant 193527-ZW. This work was supported in part by the University of Colorado Cancer Center Genomics and Microarray Core Shared Resource funded by NCI grant P30CA046934, and by NIH/NCATS Colorado CTSA Grant Number UL1TR002535. Histone profiling was performed by the Northwestern University Proteomics Core, supported by NCI CCSG P30CA060553 and by P41GM108569.
Ethics approval and consent to participate
The TMA was approved by the University of Colorado Multiple Institutional Review Board (COMIRB #17-7788). All participants consented to use of tissue specimens. Mouse PDX studies were approved by the University of Colorado IACUC (protocol #347) and were conducted in accordance with appropriate guidelines for ethical and humane treatment of animals.
Consent for publication
The authors declare that they have no competing interests.
- 4.Domchek SM, Aghajanian C, Shapira-Frommer R, Schmutzler RK, Audeh MW, Friedlander M, et al. Efficacy and safety of olaparib monotherapy in germline BRCA1/2 mutation carriers with advanced ovarian cancer and three or more lines of prior therapy. Gynecol Oncol. 2016;140(2):199–203.CrossRefGoogle Scholar
- 7.Christie EL, Fereday S, Doig K, Pattnaik S, Dawson SJ, DDL B. Reversion of BRCA1/2 Germline mutations detected in circulating tumor DNA from patients with high-grade serous ovarian cancer. J Clin Oncol. 2017.Google Scholar
- 21.Kang J, Shin SH, Yoon H, Huh J, Shin HW, Chun YS, et al. FIH is an oxygen sensor in ovarian cancer for G9a/GLP-driven epigenetic regulation of metastasis-related genes. Cancer Res. 2017.Google Scholar
- 23.Yamamoto TM, McMellen A, Watson ZL, Aguilera J, Ferguson R, Nurmemmedov E, et al. Activation of Wnt signaling promotes olaparib resistant ovarian cancer. Mol Carcinog. 2019;58:1770–1782.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.