The emergence of tumor cells with certain stem-like characteristics, such as high aldehyde dehydrogenase (ALDH) activity due to ALDH1A1 expression, contributes to chemotherapy resistance and tumor relapse. However, clinically applicable inhibitors of ALDH activity have not been reported. There is evidence to suggest that epigenetic regulation of stem-related genes contributes to chemotherapy efficacy. Here, we show that bromodomain and extraterminal (BET) inhibitors suppress ALDH activity by abrogating BRD4-mediated ALDH1A1 expression through a super-enhancer element and its associated enhancer RNA. The clinically applicable small-molecule BET inhibitor JQ1 suppressed the outgrowth of cisplatin-treated ovarian cancer cells both in vitro and in vivo. Combination of JQ1 and cisplatin improved the survival of ovarian cancer–bearing mice in an orthotopic model. These phenotypes correlate with inhibition of ALDH1A1 expression through a super-enhancer element and other stem-related genes in promoter regions bound by BRD4. Thus, targeting the BET protein BRD4 using clinically applicable small-molecule inhibitors, such as JQ1, is a promising strategy for targeting ALDH activity in epithelial ovarian cancer. Cancer Res; 76(21); 6320–30. ©2016 AACR.

Chemotherapeutic drugs, such as cisplatin, have had a major impact on the therapeutic management of many tumors, and in particular in epithelial ovarian cancer (EOC; ref. 1). However, chemotherapy resistance is a major cause of cancer morbidity and mortality. For example, although EOC, in particular high-grade serous carcinoma (HGSC), initially respond well to platinum-based chemotherapy, relapse often occurs with decreased chemotherapy sensitivity (2). The hypothesis that this occurs due to cancer stem-like cells (CSC) remains controversial. However, there is substantial evidence in the literature that cells with CSC characteristics contribute to chemotherapy resistance and tumor relapse (3). Putative EOC CSCs are typically characterized by increased aldehyde dehydrogenase (ALDH) activity with concomitant upregulation of ALDH1A1 (4–6). ALDH activity is functionally important in EOC, as suppressing ALDH activity by knocking down ALDH1A1 has been shown to sensitize EOC cells to chemotherapy (4). In addition, a population of normal ovarian stem cells also has increased ALDH activity (7), further supporting its functioning in putative ovarian CSCs. Despite the mounting evidence on the critical role of ALDH1A1 in regulating CSCs (8), the molecular mechanisms underlying its regulation remain poorly understood. Notably, clinically applicable inhibitors of ALDH activity or ALDH1A1-targeting approaches have not been reported.

Recent genome-wide next-generation sequencing studies in human cancers have revealed frequent alterations in genes and proteins that are critical in regulating the epigenetic landscape of chromatin (9, 10). This suggests that proteins encoded by these genes may be cancer therapeutic targets. Accordingly, small-molecule inhibitors targeting chromatin-regulating epigenetic enzymes have been developed (11). The bromodomain and extraterminal (BET) family of proteins recognize acetylated lysine on histones through their bromodomains (12). BET proteins control the transcription of their target genes either directly by recruiting transcriptional machinery or indirectly through involving enhancer elements in a lineage and context-specific manner (12). Highly specific BET inhibitors are in clinical trials (13). Pharmacologic inhibitors of BET proteins have shown efficacy in the clinic in a number of pathologies, most notably in cancer. There is evidence to suggest that epigenetic regulation of stem-related genes contributes to chemotherapy efficacy (14, 15). Using unbiased approaches, here, we have identified BET inhibitors as suppressors of ALDH activity that potentiate the antitumor effects of cisplatin in EOC.

Cell lines and epigenetic small-molecule screen

Human EOC cell lines were obtained from ATCC within 3 years and were reauthenticated by The Wistar Institute's Genomics Facility at the end of the experiments within last 3 months using short tandem repeat profiling using AmpFLSTR Identifiler PCR Amplification Kit (Life Technologies) and cultured as described previously (16). The Structural Genome Consortium generously provided the epigenetic compound library. OVCAR3 cells were plated in 384-well plates and treated with serial dilutions (0–20 μmol/L) of 24 epigenetic compounds with or without IC20 cisplatin. Data were analyzed with GraphPad (Prism).

In situ 3C assay

In situ chromosome conformation capture (3C) samples were prepared as described previously with modifications (17). Briefly, cells (5 × 106) were cross-linked with 1% formaldehyde and quenched by 2.5 mol/L glycine. Cells were collected and resuspended in Hi-C lysis buffer (10 mmol/L Tris-HCl pH 8.0, 10 mmol/L NaCl, 0.2% NP-40) with proteinase inhibitor (Sigma). The cell suspension was incubated on ice, washed with Hi-C buffer, resuspended in 0.5% SDS, and incubated at 65°C for 5 minutes. After quenching the SDS, chromatin was digested overnight by MboI and then ligated. Ligated DNA was purified using Wizard SV Gel and PCR Clean-Up System (Promega). Quantitative PCR was performed by using Quantitect Probe PCR Master Mix (Qiagen) with custom probe and primers as described previously (18). Probe and primer sequences are indicated in Supplementary Table S1.

Nascent RNA sequencing and chromatin immunoprecipitation sequencing

For nascent RNA sequencing (RNA-seq), cells were incubated with 0.5 mmol/L ethidium uridine (EU) and treated with 125 nmol/L JQ1 or vehicle control for 40 minutes. Total RNA was extracted with TRIzol reagent and RNeasy Mini Kit. The EU-labeled RNAs were biotinylated and precipitated by using the Click-it Nascent RNA Capture Kit (Life Technologies) following the manufacturer's instructions. Briefly, 5 μg EU-labeled RNA was biotinylated with 0.25 mmol/L biotin azide in Click-it Reaction Buffer. Biotinylated RNAs were ethanol precipitated and resuspended in ultrapure water. Biotinylated RNAs were incubated with Dynabeads MyOne Streptavidin T1 magnetic beads in Click-it RNA and subjected to library preparation. Libraries for RNA-seq were prepared with Ovation Human FFPE RNA-Seq Multiplex System 1–8 (NuGEN) and sequenced on an Illumina NextSeq 500.

For chromatin immunoprecipitation sequencing (ChIP-seq), cells were cross-linked with 1% formaldehyde for 10 minutes, followed by quenching with 125 mmol/L glycine for 5 minutes. Fixed cells were resuspended in cell lysis buffer (10 mmol/L Tris-HCl, pH 7.5, 10 mmol/L NaCl, 0.5% NP-40) and incubated on ice for 10 minutes. The lysates were washed with MNase digestion buffer (20 mmol/L Tris-HCl, pH 7.5, 15 mmol/L NaCl, 60 mmol/L KCl, 1 mmol/L CaCl2) once and incubated for 20 minutes at 37°C in the presence of 1,000 gel units of MNase (NEB, M0247S) in 250 μL reaction volume. After adding the same volume of sonication buffer (100 mmol/L Tris-HCl, pH 8.1, 20 mmol/L EDTA, 200 mmol/L NaCl, 2% Triton X-100, 0.2% sodium deoxycholate), the lysates were sonicated for 5 minutes (30 seconds on/off) in a Diagenode Bioruptor and centrifuged at 15,000 rpm for 10 minutes. The cleared supernatant equivalent to 2–4 × 106 cells was incubated with 2 μg of anti-BRD4 antibody (Bethyl, A301-985A) on a rocker overnight. Bound chromatin was eluted and reverse cross-linked at 65°C overnight. For next-generation sequencing, ChIP-seq libraries were prepared from 10 ng of ChIP and input DNAs with the Ovation Ultralow DR Multiplex system (NuGEN). The ChIP-seq libraries were sequenced in a 51 base pairs paired end run using the Illumina HiSeq 2000.

In vivo orthotopic xenograft mouse model

The Institutional Animal Care and Use Committee (IACUC) at The Wistar Institute (Philadelphia, PA) approved all animal protocols described in this study. NOD/scid gamma (NSG) mice were injected intraperitoneally with OVCAR3 luciferase cells (5 × 106). Tumors were allowed to establish for 3 weeks and randomized into four groups: control (n = 12), JQ1 (n = 11), cisplatin (n = 12), and cisplatin/JQ1 (n = 13). Tumor growth was followed by noninvasive imaging as described previously (19). Briefly, tumors were visualized by injecting luciferin (4 mg/mice i.p.) resuspended in PBS, and imaged with an IVIS Spectrum. JQ1 was resuspended in 10% 2-hydroxypropyl-β-cyclodextrin solvent (Sigma-Aldrich) as described previously (20). Cisplatin was purchased from SelleckChem and dissolved in PBS. Mice were treated daily with intraperitoneal injections of vehicle controls and/or JQ1 (20 mg/kg) and/or biweekly with cisplatin (750 μg/kg). Tumor cells collected from peritoneal washes were incubated with ammonium chloride to lyse erythrocytes and then used for the ALDEFLUOR assay and stained with PE-anti-mouse CD45 (BD Biosciences) antibody to exclude mouse-derived hematopoietic cells. Survival of tumor-bearing mice was evaluated on the basis of IACUC criteria.

Analysis using primary human ovarian tumors

For analyzing ALDH activity, the protocol for using the primary human ovarian tumor specimen was approved by the Wistar Institute Institutional Review Board (IRB). For expression of enhancer RNA (eRNA), BRD4 and ALDH1A1, the protocol using human ovarian tumor specimens was approved by the IRB at The M.D. Anderson Cancer Center (Houston, Texas). RNA was extracted from 26 high-grade serous ovarian tumors using the mirVana RNA Isolation Kit (Thermo Fisher Scientific) according to the manufacturer's protocol. Analysis of RNA levels was performed on a 7500 Fast Real-Time PCR System (Applied Biosystems) with SYBR Green–based real-time PCR using the primers as detailed in Supplementary Methods. Expression of β-actin was used as a housekeeping gene control. Analysis was performed using the 7500 Real-Time PCR software.

Bioinformatics and statistical analysis

For ChIP-seq, alignment was done versus hg19 version of human genome using bowtie algorithm. BRD4 ChIP-seq for vehicle control–treated cells was compared versus input and versus JQ1 using HOMER algorithm with “-histone” option. FDR <1% was set as a significance threshold. RNA-seq data were aligned using bowtie2 algorithm, and RSEM was used for estimating number of reads for each gene. EdgeR was used to test for differential expression and FDR <10% was used as a significance threshold unless stated otherwise. Ingenuity Pathway Analysis software was used to test gene sets for enrichment of cellular functions and canonical pathways, and Ingenuity Knowledge Base was used to create regulation and protein–protein interaction network for stem-related genes. Differences in percentage between different classes were tested using Fisher exact test, with P < 0.05 used as a significance threshold. H3K4Me1 and H3K27Ac broad peaks were downloaded from ENCODE for GM12878, H1-hESC, HSMM, HUVEC, K562, NHEK, and NHLF cell lines for overlap with BRD4 peaks. To determine the effect of combination treatment, CI (combination index) values were calculated by using Compusyn software (21). CIs <1, 1, and >1 represent synergism, additive effect, and antagonism, respectively.

BET inhibitors suppress ALDH activity and inhibit ALDH1A1 expression

As ALDH activity regulates the putative ovarian CSCs and stem-related genes are subjected to epigenetic regulation (4–6, 14, 15), we evaluated a panel of 24 small-molecule inhibitors known to target epigenetic regulators obtained from The Structure Genomics Consortium on their ability to suppress ALDH activity (Fig. 1A; Supplementary Table S2). We examined the expression of ALDH1A1, the major determinant of ALDH activity (4, 8), in a panel of high-grade serous EOC cell lines (Supplementary Fig. S1A and S1B; ref. 22). We performed the evaluation of ALDH activity in OVCAR3 cells because these cells have high ALDH1A1 expression (Supplementary Fig. S1B). To limit the potential bias introduced by different growth inhibition potential among the small-molecule inhibitors, we established a growth inhibition curve for each small-molecule inhibitor and based the dose of each small-molecule inhibitor on the established IC20 value (Supplementary Table S2). The highest tested dose (20 μmol/L) was used for those inhibitors whose IC20 was not achieved. Validating our experimental design, a previously reported positive regulator of ALDH activity, an HDAC inhibitor, was identified (23). We identified four small-molecule inhibitors that significantly suppressed ALDH activity (Fig. 1A and B). Notably, all three BET inhibitors in the panel scored as “hits” that significantly suppressed ALDH activity. As JQ1 is clinically applicable (known as TEN-010 in clinical trials), we performed further validation on this inhibitor. We validated that JQ1 decreased ALDH activity in a dose-dependent manner (Fig. 1C and Supplementary Fig. S1C) and in primary EOCs (Fig. 1D). Similar results were also obtained by using I-BET 762, another BET inhibitor that is now in clinical development (Fig. 1E). We further validated that JQ1 decreases ALDH activity of EOC cells in vivo in an orthotopic xenograft mouse model (Fig. 1F and G). As a positive control, cisplatin increased ALDH activity in vivo as reported previously (4). Notably, both ALDH1A1 mRNA and ALDH1A1 protein levels were decreased by JQ1 treatment in a dose-dependent manner in multiple cell lines (Fig. 1H and I and Supplementary Fig. S1D and S1E). This suggests that JQ1 decreases ALDH activity by suppressing ALDH1A1 expression at the transcriptional level.

Figure 1.

BET inhibition decreases ALDH enzymatic activity and suppresses ALDH1A1 expression. A, plot of ratio of the quantified ALDH-positive cells OVCAR3 cells treated with the IC20 dose of the indicated epigenetic inhibitors or vehicle controls. For epigenetic inhibitors whose IC20 dose was not achieved, the highest dose tested (20 μmol/L) was used in the assay. Error bars, SEM of three independent experiments. B, representative changes in ALDH activity in OVCAR3 cells treated with the indicated positive “hits” identified in the evaluation. DEAB-treated cells were used as a negative control for ALDH activity. C, OVCAR3 cells were treated with the indicated doses of the BET inhibitor JQ1, and ALDH activity was measured by FACS. The percentages of positive cells are indicated. D, same as C, but for primary ovarian cancer cells isolated from a serous histosubtype ovarian tumor. E, same as C, but for the BET inhibitor I-BET-762. F, JQ1 inhibits ALDH activity in vivo in an intraperitoneal xenograft model using OVCAR3 cells. Percentages of ALDH activity–positive cells collected from peritoneal washes of the indicated treatment groups are indicated. Please see Materials and Methods for experimental details. G, quantification of F. Error bars, SEM. H, same as C, but examined for ALDH1A1 mRNA expression by qRT-PCR. Mean of three independent experiments with SEM. *, P < 0.03. I, same as C, but examined for ALDH1A1 protein expression by immunoblotting.

Figure 1.

BET inhibition decreases ALDH enzymatic activity and suppresses ALDH1A1 expression. A, plot of ratio of the quantified ALDH-positive cells OVCAR3 cells treated with the IC20 dose of the indicated epigenetic inhibitors or vehicle controls. For epigenetic inhibitors whose IC20 dose was not achieved, the highest dose tested (20 μmol/L) was used in the assay. Error bars, SEM of three independent experiments. B, representative changes in ALDH activity in OVCAR3 cells treated with the indicated positive “hits” identified in the evaluation. DEAB-treated cells were used as a negative control for ALDH activity. C, OVCAR3 cells were treated with the indicated doses of the BET inhibitor JQ1, and ALDH activity was measured by FACS. The percentages of positive cells are indicated. D, same as C, but for primary ovarian cancer cells isolated from a serous histosubtype ovarian tumor. E, same as C, but for the BET inhibitor I-BET-762. F, JQ1 inhibits ALDH activity in vivo in an intraperitoneal xenograft model using OVCAR3 cells. Percentages of ALDH activity–positive cells collected from peritoneal washes of the indicated treatment groups are indicated. Please see Materials and Methods for experimental details. G, quantification of F. Error bars, SEM. H, same as C, but examined for ALDH1A1 mRNA expression by qRT-PCR. Mean of three independent experiments with SEM. *, P < 0.03. I, same as C, but examined for ALDH1A1 protein expression by immunoblotting.

Close modal

BRD4 inhibition suppresses ALDH activity and inhibits ALDH1A1 expression

As high ALDH activity is implicated in chemotherapy response (8), we determined whether BET inhibitors synergize with cisplatin by inhibiting ALDH activity. Indeed, JQ1 displayed a synergistic effect with cisplatin in multiple EOC cell lines (Fig. 2A). In addition, JQ1 displayed a synergistic effect with cisplatin in the in vitro–derived cisplatin-resistant EOC cell line A2780 CP70 (Fig. 2B). Furthermore, in colony formation assays, JQ1 significantly suppressed the outgrowth of EOC cells after cisplatin treatment in multiple EOC cell lines (Fig. 2C and D and Supplementary Fig. S2A and S2B). This correlated with inhibition of ALDH activity (Fig. 2E) and suppression of the upregulated ALDH1A1 induced by cisplatin (Fig. 2F and Supplementary Fig. S2C). We observed an increase in apoptotic markers, such as cleaved caspase-3, cleaved lamin A and cleaved PARP p85 and Annexin V in the cells treated with JQ1 and cisplatin in combination compared with either treatment alone (Fig. 2G and Supplementary Fig. S2D and S2E). Notably, JQ1 significantly decreased anchorage-independent sphere formation in ALDH-positive cells, a characteristic of putative ovarian CSCs (24), to a degree that is comparable with those observed in ALDH-negative cells (Fig. 2H–J).

Figure 2.

JQ1 synergizes with cisplatin, which correlates with inhibition of ALDH activity. A, synergy analysis for JQ1 and cisplatin in the indicated ovarian cancer cell lines. Cells were treated with the indicated concentration of JQ1 and cisplatin for 72 hours. The combination index (CI) value was calculated. Combination index values: <1, synergism; 1, additive effect; >1, antagonism. Error bars, SEM and n = 3. B, logarithmic combination index plot of JQ1 (200 nmol/L) is generated in combination with cisplatin in cisplatin-resistant CP70 ovarian cancer cells. C, OVCAR3 cells treated with 125 nmol/L JQ1, 250 nmol/L cisplatin, or in combination for 12 days were assayed for colony formation. D, quantification of C. Mean of three independent experiments with SEM E, same as C, but cells were only treated for 72 hours and examined for the percentage of ALDH activity–positive cells by FACS. F, same as E, but examined for ALDH1A1 expression by immunoblotting. G, same as E, but examined for the indicated markers of apoptosis. H, ALDH1A1 protein expression in FACS-sorted ALDH activity–positive and negative cells determined by immunoblotting. I, sphere formation by the indicated ALDH activity–negative cells or ALDH activity–positive cells treated with or without JQ1. Scale bar, 40 μm. J, quantification of I. Mean of three independent experiments with SEM.

Figure 2.

JQ1 synergizes with cisplatin, which correlates with inhibition of ALDH activity. A, synergy analysis for JQ1 and cisplatin in the indicated ovarian cancer cell lines. Cells were treated with the indicated concentration of JQ1 and cisplatin for 72 hours. The combination index (CI) value was calculated. Combination index values: <1, synergism; 1, additive effect; >1, antagonism. Error bars, SEM and n = 3. B, logarithmic combination index plot of JQ1 (200 nmol/L) is generated in combination with cisplatin in cisplatin-resistant CP70 ovarian cancer cells. C, OVCAR3 cells treated with 125 nmol/L JQ1, 250 nmol/L cisplatin, or in combination for 12 days were assayed for colony formation. D, quantification of C. Mean of three independent experiments with SEM E, same as C, but cells were only treated for 72 hours and examined for the percentage of ALDH activity–positive cells by FACS. F, same as E, but examined for ALDH1A1 expression by immunoblotting. G, same as E, but examined for the indicated markers of apoptosis. H, ALDH1A1 protein expression in FACS-sorted ALDH activity–positive and negative cells determined by immunoblotting. I, sphere formation by the indicated ALDH activity–negative cells or ALDH activity–positive cells treated with or without JQ1. Scale bar, 40 μm. J, quantification of I. Mean of three independent experiments with SEM.

Close modal

The BET family is composed of BRD2, BRD3, BRD4, and the testis-specific BRDT proteins (12). BRD4 is often amplified in EOC (25). BRD4 amplification predicts a worse overall/disease-free survival in EOC patients (26). Consistent with a previous report (27), we showed that BRD4 is expressed in both EOC cell lines and primary high-grade serous EOC specimens, and BRD4 knockdown suppressed the growth of EOC cells (Supplementary Fig. S3A–S3D). Notably, cisplatin did not affect BRD4 expression (Supplementary Fig. S3E). BRD4 knockdown by multiple shRNAs and in multiple EOC cell lines suppressed ALDH1A1 expression and consequently decreased ALDH activity (Fig. 3A–D and Supplementary Fig. S3F–S3I). The oncogene c-MYC is a well-established target gene of BRD4 (27). JQ1-induced suppression of ALDH activity and ALDH1A1 downregulation is not a consequence of c-MYC downregulation, as c-MYC knockdown did not affect either ALDH1A1 expression or ALDH activity (Supplementary Fig. S3J and S3K). In contrast to BRD4 knockdown, knockdown of BRD2 or BRD3 did not suppress ALDH activity (Supplementary Fig. S3L and S3M), suggesting that BRD4 plays a major role in the observed suppression of ALDH activity by BET inhibitors, such as JQ1.

Figure 3.

BRD4 regulates ALDH1A1 expression and ALDH activity. A–C, OVCAR3 cells were infected with lentivirus encoding the indicated short hairpin RNA to the human BRD4 gene (shBRD4) or control. The drug-selected cells were examined for the expression of BRD4 (A) and ALDH1A1 (B) mRNA by qRT-PCR or for the expression of BRD4 and ALDH1A protein expression by immunoblotting (C). Mean of three independent experiments with SEM. *, P < 0.002. D, same as A, but examined for ALDH activity by FACS. The percentage of ALDH activity–positive cells is indicated.

Figure 3.

BRD4 regulates ALDH1A1 expression and ALDH activity. A–C, OVCAR3 cells were infected with lentivirus encoding the indicated short hairpin RNA to the human BRD4 gene (shBRD4) or control. The drug-selected cells were examined for the expression of BRD4 (A) and ALDH1A1 (B) mRNA by qRT-PCR or for the expression of BRD4 and ALDH1A protein expression by immunoblotting (C). Mean of three independent experiments with SEM. *, P < 0.002. D, same as A, but examined for ALDH activity by FACS. The percentage of ALDH activity–positive cells is indicated.

Close modal

BRD4 targets the promoters of stem-related genes

BRD4 transcriptionally regulates its target gene expression (12). BRD4 is also known to regulate lineage-specific gene expression through enhancer elements (28–30), which contributes to the observed specificity and selectivity of BET inhibitors. We determined whether the observed phenotypes induced by JQ1 are due to changes in BRD4 target gene expression. Nascent transcript RNA-seq in OVCAR3 cells treated with or without JQ1 for 40 minutes was performed to identify early changes in the gene expression that are likely directly dependent on BRD4 inhibition (Fig. 4A). In addition, BRD4 ChIP followed by next-generation sequencing (ChIP-seq) analysis was performed in OVCAR3 cells treated with or without JQ1 to identify genome-wide changes in BRD4 association induced by JQ1 (Fig. 4A and Supplementary Fig. S4A and S4B). The nascent RNA-seq and ChIP-seq data are available in the Gene Expression Omnibus database (accession number GSE77568). ChIP-seq analysis indicated that BRD4 predominantly occupied promoter regions within 1 kb from transcription starting sites (Supplementary Fig. S4A and S4B). Cross-referencing of the RNA-seq and BRD4 ChIP-seq revealed that BRD4 direct target genes regulated by JQ1 treatment are significantly enriched for putative stem-related genes (Fig. 4B–C; Supplementary Fig. S4C and S4D). We validated three genes known to be implicated in CSCs, namely LIF (31), HES1 (32), and WNT5A (33), as direct BRD4 targets that are downregulated by JQ1 (Fig. 4D–E and Supplementary Fig. S4E). This observation correlated with a decrease in the association of BRD4 and RNA polymerase II with the promoter regions of these genes after JQ1 treatment (Fig. 4F). These results support the notion that JQ1 may affect putative ovarian CSCs by regulating BRD4 binding to the promoters of the identified stem-related genes.

Figure 4.

JQ1 suppresses stem-related genes. A, diagram of the strategies used for identifying direct BRD4 target genes as an overlap between BRD4 differentially occupied genes and genes differentially expressed in response to JQ1. A total of 8,049 BRD4-binding sites showed significant reduction after JQ1 treatment (FDR < 1%). A total of 129 genes were significantly altered by JQ1 (FDR < 10%). Twenty-one of 44 direct BRD4 target genes whose expression was affected by JQ1 are stem related, which are all downregulated by JQ1. B, regulation and protein–protein interaction network for stem-related genes identified in A. C, enrichment analysis of direct BRD4 target genes shows a significant enrichment of stem-related genes among direct BRD4 target genes affected by JQ1 (21/51 affected stem-related genes were BRD4 direct targets compared with 282/1,196 of all genes changed, identified on the basis of the significance threshold of P < 0.05). D, BRD4 ChIP-seq and nascent RNA-seq tracks from control and JQ1-treated cells were aligned using bowtie and bowtie 2 algorithm. LIF, HES1, and WNT5A genomic locus ChIP-seq and nascent RNA-seq are displayed. E, validation of LIF, HES1, and WNT5A mRNA downregulation by JQ1. Relative mRNA expression level of the indicated stem-related genes was measured by qRT-PCR with or without 125 nmol/L JQ1 treatment for 24 hours. n = 3; *, P < 0.001. F, JQ1 reduces the association of BRD4 and Pol II with the promoters of the indicated stem-related genes. ChIP analysis of OVCAR3 cells treated with control vehicle or JQ1 (125 nmol/L) using antibodies against BRD4 or RNA Pol II for the human LIF, HES1, and WNT5A gene promoter. An isotype-matched IgG was used as a control (n = 3; *, P < 0.05). Error bars, SEM.

Figure 4.

JQ1 suppresses stem-related genes. A, diagram of the strategies used for identifying direct BRD4 target genes as an overlap between BRD4 differentially occupied genes and genes differentially expressed in response to JQ1. A total of 8,049 BRD4-binding sites showed significant reduction after JQ1 treatment (FDR < 1%). A total of 129 genes were significantly altered by JQ1 (FDR < 10%). Twenty-one of 44 direct BRD4 target genes whose expression was affected by JQ1 are stem related, which are all downregulated by JQ1. B, regulation and protein–protein interaction network for stem-related genes identified in A. C, enrichment analysis of direct BRD4 target genes shows a significant enrichment of stem-related genes among direct BRD4 target genes affected by JQ1 (21/51 affected stem-related genes were BRD4 direct targets compared with 282/1,196 of all genes changed, identified on the basis of the significance threshold of P < 0.05). D, BRD4 ChIP-seq and nascent RNA-seq tracks from control and JQ1-treated cells were aligned using bowtie and bowtie 2 algorithm. LIF, HES1, and WNT5A genomic locus ChIP-seq and nascent RNA-seq are displayed. E, validation of LIF, HES1, and WNT5A mRNA downregulation by JQ1. Relative mRNA expression level of the indicated stem-related genes was measured by qRT-PCR with or without 125 nmol/L JQ1 treatment for 24 hours. n = 3; *, P < 0.001. F, JQ1 reduces the association of BRD4 and Pol II with the promoters of the indicated stem-related genes. ChIP analysis of OVCAR3 cells treated with control vehicle or JQ1 (125 nmol/L) using antibodies against BRD4 or RNA Pol II for the human LIF, HES1, and WNT5A gene promoter. An isotype-matched IgG was used as a control (n = 3; *, P < 0.05). Error bars, SEM.

Close modal

BRD4 regulates ALDH1A1 expression through a super-enhancer element and its associated eRNA

JQ1 decreased ALDH1A1 mRNA expression (Fig. 1). However, JQ1 did not affect BRD4 binding to the ALDH1A1 promoter region (Supplementary Fig. S5A). This suggests that JQ1 regulates ALDH1A transcription through a distal regulatory element. As BRD4 is known to regulate super-enhancer elements (28–30), we examined the role of JQ1 in regulating super-enhancers through BRD4 (Fig. 5A). To do so, we focused on BRD4-binding regions that were enriched at least 4-fold compared with input DNA and were significantly reduced by JQ1 treatment (>2-fold, P < 0.05). In addition, we prioritized the list by focusing on BRD4-binding regions that spanned >10 kb and with no known genes located within at least 100 kb. We then overlapped these regions with ENCODE ChIP-seq data and considered only BRD4-binding regions that overlapped with the enhancer H3K4Me1/H3K27Ac histone marks. Our prioritization resulted in a list of 11 candidate BRD4-binding sites similar to those previously described for super-enhancers (Supplementary Table S3; ref. 28). Interestingly, one of the potential super-enhancers is 491 kb upstream of the ALDH1A1 gene (Supplementary Table S3). On the basis of RNA-seq analysis, this region is bidirectionally transcribed into RNA albeit with low reads (Fig. 5B), which is also a known feature of super-enhancers (34, 35). We validated that JQ1 treatment decreased the expression of the RNA transcribed from the super-enhancer element (eRNA; Fig. 5C). This decrease in eRNA expression significantly correlated with the decrease in ALDH1A1 mRNA expression (Fig. 5C). We validated that the super-enhancer region is enriched in BRD4 and Pol II binding, another feature of super-enhancers (Fig. 5D; ref. 35). Furthermore, we validated the enrichment of H3K27Ac and H3K4Me1 epigenetic histone modifications in the putative super-enhancer regions (Fig. 5E). Finally, knockdown of BRD4 expression was sufficient to decrease the eRNA expression, which correlated with the decrease in ALDH1A1 mRNA (Fig. 5F). To directly determine whether the eRNA regulates ALDH1A1 mRNA expression, we knocked down the eRNA expression using siRNAs (36, 37). Knockdown of the eRNA downregulated ALDH1A1 mRNA expression (Fig. 5G), which correlated with a decrease in ALDH activity (Fig. 5H). Notably, there was a significant positive correlation between BRD4, eRNA, and ALDH1A1 expression in a panel of 26 cases of HGSC specimens (Fig. 5I–K). This further highlights the established regulation of eRNA by BRD4 and subsequent ALDH1A1 expression by eRNA.

Figure 5.

BRD4 regulates ALDH1A1 expression through a super-enhancer and its associated enhancer RNA. A, flow diagram of the strategies used for identifying the putative super-enhancer for ALDH1A1 gene. B, BRD4 ChIP-seq and nascent RNA-seq tracks from control and JQ1-treated cells were aligned. The putative super-enhancer loci are displayed together with enhancer histone marks H3K4Me1 and H3K27Ac tracks from ENCODE database from the indicated cell lines. C, validation of downregulation of the eRNA and ALDH1A1 mRNA by JQ1. OVCAR3 cells were treated with or without 125 nmol/L JQ1 for 24 hours, and the expression of ALDH1A1 mRNA and the eRNA expression was determined by qRT-PCR. D, same as C, but validated for a decrease in the association of BRD4 and Pol II with the enhancer locus by ChIP analysis. An isotype-matched IgG was used as a control. n = 3; *, P < 0.0001. E, validation of H3K27Ac and H3K4Me1 enhancer histone marks’ association with the enhancer loci by ChIP analysis. n = 3; *, P < 0.002. F, BRD4 knockdown reduces the levels of eRNA expression and suppresses ALDH1A1 expression. OVCAR3 cells were infected with lentivirus encoding the indicated shBRD4 or control. Drug-selected cells were examined for the expression of BRD4 mRNA, eRNA, and ALDH1A1 mRNA by qRT-PCR. n = 3; *, P < 0.0001. G, knockdown of the eRNA suppresses ALDH1A1 expression. OVCAR3 cells were transfected with two independent siRNAs to the eRNA for 72 hours, and expression of the eRNA and ALDH1A1 mRNA was determined by qRT-PCR. n = 3; *, P < 0.002. H, same as G, but examined for ALDH activity. The percentage of ALDH-positive cells is indicated. I–K, positive correlation between BRD4 and eRNA (I), between eRNA and ALDH1A1 (J), or between BRD4 and ALDH1A1 (K) in a panel of 26 cases of HGSOC. Expression of BRD4, eRNA, and ALDH1A1 was determined by qRT-PCR, and correlation was determined by Spearman statistical analysis. L, diagrams of ALDH1A1 genomic regions with its enhancer (black box). Arrowheads, position of primers used for detection of chromatin looping; stick bars, Mbo1 enzyme digestion sites (a–h). Constant primer at the anchor point is also indicated. TSS, ALDH1A1 gene transcription-starting site. M, 3C-quantitative PCR analysis of the looping events between the enhancer and the ALDH1A1 promoter region were detected at f and g sites, which were reduced by JQ1 (125 nmol/L) treatment for 24 hours. The relative cross-linking frequency was normalized to the closest Mbo1 digestion site E1. x-axis, distance from ALDH1A1 transcription start site (TSS).

Figure 5.

BRD4 regulates ALDH1A1 expression through a super-enhancer and its associated enhancer RNA. A, flow diagram of the strategies used for identifying the putative super-enhancer for ALDH1A1 gene. B, BRD4 ChIP-seq and nascent RNA-seq tracks from control and JQ1-treated cells were aligned. The putative super-enhancer loci are displayed together with enhancer histone marks H3K4Me1 and H3K27Ac tracks from ENCODE database from the indicated cell lines. C, validation of downregulation of the eRNA and ALDH1A1 mRNA by JQ1. OVCAR3 cells were treated with or without 125 nmol/L JQ1 for 24 hours, and the expression of ALDH1A1 mRNA and the eRNA expression was determined by qRT-PCR. D, same as C, but validated for a decrease in the association of BRD4 and Pol II with the enhancer locus by ChIP analysis. An isotype-matched IgG was used as a control. n = 3; *, P < 0.0001. E, validation of H3K27Ac and H3K4Me1 enhancer histone marks’ association with the enhancer loci by ChIP analysis. n = 3; *, P < 0.002. F, BRD4 knockdown reduces the levels of eRNA expression and suppresses ALDH1A1 expression. OVCAR3 cells were infected with lentivirus encoding the indicated shBRD4 or control. Drug-selected cells were examined for the expression of BRD4 mRNA, eRNA, and ALDH1A1 mRNA by qRT-PCR. n = 3; *, P < 0.0001. G, knockdown of the eRNA suppresses ALDH1A1 expression. OVCAR3 cells were transfected with two independent siRNAs to the eRNA for 72 hours, and expression of the eRNA and ALDH1A1 mRNA was determined by qRT-PCR. n = 3; *, P < 0.002. H, same as G, but examined for ALDH activity. The percentage of ALDH-positive cells is indicated. I–K, positive correlation between BRD4 and eRNA (I), between eRNA and ALDH1A1 (J), or between BRD4 and ALDH1A1 (K) in a panel of 26 cases of HGSOC. Expression of BRD4, eRNA, and ALDH1A1 was determined by qRT-PCR, and correlation was determined by Spearman statistical analysis. L, diagrams of ALDH1A1 genomic regions with its enhancer (black box). Arrowheads, position of primers used for detection of chromatin looping; stick bars, Mbo1 enzyme digestion sites (a–h). Constant primer at the anchor point is also indicated. TSS, ALDH1A1 gene transcription-starting site. M, 3C-quantitative PCR analysis of the looping events between the enhancer and the ALDH1A1 promoter region were detected at f and g sites, which were reduced by JQ1 (125 nmol/L) treatment for 24 hours. The relative cross-linking frequency was normalized to the closest Mbo1 digestion site E1. x-axis, distance from ALDH1A1 transcription start site (TSS).

Close modal

An important component of enhancer function is the formation of chromatin looping, allowing enhancer and promoter interaction (36, 38, 39). We directly examined chromatin looping between the super-enhancer and ALDH1A1 gene promoter using 3C in cells with or without JQ1 treatment. We observed a robust association between the super-enhancer and the promoter region of the ALDH1A1 gene (Fig. 5L and M). Remarkably, JQ1 treatment abrogated the chromatin looping between the super-enhancer and the promoter of ALDH1A1 gene (Fig. 5M). These results support the notion that JQ1 regulates transcription of ALDH1A1 through the newly identified super-enhancer.

JQ1 inhibits expression of ALDH1A1 and its associated eRNA induced by cisplatin in vivo and combination of JQ1 and cisplatin improves survival

BET inhibitors have been proven safe in patients (40). ALDH-positive cells contribute to tumor progression and relapse after initial response to chemotherapy (3, 8). To determine the effects of BET inhibitor on tumor relapse after cisplatin treatment, we orthotopically transplanted luciferase-expressing OVCAR3 cells into the peritoneal cavity of immunocompromised NSG female mice. The injected cells were allowed to grow for 3 weeks to establish tumors. We randomly assigned mice into four groups and treated mice with vehicle control (n = 12), cisplatin (750 μg/kg every 2 weeks, n = 12), JQ1 (20 mg/kg daily, n = 11), and a combination of cisplatin and JQ1 (n = 13) by intraperitoneal injection for an additional 4 weeks. Doses of JQ1 and cisplatin used were determined on the basis of suppression of ALDH1A1 expression by JQ1 and regression of ovarian tumor in a pilot experiment (Supplementary Fig. S6A–S6E). Notably, the survival of the combination-treated mice was significantly extended compared with mice treated with cisplatin alone (Fig. 6A). We followed the tumor outgrowth/relapse in mice treated with cisplatin with or without JQ1 combination after stopping drug treatment. Indeed, the outgrowth of the tumors in the combination treatment group was significantly slower compared with the cisplatin only treatment group (Fig. 6B). RNA from tumors harvested from the control and the three different treatment groups was utilized for qRT-PCR analysis. We observed that eRNA expression was induced by cisplatin alone (Fig. 6C), whereas JQ1 treatment suppressed the cisplatin-induced eRNA expression (Fig. 6C). This correlated with changes in ALDH1A1 expression in these treatment groups (Fig. 6D and Supplementary Fig. S6F). ALDH1A1 mRNA expression was also significantly downregulated in JQ1-treated tumors (Supplementary Fig. S6G). In addition, stem-related genes, such as LIF and WNT5A, were downregulated in JQ1/cisplatin-treated tumors (Supplementary Fig. S6H). However, BRD4 expression was not significantly changed in treatment groups (Supplementary Fig. S6I). Together, we conclude that a combination of JQ1 and cisplatin improves the survival of EOC-bearing mice, which correlates with the suppression of expression of ALDH1A1 and its eRNA.

Figure 6.

The combination of JQ1 and cisplatin improves survival of tumor-bearing mice. A, combination of JQ1 and cisplatin improves survival of tumor-bearing mice. Kaplan–Meier survival curves of mice in the indicated groups posttreatment were plotted using Prism software. n = 7 for the indicated groups except in the combination group, n = 8. B, quantification of tumor growth in the indicated groups after stopping drug treatment. C, same as A. Tumors from the indicated treatment groups were examined for eRNA expression by qRT-PCR at the end of the treatment. D, same as C. Tumors were sectioned and subjected to immunohistochemical staining using antibodies against ALDH1A1. Scale bar, 100 μm. E, a model for the mechanism underlying the observed synergy between BET inhibitor and cisplatin.

Figure 6.

The combination of JQ1 and cisplatin improves survival of tumor-bearing mice. A, combination of JQ1 and cisplatin improves survival of tumor-bearing mice. Kaplan–Meier survival curves of mice in the indicated groups posttreatment were plotted using Prism software. n = 7 for the indicated groups except in the combination group, n = 8. B, quantification of tumor growth in the indicated groups after stopping drug treatment. C, same as A. Tumors from the indicated treatment groups were examined for eRNA expression by qRT-PCR at the end of the treatment. D, same as C. Tumors were sectioned and subjected to immunohistochemical staining using antibodies against ALDH1A1. Scale bar, 100 μm. E, a model for the mechanism underlying the observed synergy between BET inhibitor and cisplatin.

Close modal

Here, we found that BET inhibitors suppress ALDH activity. This correlates with the suppression of ALDH1A1 expression by a BRD4-regulated super-enhancer and downregulation of its encoded eRNA (Fig. 6E). BET inhibitors are now in clinical development and are safe. This suggests that BET inhibitors can be repurposed to target ALDH activity for improving platinum-based chemotherapy by inhibiting tumor relapse, a major challenge in the clinical management of EOC. Notably, BRD4 amplification/overexpression is often mutually exclusive with “BRCAness” in EOC (25). Thus, there is an even greater need for novel therapeutic strategies for this patient population given the limited therapeutic options available (2). Our experiments clearly show that BET inhibitors, an existing class of epigenetic targeting drugs, target ALDH activity, potentiate the tumor suppression induced by cisplatin, and improve survival of EOC-bearing mice in vivo. These findings will facilitate the rapid evaluation of this new strategy in the clinic for EOC.

BRD4 is a general transcriptional regulator that controls global gene expression patterns (12). Investigation of genes hypersensitive to BET inhibition revealed that such genes typically exhibit BRD4 occupancy at super-enhancer elements (28–30). This raises the possibility that BET inhibition is selective in gene regulation and thus confers relative specificity in a cell context–dependent manner. Our findings revealed that in response to JQ1 treatment, BRD4 assumes a key role in transcriptional control of the ALDH1A1 gene through regulating its super-enhancer and the associated eRNA. Although BRD4 plays a key role in regulating ALDH1A1 transcription, there are potentially other mechanisms than BRD4 expression levels that regulate ALDH1A1 expression (12). BRD4 plays a key role in CSCs by selectively regulating the ALDH1A1 super-enhancer. In this context, BET inhibitors may selectively target CSCs by their effect on the ALDH1A1 super-enhancer. In addition to suppressing ALDH1A1 expression and ALDH activity, JQ1 also directly suppresses the expression of stem-related genes through reducing BRD4′s association with their promoters (Fig. 4). Thus, the mode of action of BET inhibitors is multifaceted and likely involves a broad range of changes in transcription and the associated signaling pathways (Fig. 6E). Given the established role of ALDH1A1 in ovarian CSCs (4), our data support the idea that the BRD4-regulated ALDH1A1 super-enhancer plays a key role in the observed phenotypes induced by BET inhibitors.

Our studies demonstrate that targeting BRD4 activity through the use of clinically applicably BET inhibitors represents a novel strategy for targeting ALDH activity. This correlates with suppression of ALDH1A1 expression via a BRD4-regulated super-enhancer and its associated eRNA. Given that there is currently no clinically applicable ALDH activity inhibitor, we expect our finding to have far-reaching implications for developing future therapeutic strategies using epigenetic targeting BET inhibitors in cancers such as EOC.

J.E. Bradner is the president at Novartis Institute of BioMedical Research. No potential conflicts of interest were disclosed by the other authors.

Conception and design: Y. Yokoyama, H. Zhu, S.Y. Wu, A. Gardini, B.G. Bitler, R. Zhang

Development of methodology: Y. Yokoyama, H. Zhu, K.C. Palozola, J.E. Bradner, A.K. Sood, T. Ordog, B.G. Bitler

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Yokoyama, H. Zhu, J.H. Lee, S.Y. Wu, L.C. Showe, K.S. Zaret, A.K. Sood, T. Ordog, B.G. Bitler

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Yokoyama, H. Zhu, A.V. Kossenkov, J.M. Wickramasinghe, X. Yin, A. Gardini, L.C. Showe, Q. Liu, D. Speicher, J.R. Conejo-Garcia, A.K. Sood, B.G. Bitler

Writing, review, and/or revision of the manuscript: Y. Yokoyama, H. Zhu, A.V. Kossenkov, S.Y. Wu, L.C. Showe, Q. Liu, D. Speicher, J.R. Conejo-Garcia, A.K. Sood, T. Ordog, B.G. Bitler, R. Zhang

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): H. Zhu, K.S. Zaret, B.G. Bitler

Study supervision: Z. Zhang, B.G. Bitler, R. Zhang

Other (performed Brd4 ChIP-seq experiments and read the manuscript): Z. Zhang

We thank The Structure Genomics Consortium for providing the epigenetic inhibitors library. We also thank Drs. Gerd Blobel and Katherine Aird for critical comments.

This work was supported by NIH/NCI grants (R01CA163377 and R01CA202919 to R. Zhang; and CA083639 to A.K. Sood), U.S. Department of Defense (OC140632P1 and OC150446 to R. Zhang), an Ovarian Cancer Research Fund (OCRF) program project (R. Zhang), and The Jayne Koskinas & Ted Giovanis Breast Cancer Research Consortium at Wistar (R. Zhang). H. Zhu is an OCRF Ann Schreiber Mentored Investigator (372953). B.G. Bitler is supported by an NIH/NCI grant (K99CA194318). S.Y. Wu is supported by the OCRF, Foundation for Women's Cancer, and by Cancer Prevention and Research Institute of Texas training grants (RP101502 and RP101489). Support of Core Facilities was provided by Cancer Center Support Grant (CCSG) CA010815 to The Wistar Institute.

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

1.
Martin
LP
,
Hamilton
TC
,
Schilder
RJ
. 
Platinum resistance: the role of DNA repair pathways
.
Clin Cancer Res
2008
;
14
:
1291
5
.
2.
Bowtell
DD
,
Bohm
S
,
Ahmed
AA
,
Aspuria
PJ
,
Bast
RC
 Jr
,
Beral
V
, et al
Rethinking ovarian cancer II: reducing mortality from high-grade serous ovarian cancer
.
Nat Rev Cancer
2015
;
15
:
668
79
.
3.
Chen
J
,
Li
Y
,
Yu
TS
,
McKay
RM
,
Burns
DK
,
Kernie
SG
, et al
A restricted cell population propagates glioblastoma growth after chemotherapy
.
Nature
2012
;
488
:
522
6
.
4.
Landen
CN
 Jr
,
Goodman
B
,
Katre
AA
,
Steg
AD
,
Nick
AM
,
Stone
RL
, et al
Targeting aldehyde dehydrogenase cancer stem cells in ovarian cancer
.
Mol Cancer Ther
2010
;
9
:
3186
99
.
5.
Steg
AD
,
Bevis
KS
,
Katre
AA
,
Ziebarth
A
,
Dobbin
ZC
,
Alvarez
RD
, et al
Stem cell pathways contribute to clinical chemoresistance in ovarian cancer
.
Clin Cancer Res
2012
;
18
:
869
81
.
6.
Choi
YJ
,
Ingram
PN
,
Yang
K
,
Coffman
L
,
Iyengar
M
,
Bai
S
, et al
Identifying an ovarian cancer cell hierarchy regulated by bone morphogenetic protein 2
.
Proc Natl Acad Sci U S A
2015
;
112
:
E6882
8
.
7.
Flesken-Nikitin
A
,
Hwang
CI
,
Cheng
CY
,
Michurina
TV
,
Enikolopov
G
,
Nikitin
AY
. 
Ovarian surface epithelium at the junction area contains a cancer-prone stem cell niche
.
Nature
2013
;
495
:
241
5
.
8.
Tomita
H
,
Tanaka
K
,
Tanaka
T
,
Hara
A
. 
Aldehyde dehydrogenase 1A1 in stem cells and cancer
.
Oncotarget
2016
;
7
:
11018
32
.
9.
Garraway
LA
,
Lander
ES
. 
Lessons from the cancer genome
.
Cell
2013
;
153
:
17
37
.
10.
Lawrence
MS
,
Stojanov
P
,
Mermel
CH
,
Robinson
JT
,
Garraway
LA
,
Golub
TR
, et al
Discovery and saturation analysis of cancer genes across 21 tumour types
.
Nature
2014
;
505
:
495
501
.
11.
Huston
A
,
Arrowsmith
CH
,
Knapp
S
,
Schapira
M
. 
Probing the epigenome
.
Nat Chem Biol
2015
;
11
:
542
5
.
12.
Shi
J
,
Vakoc
CR
. 
The mechanisms behind the therapeutic activity of BET bromodomain inhibition
.
Mol Cell
2014
;
54
:
728
36
.
13.
Filippakopoulos
P
,
Knapp
S
. 
Targeting bromodomains: epigenetic readers of lysine acetylation
.
Nat Rev Drug Discov
2014
;
13
:
337
56
.
14.
Matei
D
,
Fang
F
,
Shen
C
,
Schilder
J
,
Arnold
A
,
Zeng
Y
, et al
Epigenetic resensitization to platinum in ovarian cancer
.
Cancer Res
2012
;
72
:
2197
205
.
15.
Brown
R
,
Curry
E
,
Magnani
L
,
Wilhelm-Benartzi
CS
,
Borley
J
. 
Poised epigenetic states and acquired drug resistance in cancer
.
Nat Rev Cancer
2014
;
14
:
747
53
.
16.
Li
H
,
Cai
Q
,
Godwin
AK
,
Zhang
R
. 
Enhancer of zeste homolog 2 promotes the proliferation and invasion of epithelial ovarian cancer cells
.
Mol Cancer Res
2010
;
8
:
1610
8
.
17.
Rao
SS
,
Huntley
MH
,
Durand
NC
,
Stamenova
EK
,
Bochkov
ID
,
Robinson
JT
, et al
A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping
.
Cell
2014
;
159
:
1665
80
.
18.
Hagege
H
,
Klous
P
,
Braem
C
,
Splinter
E
,
Dekker
J
,
Cathala
G
, et al
Quantitative analysis of chromosome conformation capture assays (3C-qPCR)
.
Nat Protoc
2007
;
2
:
1722
33
.
19.
Bitler
BG
,
Aird
KM
,
Garipov
A
,
Li
H
,
Amatangelo
M
,
Kossenkov
AV
, et al
Synthetic lethality by targeting EZH2 methyltransferase activity in ARID1A-mutated cancers
.
Nat Med
2015
;
21
:
231
8
.
20.
Filippakopoulos
P
,
Qi
J
,
Picaud
S
,
Shen
Y
,
Smith
WB
,
Fedorov
O
, et al
Selective inhibition of BET bromodomains
.
Nature
2010
;
468
:
1067
73
.
21.
Chou
TC
. 
Drug combination studies and their synergy quantification using the Chou-Talalay method
.
Cancer Res
2010
;
70
:
440
6
.
22.
Domcke
S
,
Sinha
R
,
Levine
DA
,
Sander
C
,
Schultz
N
. 
Evaluating cell lines as tumour models by comparison of genomic profiles
.
Nat Commun
2013
;
4
:
2126
.
23.
Debeb
BG
,
Lacerda
L
,
Xu
W
,
Larson
R
,
Solley
T
,
Atkinson
R
, et al
Histone deacetylase inhibitors stimulate dedifferentiation of human breast cancer cells through WNT/beta-catenin signaling
.
Stem Cells
2012
;
30
:
2366
77
.
24.
Wang
Y
,
Cardenas
H
,
Fang
F
,
Condello
S
,
Taverna
P
,
Segar
M
, et al
Epigenetic targeting of ovarian cancer stem cells
.
Cancer Res
2014
;
74
:
4922
36
.
25.
Goundiam
O
,
Gestraud
P
,
Popova
T
,
De la Motte Rouge
T
,
Fourchotte
V
,
Gentien
D
, et al
Histo-genomic stratification reveals the frequent amplification/overexpression of CCNE1 and BRD4 genes in non-BRCAness high grade ovarian carcinoma
.
Int J Cancer
2015
;
137
:
1890
900
.
26.
Zhang
Z
,
Ma
P
,
Jing
Y
,
Yan
Y
,
Cai
MC
,
Zhang
M
, et al
BET bromodomain inhibition as a therapeutic strategy in ovarian cancer by downregulating FoxM1
.
Theranostics
2016
;
6
:
219
30
.
27.
Baratta
MG
,
Schinzel
AC
,
Zwang
Y
,
Bandopadhayay
P
,
Bowman-Colin
C
,
Kutt
J
, et al
An in-tumor genetic screen reveals that the BET bromodomain protein, BRD4, is a potential therapeutic target in ovarian carcinoma
.
Proc Natl Acad Sci U S A
2015
;
112
:
232
7
.
28.
Loven
J
,
Hoke
HA
,
Lin
CY
,
Lau
A
,
Orlando
DA
,
Vakoc
CR
, et al
Selective inhibition of tumor oncogenes by disruption of super-enhancers
.
Cell
2013
;
153
:
320
34
.
29.
Whyte
WA
,
Orlando
DA
,
Hnisz
D
,
Abraham
BJ
,
Lin
CY
,
Kagey
MH
, et al
Master transcription factors and mediator establish super-enhancers at key cell identity genes
.
Cell
2013
;
153
:
307
19
.
30.
Hnisz
D
,
Abraham
BJ
,
Lee
TI
,
Lau
A
,
Saint-Andre
V
,
Sigova
AA
, et al
Super-enhancers in the control of cell identity and disease
.
Cell
2013
;
155
:
934
47
.
31.
Penuelas
S
,
Anido
J
,
Prieto-Sanchez
RM
,
Folch
G
,
Barba
I
,
Cuartas
I
, et al
TGF-beta increases glioma-initiating cell self-renewal through the induction of LIF in human glioblastoma
.
Cancer Cell
2009
;
15
:
315
27
.
32.
Liu
ZH
,
Dai
XM
,
Du
B
. 
Hes1: a key role in stemness, metastasis and multidrug resistance
.
Cancer Biol Ther
2015
;
16
:
353
9
.
33.
Povinelli
BJ
,
Nemeth
MJ
. 
Wnt5a regulates hematopoietic stem cell proliferation and repopulation through the Ryk receptor
.
Stem Cells
2014
;
32
:
105
15
.
34.
Kim
TK
,
Hemberg
M
,
Gray
JM
,
Costa
AM
,
Bear
DM
,
Wu
J
, et al
Widespread transcription at neuronal activity-regulated enhancers
.
Nature
2010
;
465
:
182
7
.
35.
Pott
S
,
Lieb
JD
. 
What are super-enhancers?
Nat Genet
2015
;
47
:
8
12
.
36.
Melo
CA
,
Drost
J
,
Wijchers
PJ
,
van de Werken
H
,
de Wit
E
,
Oude Vrielink
JA
, et al
eRNAs are required for p53-dependent enhancer activity and gene transcription
.
Mol Cell
2013
;
49
:
524
35
.
37.
Schaukowitch
K
,
Joo
JY
,
Liu
X
,
Watts
JK
,
Martinez
C
,
Kim
TK
. 
Enhancer RNA facilitates NELF release from immediate early genes
.
Mol Cell
2014
;
56
:
29
42
.
38.
Mousavi
K
,
Zare
H
,
Dell'orso
S
,
Grontved
L
,
Gutierrez-Cruz
G
,
Derfoul
A
, et al
eRNAs promote transcription by establishing chromatin accessibility at defined genomic loci
.
Mol Cell
2013
;
51
:
606
17
.
39.
Sanyal
A
,
Lajoie
BR
,
Jain
G
,
Dekker
J
. 
The long-range interaction landscape of gene promoters
.
Nature
2012
;
489
:
109
13
.
40.
Herait
PE
,
Berthon
C
,
Thieblemont
C
,
Raffoux
E
,
Magarotto
V
,
Stathis
A
, et al
BET-bromodomain inhibitor OTX015 shows clinically meaningful activity at nontoxic doses: interim results of an ongoing phase I trial in hematologic malignancies
. In:
Proceedings of the 105th Annual Meeting of the American Association for Cancer Research
; 
2014
Apr
5
9
;
San Diego, CA
.
Philadelphia (PA)
:
AACR
; 
2014
.
Abstract nr CT231
.