Inhibition of EZH2 suppresses self-renewal and induces radiation sensitivity in atypical rhabdoid teratoid tumor cells

Abstract

Introduction. Overexpression of the Polycomb repressive complex 2 (PRC2) subunit Enhancer of Zeste 2 (EZH2) occurs in several malignancies, including prostate cancer, breast cancer, medulloblastoma, and glioblastoma multiforme. Recent evidence suggests that EZH2 may also have a role in rhabdoid tumors. Atypical teratoid/rhabdoid tumor (ATRT) is a rare, high-grade embryonal brain tumor that occurs most commonly in young children and carries a very poor prognosis. ATRTs are characterized by absence of the chromatin remodeling protein SMARCB1. Given the role of EZH2 in regulating epigenetic changes, we investigated the role of EZH2 in ATRT.

Methods. Microarray analysis was used to evaluate expression of EZH2 in ATRT tumor samples. We used shRNA and a chemical inhibitor of EZH2 to examine the impact of EZH2 inhibition on cell growth, proliferation, and tumor cell self-renewal.

Results. Here, we show that targeted disruption of EZH2 by RNAi or pharmacologic inhibition strongly impairs ATRT cell growth, suppresses tumor cell self-renewal, induces apoptosis, and potently sensitizes these cells to radiation. Using functional analysis of transcription factor activity, we found the cyclin D1-E2F axis to be repressed after EZH2 depletion in ATRT cells.

Conclusions. Our observations provide evidence that EZH2 disruption alters cell cycle progression and may be an important new therapeutic target, particularly in combination with radiation, in ATRT.

Atypical teratoid/rhabdoid tumors (ATRT) are aggressive and highly malignant embryonal tumors of the central nervous system (CNS) that usually occur in children <3 years of age.^1–3 The overall survival among patients with ATRT is poor, compared with the survival among patients with other pediatric cancers. Patients with pediatric CNS tumors have a 5-year overall survival of 60%–70%; however, ATRT survival is historically only an estimated 20%–25%.⁴ More recently, therapy has included either high-dose chemotherapy with autologous stem cell rescue or use of a sarcoma-based intensive regimen.^5,6 Despite a small increase in overall survival with these approaches, therapy-related toxicity remains a critical problem in this age group. ATRT is often associated with a mutation or deletion of the SMARCB1/hSNF5/INI1 gene found on chromosome 22q11.2.^7,8 SMARCB1 is a tumor suppressor, and its absence results in cyclin D1 overexpression and aberrant cell cycle progression.⁹ Because SMARCB1 is a member of the SWI/SNF chromatin remodeling complex, therapeutic agents that alter the epigenome and chromatin structure may prove to be effective in ATRT.

Enhancer of Zeste 2 (EZH2), the catalytic subunit of the Polycomb repressive complex 2 (PRC2), is highly expressed in different types of cancer, including lymphomas, glioblastoma multiforme (GBM), and ovarian, breast, and metastatic prostate cancers.^10–16 In these tumors, EZH2 is associated with tumor progression, invasive growth, and poor prognosis.¹⁴ EZH2 is responsible for the trimethylation of lysine 27 of histone H3 (H3K27me3), a repressive chromatin mark.^17–21 Given that EZH2 increases tumor cell proliferation and is associated with aggressive subgroups in several cancers, it may be a candidate for targeted therapy.²² In addition, the development of drugs that target epigenetic processes is a potential novel approach for cancer therapy.²³ To that end, a novel chemical inhibitor, 3-deazaneplanocin A (DZNep), has been found to suppress EZH2-mediated H3K27 methylation.²⁴ DZNep is one of the most potent S-adenosylhomocysteine hydrolase inhibitors. It depletes EZH2 indirectly by increasing levels of adenosylhomocysteine, which in turn, inhibits EZH2 and its methyltransferase activity.²⁵

Although the mechanism of action of EZH2 has been evaluated in several malignancies, its role in ATRT is not yet clear. Previous studies have assessed the relationship between SMARCB1 loss, increased EZH2 expression, and PcG signatures but have not evaluated the biological and molecular activity of DZNep on ATRT.²⁶ Here, we show that inhibition of EZH2 by RNAi or chemical inhibition with DZNep potently suppresses ATRT cell growth, leads to cell cycle alterations, increases apoptosis, potentiates the effect of radiation, decreases key cellular signal transduction pathways, and inhibits tumor sphere formation.

Materials and Methods

ATRT Patient Tumor Samples and Gene Expression Analysis

Primary ATRT patient tumor samples were collected for gene expression analysis as previously described.²⁷ In brief, samples were collected at the time of surgery and snap-frozen in liquid nitrogen. All tissue samples were obtained under an institutional review board–approved protocol (COMIRB 95–500). ATRT samples were obtained from patients <16 years of age. Control cerebellum samples (age, 7–16 years) were obtained from the Pediatric Co-operative Human Tissue Network (Columbus, OH). RNA was extracted from each sample using the RNeasy kit (Qiagen) and hybridized to HG-U133 Plus 2.0 GeneChips (Affymetrix) according to the manufacturer's instructions. Data analysis was performed in R (http://www.r-project.org), using packages publicly available through Bioconductor (http://www.bioconductor.org). The scanned microarray data were background-corrected and normalized using the gcRMA algorithm, resulting in log₂ gene expression values. Multiple probe sets for a gene were then collapsed to 1 entry per gene based on the mean best-expressed probe set for that gene.

Gene set enrichment analysis (GSEA)²⁸ was performed on the microarray data from 18 ATRT patient samples to evaluate for enrichment of Polycomb repressive complex 2 (PRC2)- and EZH2-responsive genes as defined by 3 recent studies. In brief, GSEA takes gene expression profiles that have been assigned a specific phenotype and creates a ranked list of genes based on the strength of association with the phenotype being interrogated.

Reagents and Cell Lines

DZNep was kindly provided by Chemical Biology Laboratory, Center for Cancer Research, National Cancer Institute at Frederick (National Institutes of Health, Frederick, MD). Antibodies used for Western blot analysis were from the following sources: histone H3-trimethyl-Lys27 (#39156, Active Motif), CDK6 (#3136, Cell Signaling), RB (4H1) (#9309, Cell Signaling), phospho-RB (Ser780) (#9307, Cell Signaling), cyclin D1 (sc-450, Santa Cruz Biotechnology), c-Myc (ab32, Abcam), actin (#MAB 1501, Chemicon International), and INI-1/anti-BAF47 (BD Biosciences).

BT12 and BT16 ATRT cell lines were kindly provided by Dr. Peter Houghton's laboratory (St. Jude Children's Research Hospital). The short-term ATRT cell culture (UPN 737) was established from a surgical sample obtained from Children's Hospital Colorado under an approved institutional review board protocol, as previously noted. Primary cell culture was generated as previously described.²⁹ Cell lines were cultured in RPMI medium (Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco) according to the supplier's recommendations. Primary cell cultures were maintained at low passage number (p2–p4) in RPMI–1640 supplemented with 20% FBS. Cells were propagated at 37°C at 5% CO₂ with humidity.

Real-Time Quantitative Reverse-transcriptase Polymerase Chain Reaction (PCR)

cDNA was generated according to the manufacturer's instructions with use of reverse transcription using mRNA and the Mastercycler personal (Eppendorf) thermal cycler. Real-time quantitative PCR was performed with the StepOne Plus detection system (Applied Biosystems). Taqman reagents (Applied Biosystems) were used according to the manufacturer's recommendations. Gene expression was determined by the ΔΔCt method. All assays were performed in triplicate. The EZH2 Taqman probe sequence was GAGAGATTATTTCTCAAGATGAAGCTGACAGAA GAGGGAAAGTGTATGATAAATACATGTGCAGCTTTCTGTTCAACTTGAACAATGATTTTGTGGTGGATGCAACCCGCAAGGGTAACAAAATTCGTT. The miR 101 primer-probe sequences was UGCC CUGGCUCAGUUAUCACAGUGCUGAUGCUGUCUAUUCUAAAGGUACAGUACUGUGAUAACUGAAGGAUGGCA.

Western Blotting

Protein expression levels were determined by Western blotting. Cells were lysed in RIPA buffer (Thermo Scientific) containing protease inhibitor cocktail (Roche), 1 mM sodium vanadate, and 0.1 mM sodium molybdate. The lysates were centrifuged for 20 min, and the supernatants were collected for protein concentration determination with Bradford reagent (Sigma). Equal amounts of cell lysates were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and Western blot analysis was performed with specific antibodies as described above.

Transfection of ATRT Cells With shEZH2

ATRT cells were transfected with control PSIF and shEZH2 plasmids (Sigma) in 6-well plates using Lipofectamine 2000 (Invitrogen). One microgram of shRNA for a 6-well plate was transfected into the medulloblastoma cell lines. The ratio used for the forward transfections was 1 µg of shRNA DNA:2 μL of Lipofectamine 2000 Transfection Reagent.

Clonogenic Assay

Control, 0.5 µM DZNep-treated, or shEZH2-transfected cells were seeded into 6-well plates in triplicate at a density of 500 cells/well in 3 mL medium containing 10% FBS. Colony growth proceeded for 10 days. Subsequently, the colonies were stained for 15 min with a solution containing 0.5% crystal violet and 25% methanol, followed by 3 rinses with water. The colony numbers (>50 cells per colony) were counted using a precise electronic counter (Heathrow Scientific) and inverted microscope.

Cellular Senescence Analysis

Five days after transfection with shEZH2 or PSIF vector control, BT12 ATRT cells were fixed and stained using a senescence-associated-β-galactosidase kit (SA-β-gal, Cell Signaling Technology) according to the manufacturer's instructions. The senescent cells (dark blue staining) were microscopically distinguished from viable cells. Four high-power fields per sample were counted in 3 independent samples to quantify the number of senescent cells.

Cell Cycle Analysis

Flow cytometric cell cycle analysis was performed to define cell cycle alterations after EZH2 disruption. Cells were seeded in 6-well plates (10⁵ cells/well) and treated with 0.5 µM DZNep and/or transfected with 1 μg PSIF vector control or shEZH2 plasmid using Lipofectamine 2000 (Invitrogen). Cells were collected 24 h later by trypsinization and fixed with chilled 70% ethanol for at least 24 h. Fixed cells were then washed and stained with propidium iodide-containing cell cycle reagent (Millipore). Flow cytometric analysis was performed on the Guava EasyCyte Plus flow cytometer (Millipore).

Assessment of Apoptosis

Apoptotic fractions of shEZH2-transfected and/or 0.5 µM DZNep-treated cells were evaluated using the Nexin assay for the Guava EasyCyte Plus System (Millipore) according to the manufacturer's instructions. After shEZH2 transfection and/or DZNep treatment, media (inclusive of de-adherent cells) and adherent cells were collected, centrifuged, and counted by ViaCount assay on the Guava EasyCyte Plus flow cytometer (Millipore). Thirty thousand cells per condition were then incubated with Nexin reagent (Annexin V-PE and 7-AAD, Millipore) for 20 min. Cell fluorescence was measured and analyzed on the Guava EasyCyte Plus flow cytometer.

Radiation Assays

Five hundred BT12 cells per well were seeded in triplicate in 6-well plates for 24 h before treatment with DZNep or transfection with shEZH2 or PSIF vector control. Cells were X-irradiated 48 h later with 0.5, 1, 2, or 4 Gy. After 7 days of additional growth, colonies were stained as described above with crystal violet and were counted. Survival curves were generated after normalizing for the amount of transfection-induced death. Radiation doses at the specified survival fractions (10%: SF_0.1; 50%: SF_0.5) were calculated by nonlinear regression analysis of survival curves. The sensitizer enhancement ratio (SER) was then calculated as follows:

Luciferase Reporter Analysis

Luciferase assays were performed using the Cignal Finder 45-Pathway Reporter Array (SA Biosciences) according to the manufacturer's instructions. Cells were seeded into 96-well plates containing luciferase reporters and PSIF control or shEZH2 vectors using Surefect transfection reagent and then incubated for 24 h. Luciferase activity was measured using the Dual Luciferase Assay system (Promega) on a Glomax multi luminometer (Promega). Firefly luciferase served as the experimental reporter and Renilla luciferase as the normalizing reporter.

Tumor Sphere Assay

Five hundred UPN 737 cells were seeded in 6-well low-attachment plates and cultured in stem cell media (neurobasal with epidermal growth factor and leukemia inhibitory factor). After small sphere formation was established, treatment with 0.5 µm DZNep was initiated. Total sphere number was determined, and sphere size was measured 1, 3, 5, and 7 days after DZNep treatment. For secondary tumor spheres, DZNep-treated cells were grown as above, dissociated on day 5, and single cells were resuspended for evaluation of secondary tumor sphere formation.

Statistical Analysis

Data obtained from these experiments were processed using Excel (Microsoft) and Prism (GraphPad), as well as gcRMA software for gene array normalization and GSEA for determination of PRC1/2-responsive gene enrichment. All experiments, with the exception of the microarray analysis, were performed, at minimum, independently in triplicate to allow for appropriate statistical analysis. Error bars, unless otherwise indicated, represent the standard error of the mean. P values were calculated using Student's t-test.

Results

EZH2 Is Overexpressed in ATRT Patient Samples and ATRT Cell Lines

EZH2 expression is upregulated in many types of cancer and is associated with tumor aggressiveness, poor prognosis, and chemoresistance. We recently showed that EZH2 is upregulated in medulloblastoma.³⁰ Previous studies have shown that EZH2 expression is increased in malignant rhabdoid tumors.²⁶ Therefore, we first evaluated expression of EZH2 in our ATRT patient samples and cell lines, compared with expression in normal pediatric brain (cortex and cerebellum). Using Affymetrix microarray analysis, we found that EZH2 was significantly upregulated in 18 ATRT patient samples (P < .001) (Fig. 1A). These data are in line with those described by Wilson et al.²⁶ In addition, quantitative reverse-transcriptase PCR analysis demonstrated EZH2 overexpression in ATRT cell lines (BT12 and BT16) and in a primary patient-derived cell culture (UPN 737, Supplementary Figure S1A, BT12: P < .001, BT16: P < .01, UPN 737: P < .001). Western blot analysis further demonstrated that EZH2 protein is expressed in ATRT cell lines but not in normal brain (Fig. 1D). Because microRNA 101 has previously been implicated in the regulation of EZH2, we examined its expression in the ATRT cell lines.³¹ Of interest, miR 101 is significantly decreased in expression in ATRT cell lines, compared with normal brain (Supplementary Figure 1B).

Fig. 1.

EZH2 expression in ATRT. (A) Scatter Plot of EZH2 expression in 18 ATRT patient tissues, compared with normal brain (cortex and cerebellum). (B) GSEA of EZH2 responsive genes in ATRT, compared with normal brain. Gene sets from three individual studies were evaluated. The enhancement score was significantly negative in all 3 cases. (C) Heat map of the top 25 most suppressed genes in ATRT, compared with normal brain. (D) Expression of EZH2 protein and SMARCB1 protein in normal brain, control GBM cell line U251G, and ATRT cell lines (BT12, BT16, and UPN 737).

EZH2 Responsive Genes Are Repressed in ATRT Patient Samples

We next sought to examine whether EZH2 over expression had any implication on global gene expression in patients with ATRT. We first performed a GSEA of EZH2 responsive genes in ATRT, compared with normal brain. GSEA is a method that examines whether a set of genes shows differences between 2 biological/tumor sample conditions. We chose to examine EZH2 responsive gene sets from 3 previous publications.^32–34 Each described gene set was different likely because of the cellular context in which the studies were done. Pereira et al. examined EZH2 function in neural progenitor cells from the cerebral cortex, and Shen et al. and Ku et al. examined EZH2 function in mouse embryonic stem cells (ESC).^32–34 In all 3 cases, genes described as responsive to EZH2 were significantly repressed in our ATRT patient samples, as shown by negative enhancement scores (ES) (Fig. 1B). The ES reflects the degree to which a gene set is upregulated (positive ES) or down regulated (negative ES). Of interest, only 145 genes were overlapping among the 3 studies. As expected, these genes were significantly suppressed in ATRT patient samples, compared with normal cerebellum (Supplementary Table S1). Figure 1C shows a heat map of the 25 most significantly downregulated EZH2 responsive genes in ATRT patient samples, compared with normal cerebellum. These data, although not identical, are similar to those described by Roberts et al., in which they performed GSEA on SMARCB1-deleted murine lymphoma and malignant rhabdoid tumors with use of data previously published by Boyer et al.³⁵

Inhibition of EZH2 Decreases Cell Growth and Clonogenic Potential of ATRT Cells

To examine whether the overexpression of EZH2 had a functional significance, we first depleted EZH2 protein from ATRT cell lines with use of shRNA and then examined proliferation. We used 2 well-characterized ATRT cell lines (BT12 and BT16) and 1 primary short-term culture from a patient with ATRT (UPN 737). All 3 cell lines were confirmed to have absence of SMARCB1 protein (Fig. 1D). Knockdown of EZH2 by shRNA decreased cell proliferation in ATRT cells, as measured by ViaCount assay, compared with the PSIF vector control (Fig. 2A). As a measure of long-term cellular growth potential, BT12 and UPN 737 cell lines were transfected with PSIF control or shEZH2 and then evaluated by a clonogenic assay. EZH2 inhibition by RNAi was associated with decreased colony formation relative to nontransfected cells and vector control (PSIF) for both cell lines (Fig. 2B and C). Statistically significant differences (P < .001 and P < .05) were achieved for both cell lines. EZH2 knockdown was confirmed by Western blotting (Fig. 2D).

Fig. 2.

EZH2 is required for ATRT cell growth. (A) RNAi-mediated inhibition of EZH2 suppresses ATRT cell proliferation. (B and C) EZH2 inhibition suppresses ATRT cell clonogenic survival. (D) RNAi-mediated inhibition of EZH2 results in decreased EZH2 protein compared to PSIF vector control. (E and F) RNAi-mediated inhibition of EZH2 results in senescence of ATRT cells. Senescent cells shown by red arrows.

Inhibition of EZH2 Induces Cellular Senescence of ATRT Cells

Previous studies have suggested that PRC2 complex binding to DNA is altered in senescent cells.³⁶ Thus, to gain further insight into the effect of EZH2 inhibition on ATRT cells, we next evaluated whether the decreased proliferation after EZH2 silencing was a result of cellular senescence. BT12 ATRT cells were transfected with shEZH2 or PSIF vector control and subsequently underwent β-galactosidase staining 5 days after transfection. β-galactosidase staining demonstrated that EZH2 inhibition by RNAi produced a greater number of senescent cells than did the PSIF control (Fig. 2E and F, P < .01).

shEZH2 Inhibits Pathways and Molecules Involved in Cell Cycle Progression

To further explore the molecular pathways involved in EZH2-regulated ATRT cell growth, we performed a luciferase reporter array that simultaneously evaluated the activity of 45 signal transduction pathways in BT12 cells after shEZH2 or PSIF transfection (Fig. 3A). Of note, the activity of crucial transduction pathways, such as E2F and c-Myc, was downregulated in response to EZH2 inhibition (Fig. 3B and C). c-Myc increases the expression of cyclin D1, which in turn, leads to retinoblastoma protein (RB) phosphorylation and resultant E2F activation and DNA synthesis.³⁷ To examine this signaling axis, we evaluated protein expression of c-Myc, cyclin D1, CDK6, RB, and phosphorylated RB by Western blotting. Expression of all these proteins (with the exception of RB) was down regulated in response to shEZH2, demonstrating the downstream signal transduction effects of EZH2 inhibition (Fig. 3D).

Fig. 3.

EZH2-regulated signaling pathways in ATRT. (A) Activity of 45 transcription factors as measured by promoter luciferase activity as a measure of signaling activity in response to RNAi inhibition of EZH2. (B and C) EZH2 inhibition suppresses activity of E2F and c-Myc. (D) Western blot analysis of c-Myc–cyclin D–RB pathway.

Pharmacologic Inhibition of EZH2 Suppresses ATRT Cell Growth

Recently, DZNep was discovered to selectively inhibit H3K27 trimethylation via inhibition of PRC2.²⁴ We evaluated whether DZNep would alter ATRT cell growth. BT12 and UPN 737 cells were treated with varying concentrations of DZNep, and cell growth was evaluated. DZNep abrogated clonogenicity of ATRT cells (Fig. 4A). DZNep decreased H3K27me3, as detected by Western blotting (Supplementary Figure S2). Of interest, the combination of shRNA-mediated knockdown and DZNep treatment was cooperative in inhibiting the methylation of H3K27 (Supplementary Figure S2).

Fig. 4.

Chemical inhibition of EZH2 in ATRT. (A) DZNep inhibits clonogenic potential of ATRT cells. Quantification of colony formation with representative wells in control and DZNep-treated cells (*P < .01, **P < .001). (B) DZNep induces a G2/M arrest in BT12 cells. Quantification of cell cycle fractions with representative flow plots. (C) Inhibition of EZH2 induces apoptosis in BT12 ATRT cells. Representative flow plots of Annexin V expression and quantification of apoptosis in control and DZNep-treated cells.

Inhibition of EZH2 Induces G2/M Phase Arrest and Apoptosis in ATRT Cells

To further evaluate the impact of EZH2 inhibition on ATRT cells, flow cytometric cell cycle analysis was performed with RNAi directed against EZH2 and pharmacologic inhibition with DZNep. In BT12 cells, 0.5 µM DZNep treatment decreased DNA synthesis (S phase) by 6%, compared with control cells. RNAi with shEZH2 decreased the S phase fraction by 4%, compared with control (Fig. 4B). Maximum decrease of the S phase fraction (7% decrease) was observed in cells undergoing both shEZH2 transfection and DZNep treatment. Transfection of BT12 cells with shEZH2 did not lead to significant differences between the number of cells in G2/M phase; however, DZNep treatment in combination with shEZH2 increased the percentage of cells in G2/M phase by >12%, compared with control (P < .001). In addition to cell cycle alterations, EZH2 interference resulted in increased apoptosis. EZH2 silencing by DZNep treatment (0.5 µM), shRNA, or a combination of shRNA and DZNep led to increased early and late apoptotic fractions in the BT12 cell line (Fig. 4C).

In UPN 737 cells, DZNep treatment decreased the S phase fraction by 10%, and in shEZH2-transfected cells, S phase was decreased by 22%, compared with control. Combination-treated UPN 737 cells displayed a similar decrease in S phase, compared with shEZH2-treated (monotherapy) cells (Supplemental Figure 3A). EZH2 inhibition via shEZH2 resulted in a 14% increased accumulation of UPN 737 cells in G2/M phase, compared with control (P < .05). DZNep treatment in conjunction with shEZH2 increased the percentage of cells in G2/M phase by up to 15%, compared with control. Evaluation of apoptotic fractions of UPN 737 cells demonstrated significantly increased apoptosis after shEZH2 transfection or a combination of shRNA and DZNep (Supplemental Figure 3B and C).

Cell cycle analysis of BT16 cells demonstrated a similar reduction in the S phase fraction in response to shEZH2 and/or DZNep treatment (Supplemental Figure 3D). In BT16 cells, the maximum accumulation of cells in G2/M phase was found in shEZH2-transfected cells (18% increase over control, P < .001). Evaluation of apoptosis in BT16 cells demonstrated an increased apoptotic fraction only in response to combination shEZH2 transfection and DZNep treatment (Supplemental Figure 3E and F).

EZH2 Silencing Increases Radiation Sensitivity of ATRT Cells

To investigate whether EZH2 inhibition affects sensitivity to ionizing radiation, we first treated BT16 and UPN 737 ATRT cells with DZNep for 24 h and, subsequently, X-irradiated the cells. Cell growth was then evaluated by clonogenic assays. Treatment with DZNep resulted in decreased surviving fractions in response to radiation for both BT16 and UPN 737 cells (Fig. 5A and B). Surviving fractions of DZNep-treated cells after 2 Gy irradiation were significantly lower than those of control cells (Fig. 5A and B). In addition, sensitizer enhancement ratios were 6.1% at 10% survival fraction (SF_0.1) and 6.6 at 50% survival fraction (SF_0.5) for BT16 cells and 2.75 (SF_0.1) and 2.9 (SF_0.5) for UPN 737 cells, demonstrating a strong potentiation of the effects of radiation after EZH2 inhibition. We subsequently confirmed the radiation sensitization effect by inhibiting EZH2 with shRNA and evaluating radiation sensitivity (Supplementary Figure 4). shEZH2 potently induced increased sensitivity to radiation in line with our chemical inhibition.

Fig. 5.

DZNep induces radiation sensitivity in ATRT cells. Line graph plotting the surviving fraction of (A) BT16 cells or (B) UPN 737 cells given different radiation doses after 24 h of treatment with DZNep or the vehicle control. Blue line represents control treated cells. Red line represents cells treated with 0.5 µM DZNep. Bar graphs show the quantification of the surviving fraction following 2 Gy irradiation.

DZNep Treatment Reduces the Self-Renewal and Tumor-Initiating Properties of ATRT Cells

EZH2 has been previously shown to be crucial for the maintenance of tumor-initiating cells in glioblastoma multiforme.³⁸ To determine whether EZH2 is similarly necessary for ATRT-initiating cells, we examined the effects of chemical inhibition of EZH2 on the ability of ATRT cells to form tumor spheres. UPN 737 cells were cultured in stem cell media (neurobasal with epidermal growth factor and leukemia inhibitory factor) and were treated with 0.5 µM DZNep after initial sphere formation was observed. Sphere formation was allowed to proceed for 7 days after DZNep treatment. Measurements of sphere size at 1, 3, 5, and 7 days after treatment demonstrated a significant decrease in sphere diameter in response to DZNep treatment (Fig. 6A and C). In addition, DZNep treatment decreased the total number of spheres relative to control (Fig. 6B). To further evaluate the impact of EZH2 inhibition on ATRT self-renewal capacity, we examined the ability of UPN 737 cells to form secondary tumor spheres after treatment with DZNep. UPN 737 cells were treated with DZNep (0.5 μM), and primary spheres formed for 5 days as described above. After 5 days, spheres were dissociated again and single cells seeded to assay for their ability to form secondary spheres. DZNep treatment substantially attenuated ability to form secondary spheres, and these spheres were significantly smaller than control-treated cells (Fig. 6D).

Fig. 6.

DZNep suppresses the tumor sphere formation of ATRT cells. (A) Representative images of UPN 737 cells grown as tumor spheres in control or DZNep-treated conditions. (B) Quantification of tumor sphere number after 5 days of DZNep treatment (*P < .01). (C) Quantification of tumor sphere size in DZNep-treated cells compared to control (*P < .01). (D) Representative images of secondary tumor spheres in DZNep-treated UPN 737 cells. DZNep-treated cells did not form any spheres >50 cells. Sphere size of DZNep-treated cells was significantly smaller (*P < .01, **P < .001).

Discussion

ATRTs are highly aggressive pediatric brain tumors with a dismal prognosis.³⁹ The key molecular feature of these tumors is the loss of the SMARCB1 gene. We have recently shown that there are 4 major subgroups within ATRT-based on genomic analysis.⁴⁰ The tumors with the BMP signaling pathway gene signature correlated with the poorest outcomes.⁴⁰ To further evaluate ATRT tumor biology, we investigated the expression of other genes that were overexpressed in our analysis. One of the top candidates was the polycomb group gene EZH2. EZH2 is a PRC2 member that mediates repression of gene transcription through its H3K27 methyltransferase activity.⁴¹ Overexpression of EZH2 correlates with increased tumor cell proliferation, aggressiveness, and invasive cancer,^14,42 suggesting a role for aberrant regulation of H3K27 in cancer.⁴³

In our experiments, we found that EZH2 is overexpressed in ATRT patient samples and in ATRT cell lines. More importantly, we found that genes shown to be responsive to EZH2 in both ESC cells and pluripotent neural progenitor cells were repressed in ATRT patient tissue samples. When we transfected ATRT cell lines (BT12, BT16, and UPN 737 short-term ATRT cell cultures) with shEZH2, we found significant reduction of cell proliferation and a decrease in clonogenic potential.

On the basis of the well-known role of EZH2 in cell cycle progression in different cancers, we evaluated the role of EZH2 inhibition on ATRT cell cycle progression. As hypothesized, disruption of functional EZH2 by RNAi led to significant inhibition of DNA synthesis and resulted in accumulation of cells in G2/M phase. These data are consistent with previous reports indicating that EZH2 overexpression results in induction of S phase.¹⁰ In addition, the inhibition of EZH2 expression by siRNA or shRNA has been shown by previous evaluations to result in cell cycle arrest in G1, G2, and G2/M, thus further demonstrating that EZH2 is also required for general cell cycle progression.^10,14,44,45

We also found that inhibition of EZH2 led to decreased activity of the oncogenic E2F and c-Myc signal transduction pathways and decreased expression of cyclin D1, CDK6, c-Myc, and phosphorylated RB. Of interest, previous studies have implicated this pathway in malignant rhabdoid tumors, including ATRT.^9,46 Indeed, Tsikitis et al. demonstrated that ablation of cyclin D1 suppressed growth of malignant rhabdoid tumors in SMARCB1-deleted mice.⁴⁷ Our data suggest that the upregulation in cyclin D1 is a result of EZH2 expression and that inhibition of EZH2 may be a way to suppress this important pathway. Thus, our data further support the concept of EZH2′s role as a mediator of oncogenesis and aberrant cell cycle progression in ATRT and corresponds to prior data indicating that EZH2 promotes epigenetic activation of oncogenic signaling cascades and inhibition of prodifferentiation pathways.⁴⁸ Given our data and the known role of EZH2 in other malignancies, further in-depth functional analysis of the role of EZH2 in ATRT signal transduction is warranted.

DZNep is a known S-adenosylhomocysteine hydrolase inhibitor. Prior studies have reported that DZNep reactivates silenced genes in cancer and that this reactivation was specific to EZH2.²⁵ However, specific evaluations of DZNep's effects on ATRT cells have not been reported. In this study, we demonstrated that DZNep inhibited growth in all ATRT cell lines tested. Our data also revealed that DZNep inhibited cell growth through cell cycle arrest and led to inhibition of DNA synthesis. Furthermore, DZNep induced apoptosis in BT12 and UPN 737 cells, although a greater effect was seen in all cell lines when DZNep treatment was combined with shEZH2. Our data correlate with previous findings that DZNep induced apoptosis in breast and colon cancer cells and disrupted the PRC2 complex, decreased EZH2 levels, and led to apoptosis in GBM.^24,38 Of most importance, EZH2 inhibition with DZNep strongly potentiated the effects of ionizing radiation on ATRT cells. To our knowledge, there are no prior studies that have evaluated these treatments in combination. Because of the young age of patients with ATRT, the ability to radiosensitize and, thus, use lower doses of radiation is of critical importance.

To evaluate the role of EZH2 in maintenance of cellular self-renewal, we examined DZNep's ability to alter tumor sphere formation. DZNep treatment suppressed the ability of ATRT cells to form spheres in stem cell conditions. Recent studies found that loss of EZH2 function in embryonic murine hepatic stem/progenitor cells decreased proliferation and self-renewal capability.⁴⁹ In addition down-regulation of EZH2 by shRNA strongly impairs GBM cells' self-renewal.^38,50 We have shown that treatment of ATRT cells with DZNep inhibits growth and represses trimethylation of histone H3K27. The mechanism of DZNep activity involved both apoptosis induction and cell cycle arrest in ATRT cells.

In summary, we describe the role of EZH2 in ATRT cells. We found that EZH2 acts on ATRT cells through mechanisms involving cell cycle, senescence, and signal transduction pathways. In addition, EZH2 inhibition effectively sensitized ATRT cells to ionizing radiation. Furthermore, we demonstrated the potential therapeutic usefulness of DZNep in ATRT, particularly in combination with ionizing radiation. Patients with ATRT are frequently very young and, thus, susceptible to potential long-term toxicities to normal tissues. In this context, recent studies showing that DZNep depletes cellular levels of PRC2 components, blocks associated histone H3K27 methylation, and induces apoptosis in malignant cells but not in normal cells are encouraging.^24,38 Thus, we have demonstrated that EZH2 is a critical regulator of ATRT cell growth and is a potential therapeutic target in this devastating brain tumor of childhood.

Supplementary Material

Supplementary material is available online at Neuro-Oncology (http://neuro-oncology.oxfordjournals.org/).

Conflict of interest statement. The authors have no conflicts of interest to disclose.

Funding

This work was supported by NINDS K08 NS059790 (to R.V.) and the Morgan Adams Foundation (to R.V. and N.K.F.).