Aberrant Epigenetic Regulation in Head and Neck Cancer Due to Distinct EZH2 Overexpression and DNA Hypermethylation

Enhancer of Zeste homologue 2 (EZH2) overexpression is associated with tumor proliferation, metastasis, and poor prognosis. Targeting and inhibition of EZH2 is a potentially effective therapeutic strategy for head and neck squamous cell carcinoma (HNSCC). We analyzed EZH2 mRNA expression in a well-characterized dataset of 230 (110 original and 120 validation cohorts) human head and neck cancer samples. This study aimed to investigate the effects of inhibiting EZH2, either via RNA interference or via pharmacotherapy, on HNSCC growth. EZH2 upregulation was significantly correlated with recurrence (p < 0.001) and the methylation index of tumor suppressor genes (p < 0.05). DNMT3A was significantly upregulated upon EZH2 upregulation (p = 0.043). Univariate analysis revealed that EZH2 upregulation was associated with poor disease-free survival (log-rank test, p < 0.001). In multivariate analysis, EZH2 upregulation was evaluated as a significant independent prognostic factor of disease-free survival (hazard ratio: 2.085, 95% confidence interval: 1.390–3.127; p < 0.001). Cells treated with RNA interference and DZNep, an EZH2 inhibitor, showed the most dramatic changes in expression, accompanied with a reduction in the growth and survival of FaDu cells. These findings suggest that EZH2 upregulation is correlated with tumor aggressiveness and adverse patient outcomes in HNSCC. Evaluation of EZH2 expression might help predict the prognosis of HNSCC patients.


Introduction
Head and neck squamous cell cancer (HNSCC) is a heterogeneous disease that potentially involves multiple sites and cellular origins within the head and neck region [1]. Despite aggressive treatments, long-term survival rates are poor and remain at~50% [2]. Therefore, studies are required to shift their focus from prognostic biomarkers and to the development of predictive biomarkers enabling patient selection for a specific therapy [3,4]. In HNSCC, epigenomic inactivation linked to tumor suppressor genes (TSGs) are more frequent than somatic mutations in cancer and may be driving tumorigenic initiation and progression [5]. Aberrant gene promoter methylation is a key event in cancer pathogenesis and has gained increasing interest in basic and translational oncology studies because of their reversible nature [6,7]. Therefore, molecular classification of HNSCCs is required to establish prognostic and potentially preventive targets of tumor recurrence in patients.
Previous studies reported that EZH2 is required for the recruitment of DNA methyltransferases and facilitates CpG methylation of EZH2-target promoters [21]. Repression of the EZH2 target genes results largely from both H3K27me3 and DNA methylation, thereby providing a basis for investigating combinations of both EZH2 and DNA methyltransferase inhibitors in patients wherein EZH2 is overexpressed [22,23]. EZH2 overexpression is correlated with increased promoter methylation across tumor types in the Cancer Genome Atlas (TCGA) [24]. However, a systematic study regarding the epigenetic and transcriptional regulation of EZH2 genes in HNSCC is still needed.
This study aimed to investigate the biological effects of EZH2 regulation in HNSCC. We assessed the mRNA expression status of EZH2 target genes in HNSCC to evaluate their clinical significance as prognostic biomarkers for recurrence risk and survival. Simultaneous analyses of the EZH2 expression and the methylation of tumor suppressor genes is important to predict tumorigenesis and the development of future targeted therapy.

EZH2 mRNA Expression Levels in Pairs of HNSCC and Normal Mucosal Tissues
EZH2 mRNA expression levels were determined in 110 HNSCC and 68 adjacent normal mucosal tissues with an independent set (original cohort). EZH2 mRNA levels were significantly greater in HNSCC tissues than in adjacent normal tissues (p = 0.003) ( Figure 1A). Furthermore, upon comparing EZH2 expression in another cohort of 120 pairs of HNSCC and normal tissues (validation cohort), EZH2 mRNA levels in HNSCC tissues were 3-fold higher than those in paired non-cancerous mucosa (p = 0.002) ( Figure 1B). EZH2 mRNA expression exhibited highly discriminative receiver-operator characteristic (ROC) curve profiles, which clearly distinguished HNSCC from normal mucosal tissues (area under the ROC = 0.641). At the cutoff value of 17.81, sensitivity was 33.3%; specifically, 95.1% ( Figure 1C).

Correlation between EZH2 mRNA Expression and Clinicopathological Assessment
Characteristics and clinicopathologic features of patients are summarized in Table 1. A total of 230 tumors showed low (<17.81) and 20 high (≥17.81) EZH2 mRNA expression levels. In the original set, EZH2 mRNA levels were significantly associated with cancer stage (p = 0.043) and recurrence (p = 0.016). EZH2 expression in patients in the validation cohort was significantly associated with recurrence (p = 0.012). Upon combining the original and validation cohorts, EZH2 was significantly correlated with recurrence (p < 0.001) ( Table 1).

Correlation between EZH2 mRNA Expression and Clinicopathological Assessment
Characteristics and clinicopathologic features of patients are summarized in Table 1. A total of 230 tumors showed low (<17.81) and 20 high (≥17.81) EZH2 mRNA expression levels. In the original set, EZH2 mRNA levels were significantly associated with cancer stage (p = 0.043) and recurrence (p = 0.016). EZH2 expression in patients in the validation cohort was significantly associated with recurrence (p = 0.012). Upon combining the original and validation cohorts, EZH2 was significantly correlated with recurrence (p < 0.001) ( Table 1).

Prognostic Value of EZH2 Gene Expression
Kaplan-Meier plots indicated that the expression status of EZH2 genes was related to disease-free survival (DFS) (Figure 2A-D). In the original cohort, a shorter DFS was also observed when EZH2 gene upregulation was compared with EZH2 gene downregulation (log-rank test, p = 0.002) (Figure 2A). EZH2 genes were upregulated in patients in the validation cohort and they had a shorter DFS than those displaying EZH2 downregulation (log-rank test, p = 0.024) ( Figure 2B). By combining the original and validation set, DFS rates in patients displaying EZH2 upregulation was 35.5%, as compared with 56.9% in those displaying EZH2 downregulation (log-rank test, p = 0.0001) ( Figure 2C). Moreover, among 152 patients with tumor stage T1, T2, and T3, the DFS rate in those displaying EZH2 gene upregulation was 39.8%, as compared with 55.6 % in those displaying EZH2 downregulation (log-rank test, p = 0.009) ( Figure 2D). Furthermore, the association between methylation and risk of recurrence was estimated via multivariate analysis using a Cox proportional hazards model adjusted for age, sex, alcohol exposure, smoking status, and clinical stage. In patients displaying EZH2 upregulation (76/230, 33.0%), the adjusted odds ratio (OR) for recurrence was 2.085 (95% confidence interval [CI]: 1.390-3.127, p < 0.001) (

Correlation between Methylation Levels of 11 Tumor Suppressor Genes and EZH2 Expression
The 11 tumor suppressor genes were defined as the number of methylated genes in each sample ( Figure 3A, Supplementary Table S3). Mean differences in the methylation index of the 11 tumor suppressor genes (TS-MI) determined on the basis of EZH2 gene expression are illustrated in Figure 3B. In particular, the TS-MI was significantly higher in patients displaying EZH2 upregulation (5.97 ± 0.25) than in those with displaying EZH2 downregulation (5.21 ± 0.22, p = 0.032) ( Figure 3B). We examined DNA methyltransferase 3α (DNMT3A) and 3β (DNMT3B) genes mRNA levels via qRT-PCR in HNSCC specimens. DNMT3A was significantly upregulated in EZH2 higher expression groups (p = 0.043) ( Figure 3C). DNMT3B expression was not associated with EZH2 expression (p = 0.628) ( Figure 3D).

External Validation of Results from TCGA
Data from 43 non-tumorous tissues and 497 HNSCC tissues were analyzed. The EZH2 gene expression status was significantly higher in HNSCC than in normal samples (p < 0.001) (Supplementary Figure S1A). The Spearman analysis revealed that EZH2 expression was positively correlated with DNMT3A and DNMT3B expression (R 2 = 0.0734, p < 0.001 and R 2 = 0.02, p = 0.002, respectively) (Supplementary Figure S1B,C).

Discussion
This study reported that EZH2 was upregulated in HNSCC cells in comparison with normal cells, being associated with a higher recurrence risk. Based on these findings, our study is the first, to our knowledge, to report a quantifiable correlation between EZH2 overexpression and DNA hypermethylation in HNSCC. Herein, we identified that DZNep, a small-molecule EZH2 inhibitor, as a putative therapeutic target of HNSCC. These results suggest that EZH2 inhibition is a promising therapeutic avenue for a substantial fraction of HNSCC patients. Clarification of the epigenetic regulation of EZH2 gene may provide insights into the underlying mechanisms of tumorigenesis and the risk of disease recurrence in HNSCC.
EZH2 expression is reportedly associated with aggressive prostate cancer [25], esophageal cancer [13], and breast cancer [25], whereas its clinical importance in HNSCC is yet unknown. EZH2 has gained interest as a marker of aggressive subgroups in several cancers [26]. DNZep is an S-adenosylhomocysteine (SAH) hydrolase inhibitor that increases intracellular SAH levels [27]. DZNep treatment induces apotosis in chondrosarcoma cells, while decreasing their migration ability [17] [28] [29]. DZNep directly induces cellular senescence, accompanied with apoptosis, in human colon cancer cells [19]. The PRC2 inhibitory effects of DZNep are important for differentiation-inducing and anticancer activities observed in HNSCC [30].
EZH2 constitutes the catalytic subunit of PRC2 and catalyzes H3K27me3 and silences target genes via local chromatin reorganization [31]. EZH2 reportedly promotes cell proliferation, migration and invasion in different in vitro cancer cell models [32]. Viré E et al. reported that EZH2 regulates CpG methylation through direct physical contact with DNA methyltransferases [21]. EZH2 overexpression has been previously correlated with increased promoter methylation across numerous tumor types in prostate cancer, small-cell lung cancer, and clear cell sarcoma of the kidney [22,24,33]. However, a systematic study of EZH2 expression and DNA promoter methylation in most human cancers is still needed. To our knowledge, this study is the first to report EZH2 expression and DNA promoter methylation in the genesis of HNSCC.
Thus, EZH2 may play a specific role depending on the cell type and different tumorigenic sites. Such a study involving human specimens and utilizing high-throughput profiling platforms may be susceptible to measurement bias from various sources. Our results suggest, for the first time, that increased EZH2 expression is correlated with tumor progression and may facilitate the accumulation of aberrantly methylated tumor suppressor genes in HNSCC. Moreover, transcriptional activation of EZH2 was associated with aberrant methylation of other tumor-related genes and DNMT3A upregulation in HNSCC. The present results suggest that the targeted inhibition of EZH2 catalytic activity may constitute a promising therapeutic intervention; however, additional prospective studies are required to prevent broad side effects.

Tumor Samples
Two independent cohorts comprised 110 and 120 (original and validation cohorts) fresh frozen samples of primary HNSCC were obtained from patients from the Hamamatsu University Hospital. Written informed consent was obtained from all patients in accordance with the tenets of the Declaration of Helsinki. The study protocol was approved by the Institutional Review Board of the

Cell Culture
The SCC cell line FaDu was purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were cultured in Dulbecco's modified Eagle's medium (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) supplemented with 10% fetal bovine serum (ThermoFisher Scientific Inc., Waltham, MA, USA) and 1% penicillin/streptomycin (FUJIFILM Wako Pure Chemical Corporation) in a humidified atmosphere containing 5% CO 2 at 37 • C.

RNA Extraction and Quantitative Reverse Transcription PCR (qRT-PCR)
Total RNA was isolated using an RNeasy Plus Mini Kit (Qiagen, Hilden, Germany); cDNA was synthesized using a ReverTra Ace qPCR RT Kit (Toyobo, Tokyo, Japan). EZH2, DNMT3A, DNMT3B, and GAPDH mRNA expression levels were measured via qRT-PCR using SYBR Premix Ex Taq (Takara Bio Inc., Tokyo, Japan), the Takara Thermal Cycler Dice Real Time System TP8000 (Takara Bio Inc.), and the primer sets presented in Supplementary Table S2. Target gene expression levels were normalized to those of GAPDH using the 2 −∆∆ Ct method. , was added to FaDu medium at 10 µM for the cell proliferation assay. For the colony forming assay, cells were seeded in 6-well plates at a density of 0.05 × 10 4 cells per well and cultured for 14 d, followed by fixation and staining 40% ethanol-1% crystal violet and enumerated in a field of >0.1 mm diameter. For the cell proliferation assay, cells (0.2 × 10 4 cells) were placed into each well in 12-well plates enumerated and at 3 time points (day 2, day 5, and day 7).

DNA Extraction, Bisulfite Treatment, and Quantitative Methylation-Specific PCR (QMSP) Analysis
Genomic DNA was extracted from tumor specimens harvested from 216 primary HSNCC patients and subjected to bisulfite conversion using the MethylEasy Xceed Rapid DNA Bisulfite Modification Kit (Takara Bio Inc.) in accordance with the manufacturer's instructions. QMSP was performed as described previously [34]. The QMSP primer sequences for CCBE1, CDH1, COL1A2, DAPK, GALR1, MGMT, p16, RASSF1A, SALL3, SST, and TAC1 are provided in Supplementary Table S2. Overall methylation levels in individual samples were determined by calculating the methylation value. TS-MI was defined as the ratio of the number of methylated genes to the number of genes analyzed in each sample [35].

Collection of Publicly Available Data from TCGA
Gene expression data were obtained from the TCGA data portal (https://tcga-data.nci.nih.gov/ tcga/) and MethHC (http://methhc.mbc.nctu.edu.tw/php/index.php) in July 2018 [36]. Expression data were obtained as processed RNA-seq data in the form of RNA-seq via Expectation Maximization.

Data Analysis and Statistics
The EZH2 expression status results and patient characteristics were compared using the students' t-test. Receiver operating characteristic (ROC) curve analysis was performed using for 120 HNSCC and 120 adjacent normal mucosal samples with the Stata/SE 13.0 system (Stata Corporation, College Station, TX, USA). Area under the ROC curve analysis indicated the optimal sensitivity and specificity cut-off levels to differentiate between gene expression levels in HNSCC and normal tissue. DFS was measured from the date of the initial treatment to the date of diagnosis of locoregional recurrence or distant metastasis. Kaplan-Meier analysis was performed to calculate survival probabilities, and the log-rank test was performed to compare survival rates. The prognostic value of EZH2 expression status was assessed via multivariate Cox proportional hazards analysis adjusted for age (≥70 versus <70 years), sex, alcohol intake, smoking status, and tumor stage (I, II, and III versus IV). Differences with p < 0.05 were considered significant. All statistical analyses were performed using the StatMate IV software (ATMS Co. Ltd., Tokyo, Japan).

Conclusions
In conclusion, the present results suggest that EZH2 is aberrantly expressed in HNSCC patients. CpG island hypermethylation is an early event in tumorigenesis, supporting the hypothesis that EZH2 plays an important role in tumorigenesis in HNSCC and serves as an important biomarker. This study is the first, to our knowledge, to report that EZH2 mRNA is downregulated in HNSCC owing to DNA methylation; this may be a critical event in HNSCC progression, which is associated with decreased survival.