Human endometrial carcinoma cell lines, Hec-1a and Ishikawa (provided by Dr Yinhua Yu; Anderson Cancer Center, Houston, TX, USA), were maintained in McCoy’s 5A medium (Jinuo Co., Ltd, Shanghai, China) supplemented with 10% fetal bovine serum (Gibco-BRL, Rockville, IN, USA). The cells were incubated at 37°C in 5% CO₂.

A total of 92 endometrial tissues (including 24 normal endometrium, 14 simple hyperplasia, 6 complex hyperplasia, 15 atypical hyperplasia and 33 endometrial cancer samples) were obtained from patients who underwent surgery between August 2008 and December 2012 at the Obstetrics and Gynecology Hospital, Fudan University (Shanghai, China). Normal endometrium and simple and complex hyperplasia samples were from patients who received dilation and curettage, whereas atypical hyperplasia and cancer tissues were from patients who received hysterectomy. All specimens were reviewed by an experienced pathologist. The patients’ demographic profiles and the pathology files were tabulated. For endometrial cancer patients, the clinicopathological factors, such as age, tumor grade, depth of myometrial invasion, lymph-vascular space invasion (LVSI) and nodal metastasis, were analyzed. The surgical pathology stage was determined by the 1998 International Federation of Gynecology and Obstetrics (FIGO) guidelines (16). All patients provided written informed consent permitting the use of their tissue for research at the time specimens were collected. This study was approved by the institutional review board of the Obstetrics and Gynecology Hospital, Fudan University.

Immunohistochemistry (IHC), antigen retrieval and antibody dilution were optimized prior to the study onset. All endometrium specimens were reviewed by experienced pathologists to confirm the diagnosis. To ensure uniformity, all sections were processed simultaneously. Four-micrometer paraffin sections adjacent to the hematoxylin and eosin sections used for histological assessment were mounted onto Superfrost Plus slides (Menzel, Braunschweig, Germany). Slides were subjected to immunoperoxidase staining for EZH2 (5246S; Cell Signaling Technology, Inc., Danvers, MA, USA). Endogenous peroxidase activity was blocked using 0.3% hydrogen peroxide. Antigen retrieval was performed by heating the sections for 30 min in a microwave oven with 10 mM sodium citrate (pH 6.0). The slides were then incubated with monoclonal antibodies against EZH2 (1:50; rabbit anti-human, -rat, -mouse and -monkey; Cell Signaling Technology, Inc.) at 4°C overnight. Slides were washed and incubated with the biotinylated secondary antibody (polyclonal goat anti-rabbit; Histostain-Plus IHC kit; Mingrui Biotech, Shanghai, China) for 45 min at 37°C and washed with phosphate-buffered saline. Slides were incubated with avidin-biotin-peroxidase (Histostain-Plus IHC kit; Mingrui Biotech) for 10 min at room temperature and incubated with diaminobenzidine (Mingrui Biotech) for 2 min. Finally, slides were counterstained with hematoxylin and evaluated at a magnification of ×200 using light microscopy. The intensity of positive cells was graded from 0 to 3 (0, negative; 1, weak; 2, medium; 3, strong). The scores were determined independently by two observers, and the average of their scores was used for evaluation. For estrogen receptor (ER) (1:150), progesterone receptor (PR) (1:160), p53 (1:500) and Ki-67(1:200) (Dako, Glostrup, Denmark) detection, antigens were unmasked by treating the slides with Target Retrieval Solution, High pH (Dako) for 30 min at 95°C. High-grade serous adenocarcinoma of the ovary was used as positive control for p53 and Ki-67, while ductal carcinoma of the breast was used as positive control for ER and PR. Expression of ER and PR was considered as positive when >1% of the nuclei of cells in the epithelium showed immunoreactivity. Expression of p53 was considered as positive when immunohistochemical staining was observed in >5% of the nuclei of cells in the epithelium. The percentage of the nuclei staining of Ki-67 was determined independently by two observers.

Cells were lysed in RIPA buffer (150 mM NaCl, 50 mM Tris-base, 5 mM EDTA, 1% NP-40, 0.25% deoxycholate, pH 7.4). Protein concentrations were measured by the BCA protein assay (23227; Thermo Fisher Scientific, Waltham, MA, USA). Equal amounts of protein were resolved by SDS-PAGE, transferred to PVDF membranes and incubated with appropriate primary antibodies (monoclonal rabbit anti-human EZH2 antibody, 5246S, 1:1000, Cell Signaling Technology, Inc.; monoclonal rabbit anti-human GAPDH antibody, 5632–1, 1:5000, Epitomics, Burlingame, CA, USA). Immune complexes were detected with a goat anti-rabbit horseradish peroxidase-conjugated secondary antibody (1:5000; SSA005; Sino Biological Inc., Beijing, China) and enhanced chemiluminescence reagent (32109; Thermo Fisher Scientific).

Hec-1a and Ishikawa cells were plated on six-well plates at a density of 2×10⁵ cells/well and grown overnight until 30–40% confluency. The cells were transfected with validated siRNA for EZH2 (sense: 5′-GUGUAUGAGUUUAGAGUCATT-3′) and a scramble siRNA-FAM (negative control) (synthesized by Jima, Co., Ltd, Shanghai, China) at a concentration of 100 nM. using Lipofectamine 2000 transfection reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s instructions. The medium was replaced with standard culture medium 12 h post-transfection. Transfection was repeated 72 h after the first transfection.

Cell proliferation was assayed using a cell proliferation kit, Cell Counting Kit-8 (CCK-8; Dojindo Molecular technologies, Inc., Kyushu, Japan) according to the manufacturer’s instructions. Hec-1a and Ishikawa cells were seeded in sextuplicate onto 96-well tissue culture plates at a density of 2×10³ cells/well the day before EZH2 siRNA transfection. Cell growth was analyzed at a wavelength of 450 nm at 0, 48, 72, 96 and 120 h after transfection using Multiskan MK3 (Thermo Fisher Scientific). Experiments were performed in triplicate.

The data are presented as the means ± standard deviation. Statistical analyses were performed using SPSS 15.0 software (SPSS, Inc., Chicago, IL, USA). Two independent samples non-parametric tests were utilized to analyze the immunohistochemistry results. Fisher’s exact tests were utilized to analyze the association between EZH2 methylation levels and clinicopathological characteristics. Paired-sample t-tests were utilized to analyze the CCK-8 results of EZH2 knockdown. P<0.05 was considered to indicate a statistically significant difference.

Immunohistochemistry was performed in 92 endometrium tissues, including 24 normal endometrium (21 proliferative endometrium, two secretory endometrium and one atrophic endometrium; age range, 38–66), 14 simple hyperplasia (age range, 30–76), six complex hyperplasia (age range, 44–62), 15 atypical hyperplasia (age range, 36–64) and 33 endometrial cancers (27 type I and six type II; age range, 37–78). The age distribution was similar among groups (P>0.05). Overexpression of EZH2 was observed in the epithelium of endometrial cancer, atypical hyperplasia and complex hyperplasia compared with the expression in simple hyperplasia and normal endometrium. No difference was seen among endometrial cancer, atypical hyperplasia and complex hyperplasia, or between simple hyperplasia and normal endometrium, in terms of epithelial EZH2 expression (Fig. 1A and B). Expression of EZH2 showed no difference in the stroma among all groups (Fig. 1C).

The association between expression of EZH2 and clinicopathological characteristics of 33 endometrial cancer tissues was analyzed. The samples were grouped based on whether they had high (IHC score, 2 or 3) or low (IHC score, 0 or 1) expression of EZH2 (Table I). Tumor grade; FIGO stage; depth of myometrial invasion; LVSI; nodal metastasis status; ER, PR and p53 expression and Ki-67 labeling index of the tumor were determined, and patients ranged in age from 37 to 78 years (medium age, 55 years). High expression of EZH2 was observed in 64% of these cases. As shown in Table I, high EZH2 expression was associated with deep myometrial invasion. In samples of low EZH2 expression, two cases (17%) were limited to the endometrium, 10 (83%) cases occupied less than half of the myometrium and none occupied more than half of the myometrium. However, in the high EZH2 expression group, 14 cases (67%) were found to occupy less than half of the myometrium, while seven (33%) cases demonstrated deep (≥1/2) myometrium infiltration (P=0.013). Cases with deeper myometrial invasion were more likely to be EZH2-overexpressing.

In addition, high EZH2 expression appeared to be associated with the presence of LVSI. Although the P-value (0.065) was not significant, this may have been due to the small sample size. No correlation was noted between EZH2 expression and other clinicopathological characteristics.

To evaluate the effect of EZH2 on cell proliferation, knockdown of EZH2 was performed in endometrial cancer cells, Hec-1a and Ishikawa, by siRNA. Subsequently, cell viability was analyzed using the CCK-8 assay. Scrambled siRNA, labeled by FAM, was transfected at the same time to observe the transfection efficiency (Fig. 2A), and the knockdown effect on EZH2 was validated by western blotting (Fig. 2B). The inhibition effect started 48 to 96 h after cell transfection and was significantly inhibited in Hec-1a and Ishikawa cells at 120h (Fig. 2C and D). Cell growth was inhibited after EZH2 knockdown in endometrial cancer cells.