SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Disclosure Statement
  8. References

In order to determine the expression pattern of EZH2 in ovarian neoplasms and assess the functions and mechanism of EZH2 in tumorigenesis in vitro and in vivo, we detected the protein and mRNA expression of EZH2 and p57 in normal, benign and malignant ovarian tissues, used shRNA to knock down EZH2 expression in ovarian cancer cells and established a nude mouse xenograft model. We found EZH2 was overexpressed in ovarian tumor with the highest level expression in malignant ovarian tissues, and the variation of EZH2 expression at different pathological type/grade and International Federation of Gynecology and Obstetrics (FIGO) stages was statistically significant. Furthermore, the EZH2 expression was inversely correlated with the p57 mRNA level in ovarian tissues. Moreover, inhibition of endogenous EZH2 increased the expression of p57 and reduced proliferation and migration of ovarian cancer cells. Loss of EZH2 suppresses ovarian tumor formation in vivo. Our results indicate that the EZH2 gene acts as an oncogene in tumorigenesis of ovarian cancer with the possible mechanism to suppress the anti-oncogene p57. EZH2 is a potential therapeutic target for treatment of ovarian cancer. (Cancer Sci 2011; 102: 530–539)

Ovarian cancer is the most common gynecological malignancy; 70% of patients with ovarian malignant tumors are in an advanced stage when diagnosed, due to the absence of symptoms in the early phase, as well as the absence of effective or sensitive screening methods.(1) Deep research into molecular and biological mechanisms leading to ovarian cancers is critical for early diagnosis and treatment in order to elevate the 5-year survival rate. Chromatin remodeling is an important regulatory mechanism of gene expression, altering chromatin structures and influencing the activity of adjacent genes through chemical modification of the extrusive tail of histone by adding groups such as acetyl and methyl.(2) The enhancer of zeste homolog 2 (EZH2), which acts as a histone methyltransferase for lysine 27 of histone H3 and can also control DNA methylation, contributes to epigenetic silencing of target chromatin.(3)

EZH2 is involved in several key regulatory mechanisms such as control of embryonal development and cell proliferation.(3,4) Moreover, there is accumulating evidence indicating that EZH2 may play a pivotal role in the etiology of several solid tumors, including prostate cancer, breast cancer and bladder cancer, and EZH2 expression is associated with proliferative and more aggressive tumor phenotypes.(5–7) Notably, EZH2 appears to be not only a potential tumor marker but also may contribute to the deregulation of cell growth. Overexpression of EZH2 conferred a cellular growth advantage in vitro and promoted invasion in mantle cell lymphoma and prostate cancer.(8,9) Vice versa, suppression of EZH2 expression resulted in growth inhibition of prostate and renal cancer cells. (5,10)

Cyclin-dependent kinase inhibitors (CDKI), which are involved in tumor suppression and deregulated in many types of human cancers by epigenetic alterations, are a large family of proteins that regulate cell cycle progression, cell proliferation and differentiation. CDK inhibitor p57 has been previously reported to be inactivated in a variety of human cancers.(11) It was reported that p57 is a direct target of EZH2 and repressed in breast cancer by multiple epigenetic mechanisms.(12)

Lu et al.(13) examined EZH2 gene expression differences in purified endothelial cells from 10 invasive epithelial ovarian cancers and five normal ovaries and found EZH2 was elevated three to 4.3-fold in tumor-associated endothelial cells. Our previous study found overexpression of EZH2 contributes to acquired cisplatin resistance in ovarian cancer cells in vitro and in vivo,(14) but EZH2 expression and mechanism of action in ovarian cancer need to be further investigated. Therefore, we analyzed EZH2 expression in normal, benign and malignant ovarian specimens and the association of EZH2 expression with clinicopathological features of ovarian carcinoma patients. In addition, we demonstrated the functional role of EZH2 in p57 gene expression regulation, proliferation control and migration of ovarian cancer cells in vitro and in vivo.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Disclosure Statement
  8. References

Tissue samples.  We obtained 103 samples of ovarian tissues, excised surgically from patients in the Department of Gynecology, Xiehe Hospital, Tongji Medical College, Huazhong University of Science and Technology, during the period from April 2002 to October 2009. Each sample was divided into two portions, one was immediately frozen in liquid nitrogen after resection and the other was paraffin-embedded after fixation by 10% formalin. There were 17 normal ovarian tissues, 15 benign cystadenomas (five serous cystadenomas and 10 mucinous cystadenomas), eight borderline cystadenomas (four serous cystadenomas and four mucinous cystadenomas), 55 malignant epithelial tumors (25 serous cystadenocarcinomas, seven mucinous cystadenocarcinomas, 11 endometrioid carcinomas, 11 clear cell tumors and one undifferentiated tumor) and eight non-epithelial ovarian cancers (one immature teratoma, three granulosa cell tumors, two dysgerminomas, one blastoma and one metastatic cancer). The normal tissues were taken from patients receiving wedge biopsy of the ovaries, or adnexectomy due to myoma or adenomyosis. All patients had never been given chemotherapy or biotherapy before, and the diagnoses were confirmed pathologically in all cases. The patients with ovarian cancer were aged from 18 to 66 years, with an average of 44.6 years, and the median was 42 years. Specimen collection and archiving of patient data was performed with written informed consent and approved by the ethical committee of the hospital.

Cell culture.  Two ovarian cancer cell lines were investigated: A2780 and SKOV3 purchased from China Center for Type Culture Collection (CCTCC, Wuhan, China). The A2780 cell line was established from tumor tissue from an untreated patient with ovarian adenocarcinoma, and the SKOV3 was an ovarian adenocarcinoma cell line obtained from malignant ascites. Both the cell lines were cultured in RPMI-1640 medium (Gibco, Carlsbad, CA, USA) with 10% fetal calf serum (FCS; Gibco) at 37°C in a humidified atmosphere of 5% CO2.

Immunohistochemistry.  Following deparaffinization, endogenous peroxidase activity was inhibited by incubation with 0.3% H2O2 for 15 min. The antigens were retrieved by microwave treatment in citrate buffer (10 mM, pH 6). The slides were then washed with phosphate-buffered saline (PBS) and blocked with 5% BSA at room temperature for 20 min. The sections were incubated with rabbit anti-EZH2 (1:100; Invitrogen, Gaithersburg, MD, USA) or rabbit anti-p57 (1:200; Invitrogen, Gaithersburg, MD, USA) at 4°C overnight, followed by 1 h incubation with biotinylated goat anti-rabbit secondary antibody (1:100; Santa Cruz, CA, USA). Immunostaining was performed with diaminobenzidine (DAB) using the streptavidin–biotin complex/horseradish peroxidase (sABC-HRP) method. Sections were counterstained with hematoxylin. The criteria for positive results was as follows: positive cells were defined as having brown-yellow granules distributed in the nucleus or cytoplasm, with stain intensity higher than the unspecific background. Five visual fields with the most concentrated positive cells (×100) were chosen, observing 100 cells in each field (×400) and finally scoring the average percentage of positive cells in the five visual fields. A percentage over 20% was defined as a positive result.

RNA isolation and real-time polymerase chain reaction.  Total RNA was extracted from tissues and cell lines using Trizol Reagent (Invitrogen, Carlsbad, CA, USA) according to the instructions. After the quality and quantity of the extracted total RNA were confirmed by a spectrophotometer (DU650; Beckman, Berkeley, CA, USA), cDNA was synthesized using a reverse transcription kit (Toyobo, Osaka, Japan) according to the manufacturer’s protocol. The EZH2 primers were as follows: upstream 5′-TTGTTGGCGGAAGCGTGTAAAATC-3′; and downstream 5′-TCCCTAGTCCCGCGCAATGAGC-3′ (NM_004456). The p57 primers were as follows: 5′-GCGGCGATCAAGAAGCTGT-3′; and downstream 5′-ATCGCCCGACGACTTC TCA-3′ (NM_000076.2). The upstream sequence of β-actin primer were as follows: forward 5′-TCCTGTGGCATCC ACGAAACT-3′ and reverse 5′-GAAGCATTTGCGGTGGACGAT-3′. All reactions were performed on an Applied Biosystems 7300 Real-time PCR system (Applied Biosystems, Foster City, CA, USA). In brief, a master mixture was prepared on ice containing 1 μL of cDNA sample, 12.5 μL of SYBR Green Real-time PCR Master Mix (Toyobo) and 1 μL of 10 μM primers. The final volume was then adjusted to 25 μL with water. Reactions were carried out under the following cycling conditions: initial denaturation at 95°C for 1 min, followed by 40 cycles of denaturation at 95°C for 15 s, annealing at 56°C for 15 s and extension at 72°C for 45 s. After amplification, the products were subjected to a temperature gradient from 68 to 95°C at 0.2°C per second with continuous fluorescence monitoring to produce a melting curve of the products. Relative quantification was performed using the Ct (2Ct) method.(15) Each PCR amplification was performed in triplicate to verify the results.

Protein extraction and western blot analysis.  Cells were washed with cold phosphate-buffered saline (PBS) twice, scraped from sub-confluent plates and lysed in cold radioimmune precipitation assay (RIPA) buffer for 30 min. Cell lyses were centrifuged at 13 000g for 10 min, and the supernatants were collected and stored in aliquots at −80°C after the protein concentration was measured using the bicinchoninic acid (BCA) assay. Fifty micrograms of total protein was denatured, separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto nitrocellulose membranes. The membranes were blocked with blocking buffer (5% nonfat dry milk in tris–buffered saline containing 0.1% [v/v] Tween 20 [TBST]) for 2 h at room temperature, incubated with mouse anti-EZH2 polyclonal antibody (1:500 dilution; Cell Signal, Beverly, MA, USA), rabbit anti-p57 polyclonal antibody (1:1000 dilution; Invitrogen) or mouse anti-β-actin polyclonal antibody (1:500 dilution; Santa Cruz, CA, USA) at 4°C overnight, and then washed three times (10 min each time) with TBST. The primary antibodies were detected using Horseradish peroxidase-conjugated anti-mouse or anti-rabbit secondary antibody (1:5000 dilution; Santa Cruz) incubated for 2 h at room temperature and visualized by an enhanced chemiluminescence kit (Pierce, Rockford, IL, USA). Protein bands were quantitated after scanning using Quantity One software (Bio-Rad, Hercules, CA, USA).

EZH2 plasmid construction.  The sequence 5′-GCACACCCAACACTTATAAG-3′ was designed to target EZH2 short hairpin RNA (shRNA). Oligonucleotides coding for the targeting sequence were generated as follows: sense: 5′-CACCGCAACACCCAACACTTATAAGTTCAA GAGACTTATAAGTGTTGGGTGTTGCTTTTTTG-3′, and anti-sense: 5′-GATCCAAAAAAGCAA CACCCAACACTTATAAGTCTCTTGAACTTATAAGTGTTGGGTGTTGC-3′. Annealed double-stranded oligonucleotides were cloned into SuperSilencing shRNA expression vector (GenePharma, Shanghai, China).

EZH2 shRNA transient transfection.  A2780 and SKOV3 cells were seeded in a six-well plate at a concentration of 1 × 106 and 1.5 × 106 cells per well. Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) was used for transfection according to the instructions. Fresh growth medium was changed 6 h after transfection, and 48 h after transfection the cells were harvested for analysis. The empty vector was used as a negative control. To verify the knockdown efficiency, mRNA and protein of transient transfected cells were collected for real-time RT-PCR and western blot analysis as described above.

EZH2 shRNA stable transfection.  A2780 cells were seeded in a 24-well plate at a concentration of 2 × 105 cells per well. Lipofectamine 2000 (Invitrogen) was used for transfection according to the instructions. Fresh growth medium was changed 6 h after transfection, and cells was passaged at a 1:10 dilution into fresh medium 24 h after transfection. Selective medium with 300 μg/mL G418 was added the following day for 3 weeks. The empty vector was used as a negative control. To verify the knockdown efficiency, mRNA and protein of stable transfected cells were collected for real-time RT-PCR and western blot analysis as described above. Verification of EZH2 knockdown was determined by normalizing the levels of EZH2 to the control.

Cell proliferation assays.  Quantification of cell proliferation was performed using a BrdU Ellisa kit (Calbiochem, San Diego, CA, USA) according to the manufacturer’s instructions. Briefly, 48 h after transient transfection, A2780 and SKOV3 cells were harvested and seeded in triplicate at a density of 5000 per well in 96-well plates at 37°C under 5% CO2. After 48 h, the BrdU label was added and cells were incubated for an additional 4 h. After cell fixation, the incorporated BrdU was detected by anti-BrdU antibody for 1 h, followed by incubation with peroxidase-labeled goat anti-mouse IgG for 30 min. After substrate solution was added to each well for 15 min, absorbance was measured with a microplate reader (Bio-Rad) at dual wavelength of 450–595 nm. The values of absorbance from shEZH2 cells were compared with the value of absorbance from the control cells.

Cell migration assay.  Cell migration assay was performed using a Boyden chamber. Cells (1 × 105 per well) were trypsinized, resuspended in serum-free RPMI 1640 medium, and then added to the transwell inserts (6.5 mm diameter, 8 μm pore size with polycarbonate membrane; Corning Costar, Cambridge, MA, USA). The lower chamber beneath the insert membrane was added with 600 μL RPMI 1640 medium with 10% FBS. The transwell chambers were then incubated for 24 h under culture conditions. Migrated cells on the lower surface of the membrane were fixed with 70% ethanol and stained with hematoxylin–eosin (HE). Cell migration was quantitated microscopically by counting the number of stained nuclei at 10 random fields per high-power field (×400). All assays were performed in triplicate.

Ovarian tumor xenograft model.  All procedures involving mice were approved by the University Committee on Use and Care of Animals (UCUCA) at the hospital and conform to their relevant regulatory standards. Five-week-old male nude athymic BALB/c nu/nu mice (Hunan, China) were used for examining tumorigenicity. To evaluate the role of EHZ2 depletion in tumor formation, the EZH2-knockdown A2780 stable cells or vector control cells were propagated and 5 × 106 cells were inoculated subcutaneously into the dorsal flank of 10 mice (= 5 per group). Tumor size was measured every week, and tumor volumes were estimated using the formula (π/6) (L × W2), where L is the length of the tumor and W is the width.

After the mice developed signs and symptoms of chachexia, they were killed and their tumors were removed. Each tumor was divided into three portions, one was extracted for total RNA and another was for protein, and RNA and protein were collected for RT-PCR and western blot analysis. The remainder of the tumor tissues was used for histopathological study. After removal of the tumors, part of the tissues were fixed in 10% buffered formalin and processed for histopathological evaluation by paraffin embedding and hematoxylin–eosin staining. The histological features studied included the degree of anaplasia, mitotic activity, presence and amount of necrosis and degree of differentiation.

Statistical methods.  Statistical analysis was performed using spss 13.0 statistic software (SPSS, Chicago, IL, USA). All numerical data were expressed as mean ± standard deviation. Differences/correlations between two groups were assessed by the Mann–Whitney U-test, Student’s t-test and Pearson’s correlation test. Multiple group comparisons were analyzed using anova with a post hoc test for subsequent individual group comparisons. The association of EZH2 protein expression with ovarian carcinoma patient’s clinicopathological features was assessed by the Chi-squared test. P < 0.05 was considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Disclosure Statement
  8. References

EZH2 was upregulated in ovarian cancer.  To determine the change of EZH2 expression during ovarian neoplasm development, we performed immunohistochemistry on malignant, benign and normal ovarian tissues obtained from 103 patients. Representative immunohistochemical staining of EZH2 is shown in Figure 1A. We found frequent nuclear accumulation of EZH2 in carcinomas, but no or very weak EZH2 expression was identified in normal human ovary tissue cells. Positive expression of EZH2 was detected in 5.9% (1/17) normal ovarian epithelium, in 40% (6/15) benign cystadenomas, in 62.5% (5/8) borderline tumors and in 85.5% (47/55) ovarian carcinomas. The frequency of EZH2 positive expression significantly increased according to pathological progression (Fisher’s exact test, P < 0.001).

image

Figure 1.  (A) Immunohistochemical expression of EZH2 and p57 in ovarian tissues (normal: normal ovarian tissues; benign: serous cystadenoma; borderline: borderline serous cystadenoma; malignant: serous cystadenocarcinoma). Images were taken at ×200 magnification (superior panels) and a zoomed-in section is shown below each image. Positively immunostained EZH2 cells are those containing dark brown granules mainly distributed in the nuclei, with or without only small stain intensity in the cytoplasm. Positively immunostained p57 cells are those containing dark brown granules mainly distributed in the cytoplasm. (B) Relative expression of EZH2 mRNA in normal, benign, borderline and malignant ovarian tissues (C) Relative expression of EZH2 mRNA in benign, borderline and malignant mucinous neoplasm tissues. (D) Relative expression of EZH2 mRNA in benign, borderline and malignant serous neoplasm tissues. (E) Relative expression of p57 mRNA in normal, benign, borderline and malignant ovarian tissues. (F) Relative expression of p57 mRNA in benign, borderline and malignant mucinous neoplasm tissues. (G) Relative expression of p57 mRNA in benign, borderline and malignant serous neoplasm tissues. The median (solid black line), interquartile range (boxed) and range (enclosed by lines) of EZH2 and p57 mRNA expression are shown (outlying cases depicted by isolated points). (H) Forty specimens were randomly selected to determine the correlation between the expression levels of EZH2 and p57 mRNA. EHZ2 and P57 mRNA are inversely expressed in ovarian tissues (Pearson correlation r = −0.596, P < 0.0001).

Download figure to PowerPoint

We compared the EZH2 protein expression in serous and mucinous neoplasms with the spectrum of malignancy. The positive expression of EZH2 was detected in 20% (1/5) benign serous cystadenomas, in 50% (2/4) serous cystic tumors with borderline malignancy and in 84% (21/25) serous cystadenocarcinomas. The frequency of EZH2 positive expression significantly increased according to the malignancy of serous neoplasms (Fisher’s exact test, P = 0.009). Positive expression of EZH2 was detected in 50% (5/10) benign mucinous cystadenomas, in 75% (3/4) mucinous cystic tumors with borderline malignancy and in 100% (7/7) mucinous cystadenocarcinomas. No significant difference in mucinous neoplasms was found (Fisher’s exact test, P = 0.061) (Table 1).

Table 1.   EZH2 and p57 expression in serous and mucinous neoplasms
VariableEZH2 proteinp57 protein
Positive (%)Negative (%)PPositive (%)Negative (%)P
  1. Fisher’s exact test.

Serous neoplasms
 Benign1 (20.00)4 (80.00)0.0095 (100.00)0 (00.00)0.000
 Borderline2 (50.00)2 (50.00)4 (100.00)0 (00.00)
 Malignant21 (84.00)4 (16.00)6 (24.00)19 (76.00)
Mucinous neoplasms
 Benign5 (50.00)5 (50.00)0.0617 (70.00)3 (30.00)0.226
 Borderline3 (75.00)1 (15.00)2 (50.00)2 (50.00)
 Malignant7 (100.00)0 (00.00)2 (28.57)5 (71.43)

In addition, we performed quantitative real-time PCR for the expression level of EZH2 mRNA by using total RNA from the malignant (n = 55), borderline (n = 8), benign ovarian tumor tissues (n = 15) and normal ovarian tissues (n = 17) obtained. The relative expression level of EZH2 mRNA is shown in Figure 1B. EZH2 expression in benign, borderline and malignant ovarian tumor tissues increased 15- (23.9), 21- (24.4) and 137-fold (27.1), respectively, compared with normal ovarian tissues. There were significant differences in EZH2 mRNA levels among malignant, benign ovarian tumor tissues and normal ovarian tissues (anova, P < 0.001). No significant difference in EZH2-mRNA between benign cystadenoma and borderline tumors was found. Similar results were observed in the 34 serous and 21 mucinous tumors; EZH2 mRNA expression was positively correlated with tumor malignancy (anova, P < 0.05; Fig. 1C,D).

Inverse correlation between EZH2 and p57 expressions in ovarian tissues.  In order to explore the EZH2 target genes, we focused on the tumor suppressor gene p57. p57 Protein (immunohistochemistry) and mRNA (qRT-PCR) were analyzed in the 103 ovarian tissue specimens. The frequency of positive p57 protein expression in normal, benign, borderline and malignant ovarian tissues was 94.1% (16/17), 80% (12/15), 75% (6/8) and 27.3% (15/55), respectively, with a significant decline (Fisher’s exact test, P < 0.001) (Fig. 1A). Positive expression of p57 was detected in 100% (5/5) serous cystadenomas, in 100% (4/4) serous tumors with borderline malignancy and in 24% (6/19) serous cystadenocarcinomas. The frequency of p57 positive expression significantly declined according to serous neoplasm malignancy (Fisher’s exact test, P < 0.001). Positive expression of p57 was detected in 70% (7/10) mucinous cystadenomas, in 50% (2/4) mucinous tumors with borderline malignancy and in 28.57% (2/7) mucinous cystadenocarcinomas. No significant difference in mucinous neoplasms with the spectrum of malignancy was found (Fisher’s exact test, P < 0.226) (Table 1).

Similarly, real-time PCR showed descending mRNA expression of p57 from normal to benign, to borderline, and the least in malignant tissues. p57 Expression in benign ovarian tumor tissues was 2.9% (2−5.1) of that in normal ovarian tissues, in borderline was 1.6% (2−6.0) and in ovarian cancer was 0.3% (2−8.3). There were significant differences between normal ovarian tissues and benign ovarian tumor tissues (P = 0.004), and between benign ovarian tumor and malignant carcinomas (P = 0.043). Differences in p57 mRNA between benign and borderline tumors or between borderline tumors and malignant carcinomas were not significant (Fig. 1E). We also compared p57 mRNA expression in serous (n = 34) and mucinous neoplasms (n = 21) with the spectrum of malignancy. There were significant differences in p57 mRNA levels between benign and malignant serous or mucinous neoplasms, as well as between borderline and malignant mucinous neoplasms (anova, P < 0.05). No significant difference between benign and borderline serous or mucinous neoplasms was found (Fig. 1F,G).

Next, we examined the association between EZH2 and p57 mRNA in 40 randomly selected ovarian tissues samples. A statistically significant inverse correlation was observed between EZH2 and p57 mRNA (Fig. 1H), with high expression of EZH2 correlating with low amounts of p57 mRNA (Pearson’s correlation test, r = −0.596, P < 0.0001).

Association of EZH2 expression with clinicopathological features of ovarian carcinoma patients.  The association between EZH2 expression in ovarian carcinomas and several known clinicopathological features was studied further. Positive expression of EZH2 was detected in 47/55 (85.5%) epithelial ovarian cancers and in 3/8 (37.5%) non-epithelial ovarian cancers (Fisher’s exact test, P = 0.0071). Additionally, we found EZH2 protein expression was positively correlated with tumor histological grade and FIGO stage. The frequency of EZH2 positive expression increased significantly during cell differentiation defect from G1 (well differentiated) 4/11 (36.36%), to G2 (moderately differentiated) 19/23 (82.61%), then to G3 (poorly differentiated) 27/29 (93.10%) (χ2=13.09, P = 0.0003). Regarding FIGO stage, higher EZH2 protein expression was found in stage III/IV, 20/21 (95.24%), than in stage I/II, 30/42 (71.43%) (Fisher’s exact test, P = 0.0447). For lymph node metastasis status, all 20 ovarian tumor specimens from patients with lymph node metastasis had detectable EZH2 protein expression (20/20, 100%), which was significantly more frequent than in ovarian cancers without lymph node metastasis (28/41, 68.29%) (Fisher’s exact test, P = 0.0024). No significant correlation was obtained between EHZ2 expression and patient age (Table 2).

Table 2.   Association of EZH2 expression with patient’s clinicopathological features in ovarian cancer
VariableEZH2 protein
Positive (%)Negative (%)χ2P
  1. †Chi-squared test. ‡ Fisher’s exact test.

Age at surgery (years)
 ≤4229 (82.86)6 (17.14)0.58640.4435†
 >4221 (75.00)7 (25.00)  
Histological type
 Epithelial ovarian cancer47 (85.55)8 (14.45)_0.0071‡
  Serous21 (84.00)4 (16.00) 0.754‡
  Mucinous7 (100.00)0 (0.00)_
  Other19 (82.61)4 (17.39) 
 Non-epithelial ovarian cancer3 (37.50)5 (62.50)  
Histological grade
 G1 (well differentiated)4 (36.36)7 (63.64)13.090.0003†
 G2 (moderately differentiated)19 (82.61)4 (17.39)  
 G3 (poorly differentiated)27 (93.10)2 (6.90)  
FIGO stage
 I/II30 (71.43)12 (28.57)_0.0447‡
 III/IV20 (95.24)1 (4.76)  
Lymph node metastasis
 Yes22 (100.00)0 (00.00)_0.0024‡
 No28 (68.29)13 (31.71)  

Similarly, EZH2 mRNA expression was significantly correlated with histological grade, FIGO stage and the lymph node metastasis status of the tumors. EZH2 was overexpressed in ovarian cancers in advanced stage (Student’s t-test, P < 0.0001, Fig. 2A), poorly differentiated tumors (G3) (anova, P < 0.0001, Fig. 2B) and tumors with lymph node metastasis (Student’s t-test, P < 0.0001, Fig. 2C). However, no significant correlation was found between EHZ2 mRNA expression and patient age or histological type (data not shown).

image

Figure 2.  Relationship between EZH2 mRNA expression and the clinicopathological features of an ovarian carcinoma patient. Box-whisker plots showing labeling indices for EZH2 in relation to (A) tumor FIGO stage (stageI/II, n = 42; stage III/IV, n = 21), (B) tumor grade of differentiation (well differentiated G1, = 11; moderately differentiated G2, = 23; poorly differentiated G3, = 29) and (C) lymph node metastasis (no metastasis, = 41; metastasis, = 22). The median (solid black line), interquartile range (boxed) and range (enclosed by lines) of EZH2 expression are shown (outlying cases depicted by isolated points).

Download figure to PowerPoint

Loss of EZH2 increased p57 expression in ovarian cancer cell lines.  To investigate the biological effect of EZH2 expression in ovarian cancer cells, we chose the strategy to specifically silence EZH2 expression by shRNA and subsequently analyzed the resulting gene expression alteration. We generated a plasmid-mediated shRNA to knock down EZH2 expression in ovarian cancer cells A2780 and SKOV3. Untransfected and empty vector-transfected cells were used as controls. An 84.3%/80.7% decrease in the EZH2 mRNA level was observed in A2780/SKOV3-shEZH2 cells when compared with that in untransfected A2780/SKOV3 cells by real-time PCR (P < 0.001), whereas no significant difference in EZH2 mRNA expression was found between vector-transfected cells and untransfected cells (Fig. 3A). The efficiency of shRNA-mediated EZH2 silencing was further monitored at the protein expression level by western blot analysis, amounting to a 79.3%/77.3% decrease compared with control transfection of A2780/SKOV3 cells (P < 0.001, Fig. 3B,C).

image

Figure 3.  Loss of EZH2 increase p57 expression in ovarian cancer cell lines. A plasmid-mediated shRNA to transient transfect A2780 and SKOV3 (shEZH2) and knockdown EZH2 expression in ovarian cancer cell lines. Untransfected A2780/SKOV3 cells (NC) and empty vector-transfected cells (shVector) were used as controls. (A) EZH2 and p57 mRNA expression was detected by qRT-PCR. The results were normalized to β-actin expression and expressed as fold change relative to the corresponding negative control. (B) Representative immunoblots show that loss of EZH2 can increase p57 protein expression. (C) The expression of EZH2 and p57 was examined by western blot. The results were normalized to β-actin protein expression and expressed as fold change relative to the corresponding negative control. All data are representative of three independent experiments.

Download figure to PowerPoint

To evaluate the effect of EHZ2 depletion on p57 expression, real-time PCR and western blot were performed to detect the expression of p57 mRNA and protein in A2780/SKOV3-shEZH2 cells. The p57 mRNA level increased 54.3%/50.7% in A2780/SKOV3-shEZH2 cells, respectively, when compared with untransfected A2780/SKOV3 cells (< 0.001; Fig. 3A). The p57 protein expression level increased, amounting to 48.7%/44.3% compared with control transfection of A2780/SKOV3 cells (P < 0.001, Fig. 3B,C).

Loss of EZH2 suppresses ovarian cancer cell proliferation and migration.  To examine the effect of EZH2 depletion on cell biological behavior, we chose the BrdU incorporation assay and transwell migration assays to analyze the resulting cell proliferation and migration. The efficiency of shRNA-mediated EZH2 silencing was monitored at the mRNA expression level by real-time PCR, amounting to an 81.0%/79.3% decrease compared with control transfection of A2780/SKOV3 cells (P < 0.001, Fig. 4A). BrdU assays were performed in transient transfected cells and control cells. The BrdU incorporation assay demonstrated that the proliferation rate of shEZH2-A2780/SKOV3 cells was significantly reduced by 39.3%/31.3% compared with shVector-A2780/SKOV3 cells 96 h after transfection (P < 0.001, Fig. 4B). In vitro transwell migration assays were subsequently performed. EZH2-knockdown A2780/SKOV3 cells demonstrated a decreased ability to migrate through 8-μm pore size polycarbonate membrane compared with control cells (Fig. 4C, P < 0.001), suggesting that EZH2 knockdown could depress the migration ability of ovarian cancer cells.

image

Figure 4.  Effects of Loss of EZH2 on cell proliferation and migration of A2780 and SKOV3 cell lines. (A) The efficiency of shRNA-mediated EZH2 silencing was monitored at the mRNA expression level by real-time polymerase chain reaction. The results were normalized to β-actin expression and expressed as fold change relative to the corresponding shVector control (shVector). (B) Cell proliferation was measured by BrdU incorporation assay after 96 h transient transfection. The results expressed as fold change relative to the corresponding shVector control. (C) In vitro transwell migration assays were subsequently performed to examine the effect of EZH2 depletion on cell migration. All data are representative of three independent experiments.

Download figure to PowerPoint

Loss of EZH2 suppresses ovarian tumor formation in vivo.  To evaluate the role of EHZ2 in tumor formation, we established a nude mouse xenograft model in vivo. We concentrated on the A2780 cell line, which was found to be readily stable transfectable by shRNA. The EZH2-knockdown A2780 cells or vector control cells were inoculated subcutaneously into the dorsal flank of 10 nude mice. Untransfected A2780 cells and empty vector-transfected cells were used as controls. An 88.7% decrease in EZH2 mRNA level was observed in A2780-shEZH2 cells than in untransfected A2780 cells by real-time RT-PCR (< 0.001, Fig. 5A), whereas no significant difference in EZH2 mRNA expression was found between vector-transfected cells and untransfected cells (= 0.098). Tumor size was measured every week, EHZ2 depletion reduces A2780 tumor growth on the mouse xenograft model at 4 and 5 weeks (Fig. 5B,C, P < 0.05). Histopathological examination of the xenografts from the different experimental groups revealed a striking difference in the mitotic activity of the tumors. Xenografts formed by untransfected A2780 cells and empty vector-transfected cells exhibited high mitotic activity (mean 6.0 and 5.4 mitoses per high-power field, ×400). In contrast, tumors derived from A2780-shEZH2 cells had a markedly decreased number of mitoses (mean 1.8 mitoses per high-power field) (Fig. 5D). All tumors were poorly differentiated and histopathological examination did not reveal significant morphological differences.

image

Figure 5.  Loss of EZH2 suppresses ovarian tumor formation in vivo. (A) Untransfected A2780 cells and empty vector-transfected cells were used as controls. An 88.7% decrease in EZH2 mRNA level was observed in the A2780-shEZH2 cells compared with untransfected A2780 cells by real-time PCR. The results were normalized to β-actin expression and expressed as fold change relative to the corresponding untransfected A2780 control cells (NC). (B) Representative xenografts show the tumor volume of the A2780-shEZH2 group (shEZH2) was significantly decreased compared with the untransfected (NC) and empty shVector (shVector) xenograft group. (C) The tumor volume of the A2780-shEZH2 group was significantly decreased compared with the untransfected and empty shVector xenograft group 4 and 5 weeks after tumor cell inoculation, P < 0.001. (D) Histopathological study of the tumors shows that tumors formed by EZH2 shRNA (shEZH2) transfected A2780 cells had decreased mitotic activity than tumors derived from untransfected (NC) and empty shVector (shVector) transfected cells. The arrows show mitotic figures (hematoxylin–eosin stain; light microscopy, ×400). The bar graph shows quantification of the mitotic activity in the xenografts. The number of mitosis per high-power field was quantified by light microscopy in at least five high-power fields per tumor in all xenografts. (E) Western blot was performed to investigate the levels of EZH2 and p57 in tumor xenografts. Stable downregulation of EZH2 and high expression of p57 was found in shEZH2 xenografts compared with shVector xenografts. (F) Real-time PCR was performed to investigate the levels of EZH2 and p57 in tumor xenografts. Stable downregulation of EZH2 and high expression of p57 was found in shEZH2 xenografts compared with shVector xenografts (P < 0.001).

Download figure to PowerPoint

Western blot and real-time PCR were performed to investigate the levels of EZH2 and p57 in tumor xenografts. Downregulation of EZH2 and high expression of p57 were found in shEZH2 xenografts compared with shVector xenografts (Fig. 5E,F).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Disclosure Statement
  8. References

Epigenetic silencing of cancer suppressor gene can prevent programmed cell death and lead to uncontrolled proliferation in human cancers.(16) EZH2 has been implicated in the pathogenesis of human cancer via epigenetic mechanisms.(17) In the present study, for the first time, the expression of EZH2 at a protein and mRNA level in normal, benign and malignant ovarian tissues was analyzed. Our results show that the EZH2 gene is overexpressed in ovarian cancer, indicating that an increase in EZH2 expression is acquired during ovarian cancer tumorigenesis.

There are plenty of documents that indicated EZH2 was related to tumorigenesis and identified EZH2 as an oncogene in several tumor types.(5,18,19) Varambally(5) found that the EZH2 transcript was significantly increased in metastatic prostate cancer with respect to clinically localized prostate cancer. In addition, it was shown that amounts of both EZH2 mRNA and the corresponding protein were apparently increased in the malignant compared with benign prostate samples. By means of real-time PCR and immunohistochemistry, Matsukawa et al.(18) reported that, compared with non-cancerous gastric mucosa tissues, EZH2 was overexpressed in gastric cancer tissues and was related to lymph node metastasis and gastric cancer clinical stages. Many experiments also indicated that overexpression of EZH2 was related to the proliferation of neoplastic cells. Kleer et al.(19) discovered overexpression of EZH2 promoted anchoring, independent growth, invasion and metastasis of breast epithelial cells. Varambally et al.(5) reported small interfering RNA (siRNA) duplexes targeted against EZH2 reduced the amounts of EZH2 protein present in prostate cells and also inhibited cell proliferation in vitro. We analyzed the association between EZH2 expression and several known clinicopathological features in ovarian carcinomas. EZH2 expression was significantly correlated with the histological grade, FIGO stage and lymph node metastasis status of tumors. Moreover, inhibition of endogenous EZH2 reduced proliferation and migration of ovarian cancer cells in vitro, and suppressed ovarian tumor formation in vivo. Our results indicate that the EZH2 gene acts as an oncogene in tumorigenesis of ovarian cancer, and EZH2 can serve as a possible new therapeutic target for ovarian cancer treatment. Studies on other cancers suggest that EZH2 may be an interesting novel diagnostic or prognostic tumor marker. For example, EZH2 overexpression can be detected in early breast cancer development even before atypia is histologically evident, and therefore may be of use in identifying patients at risk of developing breast cancer.(6) Similarly, in the present study, expression of EZH2 was found significantly increased in benign cystadenomas. In addition, increased EZH2 expression has been linked to more aggressive tumor behavior and poor prognosis in cancers of the prostate, endometrium and breast.(5,6,19) Our results also showed EZH2 expression was significantly correlated with advanced FIGO stage and positive lymph node metastasis status of ovarian cancer. Thus, our findings should form a basis for future studies about the potential of EZH2 to serve as a diagnostic or prognostic marker for ovarian cancer.

The mechanisms by which the EZH2 protein can augment cellular proliferation and migration are not clear. p57 has been previously reported to be inactivated in a variety of human cancers and be implicated in regulation of cell cycle progression and proliferation.(20) Recently, Yang et al.(12) reported that p57 is a direct target of EZH2 and is repressed in breast cancer by multiple epigenetic mechanisms. Therefore, we detected the expression of p57 in ovarian tissue specimens, and a statistically significant inverse correlation was observed between EZH2 and p57. To investigate the biological relationship between EZH2 and p57 in ovarian cancer cells, we found EHZ2 depletion heightened p57 expression, indicating that p57 may be a target gene of EZH2 in ovarian cancer. Besides, the mechanisms may include interference with retinoic acid receptor signaling(21), activation of E2F-regulated genes,(8) repression of tumor suppressor genes, such as p16,(12,22) and regulation of transforming growth factor-b1.(23)

Which factors lead to EZH2 upregulation in human cancers? Varambally et al.(24) found that genomic loss of microRNA-101 leads to EZH2 overexpression in prostate cancer. Recent studies have indicated that EZH2 is a target protein of microRNA-101, which may be a keypoint factor of the EZH2 upstream signaling pathway.(24–26) EZH2 acts as a histone methyltransferase for lysine 27 of histone H3 and can also control DNA methylation. Acharyya et al.(27) showed that TNFalpha stimulated the recruitment of EZH2 and Dnmt-3b to coordinate histone and DNA methylation, and then affected the expression of EZH2. ANCCA is an AAA+ ATPase-containing nuclear coactivator that is crucial for the assembly of chromatin-modifying complexes. Recent biological and mechanistic investigations revealed that ANCCA overexpression correlates strongly with EZH2 in tumors and ANCCA controls expression of histone methyltransferase EZH2.(28) The predominant mechanisms of EZH2 upregulation in ovarian cancer need to be explored in advanced studies.

Besides the present study, similar results of EZH2 overexpression in ovarian carcinoma were also recently reported by Rao et al.(23) They found that high expression of EZH2 was positively correlated with an ascending histological grade and advanced stage of disease examined by immunohistochemisty.(23) Our results confirmed this conclusion on both protein and mRNA levels by immunohistochemisty and real-time PCR. Moreover, in the current study, EZH2 expression was found significantly increased in benign cystadenomas and borderline tumors, hinting the possibility of identifying patients at risk of developing ovarian tumors according to the EZH2 expression level. In the ovarian cancer lines we found EZH2 knockdown reduced cell proliferation and inhibited cell migration. Rao et al.(23) showed the same results in vitro. In addition, we established a nude mouse xenograft model and demonstrated EZH2 promoted ovarian tumor progression in vivo.

In conclusion, EZH2 expression, upregulated in ovarian cancer, was significantly correlated with poor cell differentiation, advanced pathoclinic stage and lymph node metastasis. EZH2 depletion increased the expression of p57 and reduced proliferation and migration of ovarian cancer cells, indicating that EZH2 acts as an oncogene in ovarian cancer by targeting p57. Our findings imply the potential of EZH2 to serve as a diagnostic marker and new therapeutic target for patients with ovarian cancer.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Disclosure Statement
  8. References

This work was supported by grant No. 30901585 from the National Natural Science Foundation of China.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Disclosure Statement
  8. References