The enhancer of zeste homolog 2 gene contributes to cell proliferation and apoptosis resistance in renal cell carcinoma cells



The enhancer of zeste homolog 2 (EZH2) gene has been recently linked to human malignancies where it may serve as a new target for cancer therapy. Here, we analyzed EZH2 expression in primary renal cell carcinoma (RCC) specimens and in nontumorous tissue samples from adult kidney. EZH2 transcripts were detectable in all RCC specimens examined. Expression levels were significantly higher in tumor tissue (p ≤ 0.0001) than in samples from normal adult kidney. Moreover, inhibition of endogenous EZH2 expression in RCC cell lines by RNA interference (RNAi) led to reduced proliferation and increased apoptosis in RCC cells. These data show that EZH2 is overexpressed in RCC. Furthermore, they indicate that the EZH2 gene plays a role for both the proliferation and the apoptosis resistance of RCC cells. Targeted inhibition of EZH2 could therefore represent a novel strategy to improve the therapeutic response of RCC. © 2008 Wiley-Liss, Inc.

Renal cell carcinoma (RCC) is estimated to account for more than 51,000 new cases and almost 13,000 cancer-related deaths in the United States in 2007, making it the second most lethal of the urological cancers.1 RCCs typically are highly resistant toward chemotherapy with a concomitant poor prognosis in advanced stages.2 Therefore, the identification of novel therapeutic targets and the development of new strategies for RCC treatment are urgently required.

The enhancer of zeste homolog 2 (EZH2) gene encodes a polycomb group (PcG) protein, which acts as a histone methyltransferase3–5 and also can directly control DNA methylation.6EZH2 is involved in several key regulatory mechanisms within eukaryotic cells, such as control of embryonal development or cell proliferation.7, 8 Moreover, there is accumulating evidence indicating that EZH2 may also play a pivotal role in the etiology of several tumor forms, which include prostate cancer9, 10 and breast cancer.10, 11 For both of these cancers, EZH2 expression is often observed in proliferative and more aggressive tumor subgroups and has diagnostic and/or prognostic value.9–11

Notably, however, EZH2 appears to be not only a potential tumor marker but may itself contribute to the deregulation of cell growth as a bona fide oncogene. Overexpression of EZH2 conferred cellular growth advantage in vitro,11–13 promoted invasion11 and exhibited oncogenic properties in nude mice.14Vice versa, inhibition of EZH2 expression by antisense constructs or RNA interference (RNAi) did result in growth inhibition of some cancer cells.9, 15 Furthermore, RNAi-mediated inhibition of EZH2 expression induced anoikis in circulating prostate carcinoma precursor cells16 and apoptotic cell death in breast cancer cells.17

A possible role of EZH2 for RCC has not been studied so far. Here, we analyzed EZH2 expression in primary RCC specimens and in nontumorous tissue. In addition, we investigated the functional role of EZH2 for the proliferation control and apoptosis regulation of RCC cells. We show that EZH2 is significantly overexpressed in primary RCC when compared with histologically normal renal tissue. Moreover, silencing of the EZH2 gene led to reduced cell proliferation and increased apoptosis in RCC cell lines. These data indicate that EZH2 plays a role both for the growth and the intrinsic apoptosis resistance of RCC tumor cells. Targeted inhibition of EZH2 function could thus represent a novel strategy to improve the therapeutic response of RCC.

Material and methods

Tissue samples

Fresh-frozen tissue samples of primary RCCs (n = 21) as well as macroscopically and histologically normal tissue derived from kidneys removed because of malignant disease (n = 12) were obtained from the tissue bank of the Department of Urology, University of Heidelberg. These specimens included 12 paired samples of tumor and adjacent nontumorous tissue. The work was covered by a votum of the ethical committee of the University of Heidelberg No. 206/2005. Written consent was obtained from each patient.

RNA extraction and quantitative real-time reverse transcription-polymerase chain reaction

RNA extraction was performed as previously described.18 In brief, RNA was isolated from homogenized tissues by phenol/chloroform extraction. The amount of RNA was quantified spectrophotometrically (NanoDrop ND 1000, NanoDrop Technologies, Wilmington, DE). Reverse transcription of 1 μg RNA was performed using the oligo-dT primer and SuperScriptIII First-Strand kit (Invitrogen, Karlsruhe, Germany). EZH2, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and hypoxanthine phosphoribosyl-transferase 1 (HPRT1) expressions were determined by real-time PCR with a 7300 Real-Time PCR System detector (Applied Biosystems, Forster City, CA), using SYBR green PCR Master Mix (Applied Biosystems), supplemented with 500 nM of each forward and reverse primer. EZH2 expression was determined using forward primer (EZH2-for): 5′-TTGTTG GCGGAAGCGTGTAAAATC-3′ (nt 205–229; NM_004456) and reverse primer (EZH2-rev): 5′-TCCCTAGTCCCGCGCAAT GAGC-3′ (nt 389–411).19GAPDH and HPRT1 primer sequences and cycling conditions have been previously described.18 The integrity of the PCR products was initially analyzed by agarose gel electrophoresis and subsequently checked by melting point analysis after each reaction. Relative quantification was performed using the comparative Ct (2−ΔΔCt) method.20 Data are presented as the fold difference in gene expression normalized to a housekeeping gene index (the geometric mean of the expression levels of GAPDH and HPRT1), as an endogenous reference, and relative to a calibrator sample. The housekeeping genes were chosen among several tested housekeeping genes for normalization of gene expression, since they exhibited equal amplification efficiencies as our gene of interest. U2OS osteosarcoma cancer cells were used as calibrator because of their documented EZH2 expression.12 A validation assay with serial dilutions of the calibrator was performed and the absolute value of the slope was less than 0.1 (data not shown).


EZH2 mRNA measurements were logtransformed to achieve data that can be assumed to be normally distributed. To compare the distributions of log-EZH2 between tumor and nontumor tissue, a mixed linear model with the patient as random factor was applied to account for paired data in 12 patients. The test for the difference in log-EZH2 between the different tissues is two-sided with a significance level of α = 0.05. Test analysis was carried out with the Statistical Analysis System, Version 9.1 for Windows (SAS Institute, Cary, NC). Statistical significances of cell proliferation and apoptosis assays were determined by applying the student t test.

Cells, synthetic small interfering RNAs and transfections

The following RCC cell lines were investigated: 769-P, 786-O, A-498, ACHN, CaKi-1, CaKi-2, A-704, KTCTL-2, KTCTL-30, KTCTL-53 and KTCTL-84. As a positive control for EZH2 expression, the osteosarcoma cell line U2OS was used. All cells were maintained in RPMI medium, supplemented with 10% fetal calf serum (FCS), except for 769-P, 786-O and U2OS cells, which were grown in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% FCS.

Small interfering RNAs (siRNAs) were chemically synthesized (Dharmacon Research, Lafayette). siEZH2 was generated against EZH2 mRNA target sequence: 5′-AAGACTCTGAATG CAGTTGCT-3′.12 siRNAs “siControl” (Dharmacon Research; contains at least 4 mismatches to all known human genes) and “siLamin” (Dharmacon Research) served as negative controls. siRNAs were transfected with oligofectamine (Invitrogen, Karlsruhe, Germany). In brief, 8 μl oligofectamine and 100 nM of individual siRNAs were diluted in Opti-MEM I reduced serum medium (Invitrogen) and mixed in a final volume of 400 μl transfection solution. Cells, plated on 6-cm dishes at 30–50% confluency, were transfected in 1.6 ml Opti-MEM I and 4 hr later supplemented with 1 ml DMEM containing 30% FCS.

Protein analyses

For Western blot analyses, 30 μg of protein extract were prepared as described previously,21 separated by 12.5% SDS-PAGE, transferred to a Immobilon-P membrane (Millipore, Billerica, MA), and analyzed by enhanced chemiluminescence (Amersham Biosciences, Piscataway, NJ). The following antibodies were used: anti-EZH2 antibody AC22 (Cell Signaling Technology, Danvers, MA), anti-Tubulin antibody CP06 (Calbiochem, Schwalbach, Germany) and anti-β-actin antibody (Sigma-Aldrich, Munich, Germany). The hybridoma producing the monoclonal anti-Lamin antibody was raised after the fusion of splenocytes from a male Balb/c mouse, which was immunized by injection of purified Lamin-hexa-His fusion protein, and the plasmacytoma line SP2/0, following standard procedures.

Cell count and cell cycle analyses

For cell count analyses, cell suspensions were diluted in CASYton (Innovatis AG, Reutlingen, Germany) 0–96 hr after transfection, as indicated. Total cells per milliliter were measured using a CASY TTC cell counter (Innovatis AG).

For cell cycle analyses, cells were trypsinized, washed in ice-cold phosphate-buffered saline (PBS) and fixed in 80% cold ethanol overnight at 4°C. Subsequently, cells were pelleted, resuspended in PBS containing 1 mg/ml RNAse A (Roche Diagnostics, Mannheim, Germany) and 2.1 μg/ml propidium iodide (Sigma-Aldrich), and then incubated for 30 min at 37°C. Cell cycle analyses were performed using a FACSCalibur (BD Biosciences, Heidelberg, Germany) with CellQuest Pro software provided by the manufacturer. Apoptotic cells were excluded and quantitation of the percentage of cells in individual cell cycle phases was performed using FlowJo software (Tree Star, Ashland, OR), applying the Dean-Jett-Fox model.22

Detection of apoptosis

TdT-mediated dUTP nick-end labeling (TUNEL) analyses for the detection of apoptosis were performed by using the in situ cell death detection kit (Roche Diagnostics). Poly-ADP-ribose-polymerase cleavage was detected by immunofluorescence employing monoclonal Cleaved PARP (Asp214) antibody (Cell Signaling Technology). Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (Roche Diagnostics). Poly-ADP-ribose-polymerase 1 cleavage, apoptotic strand breaks and total DNA were visualized by epifluorescence microscopy (Olympus Vanox-T, Hamburg, Germany).


Analysis of EZH2 expression in RCC, nontumorous kidney and RCC cell lines

To compare the expression of EZH2 gene in RCC and normal tissue, specimens from primary RCC (n = 21) as well as from macroscopically and histologically normal kidney (n = 12) were analyzed by qRT-PCR. These specimens included 12 paired samples of tumor and adjacent nontumorous tissue. EZH2 transcripts were detected in all tissue samples examined. Notably, however, EZH2 mRNA levels were significantly higher (p ≤ 0.0001) in tumor tissue than in samples from histologically normal adult kidney, also when analyzed for paired samples from identical patients (Fig. 1a). These data show that EZH2 is overexpressed in RCC at highly significant levels.

Figure 1.

EZH2 expression in RCC. (a) EZH2 mRNA expression levels were measured by qRT-PCR analysis of tissue specimens derived from RCC (n = 21) and from nontumorous kidney tissue adjacent to tumor (n = 12). This analysis included 12 paired samples of tumor and corresponding nontumorous tissue. Values were logtransformed. A mixed linear model with the patient as random factor was applied to account for paired data in 12 patients. Log-EZH2 measurements were visualized in box plots with upper whiskers drawn up to the maximum value below 3rd quartile + 1.5* (interquartile range), lower whiskers defined accordingly. The distributions of log-EZH2 values differed significantly between tumorous and nontumorous tissue (p ≤ 0.0001). (b) Western blot analysis of EZH2 protein levels in 10 tumor-derived RCC cell lines. Tubulin: loading control.

To allow subsequent functional analyses of the EZH2 gene in RCC cells, we further analyzed 10 tumor-derived RCC cell lines for their EZH2 expression. As shown in Figure 1b, EZH2 protein levels were readily detectable in all RCC cell lines investigated (Fig. 1b).

Silencing of EZH2 gene expression in RCC cells by RNAi

To investigate the biological effects of EZH2 expression in RCC cells, we chose the strategy to specifically silence EZH2 expression by RNAi and subsequently analyze the resulting phenotypic consequences. We employed a synthetic siRNA (siEZH2) that has been previously well-characterized as a specific inhibitor of EZH2 expression.12 The efficiency of siRNA-mediated EZH2 silencing was monitored at the protein expression level by Western blot analysis. We concentrated on 3 RCC cell lines (769-P, 786-O and CaKi-1), which were found to be readily transfectable by siRNA. U2OS osteosarcoma cells that express functionally relevant EZH2 protein amounts12 served as positive control. As shown in Figure 2, transfection of siEZH2 led to an almost complete depletion of intracellular EZH2-protein when compared with control transfections (siControl or siLamin), indicating high efficiency of RNAi-mediated silencing of the EZH2 gene.

Figure 2.

RNAi-mediated repression of EZH2 expression in RCC cell lines. 769-P, 786-O, Caki-1 and U2OS control cells were transfected with synthetic siRNAs, as indicated. EZH2 and Lamin protein levels were determined by immunoblotting. Beta-Actin: loading control.

EZH2 stimulates proliferation of RCC cells

Next, we investigated whether EZH2 plays a role for the proliferation capacity of RCC cells. Cell count analyses of 769-P and 786-O cells showed a significant (p < 0.05) reduction in cell numbers upon transfection with siEZH2 when compared with the respective control-transfected cells (Fig. 3a). Similar data was obtained for CaKi-1 and U2OS cells (not shown). These results indicate that siRNA-mediated inhibition of EZH2 expression can decrease the growth rate of RCC cell lines.

Figure 3.

RNAi-mediated silencing of EZH2 expression blocks growth of RCC cells. (a) Cell count analyses of 786-O and 769-P cells, treated with siControl or with siEZH2. Depicted are the relative increases in cell numbers at different time points, as indicated. Initial cell numbers (time point 0) were set at 1.0. Experiments were performed in triplicate, standard deviations are indicated. * = statistically significant (p < 0.05); n.s. = non-significant. (b) Cell cycle analyses of 786-O, 769-P and CaKi-1 cells, treated with siControl (left panels) or with siEZH2 (right panels). Percentages of cells in the G1, S or G2 phases of the cell cycle are indicated.

To investigate whether growth inhibition upon intracellular EZH2 depletion is related to alterations in the cell cycle profile of RCC cells, we performed fluorescent-activated cell sorting (FACS) analyses. We observed that interference with EZH2 expression was associated with induction of G1 arrest in all 3 RCC cell lines investigated. This was indicated by an increase in the number of cells in the G1 phase and a reduction in the number of cells in S phase, following siEZH2 treatment (Fig. 3b). Taken together, these data show that EZH2 is a significant determinant for the proliferation of renal cancer cells. Furthermore, EZH2 contributes to cell cycle progression from G1 to S phase in RCC cells.

Silencing of endogenous EZH2 gene expression in RCC cells can lead to apoptosis

Finally, we investigated whether EZH2 may also influence the apoptosis rate in RCC cell lines. To this end, we silenced EZH2 expression by RNAi and subsequently performed TUNEL analyses to stain for apoptotic cells. We observed that intracellular depletion of the EZH2 protein resulted in a significantly (p < 0.05) increased apoptotic rate in 786-O and CaKi-1 renal cancer cells, as well as in U2OS osteosarcoma cells (Fig. 4a). On the contrary, 769-P renal cancer cells did not show a significant increase in the number of apoptotic cells under the same experimental conditions. Percentages of apoptotic cells after EZH2 depletion by RNAi were quantified in Figure 4a, exemplary data for TUNEL analyses of 786-O cells (increase in apoptosis) and 769-P cells (no increase in apoptosis) are depicted in Figure 4b. As an independent assay to detect apoptosis, we also stained for cleavage of poly-ADP-ribose-polymerase (Fig. 4c). Again, EZH2 depletion resulted in increased apoptosis in 786-O, CaKi-1 and U2OS cells, but not in 769-P cells. These results show that interference with EZH2 expression can increase the apoptosis rate of RCC cells, thus indicating that the EZH2 gene contributes to their apoptotic resistance. However, the absence of apoptosis in 769-P cells suggests that the proapoptotic effect of EZH2 depletion is not a general phenomenon in RCC cells.

Figure 4.

RNAi-mediated silencing of EZH2 expression leads to apoptosis in RCC cells. (a) Percentage of TUNEL-positive cells, 72 hr after transfection of either siEZH2 or siControl. Experiments were performed at least 3 times, 3 microscopic fields were examined for each sample, standard deviations are indicated. * = statistically significant (p < 0.05); n.s. = non-significant. (b) Exemplary TUNEL analyses of 786-O and 769-P cells. Nuclei were stained with DAPI. Microscopic fields were selected in order to allow a comparison of similar numbers of cells. (c) Detection of Poly-ADP-Ribose-Polymerase (PARP) cleavage in 786-O, 769-P, CaKi-1 and U2OS cells, 72 hr after transfection of either siEZH2 or siControl. Scale bars represent 100 um.


Our study shows, for the first time, that the EZH2 gene is overexpressed in RCC. Moreover, our findings indicate that EZH2 expression is functionally relevant for this tumor, since analyses of RCC-derived cell lines show that EZH2 can contribute to their proliferation rate and to their apoptotic resistance.

These results extend the spectrum of tumors associated with EZH2 overexpression to RCC, a major killer worldwide. To the best of our knowledge, only 1 previous study dealt with the analysis of EZH2 expression in renal cell cancer. In this in situ hybridization analysis of 454 tissues derived from various cancer forms, a total of 34% stained moderate or strong for EZH2 expression, but none of the 8 kidney tumors analyzed.12 The greater sensitivity of our detection method could be one explanation for readily detecting EZH2 expression in all 21 RCC specimens investigated. Moreover, we found that RCCs exhibit significantly higher EZH2 expression levels than histologically normal kidney, indicating that an increase in EZH2 expression is acquired during RCC tumorigenesis.

Our functional analyses in RCC-derived cell lines support the notion that EZH2 expression is relevant for the growth rate of RCC cells. The mechanisms by which the EZH2 protein can augment cellular proliferation are still ill defined, but may include activation of E2F-regulated genes,12 interference with retinoic acid receptor signaling,23 stimulation of Wnt signaling24 and repression of tumor suppressor genes, such as p16.25, 26 Growth inhibition upon interference with EZH2 expression has been linked to G1 arrest in some tumor cell lines,12, 24 whereas G2/M arrest has been observed in others, either in combination with G1 arrest12 or alone.9, 15 In RCC cells, we detected an increase in G1 and a decrease in S phase populations upon intracellular EZH2 depletion, without evidence for an increase in G2 populations. Thus, our data indicates that the EZH2 protein stimulates progression primarily at the G1/S boundary in RCC cells.

There is, thus far, only very limited and partly contradictory data available about possible effects of EZH2 on apoptosis. No obvious signs for apoptosis were observed in a study of primary human embryonic lung fibroblasts, upon RNAi-mediated silencing of EZH2 expression.15 Similarly, siRNA treatment of nontumorigenic breast epithelial MCF-10A cells did not induce apoptosis, however, it led to increased apoptosis in MCF-7 breast cancer cells.17 This latter finding raises the possibility that apoptosis induction upon inactivation of EZH2 may be selective for tumor cells.17 Yet, siRNA-mediated inhibition of EZH2 expression did not affect the viability of adherently growing prostate-carcinoma precursor cells.16 Our results in 786-O and CaKi-1 cells support the notion that interference with EZH2 function can lead to increased apoptosis. However, the lack of detectable apoptosis in equivalently treated 769-P cells which—alike 786-O and CaKi-1 cells—are of clear cell carcinoma origin indicates that the apoptotic response may vary in RCC, even among tumor cells with the same histological background.

Studies in other cancers suggest that EZH2 may be an interesting novel diagnostic or prognostic tumor marker. For example, upregulation of EZH2 expression can be detected in early breast cancer development even before atypia is histologically evident and may therefore be of use to identify patients at risk for developing breast cancer.27 In addition, increased EZH2 expression has been linked to more aggressive tumor behavior and poor prognosis for some tumor entities, such as malignant melanoma and cancers of the breast, prostate and endometrium.9–11, 28 Thus, our finding that EZH2 is overexpressed in RCC should form a basis for future studies about the potential of EZH2 to serve as a diagnostic or prognostic marker for RCC, which would require higher patient numbers.

The observation in this study that silencing of EZH2 expression has antiproliferative and proapoptotic effects in RCC cells define EZH2 as a possible new therapeutic target for RCC treatment. In this context, it is noteworthy that the targeted inhibition of EZH2, and other PcG factors, has been proposed to represent a novel approach for cancer therapy.17 In line, there is recent preclinical data that interference with EZH2 is not only growth-inhibitory and proapoptotic in vitro, but can also exert antitumor effects in mouse models. For example, siRNAs against EZH2 led to significant tumor regression of established hepatocellular carcinomas,29 whereas stable siRNA treatment of human prostate cancer cells led to a greatly diminished potential to form tumors in vivo.16 Future strategies to block EZH2 function for therapeutic purposes may include the application of 3-Deazaneplanocin A, an agent which is able to disrupt EZH2 containing protein complexes,17 or the further development of nucleic acid-based approaches blocking EZH2 expression.16, 29 The observation that EZH2 is significantly overexpressed in RCC, when compared with normal kidney, may provide a therapeutic window for EZH2 inhibitors to preferentially attack RCC tumor cells.


N.W. was supported by a “Gerok” scholarship from the German Cancer Research Center.