PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Mol Cancer Res. Author manuscript; available in PMC 2010 September 1.
Published in final edited form as:
PMCID: PMC2794652
NIHMSID: NIHMS159142

Polycomb Group Protein EZH2 is an Oncogene that Promotes the Neoplastic Transformation of a Benign Prostatic Epithelial Cell Line

Abstract

Polycomb group protein EZH2 is a master-regulatory protein that plays a critical role in development as part of the Polycomb Repressive Complex 2 (PRC2). PRC2 controls numerous cell cycle and regulatory genes through tri-methylation of Histone 3, which results in chromatin condensation and transcriptional silencing. EZH2 overexpression has been correlated with high incidence of more aggressive, metastatic prostate cancers. While this correlation means EZH2 could prove valuable as a biomarker in clinical settings, the question remains whether EZH2 is actually responsible for the initiation of these more aggressive tumor types. In this study, EZH2-mediated neoplastic transformation of the normal prostate epithelial cell line BPH1 was confirmed by in vivo tumor growth and in vitro colony formation. Furthermore, EZH2 transformation resulted in increased invasive behavior of BPH1 cells, indicating that EZH2 may be responsible for aggressive behavior in prostate cancers. BPH1 was also transformed with the classic oncogenes myristoylated-Akt and activated Ras(V12) to allow phenotype comparisons with the EZH2 transformed cells. This study marks the first demonstration of neoplastic transformation in prostate cells mediated by EZH2, and establishes that EZH2 possesses stronger transforming activity than Akt, but weaker activity than activated Ras.

Keywords: EZH2, Prostate, Cancer, Initiation, Transformation

Introduction

Polycomb Group (PcG) protein Enhancer of Zeste 2 (EZH2) was first identified for its master regulatory role over the homeobox genes during development. By controlling spacial and temporal expression of various developmental genes, EZH2 and its family members determine body patterning and cell fate [1, 2]. Within the PcG family exist two complexes: Polycomb Repressive Complex 1 (PRC1) and Polycomb Repressive Complex 2 (PRC2), the latter of which is comprised of EZH2 and its binding partners EED and Su(z)12 [3, 4]. PRC2 is expressed by proliferating cells, and is responsible for silencing target genes through trimethylation of lysine 27 on histone 3 [59]. EZH2 is the member of PRC2 responsible for the methyltransferase activity via the C-terminal SET domain [10].

EZH2 was implicated in cancer aggression when it was found to be expressed at very high levels in proliferating mantle cell lymphoma samples [11]. Soon thereafter, a gene array comparing benign and metastatic prostate cancer samples found that EZH2 was consistently overexpressed in metastatic cancer [12]. Furthermore, EZH2 expression levels were found to be predictive of metastatic behavior in early stage, organ-confined prostate cancers. Subsequently, EZH2 was found to be overexpressed in the more aggressive forms of breast [13, 14], endometrium [15], melanoma [16], myeloma [17], and gastric [18] cancers, to name a few [19].

Additional evidence regarding EZH2 and its relationship to prostate cancer aggression continued to inundate the field. In situ hybridization experiments on advanced-stage prostate cancer samples found that in many cases, EZH2 overexpression was possibly due to gene amplification [20] or to a loss of microRNA mediated inhibited [21]. EZH2 was also validated as a biomarker that could be used to determine risk of prostate cancer recurrence in patients [2224]. Furthermore, it was confirmed that EZH2 was involved in maintaining proliferation and invasive behavior of some prostate cancer cell lines [25]. Yet despite this abundance of data on prostate cancer and prostate cancer cell lines, little work has been done to examine the role of EZH2 in cancer initiation. One study demonstrated that EZH2 promoted transformation of breast epithelial cells [26], but a parallel work in prostate epithelial cells has not been performed. However, tissue-specific differences between breast and prostate cells indicate that a transformation study in prostate cells is warranted. For instance, Androgen Receptor (AR), which is critical for the growth of prostate cells and not expressed by breast cells, is recruited to target genes by the histone demethylase responsible for reversing the histone modification made by EZH2 [27], indicating the strong possibility of alternative pathways and mechanisms that may be activated in prostate cells.

Benign Prostate Hyperplasia 1 (BPH1) is an epithelial cell line that was derived from a tissue biopsy and immortalized using SV40 Large T-antigen [28]. Following immortalization, BPH1 remained non-transformed and has been used extensively as a cell line representing a more normal prostate epithelium [29, 30]. BPH1 has become a widely accepted model in which to study the initiation of prostate cancer. The cell line has been transformed using co-culture with Cancer Associated Fibroblasts (CAF) [31, 32] and with urogenital sinus mesenchyme treated with testosterone and estradiol [33]. The transformed sublines of BPH1 have then been studied as early- stage versions of prostate cancer.

Two well known and classical oncogenes are myristoylated-Akt and Ras(V12), which are both constitutively active. Akt has been confirmed to play a signaling role in prostate cancer growth [34, 35], and promotes the development of precancerous prostatic intraepithelial neoplasia in a transgenic model [36]. Ras is most often found in cancers in a mutated, constitutively active form that provides constant mitogenic and growth signaling [37], reviewed in [38, 39]. While this is also predominantly true for prostate cancer, some studies have determined that simply overexpressing Ras can cause cancer phenotypes [40, 41]. Continuous Ras-pathway signaling, either by mutation or by overexpression, results in less dependency on AR signaling in prostate cancer cells [42]. This most likely contributes to development of late-stage Hormone-Refractory Prostate Cancer (HRPC) [43].

In this study, EZH2 was overexpressed in the normal prostatic epithelial cell line BPH1 to investigate the impact of EZH2 on prostate cancer initiation. The resulting data suggests that EZH2 is in fact a transforming factor for BPH1 cells, leading to a loss of contact-inhibition, an increase in invasive behavior, and tumor growth in vivo. Furthermore, this study directly compares EZH2 to the classic oncogenes Akt and Ras, thus allowing the strength of EZH2 as an oncogene in prostate cancer to be rated.

Results

Overexpressed EZH2 in BPH1 alters H3K27Me3 levels

The role of EZH2 in prostate cancer aggression has been repeatedly confirmed; however, no data exists on whether or not EZH2 is involved in the initiation of prostate cancer. It has previously been shown that prostate cancer cell lines express anywhere from 10- to 80-fold more EZH2 compared to normal prostatic epithelia [44]. The immortalized but non-transformed epithelial prostate cell line BPH1 was found to express lower levels of EZH2 than various prostate cancer cell lines by both quantitative Real-Time PCR (qRT-PCR) and by western blot (data not shown). Because EZH2 stability and activity is dependent upon binding to EED, the ratio between endogenous EZH2 and EED was also examined. BPH1 cells had nearly 3-fold higher expression of EED than EZH2 (EED/EZH2 ratio is 2.76; Figure 1B), implying the presence of unbound EED theoretically available to stabilize additional EZH2 activity. Taken together, this data indicated that BPH1 was an excellent cell line in which to study the impact of EZH2 overexpression.

Figure 1
Lentiviral schematics and EZH2 overexpression in BPH1

In order to overexpress EZH2 in a stable manner, Self-Inactivating (SIn) lentiviral vectors were used [45]. SIn lentiviruses are useful for overexpression studies because of their safety and stable integration into the genomes of infected cells. The control lentivirus for this study contained a Cytomegalovirus Immediate Early (CMV) promoter followed by the Encephalomyocarditis virus (EMCV) internal ribosomal entry site (IRES) driving expression of Enhanced Green Fluorescent Protein (EGFP; Figure 1A). Hereafter, the control virus will be referred to as GFP. For EZH2 overexpression, HA-tagged EZH2 was inserted into the lentiviral vector under control of the CMV promoter (Figure 1A). Following lentiviral transduction, BPH1-EZH2 had nearly 40 times higher expression of EZH2 compared to BPH1-GFP (EED/EZH2 ratio of 2.2) by qRT-PCR (Figure 1B). Expression of exogenous EZH2 was also confirmed by western blot, although significant degradation of the protein product was observed due to saturation of the available EED (EZH2/EED ratio of 10.4, Figure 1C, Supplemental Figure 1). By western blot densitometry, a 65% increase in EZH2 protein over endogenous levels was achieved in BPH1-EZH2 cells compared to BPH1-GFP cells.

To examine the functional consequence induced by EZH2 overexpression, immunocytochemical staining for Histone 3 Lysine 27 tri-methylation (H3K27Me3), a unique histone modification attributed to the EZH2 complex, was performed (Figure 2). While all cells stained positive for H3K27Me3 due to endogenous EZH2, cells that were overexpressing EZH2 had much higher levels of H3K27Me3 (indicated by brighter red staining, Figure 2). These results indicate that the overexpression of EZH2 caused a discernible epigenetic modification in the transduced BPH1 cells.

Figure 2
Immunocytochemistry confirmed increase in Histone 3 Lysine 27 Tri-Methylation in EZH2-overexpressing BPH1

EZH2 is a Stronger Transforming Factor than Akt

The transforming activity of EZH2 was compared to either the constitutively active myristoylated-Akt [46], or activated K-Ras(V12) mutant [47] to evaluate the strength of EZH2 as a transforming factor. Given that Ras is upstream of Akt, with considerably more downstream effectors, it was expected to be a stronger transforming factor than Akt. For these positive controls, BPH1 cells were marked at an MOI of 1 with either Akt or K-Ras overexpression lentivirus. Following lentiviral infection, BPH1-Akt cells expressed 30-fold more Akt and BPH1-Ras expressed 60-fold more Ras than uninfected BPH1 cells.

Because prior reports have implicated EZH2 in prostate cancer cell line proliferation [25], the mitogenic effects of EZH2 overexpression on BPH1 were investigated. CCK8 proliferation assay, which measures the number of live, metabolizing cells present in a sample, was performed on all BPH1 sublines (BPH1-GFP, BPH1-EZH2, BPH1-Ras, and BPH1-Akt). By this method, no discernable differences in proliferative rates were observed (Supplemental Figure 2). To evaluate the rate of cell death in each BPH1 subline population, Lactase Dehydrogenase (LDH) assays were performed. LDH is a stable cytoplasmic enzyme released into the culture medium upon plasma membrane damage in an apoptotic or damaged cell. Uninfected BPH1 and BPH-GFP both had very low levels of cell death by this assay. BPH1-EZH2, BPH1-Akt, and BPH1-Ras all had statistically significant (p=0.0006, 0.006, and 0.000034, respectively) increases in LDH activity, indicating increased cell turn over commonly associated with a transformed phenotype (Supplemental Figure 2). Taken in conjunction with the growth assay, it appears that the transformed cells may in fact proliferate more rapidly than the untransformed controls. However, the more rapid proliferation is balanced by a more rapid rate of cell turn over, resulting in a steady state of live cells in BPH1-EZH2, BPH1-Akt, and BPH1-Ras equivalent to the unmarked BPH1.

All BPH1 sublines were plated in a soft agar transformation assay. Uninfected BPH1 cells plated in soft agar resulted in very few colonies, verifying their non-transformed state (Supplemental Figure 3). Those spots that were visible were attributed to cells that were clustered at the time of plating. As predicted, both BPH1-Akt and BPH1-Ras cells grew colonies in soft agar. The BPH1-Ras grew into larger (Figure 3C) and more numerous (Figure 3B) colonies than BPH1-Akt (Figure 3A). BPH1-GFP showed no increase in colony forming activity over the uninfected BPH1 (Figure 3A, Supplemental Figure 3). When BPH1-EZH2 was used in the soft agar transformation assay, the cells were capable of growing sizable colonies. BPH1-EZH2 colonies were larger (Figure 3C) and more numerous (Figure 3B) than BPH1-Akt, but smaller and less numerous than BPH1-Ras. It was therefore concluded that EZH2 is an oncogene in the sense that it alone is sufficient to cause the neoplastic transformation of an otherwise benign prostate epithelial cell line. Furthermore, EZH2 can be placed within the spectrum of known oncogenes as stronger than Akt, but weaker than Ras in transforming capability.

Figure 3
EZH2 overexpression was sufficient to transform BPH1 cells

EZH2 Overexpression Causes an Increase in Invasion by BPH1 Cells

To explore the role of EZH2 in aggression, BPH1 sublines were assayed for invasive behavior towards either media containing 10% FBS (Figure 4A) or 3T3-conditioned media (Figure 4B). For all cell types, the 3T3-conditioned media stimulated more invasive behavior than the 10% FBS media (quantified in Table 1). Uninfected BPH1 and BPH1-GFP showed the least and BPH1-Ras showed the greatest amount of invasion (Figure 4A, B, Supplemental Figure 3). BPH1-Akt showed a significant increase in invasive behavior compared to BPH1-GFP in the FBS but not the 3T3 assay (Figure 4C,D). BPH1-EZH2 cells were more invasive than control cells towards both FBS and 3T3 media (Figure 4A,B). Collectively, the invasive behavior of EZH2-transformed BPH1 was comparable to BPH1-Akt and slightly lower than BPH1-Ras in both 3T3 and FBS assays (Figure 4A–D).

Figure 4
Overexpression of EZH2 increased invasive behavior of BPH1 cells
Table 1
Quantification of invasion by BPH1 sublines.

Oncogene-Induced Tumor Growth in SCID Mice

A more stringent assay for tumorigenicity is the ability to form a tumor in an in vivo environment. Prior to implantation into SCID mice, all BPH1 sublines were additionally marked with a Renilla Luciferase (RLuc) expressing lentivirus to facilitate monitoring of tumor growth. RLuc signal was verified via optical imaging on Day 0 immediately following implantation to confirm that each mouse received an equivalent number of cells (Figure 5A). Tumor growth was then monitored by optical imaging (data not shown) and caliper measurements (Figure 5B) until they reached one cm or for 12 weeks, whichever occurred first (Figure 5A). BPH1-Ras tumors (n=4) grew to one cm in three weeks, and BPH1-Akt tumors (n=3) in 12 weeks (Figure 5C). EZH2-transformed BPH1 tumors (n=8), however, were approximately 0.6 cm at 12 weeks. Consequently, one subgroup of BPH1-EZH2 tumor bearing mice (n=4) were monitored through 28 weeks of growth, when the tumors reached one cm (Figure 5C). At the 12 week endpoint, BPH1-Ras, BPH1-Akt, and BPH1-EZH2 tumor growth was confirmed by an increase in mass compared to the BPH1-GFP control (n=4) group (Figure 5D).

Figure 5
EZH2-mediated transformation of BPH1 cell line resulted in tumor growth in SCID mice

To confirm the functionality of lentivirally-introduced EZH2, qRT-PCR was performed on tumors from the 12 week endpoint. HoxA9, an unrelated developmental gene, is a known target of EZH2. While BPH1-GFP and BPH1-Ras tumors showed no change in HoxA9 transcript levels, BPH1-EZH2 tumors showed a marked decrease in HoxA9 expression (Figure 5E). Interestingly, in agreement with reports of Akt's negative regulation of EZH2 function [48], BPH1-Akt tumors showed a significant increase in HoxA9 expression (Figure 5E). EZH2 was also found to regulate Adrenergic Receptor β-2 (ADRB2), which, in turn, regulates the adhesion molecules β-catenin and integrin β4 [49]. The connection between EZH2 and ADRB2 regulation provides a plausible mechanism for EZH2-mediated cancer aggression. QRT-PCR was performed to examine ADRB2 levels in the BPH1 tumors. As with HoxA9, ADRB2 levels decreased with increased EZH2 activity and increased with decreased EZH2 activity (Figure 5F). However, changes in ADRB2 expression were not as dramatic as those seen for HoxA9.

When tumors were removed, it was noted that BPH1-Ras and BPH1-Akt tumors were more vascularized than EZH2-transformed BPH1 tumors (data not shown). Therefore, qRT-PCR was performed to evaluate the levels of endothelial growth factor VEGF-A (Figure 5G). Consistent with the visual inspection, elevated levels of VEGF-A mRNA were observed in the BPH1-Ras and BPH1-Akt samples but not BPH1-EZH2 (Figure 5G). This finding suggests that the failure of the EZH2 BPH1 tumors to recruit and establish adequate vasculature resulted in the very slow growth of these tumors. Consequently, BPH1-EZH2 cells were additionally transduced with a VEGF-A expressing lentivirus to stimulate angiogenesis. BPH1-EZH2/VEGF-A tumors grew significantly faster than BPH1-EZH2 tumors while there was no change in the behavior of the control BPH1-GFP/VEGF-A tumors (Figure 5B, C). In addition, BPH1-EZH2/VEGF-A tumors were significantly larger by mass than BPH1-GFP/VEGF-A tumors (p=0.002; Figure 5D, Table 2). Clear differences in vasculature were evident between BPH1-EZH2 and BPH1-EZH2/VEGF-A tumors upon tumor excision (Supplemental Figure 4) and were confirmed by histological evaluation (Figure 5H, Supplemental Figure 5). Most importantly, once neovascularization was induced in the BPH1-EZH2 tumors through VEGF-A expression, the EZH2-transformed BPH1 grew faster than BPH1-Akt tumors, but slower than BPH1-Ras tumors (Figure 5B). This data confirms the transforming ability of EZH2, and suggests that EZH2 is a stronger transforming factor than Akt, but a weaker transforming factor than Ras.

Table 2
p-values for BPH1 tumor measurements.

Discussion

In this study, EZH2 overexpression was sufficient to transform the prostate epithelial cell line BPH1 in both in vitro and in vivo assays. EZH2 was determined to be a stronger transforming factor than constitutively active, myristoylated Akt, but a weaker transforming factor than constitutively active Ras(V12). This study marks the first demonstration of EZH2-mediated neoplastic transformation of a prostate cell line. Furthermore, this is the first direct comparison between EZH2 and other, more classic oncogenes to score the strength of EZH2 as a transforming factor. EZH2 overexpression also produced an invasive phenotype in BPH1 cells, indicating that EZH2 alone is likely sufficient to promote prostate cancer aggression. EZH2 tumors, however, were poorly vascularized and grew very slowly. This phenotype was relieved by co-expression of VEGF-A, with the result of very rapid EZH2-mediated BPH1 tumor formation.

It has been well established that EZH2 is overexpressed in numerous cancer types, and that its overexpression correlates with a more metastatic phenotype. A great deal of evidence has been accumulating on the mechanism by which EZH2 becomes overexpressed in cancers. EZH2 gene expression was shown to be regulated by the Rb-E2F pathway, p16INK4a, by p53 [5053], and most recently by microRNA mediated repression [21], all of which are interrupted or damaged in most cancers. Therefore, EZH2 expression may increase due to loss of negative regulation by these pathways. However, the vast majority of cancers have lost functionality of p53 and RB and only a minority of cancers overexpress EZH2. Alternatively, in situ hybridization experiments on late-stage prostate cancer samples revealed amplification of the EZH2 gene locus that correlates with EZH2 overexpression and cancer aggression [20]. Although our system utilizes an artificial overexpression system, the end result of lentiviral gene introduction mimics the gene duplication observed in late-stage prostate cancer. It is, therefore, a relevant model in which to study the effects EZH2 overexpression on prostate cancer initiation.

Although EZH2 caused the transformation and invasion of the prostate epithelial cell line BPH1, it failed to change the proliferation rate. One study knocked down EZH2 using RNAi, and saw a decrease in the proliferation rate, concluding that EZH2 was critical for cell proliferation [25]. However the same study attempted to overexpress EZH2 in prostate cancer cell lines and failed to see an increase in proliferation rate. Considering that EZH2 controls expression levels of various cell cycle genes such as cyclin A and p16INK4a [54, 55], one would expect that overexpressing EZH2 would result in a significant effect on cellular proliferation. Instead, it would seem that EZH2 is necessary to maintain existing cellular proliferation, but is not sufficient to boost proliferation beyond the existing rate. Further investigation is needed to elucidate the mechanisms at work in regulating the cell cycle in the presence of EZH2 overexpression.

Recent studies provided some insight into possible mechanisms of EZH2-mediated transformation. EZH2 was found to regulate actin polymerization in prostate cancer cells [56], implying a system whereby EZH2 could provoke an increase in cell motility. EZH2 was also found to regulate ADBR2 in prostate cancer cells, which in turn regulates various adhesion molecules. Knocking down EZH2 in prostate cancer cell lines restored levels of ADRB2 and subsequently decreased invasive behavior [49]. The observed down-regulation of ADRB2 in our BPH1-EZH2 tumor model is consistent with the proposed mechanism of EZH2. Furthermore, EZH2 was found to negatively regulate the expression of p16INK4a through Rb(2) [57]. As p16 is a critical gene in G1-to-S cell cycle control, this down regulation by EZH2 is a possible step towards the transformation of benign cells. Interestingly, because expression levels of EZH2 are negatively regulated by p16INK4a through E2F, this down regulation also creates a positive feedback loop by which EZH2 upregulates its own expression [10, 50, 52]. Although in normal cells EZH2 expression should be controlled through Rb and p53, this regulation may be dysfunctional in our system because BPH1 cells were immortalized using SV40 large T-antigen, which antagonizes both proteins [28]. The loss of active p53 through the large T-antigen, combined with loss of p16INK4a expression initiated by the overexpression of exogenous EZH2, could explain how BPH1 cells were susceptible to transformation by EZH2. The question then remains whether EZH2 overexpression alone is adequate to drive tumor formation as a single hit, or if EZH2 overexpression must be combined with other mutations in a multiple hit model to result in transformation. Additional studies examining the effects of EZH2 overexpression in primary prostate cells will be beneficial to further elucidate this issue.

A similar study on the transforming capabilities of EZH2 was performed in the immortalized breast epithelial cell line H16N2 [26]. These cells were immortalized by HPV-16, which acts through the E6 and E7 genes to bind to and inhibit both Rb and p53 and thus present a similar system to the BPH1 cell line presented here. In the breast study, EZH2 was overexpressed through an adenoviral vector and shown to increase invasive behavior and soft agar colony formation. However, this study did not pursue the question of in vivo transformation through a tumor growth assay. This may have been due to the limitation presented by adenoviral-mediated overexpression of EZH2. Our study was able to present the tumorigenic properties of EZH2 in the context of in vivo tumor growth because the lentiviral overexpression vector permanently and stably expressed EZH2. This contributes a valuable and previously unpublished aspect of EZH2 transformation capabilities. Furthermore, because of the tissue specific characteristics unique to breast and prostate tissues, and the resulting signaling differences that must result from these characteristics, we considered it prudent to address the issue of EZH2-mediated neoplastic transformation directly in prostate cells rather than through inference since hormone receptor signaling may alter the impact of the overexpressed gene.

EZH2 overexpression has been repeatedly implicated in the aggressive behavior of prostate and numerous other cancers. Here, we have demonstrated that EZH2 overexpression may also be responsible for the initiating events of prostate cancer. It is critical that the effect of EZH2 overexpression in prostate tissue continues to be explored. Only a more thorough understanding of tissue-specific protein behavior will facilitate the development of treatments that can specifically target the aggressive subset of EZH2 overexpressing tumors.

Materials and Methods

Constructs

Lentiviruses were constructed in pCCL [58]. HA-EZH2 [59] was inserted into pCCL-CMV-IRES-EGFP to create pCCL-CMV-EZH2-IRES-EGFP. Empty vector was used as GFP control.

pCCL-CMV-RLuc [60], pCCL-CMV-VEGFA-IRES-EGFP [60], pRRL-CMV-Ras(V12)-IRES-EGFP [47], and FUW-Akt-IRES-GFP [46] were previously described.

Tissue Culture

Cells were cultured in media with 10% Fetal Bovine Serum (FBS) and 1% Penicillin-Streptomycin at 37°C with 5% CO2 and humidity. HEK-283T cells were cultured in Iscove's Modification of DMEM with L-Glutamine and 25 mM HEPES, without α-Thioglycerol and β-Mercaptoethanol, 3T3 cells were cultured in DMEM with 4.5 g/L glucose, L-glutamine and sodium pyruvate, and BPH1 cells were cultured in RPMI 1640 with L-Glutamine (Mediatech, Manassas, VA).

Lentivirus was produced by triple transfections into HEK-293T cells using calcium phosphate transfection protocol [61].

Lentiviral transductions were performed at an MOI of 1 for 6 hours with 8 μg/mL Polybrene (Sigma, St. Louis, MO). On Day 4, transduced cells were assayed for lentiviral gene expression and seeded for additional assays.

Viable cells were measured by the formazan dye-based CCK8 assay. Briefly, cells were plated in triplicate on Day 0 at 1×103 cells per well in 100 μL of media in a 96-well plate. On Days 1, 3, and 5, 10 μL of CCK8 assay reagent (Dojindo, Japan) was added to the wells and incubated for 2 hours. Plates were read at an absorbance of 450 nm on a Bio-Tek PowerwaveXS Plate Reader (Bio-Tek, Winooski, VT) and analyzed using KC Junior Software (Bio-Tek, Winooski, VT).

Western Blot

Samples were lysed in Whole Cell Lysis Buffer (50 mM HEPES, 150 mM NaCl, 1.5 mM MgCl2, 0.5 mM EDTA, 10% Glycerol, 1% Triton-X100, 10 mM NaF, 1 mM DTT, 1 mM PMSF, pH 7.0) by 3 freeze-thaw cycles followed by 30 minutes on ice. Lysates were spun down at 4000xg for 5 minutes, and supernatants were transferred to clean tubes. 25 μg of total protein per sample was separated by electrophoresis on 4–20% Tris-HCl SDS-PAGE. Antibodies used are as follows: HRP-conjugated anti-HA (1:100; Roche, Indianapolis, IN), anti-EZH2 (1:500; Upstate, Billerica, MA), anti-GFP (1:1000; Invitrogen, Carlsbad, CA), anti-β-actin (1:5000; Sigma, St. Louis, MO), HRP-conjugated anti-rabbit (1:20,000; Santa Cruz Biotechnology, Santa Cruz, CA), and HRP-conjugated anti-mouse (1:20,000; Santa Cruz Biotechnology, Santa Cruz, CA). Quantifiation of western blots performed by densitometry using ImageJ.

Microscopy

All cells and tissues were photographed using an Olympus BX41 fluorescent microscope (Olympus, Center Valley, PA) fitted with a Q-Imaging QICAM FAST 1394 camera (Surrey, BC Canada). Images were captured using the software QCapture Pro Version 5.1 (Media Cybernetics, Bethesda, MD) and processed using Adobe Photoshop CS (Adobe Systems, Inc., San Jose, CA) or ImageJ.

Immunocytochemistry and Immunohistochemistry

Cells for Immunocytochemistry were plated at 2.5×105 on growth treated, sterile glass coverslips in a 6 well plate [62] and allowed to attach and grow for 36 hours. Coverslips were washed in PBS-CM (PBS with 100 μM CaCl2 and 1 mM MgCl2) and fixed in 3% Paraformaldehyde (PFA) in PBS-CM for 20 minutes at room temperature. Cells were permeabilized for 5 minutes in 3% PFA-PBS-CM with 0.1% Triton-X100 and washed 3 times in PBS with 0.5% Bovine Serum Albumin (BSA). Coverslips were blocked for 20 minutes at room temperature in PBS with 3% BSA and 1% Normal Goat Serum (NGS). Antibodies were diluted in PBS with 0.1% BSA and 1% NGS. Coverslips were incubated in a humidity chamber with antibody either overnight at 4°C. Coverslips were mounted on a glass slide using Vectashield Hardmount with DAPI (Vector Labs, Burlingame, CA). Antibodies used were anti-Histone 3 Lysine 27 tri-methylation (1:750, Upstate, Billerica, MA) and Alexa Fluor 594 F(ab')2 fragment of goat anti-mouse IgG (1:1000, Invitrogen, Carlsbad, CA). Images shown in Figure 2 were captured using a 40x objective lens.

One cm tumors used for immunohistochemistry were fixed overnight in 3% PFA at 4C, followed by 5 minutes of washing and storage in 50% Ethanol. Tissues were embedded in paraffin and sectioned at the UCLA Translational Pathology Core Laboratory. Sections were subsequently processed as previously described [63]. Antibodies used were biotinylated anti-CD31 (1:300; BD Pharmingen, San Diego, CA), anti-GFP (1:100; Invitrogen, Carlsbad, CA), anti-HA (1:100, Roche, Indianapolis, IN), biotinylated anti-rabbit (1:100; Vector Labs, Burlingame, CA), biotinylated anti-rat (1:100; Vector Labs, Burlingame, CA), streptavidin-HRP (1:100, Perkin Elmer, Boston, MA), streptavidin-Cy3 (Jackson Labs, Westgrove, PA) and streptavidin-FITC (1:100; Invitrogen, Carlsbad, CA). Images shown in Figure 5 and Supplemental Figure 4 were captured using a 4x or 10x objective lens.

Matrigel Invasion Assay

The Invasion Assay protocol was adapted from previously reported studies [64]. Briefly, 24-well plate inserts with 8 micron membrane pores (BD Falcon, Franklin Lakes, NJ) were coated evenly with 20 μL of 1:6 Matrigel:Serum-Free Media dilution (BD Biosciences, Franklin Lakes, NJ), then allowed to set for 30 minutes at 37°C. 1×105 cells were plated in the top chamber in 500 μL of media containing 0.5% FBS. The bottom chamber was filled with 500 μL of either media containing 10% FBS or 3T3 conditioned serum-free media. Cells were allowed to invade for 48 hours.

Invaded cells were fixed in 0.5% Glutaraldehyde in 1x DPBS for 20 minutes, then the Matrigel layer was removed with a cotton swab. The membrane was cut out of the chamber using a No. 11 scalpel blade and mounted on a slide under a coverslip using Vectashield Hardmount with DAPI (Vector Labs, Burlingame, CA). Each membrane was quantified by capturing 5 independent fields under the 10x lens of the Olympus BX41 fluorescent microscope and determining the total area covered by cells in the field using ImageJ. The five fields were combined to obtain total membrane coverage. Images shown in Figure 4 were captured using a 4x objective lens.

Soft Agar Transformation Assay

The base layer was made by combining molten 1% agar equalized to 40°C with 2x RPMI/20%FBS in a 1:1 dillution and plating 100 μL in each well of a 96-well plate. The base layer was allowed to set for 20 minutes at room temperature. 5×103 cells per well were resuspended in 25 μL of 2x RPMI/20% FBS and mixed with 25 μL of molten 0.7% sterile agarose equalized to 40°C, then plated on the base layer. The top layer was allowed to set for 20 minutes, and plates were placed at 37°C with 5% CO2 and humidity. Colonies grew for 14 days then stained overnight with 0.1% INT-violet dye (Sigma, St. Louis, MO).

Plates were assayed at the UCLA Immunology Core on an Immunospot Series 1 Imager (Cellular Technologies Limited, Shaker Heights, OH). Colonies were photographed, analyzed, and counted using ImmunoSpot 4.0 Professional by CTL.

Animal Work and Optical Imaging

Animal care and procedures were performed in accordance with the University of California Animal Research Committee guidelines. Age-matched male SCID mice from Charles Rivers (Wilmington, MA) were used. Mice were implanted on the flank with 1×106 BPH1 cells marked with each respective lentivirus (at an MOI of 3) and subsequently with Renilla Luciferase lentivirus (at an MOI of 1). For BPH1-EZH2/VEGF-A and BPH1-GFP/VEGF-A tumors, the original cultures of BPH1-EZH2 or BPH1-GFP cells were additionally marked with pCCL-CMV-VEGF-A-IRES-EGFP lentivirus at an MOI of 1. These cells were not marked with pCCL-RLuc, so the overall lentiviral load on the cells remained unchanged from the original groups.

For each imaging session, mice were anesthetized with ketamine/xylazine (4:1). In vivo luciferase expression was monitored over time using a cooled IVIS CCD camera (Xenogen, Alameda, CA). Mice were given a tail-vein injection of coelenterazine at a dose of 1 mg/kg for Renilla Luciferase Imaging. Images were analyzed with IGOR-PRO Living Image Software (Xenogen, Alameda, CA). Tumor volumes were calculated using the formula: V = a × 2b × π/6, where a is the largest diameter and b is the smallest diameter [65, 66]. All animals were sacrificed when the largest diameter of the tumor reached 10mm or at 12 weeks post-implantation. Upon removal, tumors were harvested and snap frozen in liquid nitrogen to preserve RNA integrity or fixed in 3% PFA overnight for histology.

RNA Extraction and quantitative Real-Time PCR

RNA was extracted using TRIzol Reagent (Invitrogen, Carlsbad, CA). Two μg of RNA was reverse transcribed using iScript cDNA Synthesis Kit (Biorad, Hercules, CA). Quantitative Real-Time PCR was performed using 1 μL cDNA (roughly 40 ng), SyBr green 2x master mix (Applied Biosystems, Foster City, CA), 10 nM Fluorescein, and 10 μM each of the following primers: β-Actin 5'-TCAAGATCATTGCTCCTCCTGAGC-3', 5'-TACTCCTGCTTGCTGATCCACATC-3'; EZH2 5'-AGCGGATAAAGACCCCACC-3', 5'-CTGCTTCCCTATCACTGTC-3'; E E D 5'-GTAGAAGGGCACAGAGATG-3', 5'-GGCCTGTTAGTTTTATTTGG-3'; HoxA9 5'-TGCAGCTTCCAGTCCAAGG-3', 5'-GTAGGGGTGGTGGTGATGGT-3'; ADRB2 5'- TTCACGAACCAAGCCTATGCCA -3', 5' - AGCGGCCCTCAGATTTGTCAAT -3 ' ; V E G F-A 5 '-TGTACCTCCACCATGCCAAGT -3', 5'- CGCTGGTAGACGTCCATGAA -3'.

Reactions were run on MyiQ iCycler Real-Time PCR machine (Biorad, Hercules, CA) under the following cycling conditions: 40 repeats of 95°C/15sec; 60°C/30sec; 72°C/30sec, and analyzed using BioRad iQ5 software. All samples were normalized to internal β-Actin levels by the comparative threshold cycle (Ct) method [67].

Supplementary Material

Supplement 1

Complete Western Blots for Figure 1. UI: Uninfected BPH1 cells. GFP: BPH1 cells marked with GFP control lenti virus. EZH2: BPH1 cells marked with EZH2 overexpression lenti virus. α-EZH2: antibody specific for endogenous and HA-labeled EZH2. α-HA: antibody specific for HA tag on exogenous EZH2. α-GFP: antibody specific for GFP. α-Actin: antibody specific for β-Actin, used as a loading control. Molecular weight markers are shown on the extreme left of the blots.

Supplement 2

A. Transduced BPH1 cells were assayed for proliferation. Five day time course shows no change among the Uninfected (UI) BPH1 control, BPH1-GFP, BPH1-EZH2, BPH1-Akt, and BPH1-Ras cells. Results of four independent experiments are shown. B. Transduced BPH1 cells were assayed for cell death by Lactase Dehydrogenase assay. Seven day time course shows no change between the Uninfected (UI) BPH1 control and BPH1-GFP. However statistically significant increase in cell death was observed in BPH1-EZH2 (p=0.0006), BPH1-Akt (p=0.006), and BPH1-Ras (p=0.000034) cells indicating an increase in the rate of turnover in these populations.. Results of three independent experiments are shown. P-values were calculated using Student's T-Test relative to GFP control.

Supplement 3

A. Soft agar assay using uninfected BPH1 (UI) cells showed little to no colony formation, similar to BPH1-GFP cells shown in Figure 3. Those spots that did appear 37 were attributed to cells that were clustered upon plating. B. Matrigel Invasion assay using uninfected BPH1 (UI) cells towards both 10% FBS and 3T3 conditioned media stimulus. Level of invasive behavior was similar to that of the BPH1-GFP control shown in Figure 4.

Supplement 4

BPH1 tumors upon excision from SCID mice. BPH1-EZH2 tumors were extremely solid and avascular. BPH1-EZH2/VEGF-A tumors were significantly larger and very vascular. BPH1-GFP/VEGF-A cells did not grow from the time of implantation, but were also significantly vascularized.

Supplement 5

A. Additional panels of BPH1-EZH2 and BPH1-EZH2/VEGFA tumors stained for GFP, and vasculature (CD31). GFP expression is indicative of cells transduced with either EZH2 or VEGFA virus (in BPH1-EZH2/VEGFA sample). CD31 expression on vasculature is greatly increased in the BPH1-EZH2/VEGFA tumor compared to the BPH1-EZH2 tumor. Scale bar is 200 microns. B. BPH1-EZH2/VEGFA tumor stained for HA and CD31 under 10x magnification. HA staining is nuclear because of the localization of EZH2. Scale bar is 100 microns.

Acknowledgments

BDWK is supported by UCLA SPORE in Prostate Cancer Career Development Award.

Our thanks to Dr. Simon Hayward, Dr. Lee Slice, and Dr. Li Xin for their generous contributions of BPH1 cell line, pRRL-CMV-Ras(V12)-IRES-EGFP, and FUW-Akt-IRES-GFP, respectively.

We also thank Drs. Maarten van Lohuizen and Thomas Jenuwein for supplying the initial HA-tagged EZH2 construct.

Flow Cytometry was performed in the UCLA Jonsson Comprehensive Cancer Center and Center for AIDS Research Flow Cytometry Core Facility that is supported by the National Institutes of Health Awards CA-16042 and AI-28697, by the Jonsson Cancer Center, the UCLA AIDS Institute, and the UCLA School of Medicine.

BioSpot was performed in the UCLA Immuno/BioSpot Core Facility that is supported by the UCLA Center for AIDS Research (CFAR) NIH/NIAID AI028697 and the David Geffen School of Medicine at UCLA.

References

1. Brock HW, van Lohuizen M. The Polycomb group--no longer an exclusive club? Curr Opin Genet Dev. 2001;11:175–181. [PubMed]
2. Orlando V. Polycomb, epigenomes, and control of cell identity. Cell. 2003;112:599–606. [PubMed]
3. Cao R, Zhang Y. SUZ12 is required for both the histone methyltransferase activity and the silencing function of the EED-EZH2 complex. Mol Cell. 2004;15:57–67. [PubMed]
4. Sewalt RG, van der Vlag J, Gunster MJ, et al. Characterization of interactions between the mammalian polycomb-group proteins Enx1/EZH2 and EED suggests the existence of different mammalian polycomb-group protein complexes. Mol Cell Biol. 1998;18:3586–3595. [PMC free article] [PubMed]
5. Bunker CA, Kingston RE. Transcriptional repression by Drosophila and mammalian Polycomb group proteins in transfected mammalian cells. Mol Cell Biol. 1994;14:1721–1732. [PMC free article] [PubMed]
6. Cao R, Wang L, Wang H, et al. Role of histone H3 lysine 27 methylation in Polycomb-group silencing. Science. 2002;298:1039–1043. [PubMed]
7. Kirmizis A, Bartley SM, Kuzmichev A, et al. Silencing of human polycomb target genes is associated with methylation of histone H3 Lys 27. Genes Dev. 2004;18:1592–1605. [PubMed]
8. Cao R, Zhang Y. The functions of E(Z)/EZH2-mediated methylation of lysine 27 in histone H3. Curr Opin Genet Dev. 2004;14:155–164. [PubMed]
9. Jenuwein T, Allis CD. Translating the histone code. Science. 2001;293:1074–1080. [PubMed]
10. Muller J, Hart CM, Francis NJ, et al. Histone methyltransferase activity of a Drosophila Polycomb group repressor complex. Cell. 2002;111:197–208. [PubMed]
11. Visser HP, Gunster MJ, Kluin-Nelemans HC, et al. The Polycomb group protein EZH2 is upregulated in proliferating, cultured human mantle cell lymphoma. Br J Haematol. 2001;112:950–958. [PubMed]
12. Varambally S, Dhanasekaran SM, Zhou M, et al. The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature. 2002;419:624–629. [PubMed]
13. Collett K, Eide GE, Arnes J, et al. Expression of enhancer of zeste homologue 2 is significantly associated with increased tumor cell proliferation and is a marker of aggressive breast cancer. Clin Cancer Res. 2006;12:1168–1174. [PubMed]
14. Raaphorst FM, Meijer CJ, Fieret E, et al. Poorly differentiated breast carcinoma is associated with increased expression of the human polycomb group EZH2 gene. Neoplasia. 2003;5:481–488. [PMC free article] [PubMed]
15. Bachmann IM, Halvorsen OJ, Collett K, et al. EZH2 expression is associated with high proliferation rate and aggressive tumor subgroups in cutaneous melanoma and cancers of the endometrium, prostate, and breast. J Clin Oncol. 2006;24:268–273. [PubMed]
16. McHugh JB, Fullen DR, Ma L, Kleer CG, Su LD. Expression of polycomb group protein EZH2 in nevi and melanoma. J Cutan Pathol. 2007;34:597–600. [PubMed]
17. Croonquist PA, Van Ness B. The polycomb group protein enhancer of zeste homolog 2 (EZH 2) is an oncogene that influences myeloma cell growth and the mutant ras phenotype. Oncogene. 2005;24:6269–6280. [PubMed]
18. Mattioli E, Vogiatzi P, Sun A, et al. Immunohistochemical analysis of pRb2/p130, VEGF, EZH2, p53, p16(INK4A), p27(KIP1), p21(WAF1), Ki-67 expression patterns in gastric cancer. J Cell Physiol. 2007;210:183–191. [PubMed]
19. Gil J, Bernard D, Peters G. Role of polycomb group proteins in stem cell self-renewal and cancer. DNA Cell Biol. 2005;24:117–125. [PubMed]
20. Saramaki OR, Tammela TL, Martikainen PM, Vessella RL, Visakorpi T. The gene for polycomb group protein enhancer of zeste homolog 2 (EZH2) is amplified in late-stage prostate cancer. Genes Chromosomes Cancer. 2006;45:639–645. [PubMed]
21. Varambally S, Cao Q, Mani RS, et al. Genomic loss of microRNA-101 leads to overexpression of histone methyltransferase EZH2 in cancer. Science. 2008;322:1695–1699. [PMC free article] [PubMed]
22. Laitinen S, Martikainen PM, Tolonen T, et al. EZH2, Ki-67 and MCM7 are prognostic markers in prostatectomy treated patients. Int J Cancer. 2008;122:595–602. [PubMed]
23. Rhodes DR, Sanda MG, Otte AP, Chinnaiyan AM, Rubin MA. Multiplex biomarker approach for determining risk of prostate-specific antigen-defined recurrence of prostate cancer. J Natl Cancer Inst. 2003;95:661–668. [PubMed]
24. Yu J, Yu J, Rhodes DR, et al. A polycomb repression signature in metastatic prostate cancer predicts cancer outcome. Cancer Res. 2007;67:10657–10663. [PubMed]
25. Bryant RJ, Cross NA, Eaton CL, Hamdy FC, Cunliffe VT. EZH2 promotes proliferation and invasiveness of prostate cancer cells. Prostate. 2007;67:547–556. [PubMed]
26. Kleer CG, Cao Q, Varambally S, et al. EZH2 is a marker of aggressive breast cancer and promotes neoplastic transformation of breast epithelial cells. Proc Natl Acad Sci U S A. 2003;100:11606–11611. [PubMed]
27. Wissmann M, Yin N, Muller JM, et al. Cooperative demethylation by JMJD2C and LSD1 promotes androgen receptor-dependent gene expression. Nat Cell Biol. 2007;9:347–353. [PubMed]
28. Hayward SW, Dahiya R, Cunha GR, et al. Establishment and characterization of an immortalized but non-transformed human prostate epithelial cell line: BPH-1. In Vitro Cell Dev Biol Anim. 1995;31:14–24. [PubMed]
29. Mitchell S, Abel P, Ware M, Stamp G, Lalani E. Phenotypic and genotypic characterization of commonly used human prostatic cell lines. BJU Int. 2000;85:932–944. [PubMed]
30. Untergasser G, Rumpold H, Hermann M, et al. Proliferative disorders of the aging human prostate: involvement of protein hormones and their receptors. Exp Gerontol. 1999;34:275–287. [PubMed]
31. Phillips JL, Hayward SW, Wang Y, et al. The consequences of chromosomal aneuploidy on gene expression profiles in a cell line model for prostate carcinogenesis. Cancer Res. 2001;61:8143–8149. [PubMed]
32. Shou J, Soriano R, Hayward SW, et al. Expression profiling of a human cell line model of prostatic cancer reveals a direct involvement of interferon signaling in prostate tumor progression. Proc Natl Acad Sci U S A. 2002;99:2830–2835. [PubMed]
33. Cunha GR, Hayward SW, Wang YZ. Role of stroma in carcinogenesis of the prostate. Differentiation. 2002;70:473–485. [PubMed]
34. Majumder PK, Sellers WR. Akt-regulated pathways in prostate cancer. Oncogene. 2005;24:7465–7474. [PubMed]
35. Manning BD, Cantley LC. AKT/PKB signaling: navigating downstream. Cell. 2007;129:1261–1274. [PMC free article] [PubMed]
36. Majumder PK, Yeh JJ, George DJ, et al. Prostate intraepithelial neoplasia induced by prostate restricted Akt activation: the MPAKT model. Proc Natl Acad Sci U S A. 2003;100:7841–7846. [PubMed]
37. Sansom OJ, Meniel V, Wilkins JA, et al. Loss of Apc allows phenotypic manifestation of the transforming properties of an endogenous K-ras oncogene in vivo. Proc Natl Acad Sci U S A. 2006;103:14122–14127. [PubMed]
38. Bos JL. ras oncogenes in human cancer: a review. Cancer Res. 1989;49:4682–4689. [PubMed]
39. Malumbres M, Barbacid M. RAS oncogenes: the first 30 years. Nat Rev Cancer. 2003;3:459–465. [PubMed]
40. Rosen DG, Yang G, Bast RC, Jr., Liu J. Use of ras-transformed human ovarian surface epithelial cells as a model for studying ovarian cancer. Methods Enzymol. 2005;407:660–676. [PubMed]
41. Traynor P, McGlynn LM, Mukhergee R, et al. An increase in N-Ras expression is associated with development of hormone refractory prostate cancer in a subset of patients. Dis Markers. 2008;24:157–165. [PubMed]
42. Bakin RE, Gioeli D, Sikes RA, Bissonette EA, Weber MJ. Constitutive activation of the Ras/mitogen-activated protein kinase signaling pathway promotes androgen hypersensitivity in LNCaP prostate cancer cells. Cancer Res. 2003;63:1981–1989. [PubMed]
43. Weber MJ, Gioeli D. Ras signaling in prostate cancer progression. J Cell Biochem. 2004;91:13–25. [PubMed]
44. Chen H, Tu SW, Hsieh JT. Down-regulation of human DAB2IP gene expression mediated by polycomb Ezh2 complex and histone deacetylase in prostate cancer. J Biol Chem. 2005;280:22437–22444. [PubMed]
45. Miyoshi H, Blomer U, Takahashi M, Gage FH, Verma IM. Development of a self-inactivating lentivirus vector. J Virol. 1998;72:8150–8157. [PMC free article] [PubMed]
46. Xin L, Lawson DA, Witte ON. The Sca-1 cell surface marker enriches for a prostate-regenerating cell subpopulation that can initiate prostate tumorigenesis. Proc Natl Acad Sci U S A. 2005;102:6942–6947. [PubMed]
47. Pham H, Eibl G, Vincenti R, et al. 15-Hydroxyprostaglandin dehydrogenase suppresses K-RasV12-dependent tumor formation in Nu/Nu mice. Mol Carcinog. 2008;47:466–477. [PubMed]
48. Cha TL, Zhou BP, Xia W, et al. Akt-mediated phosphorylation of EZH2 suppresses methylation of lysine 27 in histone H3. Science. 2005;310:306–310. [PubMed]
49. Yu J, Cao Q, Mehra R, et al. Integrative genomics analysis reveals silencing of beta-adrenergic signaling by polycomb in prostate cancer. Cancer Cell. 2007;12:419–431. [PubMed]
50. Bracken AP, Pasini D, Capra M, et al. EZH2 is downstream of the pRB-E2F pathway, essential for proliferation and amplified in cancer. Embo J. 2003;22:5323–5335. [PubMed]
51. Muller H, Bracken AP, Vernell R, et al. E2Fs regulate the expression of genes involved in differentiation, development, proliferation, and apoptosis. Genes Dev. 2001;15:267–285. [PubMed]
52. Reynolds PA, Sigaroudinia M, Zardo G, et al. Tumor suppressor p16INK4A regulates polycomb-mediated DNA hypermethylation in human mammary epithelial cells. J Biol Chem. 2006;281:24790–24802. [PubMed]
53. Tang X, Milyavsky M, Shats I, et al. Activated p53 suppresses the histone methyltransferase EZH2 gene. Oncogene. 2004;23:5759–5769. [PubMed]
54. Attwooll C, Oddi S, Cartwright P, et al. A novel repressive E2F6 complex containing the polycomb group protein, EPC1, that interacts with EZH2 in a proliferation-specific manner. J Biol Chem. 2005;280:1199–1208. [PubMed]
55. Tonini T, Bagella L, D'Andrilli G, Claudio PP, Giordano A. Ezh2 reduces the ability of HDAC1-dependent pRb2/p130 transcriptional repression of cyclin A. Oncogene. 2004;23:4930–4937. [PubMed]
56. Bryant RJ, Winder SJ, Cross SS, Hamdy FC, Cunliffe VT. The Polycomb Group protein EZH2 regulates actin polymerization in human prostate cancer cells. Prostate. 2008;68:255–263. [PubMed]
57. Kotake Y, Cao R, Viatour P, et al. pRB family proteins are required for H3K27 trimethylation and Polycomb repression complexes binding to and silencing p16INK4alpha tumor suppressor gene. Genes Dev. 2007;21:49–54. [PubMed]
58. Dull T, Zufferey R, Kelly M, et al. A third-generation lentivirus vector with a conditional packaging system. J Virol. 1998;72:8463–8471. [PMC free article] [PubMed]
59. Laible G, Wolf A, Dorn R, et al. Mammalian homologues of the Polycomb-group gene Enhancer of zeste mediate gene silencing in Drosophila heterochromatin and at S. cerevisiae telomeres. Embo J. 1997;16:3219–3232. [PubMed]
60. Burton JB, Priceman SJ, Sung JL, et al. Suppression of prostate cancer nodal and systemic metastasis by blockade of the lymphangiogenic axis. Cancer Res. 2008;68:7828–7837. [PMC free article] [PubMed]
61. Soneoka Y, Cannon PM, Ramsdale EE, et al. A transient three-plasmid expression system for the production of high titer retroviral vectors. Nucleic Acids Res. 1995;23:628–633. [PMC free article] [PubMed]
62. Aschenbrenner L, Lee T, Hasson T. Myo6 facilitates the translocation of endocytic vesicles from cell peripheries. Mol Biol Cell. 2003;14:2728–2743. [PMC free article] [PubMed]
63. Brakenhielm E, Burton JB, Johnson M, et al. Modulating metastasis by a lymphangiogenic switch in prostate cancer. Int J Cancer. 2007;121:2153–2161. [PMC free article] [PubMed]
64. Lochter A, Werb Z, Bissell MJ. Transcriptional regulation of stromelysin-1 gene expression is altered during progression of mouse mammary epithelial cells from functionally normal to malignant. Matrix Biol. 1999;18:455–467. [PMC free article] [PubMed]
65. Sorensen AG, Patel S, Harmath C, et al. Comparison of diameter and perimeter methods for tumor volume calculation. J Clin Oncol. 2001;19:551–557. [PubMed]
66. Tomayko MM, Reynolds CP. Determination of subcutaneous tumor size in athymic (nude) mice. Cancer Chemother Pharmacol. 1989;24:148–154. [PubMed]
67. Winer J, Jung CK, Shackel I, Williams PM. Development and validation of real-time quantitative reverse transcriptase-polymerase chain reaction for monitoring gene expression in cardiac myocytes in vitro. Anal Biochem. 1999;270:41–49. [PubMed]