Recent studies have shown that small interfering RNA (siRNA) silences genes at the transcriptional level in human cells. However, the therapeutic potential of siRNA-mediated transcriptional gene silencing remains unclear. Here, we show that siRNA targeted to the urokinase plasminogen activator (uPA) promoter induced epigenetic transcriptional silencing in human prostate cancer cells. This silencing resulted in a dramatic reduction of tumor cell invasion and angiogenesis in vitro. Furthermore, the results from a bioluminescence tumor/metastasis model showed that the silencing of uPA significantly inhibits prostate tumor growth and the incidence of lung metastasis. Our findings represent a potentially powerful new approach to not only epigenetic silencing of metastasis or growth-promoting genes as a cancer therapy, but also as a means to shed light on how aberrant de novo methylation during cancer progression might be targeted to specific sequences. [Cancer Res 2007;67(14):6637–46]

The silencing of gene expression using small interfering RNA (siRNA) is rapidly becoming a powerful technique for functional genomic analysis and represents a potential strategy for therapeutic product development (1, 2). RNA interference (RNAi) induced by siRNA was initially characterized as posttranscriptional gene silencing in a wide variety of organisms (35). Recent studies provide evidence that siRNA can also direct DNA methylation and transcriptional gene silencing in human cells (69), thereby suggesting a potential additional mechanism for transcriptional regulation in mammals. These findings provide hope for using RNAi technology in disease control.

Several studies have used siRNA as an experimental tool to dissect the cellular pathways that lead to uncontrolled cell proliferation and cancer. To develop siRNA for cancer therapy, several researchers have investigated the effects of siRNA in animal models (1014). However, reports of RNAi-based strategies for metastatic cancer are very limited. For example, in advanced prostate cancer, the sites most frequently affected by metastasis are the bones and lungs. Patients with these metastases suffer pain and rarely benefit, making it extremely important to explore avenues of treating such metastases. The power of RNAi as a therapeutic tool would be greatly enhanced by the ability to engineer stable silencing of gene expression. The development of promoter-targeted siRNA may enable heritable epigenetic silencing of metastasis or tumor-promoting genes for prostate cancer therapy.

In human cancers, alteration of urokinase plasminogen activator (uPA) expression is involved in tumor invasion and metastasis (15). Indeed, overexpression of uPA has been found in several types of cancer, including prostate, glioblastoma, melanoma, breast, colon, and lung (1623). Notably, in most of these cases, the increased expression of uPA is associated with increased metastatic potential and poor survival (2426). We previously showed that uPA promotes invasiveness of prostate cancer cells in vitro and tumorigenicity in vivo (27). Recently, we also reported that DNA demethylation is a common mechanism underlying the abnormal expression of uPA and is a critical contributing factor to the malignant progression of human prostate tumors (28). Thus, inhibition of uPA demethylation may suppress the metastatic potential of prostate tumor cells.

In the present study, we used the IVIS imaging system, a noninvasive system able to detect luciferase activity in vivo, to monitor metastatic lesions and the effects of uPA promoter–targeted siRNA on tumor regression. The results indicate that promoter-targeted siRNA significantly inhibits uPA expression and tumor cell invasion in vitro and tumor growth and incidence of lung metastasis in vivo. Our findings suggest that the siRNA-directed transcriptional silencing is a promising tool for the stable suppression of uPA gene function as well as illustrating the potential of its application in cancer therapy.

Cell lines and culture conditions. Prostate cancer cell lines PC3 and DU145 were obtained from the American Type Culture Collection and maintained in DMEM supplemented with 10% fetal bovine serum and antibiotics. Cells were cultured in a 37°C incubator with a 5% CO2 humidified atmosphere. For inhibition of DNA methylation, cells transfected with uPA siRNA were incubated in culture medium supplemented with 10 μmol/L 5-azacytidine and 100 nmol/L trichostatin for 3 days. At every 24 h interval, fresh medium containing the drug was added.

Small interfering RNA. Sets of four promoter-directed siRNA for uPA and siMM nontargeting siRNA pool for control were synthesized and annealed. Transfection of siRNA pool and control pool in six-well plates was done using LipofectAMINE 2000 (Invitrogen) according to the manufacturer's instructions. The siRNA sequences that we used are listed in Supplementary Table S1.

Methylation analysis. Genomic DNA was extracted with the DNeasy Tissue Kit (Qiagen) as per the manufacturer's instructions. Sodium bisulfite modification of genomic DNA was done as described previously (29). We carried out methylation-specific PCR (MSP) on the uPA promoter with a set of previously described primers spanning the CpG island of this region (28). MSP products were resolved on 2% agarose gel with ethidium bromide. Positive controls used for MSP included DNA from MDA-231 cells as unmethylated DNA control and CpGenome Universal methylated DNA as methylated DNA control (Chemicon). Negative control MSP reactions were done using water only as a template.

For bisulfite sequencing, PCR was done with a set of previously described primers common to the methylated and unmethylated DNA sequences using 1 μL of the DNA solution as template (28). Amplified PCR products were purified with a gel extraction kit (Qiagen) and ligated into pCR2.1-TOPO plasmid vector with a TA-cloning system (Invitrogen). Plasmid-transformed Escherichia coli were cultured and plasmid DNA was isolated with Fast Plasmid mini-prep kit (Eppendorf). Purified plasmid DNA containing the uPA sequence was sequenced with an automated DNA sequencer.

Nuclear run-on assay. Nuclear run-on transcription analysis was done according to a modified protocol that uses digoxigenin-labeled UTP (3032). Briefly, nuclei (2 × 107) from control and cells transfected with siRNA were incubated with digoxigenin-UTP (Roche Applied Science). Purified mRNA from each sample was hybridized to human uPA and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA that was immobilized on a nylon membrane (Amersham Biosciences). After hybridization, the digoxigenin-chemiluminescent detection procedure was done according to the manufacturer's instructions (Roche Applied Science). Specific hybridization to uPA was quantified using a PhosphorImager and then normalized to that of GAPDH.

Immunoblot analysis. Cells were lysed in radioimmunoprecipitation assay buffer and proteins were quantified using a bicinchoninic acid assay (Pierce). Equal amounts were separated on SDS-PAGE gels. Membranes were probed with antibodies against uPA (Biomeda) and GAPDH (Abcam). Horseradish peroxidase–conjugated secondary antibodies (Biomeda) were used for detection of immunoreactive proteins by chemiluminescence (Amersham Biosciences).

Reverse transcription-PCR analysis. Total RNA was isolated from prostate cancer cells and tissue samples using RNeasy mini kit (Qiagen) and reverse-transcribed using Superscript III reverse transcriptase (Invitrogen) according to the manufacturer's instructions. Quantitative reverse transcription-PCR (RT-PCR) was done using the iCycler IQ real-time PCR system (Bio-Rad) with the SYBR Green Mastermix as per manufacturer's instructions. The following primers were used: uPA-sense, 5′-TGCGTCCTGGTCGTGAGCGA-3′, and uPA-antisense, 5′-CTACAGCGCTGACACGCTTG-3′; GAPDH-sense, 5′-CGGAGTCAACGGATTTGGTCGTAT-3′, and GAPDH-antisense, 5′-AGCCTTCTCCATGGTGGTGAAGAC-3′. Data were normalized to GAPDH expression. All reactions were done in triplicate and with no reverse transcriptase or no template as negative controls.

Fibrin zymography. The enzymatic activity and molecular weight of electrophoretically separated forms of uPA were determined in conditioned medium of prostate cancer cells by SDS-PAGE as described previously (23). Briefly, the SDS-PAGE gel contains acrylamide to which purified plasminogen and fibrinogen were substrates before polymerization. After polymerization, equal amounts of proteins in the samples were electrophoresed and the gel was washed and stained as described previously (23).

Matrigel invasion assay. We used 6.5-mm-diameter Transwell inserts (Costar) with the 8-μm-pore membranes coated with Matrigel (Becton Dickinson) to assess the invasive potentials of prostate cancer cells after transfection with siRNA. Cells were detached, washed twice in PBS, and resuspended in serum-free DMEM. A total of 5 × 105 cells in 0.2 mL were placed in the upper chamber of a Transwell insert and the lower chamber was filled with 400 μL of DMEM/10% fetal bovine serum. After a 24-h incubation period, the cells in the upper chamber that did not migrate were gently scraped away and adherent cells present on the lower surface of the insert were stained with Hema-3 and photographed.

In vitro angiogenesis assay. The endothelial tube–like formation assay was done as described previously (12). Human dermal endothelial cells were seeded on 8-well chamber slides at 4 × 104 per well. After cells adhered to the well (12 h), growth medium was replaced by conditioned medium. The conditioned medium collected from PC3 prostate cell cultures were pretreated with uPA or vascular endothelial growth factor (VEGF) neutralized antibodies at the indicated concentrations for 2 h at 37°C before being added to human dermal endothelial cells. After a 48 h culture in the treated medium, dead floating cells were washed away. To assess tube formation, the attached living cells were stained with anti–von Willebrand factor (vWF; Dako). We quantified tube formation by measuring the number of branch points and the total number of branches per point in triplicate wells.

Dorsal skin-fold chamber model. Tumor angiogenesis was quantified in vivo using the transparent dorsal skin-fold chamber model (33). Briefly, after total body anesthesia with ketamine (50 mg/kg) and xylazine (10 mg/kg), a dorsal air sac was made in the mouse by injecting 10 mL of air. After dorsal skin-fold preparation, chambers were inoculated with 2 × 106 PC3 cells. Ten days later, the animals were anesthetized and sacrificed. The animals were carefully skinned around the implanted chambers, and the implanted chambers were removed from the s.c. air fascia. The skin fold covering the chambers was photographed under visible light. The number of blood vessels within the chamber in the area of the air sac fascia was counted and their lengths were measured.

Orthotopic mouse prostate tumor/metastasis model. Athymic male nude mice (nu/nu; 6–8 weeks old) were obtained from Harlan Sprague-Dawley. Animal handling and experimental procedures were approved by the University of Illinois College of Medicine Institutional Animal Care and Use Committee. Orthotopic implantation was carried out as described previously with minor modifications (27). Briefly, after total body anesthesia with ketamine (50 mg/kg) and xylazine (10 mg/kg), a small incision was made to reveal the prostate gland. PC3-luc cells transfected with either siuPA pool or siMM pool were harvested from culture and resuspended in PBS. A cellular suspension of each condition (106 cells in 50 μL of PBS) was injected into the lateral lobe of the prostate and the wound was closed with surgical metal clips. This cell concentration was necessary to achieve consistent local tumor growth within 7 days of implantation. Prostate tumor growth and subsequent metastasis to lungs were assessed weekly using the in vivo imaging system coupled to Living Image acquisition and analysis software (Xenogen) according to the manufacturer's instructions. All mice were sacrificed at either 6 weeks postimplantation or when clinical conditions dictated euthanasia. We then excised, measured, and weighed the prostate tumors. We isolated the lungs and counted the surface metastases using a dissecting microscope. Part of the tumor was snap frozen for RT-PCR analysis, and the remaining portion was placed in phosphate-buffered formalin and then ethanol before preparing paraffin sections for H&E staining.

Histologic analysis and immunohistologic staining. Tissues were fixed with formalin, embedded in paraffin, sectioned (5 μm), and stained with H&E. Immunostaining was done according to standard techniques using a polyclonal antibody specific for human vWF (Dako). Visualization of antibody binding was done by the second antibody tagged with horseradish peroxidase following reaction to 3,3′-diaminobenzidine (Sigma). Cells positive for vWF were counted in five high-power fields.

Statistical analysis. Statistical comparisons were done using ANOVA for analysis of significance between different values using GraphPad Prism software. Values are expressed as mean ± SD from at least three separate experiments and differences were considered significant at a P value of <0.05.

siRNA targeting the uPA promoter–induced DNA methylation and transcriptional gene silencing. To determine whether promoter-directed double-stranded siRNA could induce DNA methylation to silence endogenous uPA expression, we designed and synthesized four siRNA targeting the uPA promoter at sequence positions −213 (uPA-1), −131 (uPA-2), −111 (uPA-3), and −50 (uPA-4) relative to the transcription start site (Fig. 1A). We used the highly metastatic prostate cancer cell line PC3, which express uPA aberrantly by promoter demethylation (28). We transfected PC3 cells with the pool of four promoter-targeted siRNA (siuPA) as well as mismatched siRNA oligonucleotides (siMM) and reagent-only (mock) as controls. We first analyzed the methylation status of the targeted regions using MSP. Only unmethylated and no methylated DNA was detected in cells transfected with controls. In contrast, only methylated and no unmethylated DNA was detected in cells transfected with siuPA (Fig. 1B,, left). In addition, bisulfite-based DNA methylation analysis using quantitative real-time PCR method also suggested that the methylation levels were higher in cells transfected with siuPA compared with control cells (Fig. 1B,, right). Next, we did bisulfite sequencing to determine the DNA methylation status of all 25 CpG sites in the targeted region. As shown in Fig. 1C, we found that nearly all CpG sites located within the CpG island of uPA were methylated in siuPA-transfected cells. In contrast, almost all of the CpG sites remained unmethylated in siMM- or mock-transfected cells. Furthermore, transfection of siuPA failed to induce DNA methylation within a nontargeted 14-3-3σ promoter (Supplementary Fig. S1), which is not normally methylated in PC3 cells (29). Induction of methylation by siuPA significantly reduced uPA expression at both the mRNA and protein levels in PC3 cells. Both uPA mRNA and protein were undetectable by day 4 of siuPA transfection compared with siMM- or mock-transfected cells (Fig. 1D). However, transfection of uPA siRNA individually failed to induce DNA methylation (data not shown) and uPA silencing (Supplementary Fig. S2). These data indicate that induction of DNA methylation and silencing of uPA gene expression requires a combination of all four uPA siRNA.

Figure 1.

siRNA targeting the uPA promoter induced DNA methylation and transcriptional silencing in PC3 prostate cancer cells. A, design of uPA siRNA. Locations of uPA siRNA nos. 1, 2, 3, and 4 are shown relative to the transcriptional start site (0). B, MSP analysis of the CpG island in the uPA promoter in mock-transfected cells and in cells transfected with siuPA or siMM (left). Bisulfite-based DNA methylation analysis using quantitative real-time PCR (right). Using bisulfite-treated genomic DNAs from siRNA-transfected and mock cells, the CpG island in the uPA promoter region was PCR amplified with methylation- or unmethylation-specific primers. Real-time PCR was used to determine the relative levels of methylated and unmethylated uPA. The PCR product amplified with methylation-insensitive primers was used for normalization. uPA-U, unmethylated PCR product; uPA-M, methylated PCR product. Positive and negative controls were described in Materials and Methods. C, bisulfite sequencing analysis of the CpG island in the uPA promoter in mock-transfected cells and in cells transfected with siuPA or siMM. CpG positions are indicated relative to the transcriptional start site (0); each circle in the figure represents a single CpG site. •, methylated CG dinucleotide; ○, unmethylated CG dinucleotide. Columns, percentage of methylation in cells transfected with mock, siMM, or siuPA. D, RT-PCR (top) and immunoblot (bottom) analysis of cells transfected with mock, siMM, or siuPA were collected on the indicated days (D2–D14). GAPDH was used as a loading control for RNA and protein analysis.

Figure 1.

siRNA targeting the uPA promoter induced DNA methylation and transcriptional silencing in PC3 prostate cancer cells. A, design of uPA siRNA. Locations of uPA siRNA nos. 1, 2, 3, and 4 are shown relative to the transcriptional start site (0). B, MSP analysis of the CpG island in the uPA promoter in mock-transfected cells and in cells transfected with siuPA or siMM (left). Bisulfite-based DNA methylation analysis using quantitative real-time PCR (right). Using bisulfite-treated genomic DNAs from siRNA-transfected and mock cells, the CpG island in the uPA promoter region was PCR amplified with methylation- or unmethylation-specific primers. Real-time PCR was used to determine the relative levels of methylated and unmethylated uPA. The PCR product amplified with methylation-insensitive primers was used for normalization. uPA-U, unmethylated PCR product; uPA-M, methylated PCR product. Positive and negative controls were described in Materials and Methods. C, bisulfite sequencing analysis of the CpG island in the uPA promoter in mock-transfected cells and in cells transfected with siuPA or siMM. CpG positions are indicated relative to the transcriptional start site (0); each circle in the figure represents a single CpG site. •, methylated CG dinucleotide; ○, unmethylated CG dinucleotide. Columns, percentage of methylation in cells transfected with mock, siMM, or siuPA. D, RT-PCR (top) and immunoblot (bottom) analysis of cells transfected with mock, siMM, or siuPA were collected on the indicated days (D2–D14). GAPDH was used as a loading control for RNA and protein analysis.

Close modal

Nuclear run-on experiments were done to determine whether loss of uPA expression by the promoter-targeted siuPA occurs at the transcriptional level. The transcription rate of the uPA gene was 96% lower in cells transfected with siuPA than in cells transfected with siMM (Supplementary Fig. S3A; P < 0.01). We then examined whether the uPA silencing is mediated by DNA methylation or by other mechanisms. Because the methylation status of CpG sites is associated with histone deacetylation (34, 35), we transfected siuPA into PC3 cells in the presence of the 5-azacytidine and trichostatin, inhibitors of DNA methylation and histone deacetylases, respectively. Treatment with a combination of 5-azacytidine and trichostatin strongly reduced the DNA methylation at CpG sites located within the CpG island of uPA promoter (data not shown) and restored uPA expression (Supplementary Fig. S3B; P < 0.01). Taken together, these results show that siRNA targeted to the uPA promoter induces transcriptional gene silencing via DNA methylation in PC3 prostate cancer cells.

Silencing of uPA-suppressed invasion and angiogenesis. It is well established that uPA plays an important role in invasion and metastasis of tumor cells (15, 16, 22, 27). To test whether altering the methylation status and expression of uPA would actually modulate the metastatic potential of invasive prostate cancer cells, we did in vitro invasion assays with siRNA-transfected cells. Invasion through a membrane (8-μm pores coated with a layer of matrigel consisting of extracellular matrix proteins) was significantly reduced by >75% in PC3 cells where endogenous uPA was transcriptionally silenced by promoter-targeted siRNA transfection (Fig. 2A; P < 0.01). However, this reduction in invasive ability was reversed by treating cells with 5-azacytidine and trichostatin. Incubation of these cells with the anti-uPA antibody blocked this acquired increase in invasive potential (Fig. 2A and B). These data provided further convincing evidence that this increase in tumor cell invasion after treatment with 5-azacytidine and trichostatin is due to the restoration of uPA expression.

Figure 2.

uPA silencing suppressed invasion and angiogenesis in PC3 prostate cancer cells. A, comparison of the in vitro invasive potentials of cells transfected with mock, siMM, or siuPA in the presence or absence of 5-azacytidine (5-aza) and trichostatin (TSA). Ab, antibody. Columns, mean of three independent experiments (P < 0.01); bars, SD. B, representative invasion photographs from cells transfected with mock, siMM, or siuPA as described in (A). C, the PC3-derived conditioned medium (CM) significantly increased endothelial cell tube-like network formation in culture, an effect nearly completely blocked by neutralizing antibodies to either uPA or VEGF. D, angiogenic activity in medium conditioned by uPA knockdown or control cell culture using the endothelial cell tube-like formation assay. Columns, mean of three independent experiments (*, P < 0.01); bars, SD. Bottom, representative angiogenic activity photographs in medium conditioned by uPA knockdown or control cell cultures.

Figure 2.

uPA silencing suppressed invasion and angiogenesis in PC3 prostate cancer cells. A, comparison of the in vitro invasive potentials of cells transfected with mock, siMM, or siuPA in the presence or absence of 5-azacytidine (5-aza) and trichostatin (TSA). Ab, antibody. Columns, mean of three independent experiments (P < 0.01); bars, SD. B, representative invasion photographs from cells transfected with mock, siMM, or siuPA as described in (A). C, the PC3-derived conditioned medium (CM) significantly increased endothelial cell tube-like network formation in culture, an effect nearly completely blocked by neutralizing antibodies to either uPA or VEGF. D, angiogenic activity in medium conditioned by uPA knockdown or control cell culture using the endothelial cell tube-like formation assay. Columns, mean of three independent experiments (*, P < 0.01); bars, SD. Bottom, representative angiogenic activity photographs in medium conditioned by uPA knockdown or control cell cultures.

Close modal

Apart from the regulation of tumor cell invasion, uPA is also involved in controlling tumor angiogenesis (36). Because tumor angiogenesis involves multiple factors, including endothelial cells and tumor cells, we first tested whether PC3 cells exerted a proangiogenic effect on neighboring endothelial cells through uPA and other angiogenic proteins. We used the endothelial cell tube-like network formation assay, an in vitro assay that correlates well with in vivo angiogenic activity (12, 37). Endothelial cell networks increased 4.2-fold after exposure to PC3 conditioned medium compared with basal medium. Further, blocking antibodies to both uPA and VEGF were able to abolish the angiogenic effects of the PC3 conditioned medium on endothelial cell network formation (Fig. 2C; P < 0.01). We could not, however, show an additive or synergistic effect of uPA and VEGF inhibition, because uPA depletion alone resulted in a strong inhibition of network formation. These data show the ability of uPA to induce endothelial cell tube-like network formation. Next, we examined whether silencing of uPA by siuPA could alter endothelial cell network formation. As shown in Fig. 2D, the siuPA-transfected PC3 cells formed endothelial cell networks ∼70% less compared with the control siRNA-transfected counterparts (P < 0.01). Taken together, these results suggest that promoter-targeted RNAi significantly interfered with the uPA-mediated invasion and angiogenesis.

Next, we transfected siuPA into the highly invasive human prostate cancer cell line DU145 to confirm that these results could be extended to another cell line. We chose this cell line because DU145 cells also contain high levels of uPA expression via demethylation of the uPA promoter (28). We isolated genomic DNA from the cells and did both MSP and bisulfite sequencing. Transfection of siuPA directed against the uPA promoter significantly induced the methylation within the targeted regions (Fig. 3A; Supplementary Fig. S4) and silenced uPA expression in DU145 cells (Fig. 3B). Furthermore, the silencing of uPA expression and activity in siuPA-transfected DU145 cells was reversed in the presence of 5-azacytidine and trichostatin (Fig. 3C and D; P < 0.01). This silencing of uPA expression and activity was accompanied by significant reductions in the invasive and angiogenic capacities of the transfected DU145 cells as determined by matrigel invasion assay and endothelial cell tube-like formation assay, respectively (Supplementary Fig. S5; P < 0.01). These experiments confirmed our hypothesis that reversal of the demethylation state by promoter-targeted RNAi results in silencing of uPA gene expression and activity, and, thus, inhibits the invasive and angiogenic capacities of metastatic prostate cancer cells.

Figure 3.

siRNA targeting the uPA promoter–induced DNA methylation and transcriptional silencing in DU145 prostate cancer cells. A, MSP analysis of the CpG island in the uPA promoter in mock-transfected cells and in cells transfected with siuPA or siMM. Positive and negative controls are described in Materials and Methods. B, RT-PCR (top) and immunoblot (bottom) analysis on day 4 of DU145 cells transfected with mock, siMM, or siuPA. GAPDH was used as a loading control for RNA and protein analysis. C, expression levels of uPA mRNA as determined by real-time RT-PCR. DU145 cells were transfected with mock, siMM, or siuPA in the presence or absence of 5-azacytidine and trichostatin. Columns, mean of three independent experiments (*, P < 0.01); bars, SD. D, fibrin zymography of cells transfected with mock, siMM, or siuPA in the presence or absence of 5-azacytidine and trichostatin. Fibrinolytic activity was detected as clear lysis bands after Amido Black staining and subsequent destaining with methanol/acetic acid.

Figure 3.

siRNA targeting the uPA promoter–induced DNA methylation and transcriptional silencing in DU145 prostate cancer cells. A, MSP analysis of the CpG island in the uPA promoter in mock-transfected cells and in cells transfected with siuPA or siMM. Positive and negative controls are described in Materials and Methods. B, RT-PCR (top) and immunoblot (bottom) analysis on day 4 of DU145 cells transfected with mock, siMM, or siuPA. GAPDH was used as a loading control for RNA and protein analysis. C, expression levels of uPA mRNA as determined by real-time RT-PCR. DU145 cells were transfected with mock, siMM, or siuPA in the presence or absence of 5-azacytidine and trichostatin. Columns, mean of three independent experiments (*, P < 0.01); bars, SD. D, fibrin zymography of cells transfected with mock, siMM, or siuPA in the presence or absence of 5-azacytidine and trichostatin. Fibrinolytic activity was detected as clear lysis bands after Amido Black staining and subsequent destaining with methanol/acetic acid.

Close modal

Silencing of uPA inhibited tumor growth and metastasis. We used the PC3 cell line to assess the effects of siuPA on tumor growth and metastasis because PC3 cells have high uPA expression and induce lung metastasis in mice (28). Furthermore, we derived PC3-luc cells by stable transfection of a plasmid expressing the luciferase gene, so that we could track tumor cells and metastases in vivo. We transfected PC3-luc cells with siRNA and implanted the cells into the prostate of immunodeficient mice. Tumor progression was monitored in mice using the in vivo imaging system (Xenogen). We obtained photon counts from the tumor region on days 10, 20, and 40 (Fig. 4A,, left). The mice injected with PC3-luc cells transfected with siuPA developed prostate tumors of significantly smaller volume by day 40 when compared with the mice inoculated with cells either transfected with siMM or mock (Fig. 4A,, right). The presence of significant growth difference in tumors of the prostate between control and uPA-silenced cell–injected mice were confirmed by autopsy after imaging (Fig. 4B and C; P < 0.01). RT-PCR analysis showed a prominent expression of uPA mRNA in the prostate tumors isolated from cells transfected with mock or siMM. In striking contrast, the results showed long-term silencing of uPA in tumors developed from the cells transfected with siuPA (Fig. 4D).

Figure 4.

Silencing of uPA expression inhibited tumor growth in an orthotopic mouse prostate tumor model. A, photon counts of orthotopic prostate tumors on days 10, 20, and 40 (left). Nude mice carrying an orthotopic prostate tumor, established from either mock or siMM-transfected PC3-luc cells, showed substantial bioluminescence signal as indicated by photon counts on day 40. In contrast, the orthotopic prostate tumors developed from PC3-luc cells transfected with siuPA showed significantly low bioluminescence signal (right). B, comparison of dissected prostate tumors in (A) from mice 40 d after cell implantation (*, P < 0.01). C, prostate tumors in (A), from mice 40 d after cell implantation, were excised and photographed. D, RNA samples extracted from PC3-luc prostate tumors (six animals per group) were analyzed using RT-PCR for uPA expression levels. The siuPA group showed the most prominent and specific knockdown of uPA. GAPDH was used as a control and remained unchanged.

Figure 4.

Silencing of uPA expression inhibited tumor growth in an orthotopic mouse prostate tumor model. A, photon counts of orthotopic prostate tumors on days 10, 20, and 40 (left). Nude mice carrying an orthotopic prostate tumor, established from either mock or siMM-transfected PC3-luc cells, showed substantial bioluminescence signal as indicated by photon counts on day 40. In contrast, the orthotopic prostate tumors developed from PC3-luc cells transfected with siuPA showed significantly low bioluminescence signal (right). B, comparison of dissected prostate tumors in (A) from mice 40 d after cell implantation (*, P < 0.01). C, prostate tumors in (A), from mice 40 d after cell implantation, were excised and photographed. D, RNA samples extracted from PC3-luc prostate tumors (six animals per group) were analyzed using RT-PCR for uPA expression levels. The siuPA group showed the most prominent and specific knockdown of uPA. GAPDH was used as a control and remained unchanged.

Close modal

We assessed whether uPA silencing in PC3-luc cells affected the ability of these cells to metastasize in vivo. The lung was dissected from each mouse and photon counts were recorded. The lungs derived from control mice injected with cells either transfected with siMM or mock had high photon counts. However, there was a marked reduction (∼4.1-fold) in the incidence of lung metastasis when primary tumor cells transfected with siuPA were seeded in host mice (Fig. 5A; P < 0.01). In addition to imaging analysis, lung metastatic nodules were counted to confirm the reduction in the metastatic capacity of siuPA-transfected cells. We detected no lung surface metastasis in mice injected with PC3-luc + siuPA (0 of 8), whereas lung metastasis was detected in mice injected with PC3-luc alone or PC3-luc + siMM (Fig. 5B and C; P < 0.01). Subsequent RT-PCR analysis confirmed stable silencing of uPA in lungs from the mice injected with PC3-luc + siuPA (Fig. 5D). These results indicate that stable silencing of uPA in prostate cancer cells significantly suppressed both primary tumor growth and the in vivo incidence of lung metastasis.

Figure 5.

Inhibition of metastatic tumor growth in lungs by promoter-directed uPA siRNA. A, bioluminescence signal in the lung, representative of lung metastasis, was recorded for each mouse. Lung images from different mice (top). Lung metastasis signals were analyzed by photon counts (bottom, P < 0.01). We observed significant inhibition of lung metastasis in PC3-luc cells transfected with siuPA relative to the control groups. B, scatter plot of the lung surface metastasis counts of mice implanted with control and uPA knockdown cells (*, P < 0.01). C, effect of uPA silencing on the metastatic potential of PC3-luc cells. D, tumors metastasized to lung in (A) from mice were examined for uPA expression by RT-PCR. GAPDH was used as a control and remained unchanged.

Figure 5.

Inhibition of metastatic tumor growth in lungs by promoter-directed uPA siRNA. A, bioluminescence signal in the lung, representative of lung metastasis, was recorded for each mouse. Lung images from different mice (top). Lung metastasis signals were analyzed by photon counts (bottom, P < 0.01). We observed significant inhibition of lung metastasis in PC3-luc cells transfected with siuPA relative to the control groups. B, scatter plot of the lung surface metastasis counts of mice implanted with control and uPA knockdown cells (*, P < 0.01). C, effect of uPA silencing on the metastatic potential of PC3-luc cells. D, tumors metastasized to lung in (A) from mice were examined for uPA expression by RT-PCR. GAPDH was used as a control and remained unchanged.

Close modal

Finally, we compared the morphology of prostate tumors harvested at the end of the experiment (Fig. 4C)—the invasive edge and degree of vascularization of the uPA-expressing control groups (siMM or mock) versus uPA silenced mice. H&E staining showed that the morphology of the uPA-silenced tumors indicated mostly normal histology, whereas tumors of control groups displayed observable change in the general architecture of the prostate gland, and H&E staining showed both tumor and host cells (Fig. 6A,, left column). Immunohistochemical staining for endothelial cells revealed prominent vWF-positive vascular structures within the tumor mass of the control groups. In contrast, significantly less vascularity was present in tumors developed from the cells transfected with siuPA (Fig. 6A , middle and right columns, B; P < 0.01).

Figure 6.

Histopathology and vWF staining of paraffin-embedded tissue sections from orthotopic prostate tumors. A, tumors were harvested from all PC3-luc prostate tumor groups as described in Fig. 4C. Tissue sections (5 μm) were prepared and stained with H&E for histopathologic analysis (left column) and with an anti-vWF for immunohistochemistry (middle and right columns). Following blinded procedures, a pathologist examined the H&E and vWF-stained sections (n = 12–15) from each tumor (n = 6 tumors from each treatment group) and determined the morphology, invasive edge, and vascularity. B, cells positive for vWF counted in five high-power fields in tumor sections with indicated groups. The siuPA group has a significantly lower mean blood vessel density (*, P < 0.01). C, tumor angiogenesis was monitored in the dorsal skin-fold chamber of an immunodeficient mouse (left). The photographs show changes in microvessel density of mice skin-folds in response to the secretions of PC3 cells either expressing uPA (mock and siMM) or experiencing uPA knockdown (siuPA). Arrowheads, newly formed blood vessels in the skin folds. Microvessels were counted under a microscope in five random fields (right). Columns, mean from six implants per group (*, P < 0.01); bars, SD. D, schematic model of RNAi-directed alteration of uPA methylation status as a key regulatory switch for metastatic potential.

Figure 6.

Histopathology and vWF staining of paraffin-embedded tissue sections from orthotopic prostate tumors. A, tumors were harvested from all PC3-luc prostate tumor groups as described in Fig. 4C. Tissue sections (5 μm) were prepared and stained with H&E for histopathologic analysis (left column) and with an anti-vWF for immunohistochemistry (middle and right columns). Following blinded procedures, a pathologist examined the H&E and vWF-stained sections (n = 12–15) from each tumor (n = 6 tumors from each treatment group) and determined the morphology, invasive edge, and vascularity. B, cells positive for vWF counted in five high-power fields in tumor sections with indicated groups. The siuPA group has a significantly lower mean blood vessel density (*, P < 0.01). C, tumor angiogenesis was monitored in the dorsal skin-fold chamber of an immunodeficient mouse (left). The photographs show changes in microvessel density of mice skin-folds in response to the secretions of PC3 cells either expressing uPA (mock and siMM) or experiencing uPA knockdown (siuPA). Arrowheads, newly formed blood vessels in the skin folds. Microvessels were counted under a microscope in five random fields (right). Columns, mean from six implants per group (*, P < 0.01); bars, SD. D, schematic model of RNAi-directed alteration of uPA methylation status as a key regulatory switch for metastatic potential.

Close modal

We next used a dorsal skin-fold chamber model to further characterize the effects of siuPA on tumor angiogenesis in vivo. Implantation of a chamber containing either mock or cells transfected with siMM resulted in tiny new vascular sprouts and branches that grew from preexisting microvessels. In contrast, we observed significantly less angiogenic response in the skin-folds of mice implanted with chambers containing uPA-silenced PC3 cells (Fig. 6C; P < 0.01). These results suggest that uPA can promote tumor growth, angiogenesis, and metastasis, but significant inhibition of these processes can be achieved by transcriptional silencing of uPA. The data also suggest that the methylation status of uPA is a key regulatory switch for metastatic potential (Fig. 6D).

The discovery that siRNA targeted to promoters could induce transcriptional silencing via DNA methylation in human cells has raised the possibility that RNA-directed DNA methylation might be a potential additional mechanism of gene silencing. This notion has been supported by numerous lines of evidence. For example, Kawasaki and Taira (7) reported that siRNA targeted to CpG islands of an E-cadherin promoter induced significant DNA methylation in human breast cancer cells and in normal mammary epithelial cells. They showed that silencing of the target gene was dependent on the DNA methyltransferases DNMT1 and DNMT3b, suggesting that it occurred at the transcriptional level as a result of DNA methylation. In a similar study, Morris et al. (8) showed that this type of transcriptional gene silencing mechanism could also occur in human kidney epithelial cells. They showed that transcription of EF-1α was suppressed in HEK293 cells by promoter-specific siRNA. The studies of Castanotto et al. (6) showed that shRNA complementary to the RASSF1A promoter can induce low levels of methylation and partial gene silencing in HeLa cells. Recent work by Murayama et al. (9) showed that expression of shRNA directed against the hIL2 promoter induced CpG methylation and dramatically reduced endogenous hIL2 expression in Jurkat cells. Taken together, these studies suggest that RNA-directed DNA methylation might be a natural mechanism for regulating gene expression at the transcriptional level in human cells. However, other attempts to achieve RNA-directed DNA methylation at homologous genes in human cells have been unsuccessful (3840). Explanations for these differences include variations in cell type and/or target site selection as well as differences in the gene being silenced.

In the present study, we showed that siRNA directed against the uPA promoter successfully inhibited the expression and activity of uPA in the human metastatic prostate cancer cell lines PC3 and DU145. Several findings presented here provide evidence that inhibition of uPA expression using promoter-directed siRNA occurs at the transcriptional level as a result of DNA methylation. First, cytosines within the target sequence were specifically methylated in cells transfected with siuPA, but not in the siMM- or mock-transfected cells. This observation was confirmed by multiple methods. Second, the link between methylation and uPA silencing was supported by the observation that siuPA were not able to inhibit expression of uPA in the presence of the epigenetic silencing modifiers, 5-azacytidine and trichostatin. Third, nuclear run-on assays revealed that RNA was not being transcribed, which is consistent with a mechanism of action that involves inhibition of transcription. Fourth, silencing of the target gene was quite robust and long-lasting after a single transfection of siuPA; potent inhibition uPA expression at both the mRNA and protein levels in PC3 cells remained detectable 14 days after siuPA transfection. Furthermore, we observed that siuPA transfection had long-term effects in vivo on uPA expression and metastasis. This prolonged silencing provides additional support for the conclusion that the mechanism of uPA knockdown by siuPA is dependent on methylation. All these data suggest that stable suppression of gene expression can result from a single exposure to a specific methylation-inducing RNAi agent targeting the uPA promoter sequence.

Successful application of RNAi as a therapeutic method requires an efficient and stable silencing system that can bypass the difficulties associated with traditional gene therapy (4143). Unlike posttranscriptional gene silencing that requires the continued presence of a siRNA molecule targeting a coding sequence, targeting a promoter sequence allows stable, long-term silencing of the desired gene, thereby allowing phenotypic changes associated with loss of that gene. In the present study, siRNA targeted to uPA promoter significantly inhibited uPA expression and tumor cell invasion in vitro and tumor growth and incidence of lung metastasis in vivo, demonstrating that promoter-targeted siRNA represents an alternative strategy for harnessing RNAi. Perhaps the most important finding is that RNAi for uPA substantially reduced angiogenesis in metastatic prostate cells in vitro as well as in vivo. Taken together, our results clearly show that a novel uPA blockade system by promoter-targeted RNAi is a valid, potential therapeutic.

The major risk faced by patients with cancer is the development of metastatic disease. Although genes associated with metastatic progression can be identified readily by screening techniques such as gene arrays (44), the validation and characterization of these genes will require sophisticated animal models and efficient gene silencing systems. Increased expression of uPA has been reported in various malignancies including prostate cancer (16, 22, 28). Numerous studies in tumor cell lines have correlated the metastatic potential of tumor cells with uPA activity and protein expression (15, 27). Furthermore, several studies of clinical tumor samples have correlated high uPA expression with tumor progression, and, in some cases, poor patient postoperative survival (45, 46). In our recent study, we documented that the activation of uPA is strongly associated with the demethylation of uPA promoter (28). In the present study, we have shown that RNAi-directed reversal of uPA demethylation blocks prostate cancer progression into the aggressive and metastatic stages of the disease.

The present study also expands our understanding of the DNA demethylation by elucidating which gene-specific activation can regulate cancer progression. Previous research emphasized the role of hypermethylation of tumor-suppressor genes in tumor growth and metastasis (47, 48), but the present results show that DNA demethylation can also directly affect metastatic progression by increasing the activity of tumor-promoting genes. Key to the demonstration of this direct effect on tumor cell biology is the use of the bioluminescent orthotopic tumor model, which provides opportunities to access the effects of alterations in methylation status in the establishment of tumor and metastasis formation. Our findings provide important clues for understanding uPA-associated tumorigenesis.

In summary, we have shown that promoter-directed siRNA for uPA can significantly suppress both primary tumor growth and the in vivo incidence of lung metastasis. The data provided here offer proof-of-concept that siRNA targeted against a promoter could be used therapeutically for human metastatic cancer. To our knowledge, our results present the first evidence that transcriptional silencing by means of promoter-specific siRNA may have therapeutic potential in the treatment of advanced human prostate cancer. We believe that this system has the potential to be developed into a useful method for inhibition of metastatic growth and may also have broad applicability in cancer therapy.

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

Grant support: National Cancer Institute grants CA 75557, CA 92393, CA 95058, CA 116708; National Institute of Neurological Disorders and Stroke grants NS47699 and NS057529; Caterpillar, Inc.; and OSF Saint Francis Medical Center, Peoria, IL (J.S. Rao).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Shellee Abraham for preparing the manuscript, Diana Meister and Sushma Jasti for manuscript review, and Noorjehan Ali for technical assistance.

1
Bumcrot D, Manoharan M, Koteliansky V, Sah DW. RNAi therapeutics: a potential new class of pharmaceutical drugs.
Nat Chem Biol
2006
;
2
:
711
–9.
2
Hannon GJ, Rossi JJ. Unlocking the potential of the human genome with RNA interference.
Nature
2004
;
431
:
371
–8.
3
Caplen NJ, Parrish S, Imani F, Fire A, Morgan RA. Specific inhibition of gene expression by small double-stranded RNAs in invertebrate and vertebrate systems.
Proc Natl Acad Sci U S A
2001
;
98
:
9742
–7.
4
Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K, Tuschl T. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells.
Nature
2001
;
411
:
494
–8.
5
Hammond SM, Caudy AA, Hannon GJ. Post-transcriptional gene silencing by double-stranded RNA.
Nat Rev Genet
2001
;
2
:
110
–9.
6
Castanotto D, Tommasi S, Li M, et al. Short hairpin RNA-directed cytosine (CpG) methylation of the RASSF1A gene promoter in HeLa cells.
Mol Ther
2005
;
12
:
179
–83.
7
Kawasaki H, Taira K. Induction of DNA methylation and gene silencing by short interfering RNAs in human cells.
Nature
2004
;
431
:
211
–7.
8
Morris KV, Chan SW, Jacobsen SE, Looney DJ. Small interfering RNA-induced transcriptional gene silencing in human cells.
Science
2004
;
305
:
1289
–92.
9
Murayama A, Sakura K, Nakama M, et al. A specific CpG site demethylation in the human interleukin 2 gene promoter is an epigenetic memory.
EMBO J
2006
;
25
:
1081
–92.
10
Filleur S, Courtin A, it-Si-Ali S, et al. siRNA-mediated inhibition of vascular endothelial growth factor severely limits tumor resistance to antiangiogenic thrombospondin-1 and slows tumor vascularization and growth.
Cancer Res
2003
;
63
:
3919
–22.
11
Gondi CS, Lakka SS, Dinh DH, Olivero WC, Gujrati M, Rao JS. RNAi-mediated inhibition of cathepsin B and uPAR leads to decreased cell invasion, angiogenesis and tumor growth in gliomas.
Oncogene
2004
;
23
:
8486
–96.
12
Lakka SS, Gondi CS, Dinh DH, et al. Specific interference of uPAR and MMP-9 gene expression induced by double-stranded RNA results in decreased invasion, tumor growth and angiogenesis in gliomas.
J Biol Chem
2005
;
280
:
21882
–92.
13
Takei Y, Kadomatsu K, Yuzawa Y, Matsuo S, Muramatsu T. A small interfering RNA targeting vascular endothelial growth factor as cancer therapeutics.
Cancer Res
2004
;
64
:
3365
–70.
14
Verma UN, Surabhi RM, Schmaltieg A, Becerra C, Gaynor RB. Small interfering RNAs directed against β-catenin inhibit the in vitro and in vivo growth of colon cancer cells.
Clin Cancer Res
2003
;
9
:
1291
–300.
15
Rao JS. Molecular mechanisms of glioma invasiveness: the role of proteases.
Nat Rev Cancer
2003
;
3
:
489
–501.
16
Gaylis FD, Keer HN, Wilson MJ, Kwaan HC, Sinha AA, Kozlowski JM. Plasminogen activators in human prostate cancer cell lines and tumors: correlation with the aggressive phenotype.
J Urol
1989
;
142
:
193
–8.
17
Lakka SS, Bhattacharya A, Mohanam S, Boyd D, Rao JS. Regulation of the uPA gene in various grades of human glioma cells.
Int J Oncol
2001
;
18
:
71
–9.
18
Look MP, Foekens JA. Clinical relevance of the urokinase plasminogen activator system in breast cancer.
APMIS
1999
;
107
:
150
–9.
19
Pyke C, Kristensen P, Ralfkiaer E, et al. Urokinase-type plasminogen activator is expressed in stromal cells and its receptor in cancer cells at invasive foci in human colon adenocarcinomas.
Am J Pathol
1991
;
138
:
1059
–67.
20
Siddique K, Yanamandra N, Gujrati M, Dinh D, Rao JS, Olivero W. Expression of matrix metalloproteinases, their inhibitors, and urokinase plasminogen activator in human meningiomas.
Int J Oncol
2003
;
22
:
289
–94.
21
Skriver L, Larsson LI, Kielberg V, et al. Immunocytochemical localization of urokinase-type plasminogen activator in Lewis lung carcinoma.
J Cell Biol
1984
;
99
:
752
–7.
22
Van Veldhuizen PJ, Sadasivan R, Cherian R, Wyatt A. Urokinase-type plasminogen activator expression in human prostate carcinomas.
Am J Med Sci
1996
;
312
:
8
–11.
23
Yamamoto M, Sawaya R, Mohanam S, et al. Expression and localization of urokinase-type plasminogen activator in human astrocytomas in vivo.
Cancer Res
1994
;
54
:
3656
–61.
24
Hsu DW, Efird JT, Hedley-Whyte ET. Prognostic role of urokinase-type plasminogen activator in human gliomas.
Am J Pathol
1995
;
147
:
114
–23.
25
Miyake H, Hara I, Yamanaka K, Arakawa S, Kamidono S. Elevation of urokinase-type plasminogen activator and its receptor densities as new predictors of disease progression and prognosis in men with prostate cancer.
Int J Oncol
1999
;
14
:
535
–41.
26
Yang JL, Seetoo D, Wang Y, et al. Urokinase-type plasminogen activator and its receptor in colorectal cancer: independent prognostic factors of metastasis and cancer-specific survival and potential therapeutic targets.
Int J Cancer
2000
;
20
:
431
–9.
27
Pulukuri SM, Gondi CS, Lakka SS, et al. RNA Interference-directed knockdown of urokinase plasminogen activator and urokinase plasminogen activator receptor inhibits prostate cancer cell invasion, survival, and tumorigenicity in vivo.
J Biol Chem
2005
;
280
:
36529
–40.
28
Pulukuri SM, Estes N, Patel J, Rao JS. Demethylation-linked activation of urokinase plasminogen activator is involved in progression of prostate cancer.
Cancer Res
2007
;
67
:
930
–9.
29
Pulukuri SM, Rao JS. CpG island promoter methylation and silencing of 14–3-3σ gene expression in LNCaP and Tramp-C1 prostate cancer cell lines is associated with methyl-CpG-binding protein MBD2.
Oncogene
2006
;
25
:
4559
–72.
30
Bentley DL, Groudine M. A block to elongation is largely responsible for decreased transcription of c-myc in differentiated HL60 cells.
Nature
1986
;
321
:
702
–6.
31
Celano P, Berchtold C, Casero RA, Jr. A simplification of the nuclear run-off transcription assay.
Biotechniques
1989
;
7
:
942
–4.
32
Merscher S, Hanselmann R, Welter C, Dooley S. Nuclear runoff transcription analysis using chemiluminescent detection.
Biotechniques
1994
;
16
:
1024
–6.
33
Leunig M, Yuan F, Menger MD, et al. Angiogenesis, microvascular architecture, microhemodynamics, and interstitial fluid pressure during early growth of human adenocarcinoma LS174T in SCID mice.
Cancer Res
1992
;
52
:
6553
–60.
34
Jones PL, Veenstra GJ, Wade PA, et al. Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription.
Nat Genet
1998
;
19
:
187
–91.
35
Nan X, Ng HH, Johnson CA, et al. Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex.
Nature
1998
;
393
:
386
–9.
36
Li H, Lu H, Griscelli F, et al. Adenovirus-mediated delivery of a uPA/uPAR antagonist suppresses angiogenesis-dependent tumor growth and dissemination in mice.
Gene Ther
1998
;
5
:
1105
–13.
37
Gondi CS, Lakka SS, Yanamandra N, et al. Adenovirus-mediated expression of antisense urokinase plasminogen activator receptor and antisense cathepsin B inhibits tumor growth, invasion, and angiogenesis in gliomas.
Cancer Res
2004
;
64
:
4069
–77.
38
Janowski BA, Huffman KE, Schwartz JC, et al. Inhibiting gene expression at transcription start sites in chromosomal DNA with antigene RNAs.
Nat Chem Biol
2005
;
1
:
216
–22.
39
Park CW, Chen Z, Kren BT, Steer CJ. Double-stranded siRNA targeted to the huntingtin gene does not induce DNA methylation.
Biochem Biophys Res Commun
2004
;
323
:
275
–80.
40
Ting AH, Schuebel KE, Herman JG, Baylin SB. Short double-stranded RNA induces transcriptional gene silencing in human cancer cells in the absence of DNA methylation.
Nat Genet
2005
;
37
:
906
–10.
41
Dykxhoorn DM, Novina CD, Sharp PA. Killing the messenger: short RNAs that silence gene expression.
Nat Rev Mol Cell Biol
2003
;
4
:
457
–67.
42
Paddison PJ, Caudy AA, Hannon GJ. Stable suppression of gene expression by RNAi in mammalian cells.
Proc Natl Acad Sci U S A
2002
;
99
:
1443
–8.
43
Tuschl T. Expanding small RNA interference.
Nat Biotechnol
2002
;
20
:
446
–8.
44
Varambally S, Dhanasekaran SM, Zhou M, et al. The polycomb group protein EZH2 is involved in progression of prostate cancer.
Nature
2002
;
419
:
624
–9.
45
Halabi S, Small EJ, Kantoff PW, et al. Prognostic model for predicting survival in men with hormone-refractory metastatic prostate cancer.
J Clin Oncol
2003
;
21
:
1232
–7.
46
Kuhn W, Pache L, Schmalfeldt B, et al. Urokinase (uPA) and PAI-1 predict survival in advanced ovarian cancer patients (FIGO III) after radical surgery and platinum-based chemotherapy.
Gynecol Oncol
1994
;
55
:
401
–9.
47
Esteller M, Corn PG, Baylin SB, Herman JG. A gene hypermethylation profile of human cancer.
Cancer Res
2001
;
61
:
3225
–9.
48
Esteller M. Aberrant DNA methylation as a cancer-inducing mechanism.
Annu Rev Pharmacol Toxicol
2005
;
45
:
629
–56.