Targeting estrogen-related receptor alpha inhibits epithelial-to-mesenchymal transition and stem cell properties of ovarian cancer cells

Abstract

Epithelial-mesenchymal transition represents a key event in cancer progression and has emerged as a promising anticancer target. Estrogen-related receptor alpha (ERRα) is frequently elevated in advanced-stage ovarian cancer, but its potential role in tumor progression is not known. Here we show that ERRα functions in epithelial-mesenchymal transition and in subsequent stem cell traits responsible for the acquisition of high degree of aggressiveness and potential for metastasis that are characteristic of ovarian cancer. Importantly, targeted inhibition of ERRα also inhibited the expression of Snail, a repressor of E-cadherin and an inducer of epithelial-mesenchymal transition. Interestingly, induction of Snail resulted from not only changes in mRNA transcription rate but also mRNA stability. We thus identified the miR-200 family as a new player in the ERRα-mediated posttranscriptional regulation of Snail, and antagonism of miR-200a/b could revert the decreased expression of Snail and reversal of epithelial-mesenchymal transition and stem cell characteristics due to ERRα depletion. Finally, we showed that RNA interference-mediated inhibition of ERRα significantly reduced tumor burden, ascites formation, and metastatic peritoneal nodules in vivo in an orthotopic model of ovarian cancer. These results suggest ERRα activation as a mechanism of tumor aggressiveness and imply that targeting ERRα may be a promising approach in ovarian cancer treatment.

Figure 1

ERRα induces EMT in ovarian cells. (a) OVCAR-3 cells transfected with empty vector pcDNA3.1 or ERRα construct or (b) SKOV-3 cells transfected with NS siRNA or ERRα siRNA for 24 hours were fixed and assessed for morphologic changes consistent with EMT. Left, the presence of spindle-shaped cells with loss of epithelial cell morphology was noted in ERRα-overexpressing and NS siRNA–treated cells. Right, scattered colonies were scored. Inset, whole-cell lysates were analyzed for the levels of ERRα by western blot analysis. (c) Expression of the epithelial cell molecule E-cadherin and expression of mesenchymal marker N-cadherin were assessed by western blotting. β-actin was included as a loading control. The band intensities were quantified by densitometric analysis and expression levels relative to that of β-actin are indicated. Results are presented as the mean ± SD and were analyzed using Mann–Whitney U-test. *P < 0.05, compared with pcDNA3.1 or NS siRNA. Bar = 50 µm. EMT, epithelial–mesenchymal transition; ERRα, estrogen-related receptor alpha; NS, nonspecific; siRNA, small interfering RNA.

Figure 2

ERRα induces the expression of Snail. (a) OVCAR-3 cells were transfected with empty vector pcDNA3.1 or ERRα construct or (b) SKOV-3 cells were transfected with nonspecific (NS) siRNA or ERRα siRNA for 24 hours. Total RNA was extracted and reverse transcription–polymerase chain reaction was performed using sequence-specific primers to ERRα, Snail, and Slug. β-actin was included as an internal control. The signal intensities were quantified by densitometric analysis and the amount was normalized for the amount of β-actin. Results are presented as the mean ± SD and were analyzed using Mann–Whitney U-test. *P < 0.05, compared with pcDNA3.1 or NS siRNA. ERRα, estrogen-related receptor alpha; siRNA, small interfering RNA.

Figure 3

ERRα activates the promoter of Snail and attenuates its mRNA degradation. (a) OVCAR-3 cells were transiently transfected with 1.5 µg of the Snail promoter and 15 ng of β-galactosidase plasmid for 24 hours. Luciferase and β-galactosidase activities were assayed, and the luciferase activity of each sample was normalized with β-galactosidase activity. The luciferase activity was calculated relative to that with promoter alone, which was arbitrarily assigned a value of one. (b) OVCAR-3 cells transfected with the empty vector pcDNA3.1 or ERRα construct or (c) SKOV-3 cells transfected with nonspecific (NS) siRNA or ERRα siRNA were incubated with actinomycin D (ActD; 5 µg/ml) over a time course of 0, 1, 2, 4, and 6 hours. Total RNA was then extracted and reverse transcription–polymerase chain reaction was performed using Snail sequence–specific primers. β-actin was included as an internal control. The signal intensities were quantified by densitometric analysis, and the amount was normalized for the amount of β-actin. Results are presented as the mean ± SD and were analyzed using Kruskal–Wallis test followed by Dunn's test for post hoc analysis. *P < 0.05, compared with pcDNA3.1. ERRα, estrogen-related receptor alpha; siRNA, small interfering RNA.

Figure 4

ERRα-induced EMT requires miR-200 family members. SKOV-3 cells were transfected with NS siRNA or ERRα siRNA for 24 hours. Total RNA was extracted and reverse transcription–polymerase chain reaction was performed using sequence-specific primers to miR-141, miR-200a, miR-200b, and miR-200c. U6 was included as an internal control. The signal intensities were quantified by densitometric analysis and the amount was normalized for the amount of U6. Results are presented as the mean ± SD and were analyzed using Mann–Whitney U-test. *P < 0.05, compared with NS siRNA. EMT, epithelial–mesenchymal transition; ERRα, estrogen-related receptor alpha; NS, nonspecific; siRNA, small interfering RNA.

Figure 5

ERRα-induced EMT requires miR-200 family members. (a) SKOV-3 cells were transfected with nonspecific (NS) siRNA or ERRα siRNA in combination with NS miRNA, anti-miR-200a, or anti-miR-200b for 24 hours. Morphologic changes were assessed by light microscopy (upper) and scattered colonies were scored (lower). (b) Total RNA was extracted and reverse transcription–polymerase chain reaction was performed using sequence-specific primers to Snail. β-actin was included as an internal control. The signal intensities were quantified by densitometric analysis and the amount was normalized for the amount of β-actin. Results are presented as the mean ± SD and were analyzed using Kruskal–Wallis test followed by Dunn's test for post hoc analysis. *P < 0.05, compared with NS siRNA. Bar = 50 µm. EMT, epithelial–mesenchymal transition; ERRα, estrogen-related receptor alpha; siRNA, small interfering RNA.

Figure 6

Knockdown of ERRα inhibits CSC phenotype. (a) SKOV-3 cells were transfected with nonspecific (NS) siRNA or ERRα siRNA for 72 hours. The number of tumor spheres generated was photographed (left) and counted (right). Results are presented as the mean ± SD and were analyzed using Mann–Whitney U-test. (b) Total RNA was extracted and reverse transcription–polymerase chain reaction was performed using sequence-specific primers to Nanog, Oct-4, and Bmi-1. β-Actin was included as an internal control. The signal intensities were quantified by densitometric analysis and the amount was normalized for the amount of β-actin. Results are presented as the mean ± SD and were analyzed using Mann–Whitney U-test. (c) The number of tumor spheres generated was photographed (left) and counted (right). Bar = 50 µm. Results are presented as the mean ± SD and were analyzed using Kruskal–Wallis test followed by Dunn's test for post hoc analysis. *P < 0.05; **P < 0.01, compared with NS siRNA. Bar = 50 µm. CSC; cancer stem cells; ERRα, estrogen-related receptor alpha; siRNA, small interfering RNA.

Figure 7

ERRα knockdown inhibits peritoneal dissemination of ovarian cancer cells in vivo. (a) NOD/SCID mice were orthotopically injected with SKOV-3 cells transduced by nonspecific (NS) shRNA or ERRα shRNA (three mice per group). At the time of sacrifice, bioluminescence imaging was again performed, and the peritoneal cavity was assessed for evidence of metastases. (b) Ascites fluid was collected and the volume was measured. (c) The metastatic lesions were excised and their total number was counted. Results are presented as the mean ± SD and were analyzed using Mann–Whitney U-test. ERRα, estrogen-related receptor alpha; NOD/SCID, nonobese diabetic/severe combined immunodeficient; shRNA, short hairpin RNA.