miR-30a inhibits endothelin A receptor and chemoresistance in ovarian carcinoma

Abstract

Drug resistance remains the major clinical barrier to successful treatment in epithelial ovarian carcinoma (EOC) patients, and the evidence of microRNA involvement in drug resistance has been recently emerging. Endothelin-1 (ET-1)/ETA receptor (ETAR) axis is aberrantly activated in chemoresistant EOC cells and elicits pleiotropic effects promoting epithelial-to-mesenchymal transition (EMT) and the acquisition of chemoresistance. However, the relationship between ETAR and miRNA is still unknown. Hence, in this study we evaluated whether dysregulation of miRNA might enhance ETAR expression in sensitive and resistant EOC cells. Based on bioinformatic tools, we selected putative miRNA able to recognize the 3'UTR of ETAR. An inverse correlation was observed between the expression levels of miR-30a and ETAR in both EOC cell lines and tumor samples. miR-30a was found to specifically bind to the 3'UTR of ETAR mRNA, indicating that ETAR is a direct target of miR-30a. Overexpression of miR-30a decreased Akt and mitogen activated protein kinase signaling pathway activation, cell proliferation, invasion, plasticity, EMT marker levels, and vascular endothelial growth factor release. Interestingly, ectopic expression of miR-30a re-sensitized platinum-resistant EOC cells to cisplatinum-induced apoptosis. Consistently, resistant EOC xenografts overexpressing miR-30a resulted in significantly less tumor growth than controls. Together our study provides a new perspective on the regulatory mechanism of ETAR gene. Interestingly, our findings highlight that blockade of ETAR regulatory axis is the mechanism underlying the tumor suppressor function of miR-30a in chemoresistant EOC cells.

Keywords: chemoresistance; endothelin A receptor; endothelin-1; miR-30a; ovarian carcinoma.

Figure 1. miR-30a is downregulated in chemoresistant EOC cells

A. Expression of miR-30 family members in sensitive and resistant 2008 cells measured by qPCR. miRNA levels are normalized using endogenous U6 snRNA. Values are the mean ± SD (n = 3; *, p < 0.005 compared to sensitive cells). Lysates from sensitive and resistant 2008 and A2780 cells transfected with mimic-miR control (Ctr) or mimic-miR-30a B. or from SKOV3, HEY and OVCA433 EOC cell lines C. are analyzed by western blotting for ET_AR expression. β-actin is used as loading control. D. Expression levels of ET_AR and miR-30a in a panel of seven EOC cell lines. The ratio of ET_AR/β-actin, evaluated by western blotting as shown in B and C, and miR-30a/U6 expression, evaluated by qPCR, is shown as bar and line, respectively. E. Scatter plot of the expression of miR-30a, as determined by qPCR, in the 39 EOC patients dichotomized into ET_AR high or ET_AR low expressing tumors. The expression levels of miR-30a were normalized to RNU44 expression. Horizontal lines indicate median n = 13 (low ET_AR), n = 26 (high ET_AR) and Mann-Whitney ran sum test was used in the statistical analysis (p = 0.0081).

Figure 2. ET_AR is a novel target of miR-30a

A. Luciferase activity in 2008 cells cotransfected with mimic-miR control (Ctr) or mimic-miR-30a and the reporter plasmid containing 3′UTR region of ET_AR (WT ET_AR 3′UTR) and its mutant (mutant ET_AR 3′UTR). Values are the mean ± SD (n = 3; *, p < 0.05 compared to Ctr). B. Western blotting for ET_AR in the lysates from 2008 CIS and A2780 CIS cells transfected for 48h with mimic-miR control, mimic-miR-30a or anti-miR-30a. β-actin is used as loading control. C. Cell viability of sensitive and resistant 2008 and A2780 cells, transfected as in B or treated with macitentan (MAC). Values are the mean ± SD (n = 3; *, p < 0.01 vs Ctr). D. Cell vitality of 2008 CIS cells, transfected with control plasmid alone (Ctr), or with mimic-miR-30a, or si-ET_AR, or with ET_AR expression plasmid alone or in combination with mimic-miR30a. Values are the mean ± SD (n = 3; *, p < 0.05 compared to Ctr; **, p < 0.05 compared to miR-30a). E. Lysates of cells treated as in D are analyzed by western blotting for ET_AR expression. β-actin is used as loading control.

Figure 3. Ectopic expression of miR-30a sensitizes EOC cells to cisplatinum-induced apoptosis

A. Lysates from A2780 and A2780 CIS cells transfected with Ctr or mimic-miR-30a and stimulated with ET-1 as indicated, are immunoblotted with anti-pMAPK, anti-MAPK, anti-pAkt and anti-Akt Abs. β-actin is used as loading control. B. Cell vitality of sensitive and resistant A2780 cells transfected with Ctr or mimic-miR-30a and treated with cisplatinum (CIS; 1 μM) for 72 h alone or in combination. Values are the mean ± SD (n = 3; *, p < 0.05 vs Ctr; **, p < 0.05 vs cisplatinum-treated cells). C. Detection of DNA fragmentation in sensitive and resistant A2780 cells cultured for 72 h as indicated in B. Values are the mean ± SD (n = 3; *, p < 0.05 vs Ctr; **, p < 0.05 vs cisplatinum-treated cells). D. TUNEL immunofluorescence staining for apoptotic cells (green) treated for 72 h as indicated in B. Cell nuclei are counterstained with DAPI (blue). Graph represents the percentage of apoptotic cell numbers. Magnification x40. Values are the mean ± SD (n = 3; *, p < 0.05 vs Ctr; **, p < 0.05 vs cisplatinum-treated cells). E. Western blotting for cleaved PARP (cl-PARP) and caspase-7 (cl-caspase-7) in sensitive and resistant A2780 cells transfected and treated as indicated in B. β-actin is used as loading control.

Figure 4. miR-30a inhibits EMT phenotype and cell invasion

A. E-cadherin, Snail and vimentin mRNA expression in 2008 or 2008 CIS cells transfected with Ctr or mimic-miR-30a evaluated by qPCR. Cyclophilin-A is used to normalize. Values are the mean ± SD (n = 3; *, p < 0.05 vs Ctr of sensitive cells; **, p < 0.05 vs Ctr of resistant cells). B. Lysates from resistant 2008 CIS cells transfected with Ctr or mimic-miR-30a are analyzed by western blotting for E-cadherin and N-cadherin expression. β-actin is used as loading control. C. Chemoinvasion assay of 2008 and 2008 CIS cells transfected with Ctr or mimic-miR-30a. The crystal violet-stained invasive cells are photographed (left) or counted (right). Magnification x40. Columns show the mean ± SD (n = 3; *, p < 0.001 vs Ctr of sensitive cells; **, p < 0.001 vs Ctr of resistant cells).

Figure 5. miR-30a inhibits the release of VEGF and vasculogenic-like tubule formation

A. VEGF release evaluated by ELISA from conditioned media of 2008 CIS cells transfected with Ctr or mimic-miR-30a in the presence or in the absence of ET-1. Values are the mean ± SD (n = 3; *, p < 0.001 vs Ctr; **, p < 0.001 vs ET-1-treated cells). Tubule-like structure formation in 2008 B. or 2008 CIS C. cells transfected with Ctr or mimic-miR-30a was photographed (left) and counted (right). Magnification x20. Graphs represent the tube length and the number of intersections. Values are the mean ± SD (n = 3; *, p < 0.05 vs Ctr).

Figure 6. Tumor-inhibitory effects of miR-30a in resistant EOC xenografts

A. Growth curve of tumors from vector control (Ctr) or miR-30a-expressing resistant 2008 CIS cells s.c. injected into the mice. Data points, averages ± SD (p<0.05). B. Tumor weight of tumors grown in vector control (Ctr) and miR-30a 2008 CIS xenografts reported as the mean ± SD. *, p < 0.05 vs Ctr. Bottom, representative tumors of 2008 CIS xenografts. C. Western blotting analysis of ET_AR, p-MAPK, MAPK, p-Akt and Akt expression in tumors of vector control (Ctr) and miR-30a 2008 CIS xenografts. β-actin is used as loading control. D. A schematic model describing the potential mechanism underlying the suppressive role of miR-30a in EOC. In chemoresistant EOC cells downregulated miR-30a expression has strong increased effects on ET_AR expression promoting enhancement of Akt and MAPK signaling activation. Therefore miR-30a downregulation results in increased cell proliferation, invasion, plasticity, VEGF release, EMT and drug resistance by inhibiting ET_AR pathway.

Figure 7. miR-30a correlates with poor survival in TCGA of ovarian cancer

A. Boxplot distribution of miR-30a expression in 567 tumors and 8 normal samples. B. Boxplot distribution of miR-30a expression in 112 patients subdivided into low or high miR-30a by using z-score higher than 1 (n = 55) and lower than −1 (n = 57). C. Kaplan-Meier curves of progression free survival for subgroup of 112 patients described in B. Univariate and multivariate analysis by Cox proportional hazard regression for subgroup of patients described in B, adjusted for stage, grade, and treatment response. D. ET_AR gene expression resulted to be inversely correlated to miR-30a expression in the TCGA cohort of HG-SOC patients (n = 303 tumoral samples). A significative inverse correlation was found in both sensitive and resistant samples.