Expression of miR-203 and IRS-1 in prostate cancer cells
miR-203 is abnormally expressed in a variety of malignancies, including prostate cancer [7, 10-17]. We first examined the expression level of miR-203 in prostate cancer cells compared to normal prostate cells (Fig. 1a). Consistent with previous reports [6], our result exhibited the dramatic reduction in prostate cancer cells while showed high expression in normal prostate cells determined by quantitative PCR. In consideration of miR-203 functions mainly through its target genes, we screened the potential targets by TargetScan6.2. The analysis data predicted that IRS-1 is a putative target gene of miR-203. Furthermore, we observed that miR-203 and IRS-1 have an opposite expression pattern among different prostate cell lines (Fig. 1b). This result further supports that IRS-1 is a potential target gene post-transcriptionally regulated by miR-203.
IRS-1 is a direct target gene regulated by miR-203
To determine whether IRS-1 was the direct target of miR-203, luciferase reporter gene constructs containing full-length IRS-1 3’-UTR were generated, together with their corresponding mutant or deletion counterpart at the binding sites of miR-203. We found that the 144 to 150nt and 2597 to 2604nt of the IRS-1 3’-UTR were two potential miR-203 binding sites. Both the binding sites were highly conserved across species (Fig. 2a). Sequence analysis showed no mutation or deletion of the IRS-1 3’-UTR in DU145 and PC-3 cells. Co-transfection of the reporters with miR-203 caused 54.3% and 71.7% decrease in luciferase activity of pGL3-IRS1-site A and pGL3-IRS1-site B constructs respectively compared with the control (Fig. 2b). Luciferase activity was recovered in cells transfected with either the mutant of pGL3-IRS1-site A and/or pGL3–IRS1-site B or deleted 3’-UTR seed sequences of pGL3-IRS1-site A and/or pGL3-IRS1-site B. As a negative control, the luciferase activity was not affected in cells transfected with reporter constructs lacking 3’-UTR sequences. These results clearly showed that miR-203 can specially bind to the 3′UTR region of the IRS-1, suggesting that IRS-1 is a direct target of miR-203.
We next examined the expression levels of IRS-1 by ectopically expressed miR-203 in PC-3 and DU145 cells (Fig. 2c). A remarkable decrease of IRS-1 protein level was detected in miR-203 overexpressing cells compared with pCDH control vector (Fig. 2e). However, there was no significant difference in IRS-1 mRNA expression after overexpression of miR-203 (Fig. 2d). These results suggest that miR-203 down-regulates the protein expression level of IRS-1 mainly by inhibiting the translation of IRS-1 mRNA instead of degrading mRNA. Together, the analysis not only identified the binding site for miR-203, but also confirmed that IRS-1 is a direct regulatory target gene of miR-203.
Down-regulation of IRS-1 inhibits cell proliferation of prostate cancer cells
We have previously demonstrated that over-expression of miR-203 significantly down-regulates IRS-1 protein expression, while previous studies have shown that over-expression of miR-203 can inhibit proliferation of prostate cancer cells [7]. Therefore, we speculate that down-regulation of IRS-1 may inhibit the proliferation of prostate cancer. To verify the assume, we first knockdown IRS-1 by shRNA in DU145 and PC-3. Both of shRNAs (shIRS1-1 and shIRS1-2) had a significant down-regulation of the expression level of IRS-1 (Fig. 3(a,b)). We then examined the effect of knocking down IRS-1 on prostate cancer cell proliferation by MTT compared with overexpression of miR-203. The results showed that knockdown of IRS-1 significantly inhibited DU145 and PC-3 proliferation (Fig. 3c), consistent with the overexpression of miR-203 (Fig. 3d). Thus, these results indicate the important role of IRS-1 in the proliferation of prostate cancer cells and also suggest that IRS-1 is most likely to be a functional target of miR-203.
miR-203 induced G0/G1 arrest of prostate cancer by down-regulating IRS-1
Cell proliferation is directly related to cell cycle. To investigate the mechanism of miR-203-mediated down-regulation of IRS-1 promoting growth inhibition, we knocked down IRS-1 in DU145 and PC-3 cells to check for subsequent cell cycle changes, compared with the change of cell cycle after overexpression of miR-203.
Flow cytometry analysis showed that knockdown of IRS-1 in DU145 and PC-3 induced cell cycle arrest at the G0/G1 phase (the G0/G1 phase increased from 66.40% to 75.16%/75.86% in DU145 and increased from 65.34% to 69.90%/70.43% in PC-3) (Fig. 4a). Overexpression of miR-203 had a similar effect, the G0/G1 phase increased from 63.39% to 68.66% in DU145 and increased from 52.22% to 56.23% in PC-3 (Fig. 4b).
To further determine the effect of miR-203 on cell cycle distribution by downregulating IRS-1, we generated cell lines DU145-miR-203 and PC-3-miR-203 that stably expressed miR-203, and constructed IRS-1 Over-expression vector. We found that restoration of IRS-1 in DU145-miR-203 and PC-3-miR-203 led to partially or completely reversed the blockade effect of miR-203 on the cell cycle (the G0/G1 phase decreased from 64.13% to 61.53% in DU145-miR-203 cells and decreased from 51.40% to 46.96% in PC-3-miR-203) (Fig. 4c), suggesting that IRS-1 can promote cell cycle progression and cell proliferation.
These results indicate that miR-203 can arrest cell cycle progression in the G0/G1 phase by down-regulating IRS-1, thereby inhibiting prostate cancer cell proliferation.
miR-203 inhibits ERK signaling pathway by targeting IRS-1
IRS-1 can transmit a variety of extra-cellular signal stimuli, acting as a scaffold to initiate intracellular signaling pathways. Previous reports have shown that AKT and ERK signaling pathways are the major signaling pathways downstream of IRS-1 [18]. To further investigate the molecular mechanism of miR-203 and IRS-1 in prostate cancer, we primarily detected the phosphorylated AKT (P-AKT) and phosphorylated ERK (P-ERK) protein levels to detect the activation on these two signaling pathways.
We found that P-ERK was significantly reduced after knocking down IRS-1 in DU145 and PC-3, while there was a little change in P-AKT (Fig. 5a), suggesting that IRS-1 mainly activates the ERK signaling pathway in prostate cancer. The effect of overexpressing miR-203 on AKT and ERK signaling pathways was similar to that of knocking down IRS-1 (Fig. 5b). Furthermore, up-regulation of P-ERK was detected after restoration of IRS-1 expression in DU145-miR-203 and PC-3-miR-203, suggesting that the restoration of IRS-1 expression may at least partially abolish the inhibitory effect of miR-203 on the ERK signaling pathway (Fig. 5c). We also detected a significant up-regulation of P-AKT after restoration of IRS-1 expression in DU145-miR-203 and PC-3-miR-203 (Fig. 5c), indicating that the expression of IRS-1 indeed affects the signaling of the AKT signaling pathway. Considering that knocking down IRS-1 does not significantly reduce the level of P-AKT, we speculate that down-regulation of IRS-1 may cause constitutive activation of AKT. That is to say, AKT can maintain its continuous activation state independent of IRS-1.
Above all, our data demonstrate that miR-203 can inhibit the signaling of ERK other than AKT by down-regulating IRS-1.
IRS-1 down-regulation inhibits prostate cancer metastasis
The progression and metastasis of prostate cancer are closely related to the stromal response (tumor-associated tissue remodeling) caused by tumor invasion into the stroma [19]. To investigate whether IRS-1 is involved in the regulation of tumor metastasis in PCa, we selected the gene chip GDS4114 from GEO database to analyze the expression of IRS-1 in invasive prostate cancer stroma and normal prostate stroma (Fig. 6a). The results showed that the expression level of IRS-1 in invasive prostate cancer stroma was significantly higher than that of normal prostate stroma (* P<0.05), suggesting that IRS-1 may be involved in the metastasis of prostate cancer.
To prove whether IRS-1 is involved in the regulation of prostate cancer metastasis, we first examined the migration ability of DU145 and PC-3 after knocking down IRS-1 or overexpressing miR-203. Compared with the control group, either knockdown of IRS-1 or overexpression of miR-203 significantly decreased cell migration, indicating that down-regulation of IRS-1 can inhibit the migration of prostate cancer cells (Fig. 6(b,c)).
The epithelial-mesenchymal transition (EMT) process can deprive cells of their ability to bind tightly to neighboring cells, allowing them to escape from orthotopic tumors and migrate throughout the body. Previous studies have demonstrated that up-regulation of miR-203 expression in prostate cancer cells results in up-regulation of E-cadherin and down-regulation of Vimentin, thereby inhibiting EMT transformation [6, 7]. Whether IRS-1, the target of miR-203, is involved in regulating the EMT process of PCa is still unknown. Hence, we examined the expression of E-cadherin and Vimentin after knocking down IRS-1. However, our data showed that IRS-1 doesn’t affect EMT conversion since there was no significant change in the expression of E-cadherin and Vimentin after knocking down IRS-1 in DU145 and PC-3 (Fig. 6d). Considering that miR-203 can also play a role in multiple steps of PCa transfer cascade by inhibiting a series of metastatic genes, including ZEB2, Runx2, Dlx5, WASF1, LASP1, CKAP2 [6, 7], we speculate that miR-203 may regulate the EMT transformation of prostate cancer cells by targeting other proteins in prostate cancer.
The above results indicate that down-regulation of IRS-1 can inhibit the migration of prostate cancer cells, while it does not seem to affect the EMT transformation process. Whether IRS-1 is involved in the regulation of prostate cancer metastasis needs further research and confirmation.