Overexpression of microRNA-9 enhances cisplatin sensitivity in hepatocellular carcinoma by regulating EIF5A2-mediated epithelial-mesenchymal transition

Abstract

We investigated the role of microRNA (miR)-9 in modulating chemoresistance in hepatocellular carcinoma (HCC) cells. MiR-9 was overexpressed or knocked down in HCC cell lines. Cell viability, cell proliferation, the expression of EIF5A2 and the epithelial-mesenchymal transition (EMT)-related proteins were examined. HCC cells overexpressing miR-9 were more sensitive to cisplatin; miR-9 knockdown yielded the opposite result. The in vivo nude mouse HCC xenograft tumors yielded the same results. EIF5A2 was identified as a potential target of miR-9, where miR-9 regulated EIF5A2 expression at mRNA and protein level. EIF5A2 knockdown reversed miR-9 inhibition-mediated cisplatin resistance. Altering miR-9 and EIF5A2 expression changed E-cadherin and vimentin expression. Furthermore, EIF5A2 mediated miR-9 EMT pathway regulation, indicating that miR-9 can enhance cisplatin sensitivity by targeting EIF5A2 and inhibiting the EMT pathway. Targeting miR-9 may be useful for overcoming drug resistance in HCC.

Keywords: EIF5A2; chemoresistance; epithelial-mesenchymal transition; miR-9.

Figure 1

miR-9 expression pattern in HCC cell lines and its relationship with cisplatin sensitivity. (A) The miR-9 expression was detected in 20 paired HCC tissues and normal tissues. (B) Real-time PCR investigation of miR-9 expression in HCC cell lines and the normal hepatic cells HL-7702. (C) CCK-8 assay of HCC cell viability in the presence of cisplatin (0, 1.5, 3.0, 4.5 μg/ml).

Figure 2

Effect of miR-9 plus cisplatin on HCC cell proliferation and apoptosis combined in vitro. (A-E) CCK-8 assay of the viability of HCC cells treated with miR-9 mimic, miR-9 inhibitor, or corresponding control RNAs in the presence of cisplatin (0, 1.5, 3.0, 4.5 μg/ml). (F-J) EdU assay of cell proliferation rate under cisplatin in HCC cell lines treated with miR-9 mimic, miR-9 inhibitor, or control RNAs. The EdU-positive cells were counted. *vs. cisplatin, ^#vs. cisplatin plus miR-9 mimic, *P < 0.05, **P < 0.01, ***P < 0.001, ^##P < 0.01, ^###P < 0.001.(K).QRT-PCR of miR-9 level in HCC cell treated with miR-9 mimic , inhibitor or combined with cisplatin, *P<0.05.

Figure 3

Effect of miR-9 plus cisplatin on HCC cell viability and proliferation in vivo. (A) Gross morphology of subcutaneous xenograft tumors. The tumors were treated with control, cisplatin, miR-9 mimic, or cisplatin plus miR-9 mimic. (B) Growth curves of xenograft tumors treated with control, cisplatin, miR-9 mimic, or cisplatin plus miR-9 mimic (n = 4 per group). *P < 0.05,***P < 0.001 vs Control. (C) The weights of mice were measured on day 0-15 which was treated with control, cisplatin, miR-9 mimic, or cisplatin plus miR-9 mimic. (D) Ki-67 staining in the treatment groups (×40 magnification) and rate of Ki-67 positive staining. Bottom panels depict the insets of the top panels (×100 magnification). (E) TUNEL of apoptosis in the treatment groups (×40 magnification). Bottom panels depict the insets of the top panels (×100 magnification). *vs. control, ^#vs. cisplatin plus miR-9 mimic, *P < 0.05, **P < 0.01, ***P < 0.001, ^##P < 0.01, ^###P < 0.001. IHC-P, immunohistochemistry-positive.

Figure 4

miR-9 targeted EIF5A2 in HCC cells. (A) The predicted miR-9 binding site in the EIF5A2 3′ UTR. (B) miR-9 and its supposed binding sequence in the 3'-UTR of EIF5A2. For seeking region of miR-9, mutant binding site was acquired in the complementary site. (C) Plasmids which carried wild-type (wt) or mutant (mt) 3'-UTR of EIF5A2 and miR-9 mimic or miR-9 inhibitor were co-transfected to HEK293T cells. Luciferase activity was measured using a dual-luciferase reporter assay system (Promega) and normalized to Renilla activity. * vs. Negative control, P < 0.05. (D) HCC cell lines were transfected with miR-9 mimic, miR-9 inhibitor, or control for 48 h. The total protein was extracted and subjected to western blotting. (E-I) HCC cell lines were transfected with miR-9 mimic, miR-9 inhibitor, or control for 48 h. The total RNA was isolated and subjected to RT-qPCR. *vs. control, ^#vs. miR-9 mimic, *P < 0.05, **P < 0.01, ***P < 0.001, ^##P < 0.01, ^###P < 0.001.

Figure 5

EIF5A2 expression pattern in HCC cell lines and its effect on cisplatin sensitivity. (A, B) Western blot (A) of EIF5A2 protein expression in HCC cell lines and its relative expression value as compared to GAPDH (B). (C) Western blot confirmation of EIF5A2 siRNA interference efficiency in HCC cell lines. (D-H) The effect of EIF5A2 on viability of HCC cells treated with EIF5A2 siRNA or negative siRNA in the presence of cisplatin (0, 1.5, 3.0, 4.5 μg/ml). *vs. Negative siRNA, *P < 0.05.

Figure 6

EIF5A2 reversed the effects of miR-9 on cell viability. (A) Western blot confirmation of EIF5A2 siRNA interference efficiency in Hep3B and SNU449 cells. (B, C) CCK-8 assay of Hep3B cells (B) and SNU449 cells (C) transfected with EIF5A2 siRNA plus synthetic miR-9 inhibitor or control in the presence of cisplatin (0, 1.5, 3.0, 4.5 μg/ml).

Figure 7

miR-9 and EIF5A2 regulated the EMT pathway. (A) Western blot comparison of E-cadherin and vimentin protein expression levels between HCC cells transfected with miR-9 mimic or miR-9 inhibitor. (B) Western blot comparison of E-cadherin and vimentin protein expression levels between HCC cells transfected with EIF5A2 siRNA or control siRNA. (C) E-cadherin and vimentin protein expression levels in Hep3B and SNU449 cells transfected with EIF5A2 siRNA alone or EIF5A2 siRNA plus miR-9 inhibitor. GAPDH was the internal control.