Aurora-A promotes chemoresistance in hepatocelluar carcinoma by targeting NF-kappaB/microRNA-21/PTEN signaling pathway

Abstract

Hepatocellular carcinoma (HCC) is highly resistant to chemotherapy. Previously, we have shown that Aurora-A mRNA is upregulated in HCC cells or tissues and silencing of Aurora-A using small interfering RNA (siRNA) decreases growth and enhances apoptosis in HCC cells. However, the clinical significance of Aurora-A protein expression in HCC and association between Aurora-A expression and HCC chemoresistance is unclear. Here, we showed that Aurora-A protein is upregulated in HCC tissues and significantly correlated with recurrence-free and overall survival of patients and multivariate analysis indicated that immunostaining of Aurora-A will be an independent prognostic factor for patients. Silencing of Aurora-A significantly increased the chemosensitivity of HCC cells both in vitro and in vivo, while overexpression of Aurora-A induced the opposite effects. Furthermore, overexpression of Aurora-A reduces chemotherapy-induced apoptosis by promoting microRNA-21 expression, which negatively regulates PTEN and then inhibits caspase-3-mediated apoptosis induction. Mechanically, we demonstrated that Aurora-A promotes expression of nuclear Ikappaβ-alpha (Iκβα) protein and enhances NF-kappa B (NF-κB) activity, thus promotes the transcription of miR-21. This study first reported the involvement of Aurora-A/NF-κB/miR-21/PTEN/Akt signaling axis in chemoresistance of HCC cells, suggesting that targeting this signaling pathway would be helpful as a therapeutic strategy for the reversal of chemoresistance in HCC.

Figure 1. Expression of Aurora-A protein in HCC tissues and its correlation with prognosis of patients

(A) Western blotting was performed to detect Aurora-A protein level in 44 paired of HCC and NTs. Three pairs of human HCC samples were exhibited. GAPDH was used as an internal control. (B) Immunohistochemical analysis. Intensive staining was observed in the cytoplasma of HCC cells, whereas less staining was showed in NTs (original magnification a, c: ×200; b, d: ×400). (C) The Kaplan-Meier survival curve of recurrence-free survival (RFS) according to Aurora-A immunostaining in 44 cases of HCC patients. (D) The Kaplan-Meier survival curve of overall survival (OS) according to Aurora-A immunostaining in 44 cases of HCC patients. Data were presented as mean ± SD of at least three independent experiments. N.S, P>0.05; *P<0.05; **P<0.01.

Figure 2. Effects of chemotherapy on the expression of Aurora-A and apoptosis-related proteins

(A) Western blotting detection of Aurora-A protein expression in three HCC cell lines (HepG2, Hep3B and SMMC-7721) and a normal human hepatocyte cell line (HH). (B) MTT analysis of IC₅₀ values of ADR and CDDP in three HCC cell lines (HepG2, Hep3B and SMMC-7721). (C) 24h after HCC cells (HepG2, Hep3B and SMMC-7721) were treated with ADR (0.5 μg/ml), Western blotting detection of Aurora-A, p-Aurora-A, PTEN, Bcl-2, Bax and cleaved caspase-3 proteins. (C) 24h after HCC cells (HepG2, Hep3B and SMMC-7721) were treated with CDDP (1.0 μg/ml), Western blotting detection of Aurora-A, p-Aurora-A, PTEN, Bcl-2, Bax and cleaved caspase-3 proteins. GAPDH was used as an internal control. Data were presented as mean ± SD of at least three independent experiments. N.S, P>0.05; *P<0.05; **P<0.01.

Figure 3. Effects of Aurora-A downregulation on in vitro chemosensitivity of HCC cells

(A) qRT-PCR and Western blotting detection of Aurora-A mRNA and protein expression in stably transfected SMMC-7721/shAurora-A or SMMC-7721/shcontrol cells, respectively. GAPDH was used as an internal control. (B) MTT analysis of IC₅₀ values of ADR and CDDP in SMMC-7721/shAurora-A or SMMC-7721/shcontrol cells, respectively. (C) The colony formation of SPC-A1 and SPC-A1/DTX cells treated with various concentrations of ADR (0.0 or 0.5 μg/ml) or CDDP (0.0 or 1.0 μg/ml). (D) Flow cytometric analysis of apoptosis in SMMC-7721/control or SMMC-7721/Aurora-A cells combined with various concentrations of ADR (0.0, 1.0 and 2.0 μg/ml) or CDDP (0.0, 1.5 and 3.0 μg/ml). Data were presented as mean ± SD of at least three independent experiments. N.S, P>0.05; *P<0.05; **P<0.01.

Figure 4. Effects of Aurora-A downregulation on in vivo chemosensitivity of HCC cells

Mice were treated with ADR (2.0 mg/kg body weight; i.p., thrice), DDP (3.0 mg/kg body weight; i.p., thrice), or with 0.1ml PBS (pH 7.4; i.p., thrice). (A) Growth of tumors in the mice injected with SMMC-7721/shAurora-A or SMMC-7721/shcontrol cells treated with ADR, DDP or PBS. The inoculation was done in eight mice. (B) Representative features of tumors 35d after inoculation using SMMC-7721/shAurora-A or SMMC-7721/shcontrol cells treated with ADR, DDP or PBS. (C) Western blotting detection of Aurora-A protein expression in tumors developed from SMMC-7721/shAurora-A or SMMC-7721/shcontrol cells treated with ADR, DDP or PBS, respectively. GAPDH was used as an internal control. (D) Immunostaining of Aurora-A, Ki-67 and PCNA protein expression in tumors developed from SMMC-7721/shAurora-A or SMMC-7721/shcontrol cells treated with ADR, DDP or PBS. Upper: H&E staining; Intermediate and lower: immunostaining; Bars, 100μm. (E) TUNEL assay detection of apoptosis in tumors developed from SMMC-7721/shAurora-A or SMMC-7721/shcontrol cells treated with ADR, DDP or PBS, respectively. Data were presented as mean ± SD of at least three independent experiments. N.S, P>0.05; *P<0.05; **P<0.01.

Figure 5. Effects of Aurora-A upregulation on in vitro chemosensitivity of HCC cells

(A) qRT-PCR and Western blotting detection of Aurora-A mRNA and protein expression in stably transfected HepG2/Aurora-A or HepG2/control cells, respectively. GAPDH was used as an internal control. (B) MTT analysis of IC₅₀ values of ADR or CDDP in HepG2/Aurora-A or HepG2/control cells, respectively. (C) The colony formation of HepG2/Aurora-A or HepG2/control cells treated with various concentrations of ADR (0.0 or 0.5 μg/ml) or CDDP (0.0 or 1.0 μg/ml). (D) Flow cytometric analysis of apoptosis in HepG2/Aurora-A or HepG2/control cells combined with various concentrations of ADR (0.0, 1.0 and 2.0 μg/ml) or CDDP (0.0, 1.5 and 3.0 μg/ml). Data are expressed as the mean±SD of three individual experiments. N.S, P>0.05; *P<0.05; **P<0.01.

Figure 6. Effects of Aurora-A upregulation on in vivo chemosensitivity of HCC cells

Mice were treated with ADR (2.0 mg/kg body weight; i.p., thrice), DDP (3.0 mg/kg body weight; i.p., thrice), or with 0.1ml PBS (pH 7.4; i.p., thrice). (A) Growth of tumors in the mice injected with HepG2/Aurora-A or HepG2/control cells treated with ADR, DDP or PBS. The inoculation was done in eight mice. (B) Representative features of tumors 28d after inoculation using SMMC-7721/shAurora-A or SMMC-7721/shcontrol cells treated with ADR, DDP or PBS. (C) Western blotting detection of Aurora-A protein expression in tumors developed from HepG2/Aurora-A or HepG2/control cells treated with ADR, DDP or PBS, respectively. GAPDH was used as an internal control. (D) Immunostaining of Aurora-A, Ki-67 and PCNA protein expression in tumors developed from HepG2/Aurora-A or HepG2/control cells treated with ADR, DDP or PBS. Upper: H&E staining; Intermediate and lower: immunostaining; Bars, 100μm. (E) TUNEL assay detection of apoptosis in tumors developed from HepG2/Aurora-A or HepG2/control cells treated with ADR, DDP or PBS, respectively. Data were presented as mean ± SD of at least three independent experiments. N.S, P>0.05; *P<0.05; **P<0.01.

Figure 7. Aurora-A inhibits chemotherapy-induced apoptosis in HCC cells by upregulation of miR-21

(A) Western blotting detection of apoptosis-related proteins (PTEN, p-Akt, total Akt, Bcl-2, Bax, cleaved caspase-3 and total caspase-3) in SMMC-7721/shAurora-A (or SMMC-7721/shcontrol) or HepG2/Aurora-A (or HepG2/control) cells. (B) Western blotting detection of total Aurora-A, p-Aurora-A and above apoptosis-related proteins in SMMC-7721 cells treated with various concentrations of MLN8237 (0.0, 0.5, 1.0 and 2.0 nM) for 24h or lengths (0, 6 and 12h) of MLN8237 (1.5 nM). (C) Flow cytometry detection of apoptosis in SMMC-7721 cells treated with various concentrations of MLN8237 (0.0, 0.5, 1.0 and 2.0 nM) for 24h or lengths (0, 6 and 12h) of MLN8237 (1.5 nM). (D) 24h after HCC cells (HepG2 and SMMC-7721) were treated with PBS, ADR (0.5 μg/ml) or CDDP (1.0 μg/ml), qRT-PCR detection of miR-21 expression. (E) qRT-PCR detection of miR-21 expression in SMMC-7721/shAurora-A (or SMMC-7721/shcontrol) or HepG2/Aurora-A (HepG2/control) cells, respectively. (F) qRT-PCR detection of miR-21 expression in SMMC-7721 cells treated with various concentrations of MLN8237 (0.0, 0.5, 1.0 and 2.0 nM) for 24h or lengths (0, 6 and 12h) of MLN8237 (1.5 nM). (G) MTT analysis of IC₅₀ values of ADR or CDDP in SMMC-7721/shAurora-A or SMMC-7721/shcontrol cells or co-transfection with miR-21 or miR-NC mimics, or in HepG2/Aurora-A or HepG2/control or or co-transfection with miR-21 or miR-NC inhibitor, respectively. (H) Western blotting detection of above apoptosis-related proteins in SMMC-7721/shAurora-A cells or co-transfection with miR-21 or miR-NC mimics. (I) Western blotting detection of above apoptosis-related proteins in HepG2/Aurora-A cells or co-transfection with miR-21 or miR-NC inhibitor. GAPDH or U6 was used as an internal control. Data were presented as mean ± SD of at least three independent experiments. N.S, P>0.05; *P<0.05; **P<0.01.

Figure 8. Aurora-A upregulates miR-21 transcription via NF-κB in HCC cells

(A) Western blotting detection of nuclear or cytoplasmic IκBα protein expression in SMMC-7721/shAurora-A (or SMMC-7721/shcontrol) or HepG2/Aurora-A (or HepG2/control) cells. Topo I or GAPDH was used as an internal control, respectively. (B) The activity of NF-κB was measured in SMMC-7721/shAurora-A (or SMMC-7721/shcontrol) or HepG2/Aurora-A (or HepG2/control) cells using a luciferase reporter system. (C) The promoter of miR-21 was activated by Aurora-A. SMMC-7721/shAurora-A (or SMMC-7721/shcontrol) or HepG2/Aurora-A (or HepG2/control) cells were treated with pGL3/miR-21-promoter or pGL3-Basic, respectively. (D) ChIP assays with anti-NF-κB/p65 antibodies showed binding of NF-κB to the promoter of miR-21 in SMMC-7721/shAurora-A (or SMMC-7721/shcontrol) or HepG2/Aurora-A (or HepG2/control) cells. The relative occupancies of NF-κB/p65 are indicated as vertical bars. The bar graphs show the averages of three independent ChIP experiments. (E) qRT-PCR and Western blotting detection of p65 mRNA and protein expression in SMMC-7721 transiently transfected with control/siRNA, p65/siRNA1 or p65/siRNA2, respectively. (F) qRT-PCR detection of miR-21 expression in SMMC-7721 cells transiently transfected with control/siRNA or p65/siRNA1, or HepG2/Aurora-A (or HepG2/control) cells or co-transfection with control/siRNA or p65/siRNA1, respectively. U6 was used as an internal control. (G) MTT analysis of IC₅₀ values of ADR or CDDP in HepG2/Aurora-A (or HepG2/control) or co-tranfection with control/siRNA or p65/siRNA1, respectively. (H) Western blotting detection of above apoptosis-related proteins in HepG2/Aurora-A cells or co-transfection with control/siRNA or p65/siRNA1, respectively. GAPDH was used as an internal control. Data were presented as mean ± SD of at least three independent experiments. N.S, P>0.05; *P<0.05; **P<0.01.

Figure 9. Expression of miR-21 and PTEN protein and their correlations with Aurora-A protein expression in HCC tissues

(A) qRT-PCR detection of miR-21 expression in 44 paired of HCC and NTs. (B) Western blotting detection of PTEN protein expression in 44 paired of HCC and NTs. (C) A statistically significant positive correlation between miR-21 and Aurora-A protein expression levels in 44 cases of LAD tissues (Spearman's correlation analysis, r = 0.896; P<0.01). (D) A statistically significant inverse correlation between Aurora-A and PTEN protein expression levels in 44 cases of LAD tissues (Spearman's correlation analysis, r = −0.816; P<0.01). (E) A statistically significant inverse correlation between miR-21 and PTEN protein expression levels in 44 cases of LAD tissues (Spearman's correlation analysis, r = −0.814; P<0.01). Data were presented as mean ± SD of at least three independent experiments. Corresponding P values analyzed by Spearman correlation test are indicated.