The significance of EphA2-regulated Wnt/β-catenin signal pathway in promoting the metastasis of HBV-related hepatocellular carcinoma

Hepatitis B virus (HBV) infection is closely associated with the malignant progression of hepatocellular carcinoma (HCC). However, the mechanism involved in the HBV-related HCC development remains poorly understood. Hence, the aim of this study is to investigate the regulatory mechanism of EphA2-induced epithelial-mesenchymal transition (EMT) in the metastasis of HBV-related HCC cells. The expression level of EphA2 was determined in HBV-related human HCC cells. Then, the effects of EphA2 silencing on the EMT-associated proteins, the Wnt/β-catenin signal pathway and the metastatic potential of HBV-related HCC cells were evaluated. Finally, the inhibitory role of Entecavir (a potent antiviral drug for HBV) on EphA2-induced EMT was explored. The present study revealed that the EphA2 expression level was increased in HBV-related HCC cells compared with non-related HCC cells. Following EphA2 knockdown, the downregulation of Vimentin, β-catenin and p-GSK-3βSer9 expressions, the upregulation of E-cadherin expression, and the suppressed migration and invasion ability of HBV-related HCC cells were found. Additionally, Entecavir was proved to have a significant inhibitory effect on EphA2-induced EMT via attenuating the Wnt/β-catenin signal pathway. In this study, we found that EphA2-induced EMT was involved in the enhanced metastatic potential of HBV-related HCC cells through the activation of the Wnt/β-catenin signal pathway.


Introduction
Hepatocellular carcinoma (HCC), which has high metastatic potential, is one of the most common malignant tumors worldwide. HBV infection has been considered to be the major cause of HCC. HBV surface antigen (HBsAg) and HBV-DNA can be detected in a large number of patients with HCC [1]. It has been reported that the HBV-DNA level was closely linked to the malignant progression of HCC cells [2]. However, the mechanism involved in the HBV-induced HCC development remains poorly understood.
EphA2, a member of the receptor tyrosine kinase family, plays an essential role in the regulation of the growth, proliferation and transformation of cells at low levels in epithelial cells of various tissues [3]. To date, EphA2 has been found to be abnormally expressed at high levels in a variety of tumors and EphA2 overexpression can promote the malignant transformation and distant metastases of tumor cells [4,5]. Epithelial-mesenchymal transition (EMT) has been recognized as a typical pathological feature in the metastasis of tumor cells [6]. Through EMT, cancer cells of epithelial origin can break away from orthotopic tumors, invade the surrounding tissues and blood vessels, and metastasize to distant sites [7,8]. EMT also plays a pivotal role in tumor immunosuppression and EMT-induced immune evasion may be associated with the activation of different immune checkpoint molecules [9]. It has been indicated that EphA2 may be a potential upstream regulator of EMT and that EphA2-induced EMT might be involved in the migration and invasiveness of different types of tumor cells [10].
Our previous study revealed that the expression of EphA2 protein in HCC patients can be influenced by HBV infection, but the detailed regulatory mechanism has not been well elucidated. Therefore, the purpose of this study is to investigate the regulatory mechanism of EphA2induced EMT in the metastasis of HBV-related HCC cells.

Cell lines and drugs
The THLE-3 (normal human liver cells), HepG2 (human hepatoma cells) and HepG2.2.15 (HBV transfected HepG2 cells) cell lines were obtained from the Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences (Beijing, China). All cell lines were authenticated by STR profiling. Entecavir was purchased from Sino American Shanghai Squib Pharmaceutical Limited (Shanghai, China) and the HepG2.2.15 cells were treated with different concentration of Entecavir (5 μg/mL, 15 μg/mL and 50 μg/mL).

EphA2-shRNA vector transfection
According to the target sequence "GAA GTT CAC TAC CGA GATC", a recombinant lentiviral expression vector of EphA2-shRNA was designed and constructed by Jikai Gene Chemical Technology Company Limited (Shanghai, China). The liver cancer cells in the logarithmic stage of growth were digested with trypsin and seeded at a density of 5 × 10 5 cell per well in 6-well plates. After the fusion degree of cells reached 30%, the EphA2-shRNA lentiviral vector or the blank vector was added into the HepG2 and HepG2.2.15 cells. Three days after transfection, the EphA2 knockdown efficiency was found to be greater than 80%.

Cell migration and invasion analysis
When the cells were grown to 70% confluence, the initial scratch width (730 μm) was generated using a 96-pin Wound Maker tool. Following washing with PBS solution, the cells were cultured with DMEM/F12 medium. For invasion analysis, the matrigel was applied to create a 3D matrix on top of the cells. Time-lapse images of the cells were obtained at 36, 48, 60, 72, 84, 96, 120 and 144 h with six samples per group, and the migration length and the invasion length (the difference between the initial scratch width and the final wound width) were determined using a real-time quantitative live-cell analysis system (IncuCyte Zoom, Essen Bioscience, MI, USA). Each test process was repeated three times.

Immunofluorescence assay
The cells were fixed in 95% alcohol at 4 °C for 4 h and were incubated with 0.3% H 2 O 2 at 37 °C for 10 min. Following washing with PBS, the cells were immersed into 1% bovine serum albumin at 37 °C for 30 min. The cells were incubated with the primary antibody including EphA2 (CST, MA, USA), E-cadherin, Vimentin, β-catenin and p-GSK-3β Ser9 (Abcam, Cambridge, UK) at 4 °C overnight, which was followed by an incubation with a rhodamine-conjugated secondary antibody at 37 °C for 30 min. Finally, the cells were counterstained with DAPI (4′,6-Diamidino-2-Phenylindole) and observed by a fluorescence microscopy. The density (mean) of each protein was then quantified using the image pro plus 6.0 (Media Cybernetics, DE, USA). Each test process was repeated three times with three samples per group.

Immunocytochemical staining
The cells were grown on coverslips and fixed in 4% polyformaldehyde at 4 °C. After washed with PBS, the cells were incubated with 0.3% H 2 O 2 for 10 min at room temperature. Following blocking with 5% goat serum, the cells were incubated with a rabbit anti-EphA2 monoclonal antibody (1:100; CST, MA, USA) at 4 °C overnight and were immersed in a solution of horseradish peroxidase-labelled secondary antibody for at 37 °C 30 min. After stained with DAB, the cells were dehydrated, cleared and mounted. The density (mean) of EphA2 was quantified using the image pro plus 6.0 (Media Cybernetics, DE, USA). Each test process was repeated three times with three samples per group.

Assessment of HBV-DNA level
After the HepG2.2.15 cells were treated with Entecavir (5 μg/mL, 15 μg/mL or 50 μg/mL) separately for 48 h with six samples per group, the load of HBV-DNA in the culture supernatant was determined using real-time fluorescence quantitative PCR. The HBV-DNA level was calculated according to the standard curve and a logarithmic transformation was applied in data analysis. Each test process was repeated three times.

Western blotting analysis
After the HepG2.2.15 cells were treated with Entecavir for 48 h, the total protein was separated in SDS-PAGE gels and electrotransferred onto PVDF membranes. The transferred membranes were blocked with 5% skim milk for 2 h and incubated with the primary monoclonal antibody including EphA2, β-catenin, p-GSK-3β Ser9 , E-cadherin and c-myc (1:1000; CST, MA, USA) at 4 °C overnight. After washed with Tris-HCl, the membranes were incubated with the secondary antibody (1:3000) at 37 °C for 1.5 h. The protein bands were quantified using a gel image analysis system (Image J). Briefly, the initial images were first transferred to gray images and then the background was subtracted. Following setting the measurement scales, all of the bands were measured. GAPDH was used as an endogenous control. Each test process was repeated three times with three samples per group.

Statistical analysis
All the data were expressed as the mean ± standard deviation and were analysed using the SPSS 11.5 software (Chicago, IL, USA). One-way analysis of variance with post hoc contrasts by Student-Newman-Keuls test and Pearson correlation analysis were performed, and P < 0.05 was considered significant.

Expression level of EphA2
The immunological staining showed that the density (mean) of EphA2 in HepG2 and HepG2.2.15 cells were both higher than that in THLE-3 cells and the density (mean) of EphA2 in HepG2.2.15 cells was higher than that in HepG2 cells ( Fig. 1A and B). The western blotting assay also showed that the expression levels of EphA2 in HepG2 and HepG2.2.15 cells were both significantly upregulated compared with the EphA2 level in THLE-3 cells and the expressional level of EphA2 in HepG2.2.15 cells was higher than that in HepG2 cells (Fig. 1C).

Evaluation of cell migration and invasion
To investigate the regulatory role of EphA2 in the metastatic potential of HBV-related HCC cells, firstly the cell migration and invasion activity was compared between HepG2 and HepG2.

Detection of EphA2-induced EMT
To explore the involvement of EphA2-induced EMT in the metastasis of HBV-related HCC cells, the levels of E-cadherin and Vimentin were compared between HepG2 and HepG2.2.15 cells. The data revealed that in HepG2.2.15 cells the expression level of E-cadherin was lower, and the expression level of Vimentin was higher than the corresponding levels in HepG2 cells. Following EphA2 knockdown, the expression of E-cadherin was upregulated, and the expression of Vimentin was downregulated in both HepG2 and HepG2.2.15 cells, but the E-cadherin level remained lower, and the Vimentin level remained higher in HepG2.2.15 cells (Fig. 4). Moreover, following EphA2 knockdown, the cells appeared to be clustered and the cell pseudopodium was less than before (Fig. 4). Fig.1 The expression level of EphA2 was determined in THLE-3, HepG2 and HepG2.2.15 cell lines using immunocytochemical staining (A) (400 ×), immunofluorescence assay (B) (400 ×) and western blotting (C). *P < 0.01 vs. THLE-3 cells, # P < 0.01 vs. HepG2 cells 1 3 Fig.2 The migration activity of HCC cells was determined using the scratch assay (100 ×

Regulation of Wnt/β-catenin signaling
To investigate the regulatory mechanism of EphA2 in the metastatic potential of HBV-related HCC cells, the levels of Wnt/β-catenin signaling molecules were compared between HepG2 and HepG2.2.15 cells. The results showed that in HepG2.2.15 cells the expression levels of p-GSK-3β Ser9 and β-catenin were higher than the corresponding levels in HepG2 cells. Following EphA2 knockdown, the expression levels of p-GSK-3β Ser9 and β-catenin were downregulated in both HepG2 and HepG2.2.15 cells, but those in HepG2.2.15 cells remained higher (Fig. 5). Furthermore, the level of p-GSK-3β Ser9 was positively correlated with the level of β-catenin (r = 0.958, P = 0.042) and the level of β-catenin was positively correlated with the level of Vimentin (r = 0.962, P = 0.038).

Effect of Entecavir on EphA2-induced EMT
The effect of the antiviral drug Entecavir on EphA2-mediated EMT was explored in the HepG2.2.15 cells. With the increase of Entecavir dosage, the average load of HBV DNA decreased gradually and correspondingly the EphA2 expression was suppressed; subsequently, the levels of p-GSK-3β Ser9 , β-catenin and c-myc were reduced and the E-cadherin level was increased gradually (The figures were present in supplementary materials).

Discussion
EphA2, which activates downstream signaling pathways in combination with its ligand EphrinA1, plays a crucial role in cell proliferation and cell transformation [11]. Recently, EphA2 has been found to be expressed at high level in tumor tissues and its overexpression is closely associated with the development of tumor cells [12][13][14]. Increasing evidence indicated that EphA2 may be a potential regulatory factor in the progression of HCC and EphA2 overexpression can lead to tumor angiogenesis, lymph node metastasis and satellite tumor lesions in HCC tissues [15,16]. Moreover, the abnormal level of EphA2 dramatically affects the degree of malignancy and the prognosis of HCC patients [17]. Thus, EphA2 should be an available biomarker for the diagnosis, treatment and judgement of prognosis in HCC patients.
Until now, chronic HBV infection has been recognized as the primary cause of HCC in China. HBV-DNA integration can cause changes in the genetic structure and function of genes in the host, which may result in alterations within the hepatic microenvironment and the malignant transformation of liver cells in humans [18]. Furthermore, HBx protein as an oncoprotein encoded by HBV DNA can promote the invasion and metastasis of HBV-related liver cancer [19]. The findings suggest that HBV participates in the development of HCC, but little is known about its regulatory mechanism and its relationship with EphA2 expression.
In the present study, the EphA2 expression level and the metastatic potential of tumor cells were determined in HBV-related HCC cells. The data revealed that the EphA2 expression level was remarkably higher, and the cell migration and invasion ability was also stronger in HBV-related HCC cells. Following EphA2 knockdown, the migration and invasion ability of HBV-related HCC cells was inhibited, and the EphA2 level was positively correlated with the cell migration and invasion ability. The above findings suggest that the abnormal expression of EphA2 is closely associated with the metastatic potential of HBV-related HCC cells.
EMT is an important characteristic in the progression of tumor metastasis and tumor cells can acquire the ability to escape from in situ tumor tissue and achieve regional and distant metastasis via EMT [20]. During the progression of EMT, the expression of epithelial-cell markers such as E-cadherin is downregulated, whereas the expression of mesenchymal-cell markers such as Vimentin is upregulated [21]. It has been reported that the overexpression of EphA2 can impair cell-cell junctions and enhance the invasion activity of tumor cells via the downregulation of E-cadherin and the upregulation of Vimentin [22]. We previously found that the deficiency of E-cadherin caused a disruption in the integration of EphA2 with its ligand EphrinA1, which suggests may be a potential upstream regulator of EMT.
To explore the involvement of EphA2-induced EMT in the metastasis of HBV-related HCC cells, the expressions of E-cadherin and Vimentin were determined in HBV-related HCC cells. The results showed that the E-cadherin level was lower, and the Vimentin level was higher in HBV-related HCC cells compared with non-related HCC cells. Following EphA2 knockdown, the E-cadherin expression was upregulated, and the Vimentin expression was downregulated in HBV-related HCC cells. Moreover, the antiviral drug for HBV had an inhibitory effect on EphA2-induced EMT in HBV-related HCC cells. The above findings indicate that EphA2-induced EMT is involved in regulation of the metastasis of HBV-related HCC cells.
Emerging evidence indicated that multiple signaling pathways are involved in the progression of EMT in tumor cells [23,24]. The Wnt/β-catenin signal pathway is well known to be a canonical Wnt pathway in which Wnt protein complex can attenuate the function of GSK-3β through increasing the level of p-GSK-3β Ser9 , inhibit the phosphorylation-dependent degradation of β-catenin, elevate the level of β-catenin in the cytoplasm and leads to the translocation of β-catenin into the nucleus, where it activates downstream target genes [25,26]. It has been proved that the aberrant activation of Wnt/β-catenin signals can cause the dysfunction of E-cadherin, the inhibition of cell adhesion, the induction of the EMT process and the metastasis of tumor cells [27,28].
To further investigate the detailed mechanism of EphA2 regulation in the metastasis of HBV-related HCC cells, the expressions of Wnt/β-catenin signaling molecules were determined in HBV-related HCC cells. The data showed the p-GSK-3β Ser9 and β-catenin levels were higher in HBV-related HCC cells compared with non-related HCC cells. Following EphA2 knockdown, the expressions of p-GSK-3β Ser9 and β-catenin were downregulated in HBV-related HCC cells. Moreover, the antiviral drug for HBV had an inhibitory effect on the activation of Wnt/β-catenin pathway. The above findings indicate that EphA2 promotes the metastasis of HBV-related HCC cells through the activation of Wnt/β-catenin signal pathway.
Increasing evidence indicates that tumor microenvironment in HCC is strongly immunosuppressive. The high levels of immunosuppressive cytokines, such as programmed death-1 (PD-1), T-lymphocyte antigen 4 (CTLA-4) and mucin domain molecule 3 (TIM-3), induce T cell inhibition and represent one of the major mechanisms of HCC immune escape [29,30]. EMT has been found to be associated with the activation of different immune checkpoint molecules and EMT-induced immune escape may provide predictive biomarkers for checkpoint inhibitor therapeutic response in HCC [9]. In this study, the data revealed that EphA2 induced EMT in the metastasis of HBV-related HCC cells and we assumed that EphA2 may play a role in tumor immunosuppression, which needs further research and exploration.
In summary, this study reveals that EphA2-induced EMT is involved in the enhanced metastatic propensity of HBV-related HCC cells through the activation of the Wnt/β-catenin signal pathway.