Immortalized normal human liver cell line L02 and HCC cell lines (Hep3B, HepG2, Huh7, SMMC7721, MHCC97L, MHCC97H, HCCLM3) were used in our study. L02, Hep3B, HepG2, SMMC7721 and Huh7 cells were purchased from the Chinese Academy of Sciences (Shanghai, China). MHCC97L, MHCC97H and HCCLM3 cells were established in the Liver Cancer Institute, Zhongshan Hospital. The cells were all cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) at 37°C with 5% CO₂.

The GP73-RNA interference lentiviral vector was constructed by GeneChem Co., Ltd. (Shanghai, China). The double-stranded oligonucleotides targeted to GP73 mRNA targeting coding sequence (5′-AGGGAATGACAGAAACATA-3′) was annealed and inserted into the shRNA expression vector pGV115-GFP. The cDNA encoding GP73 was amplified by reverse transcription polymerase chain reaction (RT-PCR) and cloned into the pGV218-GFP vector (the cells stably expressed GP73 shRNA/GP73 proteins). Cells only expressing vectors (Mock cells) were used for the negative control. The lentivirus was generated and harvested (Shanghai GeneChem Co., Ltd.). Then, the lentivirus was transfected into targeted cells with a multiplicity of infection (MOI) of 10 to 50 (optimal MOI is 20).

Cell migration and invasion assays were performed using a 24-well Transwell (8.0-µm pore size; Millipore, USA) precoated without or with Matrigel (BD Biosciences, USA). SMMC7721, MHCC97H or HCCLM3 cells (5×10⁴) were suspended in 1.5 ml serum-free DMEM and transferred into the inside chamber of a 24-well cell culture insert with a 8.0-µm pore size. An amount of 600 µl media with 20% FBS was added into the outside well. After incubation for 24 h, the cells remaining on the upper side of the filters were cleaned with cotton-tipped swabs. Cells on the lower surface of the membrane were fixed with methanol and subjected to Giemsa staining. The cells on the underside of the filters were counted in five randomly selected fields (at ×200 magnification), and the average cell number per view was calculated. All experiments were performed in triplicate.

Twenty-two HCC tissues and their paired adjacent non-tumor tissues were collected from patients undergoing resection at the Liver Cancer Institute, Zhongshan Hospital. The tissues were used for western blot analysis. General characteristics regarding these 22 HCC patients are described in Table I. The study was approved by the Research Ethics Committee of Zhongshan Hospital, and informed consent was obtained from each patient.

Protein concentrations were determined using the bicinchoninic acid (BCA) method. Aliquots (20 µg) of proteins were loaded and resolved by 10% SDS-PAGE and transferred to PVDF membranes using a Bio-Rad SemiDry apparatus. After being blocked for non-specific binding sites, the membranes were incubated with the indicated primary antibodies: anti-GP73 (1:200 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA), E-cadherin (1:200 dilution; Abcam, Hong Kong), N-cadherin (1:100 dilution; Invitrogen, Carlsbad, CA, USA) and GAPDH (1:10,000 dilution; Kang-Cheng, Shanghai, China) overnight at 4°C, followed by HRP-conjugated secondary antibodies for 1 h at room temperature. After washing three times in Tris-buffered saline with 0.1% Tween-20 (TBST), immunoreactive protein bands were visualized using an enhanced chemiluminescence (ECL) detection system (GE Healthcare, Piscataway, NJ, USA).

For immunofluorescence staining, cells grown on glass coverslip were fixed in 4% paraformaldehyde and permeabilized using 0.5% Triton X-100. Non-specific binding sites were blocked with normal goat or rabbit serum. The cells were then incubated with the primary antibodies against GP73 (1:50 dilution), E-cadherin (1:100 dilution), N-cadherin (1:100 dilution) overnight at 4°C. After thorough washing, the cells were then incubated with Alexa-Fluor 555 anti-mouse IgG (1:1,000 dilution; Cell Signaling Technology, Danvers, MA, USA) or anti-goat IgG (1:1,000 dilution; Abcam). Finally, the cells were washed and stained with DAPI. Images were captured using a Leica fluorescence microscope.

Tissue samples of 48 HCC patients with or without metastasis were obtained from the Department of Hepatobiliary Surgery, The First Affiliated Hospital of Guangxi Medical University (Nanning, China). HCC diagnosis was based on World Health Organization criteria. Ethical approval was obtained from the Research Ethics Committee of The First Affiliated Hospital of Guangxi Medical University, and written informed consent was obtained from each patient. Tissue microarrays were constructed using formalin-fixed, paraffin-embedded tissue samples. Primary antibody against GP73 (1:50 dilution) and donkey anti-goat secondary antibody was used for immunohistochemical staining. Then, the integrated optical density of the tissue microarray derived from 48 HCC patients was evaluated by IPP software (Image-Pro Plus 5.1).

Staining for GP73 in the tissue microarray of 90 HCC patients was assessed using a previously described scoring method (20). The 90 HCC samples were obtained from the Liver Cancer Institute, Zhongshan Hospital. Ethical approval was obtained from the Research Ethics Committee of Zhongshan Hospital, and written informed consent was obtained from each patient. The staining intensity was scored on a scale of 0 to 3 as negative, weak, medium and strong, respectively. The stained area, which was calculated as the percentage of positively stained cells relative to the total cells, was scored on a scale of 0 to 4: 0 (0%), 1 (1–25%), 2 (26–50%), 3 (51–75%) and 4 (76–100%). The overall score was calculated by multiplying the intensity score and the staining area score. Samples were categorized into four grades: an overall score equal to 0 was graded as '−'; an overall score equal to 1, 2, 3 or 4 was graded as '+'; an overall score equal to 5, 6, 7 or 8 was graded as '++'; an overall score equal to 9, 10, 11 or 12 was graded as '+++'. The stained tissue sections were analyzed by two pathologists without any knowledge regarding the patient clinical information. Based on the immunohistochemical grades, the patients were divided into two groups: the high expression group, which included patients with grades '++' or '+++', and the low expression group, including patients graded as '−' or '+'. General characteristics of these HCC patients for the tissue microarray are described in Table II.

Statistical analysis was performed with SPSS 15.0 for Windows (SPSS, Chicago, IL, USA). Data are presented as the mean ± SD unless otherwise indicated. The Student's t-test (two-tailed) was used to compare two groups of parametric variants, and Spearman's Rho test or Chi-square test was used to analyze non-parametric variants. p<0.05 was considered to indicate a statistically significant result.

Expression levels of GP73 in various human cell lines (L02, Hep3B, HepG2, Huh7, SMMC7721, MHCC97L, MHCC97H and HCCLM3) were investigated. As shown in Fig. 1A, increased expression of GP73 was observed in HCC cell lines with high metastatic potential (MHCC97L, MHCC97H, HCCLM3) compared with the levels in the low or non-metastatic cell lines (L02, Hep3B, HepG2, Huh7, SMMC7721). To elucidate the effects of GP73 knockdown and overexpression on HCC cell behavior, lentiviral-mediated shRNA was used to knock down the expression of GP73 in the MHCC97H and HCCLM3 cells, while a GP73 cDNA expression vector was introduced into the SMMC7721 cells for overexpression of GP73. GP73 expression in the transfected cells was confirmed by western blotting (Fig. 2A). To examine whether GP73 promotes EMT, cellular morphology was observed. Knockdown of GP73 led to marked morphological changes from a mesenchymal-like phenotype to an epithelial-like phenotype in the MHCC97H and HCCLM3 cells (Fig. 3A and B). While in the SMMC7721 cells, overexpression of GP73 altered cells from an epithelial-like phenotype to a mesenchymal-like phenotype (Fig. 3C). Western blot and cell immunofluorescence analyses were used to detect expression of EMT markers in the GP73-shRNA-treated MHCC97H and HCCLM3 cells and the GP73-overexpressing SMMC7721 cells. In the GP73-shRNA-treated cells, the expression of E-cadherin (an epithelial marker) was increased and the expression of N-cadherin (a mesenchymal marker) was decreased. In contrast, the protein level of E-cadherin was downregulated and N-cadherin was upregulated in the GP73-overexpressing cells (Fig. 2). These findings indicated that GP73 was involved in EMT in HCC cell lines.

We also examined whether the change in GP73 expression affects the proliferation, migration and invasion of the HCC cells using the transfected (knockdown or overexpression of GP73) cell lines. MHCC97H Mock and HCCLM3 Mock cells (cells transfected with scrambled shRNA) and MHCC97H GP73 shRNA and HCCLM3 GP73 shRNA cells (cells transfected with GP73 shRNA), grew at similar rates (Fig. 4A and B). The same results (Fig. 4C) were obtained for the GP73-overexpressing SMMC7721 cells, indicating that GP73 was not required for the proliferation of HCC cells. Cell migration using a Transwell assay chamber was also assessed. Cells that migrated into the lower compartment of the migration chamber were fixed and then stained with Giemsa. Effective knockdown of GP73 in both MHCC97H and HCCLM3 cells markedly decreased cellular motility, as the number of migrated cells was significantly less than that for the MHCC97H Mock and HCCLM3 Mock cells. The invasive activity caused by knockdown of GP73 was determined using invasion chamber assays with Matrigel. Consistent with the cell migration results, knockdown of GP73 significantly decreased cell invasion capacity compared with the mock cells (Fig. 5A and B). Taken together, these results indicated that knockdown of GP73 in MHCC97H and HCCLM3 cells inhibited cell migratory and invasive abilities.

Cell migration and invasion abilities in the SMMC7721 LV-GP73 cells (cells overexpressing GP73) were assessed. The results showed that overexpression of GP73 markedly increased cell migratory and invasive abilities, as the numbers of migrated SMMC7721 LV-GP73 cells were significantly higher compared to the SMMC7721 Mock cells (cells trans-fected with an empty vector) (Fig. 5C). Consistent with the results of GP73 knockdown in the MHCC97H and HCCLM3 cells, GP73 overexpression in SMMC7721 cells enhanced cell invasion and metastasis.

GP73 was significantly increased in the HCC tissues compared with the paired non-cancerous tissues in 22 patients by western blot analysis (Fig. 1B). Furthermore, HCC tissues from 90 patients with survival information from an 80-month follow-up period were collected for production of a tissue microarray. The typical images of negative and positive staining of GP73 are shown in Fig. 6A. Kaplan-Meier survival analysis showed that the survival of patients with high GP73 expression was significantly poorer than the survival of those with low GP73 expression (Fig. 6B). The GP73 level in the HCC tissues derived from the patients with metastasis was obviously increased in comparison with that in the HCC tissues from patients without metastasis by immunohistochemical analysis, indicating the possible role of GP73 in HCC metastasis and its aberrant expression was indicative of poor outcomes in HCC (Fig. 7).