Sixty fresh tumor samples for qRT-PCR and immunohistochemistry were randomly collected from HCC patients who underwent curative resection between 2008 and 2010 at the First Affiliated Hospital of Medical College, Xi’an Jiaotong University (Xi’an, China). For western blot analysis, 10 invasive tumor tissues were selected from the 60 cases. Resected tumors and corresponding non-tumor tissues specimens (>2 cm away from the tumor) were immediately cut from the resected liver and subdivided into 2 portions. One portion was rapidly frozen in liquid nitrogen and stored at −80°C until use and the other was fixed in buffered paraformaldehyde for immunohistochemical staining. None of the patients had received radiotherapy or chemotherapy prior to sampling. Clinical data were obtained from medical records. Histopathological Edmonson classification, clinical tumor-node-metastasis (TNM) grading, maximum tumor diameter and normal tumor-adjacent tissues were all confirmed by an experienced pathologist who was blinded to the clinical information. Written informed consent was obtained from all patients. The Xi’an Jiaotong University Ethics Committee approved all protocols according to the Helsinki Declaration of 1975.

Immunohistochemical staining was performed on paraformaldehyde-fixed paraffin sections (4 μm), as described previously (19). In brief, sections were dewaxed and rehydrated, and antigens were retrieved in citrate buffer. Endogenous peroxidase activity was blocked for 10 min using 3.0% hydrogen peroxide. Goat serum (10%) was applied to the sections to prevent non-specific binding. The sections were then separately incubated with the following primary antibodies at 4°C overnight: rabbit anti-FAK (1:200; Cell Signaling Technology, Inc., Danvers, MA, USA), rabbit anti-KLF8 (1:200; Aviva Systems Biology, San Diego, CA, USA), mouse anti-E-cadherin (1:100) and rabbit anti-MMP-9 (1:200; both Santa Cruz Biotechnology, Santa Cruz, CA, USA). The primary antibody was detected using biotinylated secondary antibodies (Beijing Zhongshan Golden Bridge Biotechnology Co., Ltd., Beijing, China) according to the manufacturer’s instructions. Sections were visualized with diaminobenzidine, counterstained with hematoxylin, dehydrated in alcohol and xylene and mounted onto glass slides. For negative controls, primary antibody was replaced with PBS.

All sections were assessed independently by two experienced pathologists. The staining results for the four proteins (FAK, KLF8, E-cadherin and MMP-9) were semi-quantitatively expressed by an immunohistochemical score combined with the percentage of liver cells revealing specific immunoreactivity. Staining intensity was expressed by four grades as follows: 0, none; 1, weak; 2, moderate; and 3, strong. The percentage of positive liver cells was expressed by the following grades: 0, <10%; 1, 10–25%; 2, 26–50%; 3, 51–75%; and 4, >75%. The staining intensity and average percentage of positive liver cells were assayed for 5 independent high magnification fields. The total score was calculated by multiplying the staining intensity and the percentage score of positive liver cells. Specimen staining was assessed as total score: negative (−), score 0; weakly positive (+), score 1–4; positive (++), score 5–8; strongly positive (+++), score 9–12.

Proteins were extracted from the portal vein cancer emboli and corresponding HCC tissues of 5 patients and lysed in RIPA buffer containing protease inhibitor cocktail and PMSF on ice. Protein concentration was quantified using a BCA Protein Assay kit (Pierce Biotechnology, Inc., Rockford, IL, USA) according to the manufacturer’s instructions. Proteins suspended in loading buffer were heated at 95°C for 5 min for denaturation. Equal amounts of proteins (20 μg/lane) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinylidene fluoride membranes (Millipore, Billerica, MA, USA). Following blocking for 1 h with 5% non-fat milk and TBST washes, the membranes were incubated overnight at 4°C with primary antibodies against FAK (1:1,000), KLF8 (1:500), E-cadherin (1:1,000), MMP-9 (1:1,000) and β-actin (1:2,000; Santa Cruz Biotechnology). Subsequently, the membranes were incubated with goat anti-mouse or anti-rabbit horseradish peroxidase-conjugated secondary antibodies (1:10,000; Santa Cruz Biotechnology) for 1 h at room temperature. Blots were then visualized using the Super Signal West Pico chemiluminescent substrate kit (Millipore) and exposed to X-ray films.

Total RNA was isolated from the frozen specimens stored in liquid nitrogen using the RNAfast 200 purification kit (Fastagen Biotech, Shanghai, China). A_260/280 optical density values of the RNA samples were between 1.8 and 2.1. RNA integrity was confirmed by the presence of intact 18S and 28S bands on 2% agarose gels. Next, total RNA was reverse transcribed at 37°C for 15 min and incubated at 85°C for 5 sec using PrimeScript RT reagent (Takara Biotechnology, Shiga, Japan). cDNA was used for qRT-PCR analysis with SYBR-Green PCR master mix (Takara Biotechnology) and the BD IQ5 cycler (BD Biosciences, Franklin Lakes, NJ, USA) according to the manufacturer’s instructions. The primer sequences used for the detection of FAK, KLF8, MMP-9, E-cadherin and β-actin are summarized in Table I. PCR was performed under the following conditions: 1 cycle at 95°C for 5 min, then 40 cycles at 95°C for 30 sec and 60°C for 30 sec. β-actin was used as an internal control. Relative mRNA levels were calculated based on the threshold cycle (CT) values, which were corrected based on β-actin expression, according to the equation: 2^−ΔCT (ΔCT = CT_{target gene} − Ct_β-actin).

Statistical analyses were performed with the SPSS 16.0 software (SPSS, Inc., Chicago, IL, USA). For continuous variables, data are expressed as the mean ± SD. KLF8 mRNA expression levels in the tumor and adjacent non-tumorous tissues were compared by the paired t-test. Clinicopathological parameters and KLF8 mRNA expression correlations were analyzed using the Mann-Whitney U test. Clinicopathological parameters and KLF8 protein expression correlations were analyzed using the Spearman’s rank correlation coefficient test. The correlation between KLF8 mRNA expression and FAK, MMP-9 and E-cadherin mRNA expression was analyzed using the Pearson correlation coefficient test and the correlation between KLF8 protein expression and FAK, MMP-9 and E-cadherin protein expression was analyzed using Spearman’s rank correlation coefficient test. P<0.05 was considered to indicate a statistically significant difference.

KLF8 mRNA expression in 60 pairs of HCC tumor and adjacent non-tumorous tissues was detected by qRT-PCR. The expression of KLF8 mRNA in HCC tissues was markedly upregulated compared with adjacent non-tumorous tissues (P<0.05; Fig. 1A). Immunohistochemical staining was performed in the same tissues to detect KLF8 protein expression. Specifically, 51.6% (n=31) of the HCC tumor tissues and 41.6% (n=25) of the adjacent non-tumorous tissues revealed KLF8-positive staining. The majority of positive cells in HCC tumor tissues revealed diffuse cytoplasmic staining and intense nuclear staining of KLF8, while adjacent non-tumorous tissues mainly exhibited diffuse cytoplasmic staining (Fig. 2). These observations indicate that KLF8 nuclear translocation is required during HCC progression. We also investigated whether the higher expression of KLF8 in HCC tissues was consistent at the mRNA and protein levels. As demonstrated in Fig. 1B, by comparing the mRNA levels of KLF8 in tumor tissues with positive staining (n=31) to those with negative (n=29) staining, as determined by immunohistochemistry, samples with positive protein staining were found to have significantly higher mRNA expression than the negative group (P<0.05), indicative of a consistency between KLF8 mRNA and protein levels during HCC development.

To assess the clinical significance of KLF8 expression in HCC, the correlation between KLF8 expression and the clinicopathological characteristics of the 60 HCC patients was statistically analyzed, and the results are presented in Table II. HCC tissues with vascular invasion exhibited higher KLF8 mRNA and protein expression levels than those without vascular invasion (P<0.05). HCC tissues with advanced TNM staging exhibited higher KLF8 protein expression levels than those with a lower TNM staging (r=0.261, P<0.05). However, no significant correlation was observed between KLF8 expression and age, gender, HBV infection, fetoprotein (AFP), tumor number and tumor size. In addition, western blot analysis revealed that KLF8 protein levels were higher in portal vein cancer emboli compared with their corresponding primary HCC tissues (Fig. 3).

KLF8 and FAK mRNA expression levels were determined by qRT-PCR in 60 HCC tissues. KLF8 mRNA expression was found to positively correlate with FAK mRNA expression levels in the HCC tissues (r=0.362, P<0.001), as assessed by Pearson correlation coefficient test (Fig. 4A). KLF8 and FAK protein expression was determined by immunohistochemical staining. FAK protein was positively stained in the cytoplasm of carcinoma cells and 53.3% (n=32) of the HCC tumor samples revealed positive staining of FAK protein. KLF8 protein expression levels positively correlated with FAK protein expression levels in the HCC tissues (r=0.377, P<0.01), as assessed by Spearman’s rank correlation coefficient test (Table III, Fig. 5).

qRT-PCR demonstrated that KLF8 mRNA expression levels positively correlated with MMP-9 mRNA expression levels in the HCC tissues (r=0.392, P<0.01), as assessed by Pearson correlation coefficient test (Fig. 4B). Immunohistochemical staining revealed that MMP-9 protein was diffuse with moderate or strong staining in the cytoplasm of tumor cells. Positive staining of MMP-9 protein was found in 73.3% (n=44) of the HCC tissues. Consistent with mRNA expression levels, KLF8 protein expression positively correlated with MMP-9 protein expression in the HCC tissues (r=0.336, P<0.01; Table III, Fig. 5).

KLF8 mRNA expression levels were found to negatively correlate with E-cadherin mRNA expression levels in the HCC tissues (r=−0.364, P<0.01), as assessed by Pearson correlation coefficient test (Fig. 2C). Immunohistochemical analysis revealed that positive E-cadherin protein staining was largely presented in the membranes of carcinoma cells. The positive staining of E-cadherin protein was found in 45% (n=27) of the HCC tissues. KLF8 protein expression levels negatively correlated with E-cadherin protein expression levels in the HCC tissues (r=−0.410, P<0.01; Table III, Fig. 5).