Tumor tissues and paired non-cancerous hepatic parenchyma were collected from 100 patients with primary HCC who received surgical resection from May 2010 to June 2011 at the Department of Hepatobiliary Surgery, Shandong Provincial Hospital Affiliated to Shandong University. Serum specimens were collected from the patients and 30 control patients without HCC. None of the patients had received preoperative chemotherapy or other treatment before surgery. Patient written informed consent was obtained, and the study protocol was approved by the Health Service Ethics Committee of Shandong Provincial Hospital Affiliated to Shandong University. HCC was histologically diagnosed by two pathologists independently and the clinical characteristics of each patient were recorded as shown in Table I.

The present study was performed in accordance with the Declaration of Helsinki and approved by the local Ethics Committee. All patients provided their informed consent.

Fresh HCC tissues were treated with TRIzol reagent for total RNA extraction (Invitrogen Carlsbad, CA, USA) and purified by phenol/CHCl₃ according to the manufacturer's instructions. Total RNA (5 µg) was reversely transcribed to cDNA using the MBI Fermantas reverse transcription kit (MBI Fermentas, Vilnius, Lithuania). The Quantitative SYBR-Green PCR kit and ABI Prism 7000 Sequence Detection System (both from ABI, USA) were applied to test the expression level of IL-32α under the following conditions: 30 cycles: 1 cycle at 95°C for 5 min, then 30 cycles at 94°C for 30 sec and 60°C for 45 sec; quantitative RT-PCR was repeated at least 3 times. β-actin expression was used for normalization. Primer sequences are listed as follows: IL-32α F, 5′-ACAGTGGCGGCTTATTATGAGGA-3′ and R, 5′-GTTGCCTCGGCACCGTAATC-3′; β-actin F, 5′-AATGCTTCTAGGCGGACTATGA-3′ and R, 5′-CAAGAAAGGGTGTAACGCAACT-3′.

Five fresh HCC and paired non-cancerous tissues were lysed by cold RIPA buffer containing protease inhibitor on ice for 30 min, and centrifuged at 12,000 × g at 4°C for 20 min. The protein concentration was determined using a BCA protein assay kit (Biocolor Biotech, Shanghai, China). Proteins suspended in loading buffer were denatured and separated using 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and electrophoretically transferred onto a polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA, USA) in transfer buffer at 40 V for 105 min. The membrane was blocked using 5% skimmed milk in Tris-buffered saline with Tween-20 (TBST) for 2 h, washed with TBST and incubated with mouse anti-IL-32 antibody (R&D Systems, Minneapolis, MN, USA) overnight at 4°C. The membrane was incubated with an anti-mouse horseradish peroxidase-conjugated secondary antibody (Dako, Glostrup, Denmark) at a dilution of 1:200 at room temperature for 1 h. Protein bands were visualized by SuperSignal West Pico Chemiluminescent Substrate kit (Pierce, Rockford, IL, USA) and exposed using Kodak X-ray film (Kodak, Rochester, NY, USA). Proteins were re-blotted with anti-GAPDH (Zymed, South San Francisco, CA, USA) as an internal control.

For immunohistochemical analysis, 4-µm tissue sections were cut from paraffin blocks and baked at 60°C for 2 h before staining with mouse anti-IL-32α antibody (dilution 1:100; R&D Systems). Endogenous peroxidase activity was blocked with 3% H₂O₂ for 30 min. Then tissue sections were pre-treated in citrate buffer using a water bath for 15 min for antigen retrieval. Goat serum (1%) was applied to prevent a non-specific reaction. The primary antibody was incubated overnight at 4°C. An anti-mouse antibody kit (Jing Mei Biotech, Shanghai, China) was applied and DAB reaction was performed following the protocol. Control IgG antibody was used as a negative control. Histomorphometric analysis was performed by Image-Pro Plus image analysis system (Media Cybernetics, Inc., Rockville, MD, USA).

A sandwich ELISA was designed for the quantification of IL-32α in human serum. A 96-well microtiter plate was coated overnight at 4°C with goat antibody (PAb; R&D Systems) to IL-32α (1 µg/ml in PBS, 100 µl/well) and rinsed with PBST. The wells were then coated with 1% BSA solution in PBS. IL-32α standard samples were prepared using a serial dilution of a recombinant human IL-32α solution. Samples were grouped into control and HCC. IL-32α ELISA was carried out according to the manufacturer's instructions as follows: assay diluent (80 µl) was added in duplicate to all wells. Each prepared standard dilution (20 µl) was added to samples and incubated at room temperature. Biotin-conjugate (100 µl) was added to all wells and incubated at room temperature. Diluted streptavidin-HRP (100 µl) was added to all wells and incubated at room temperature. The enzyme reaction was stopped by quickly pipetting 100 µl of stop solution into each well. Absorbance of the reaction product was measured at 490 nm on an ELISA reader (Molecular Devices, Sunnyvale, CA, USA).

HCC cell lines Hu7 and HepG2 were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 100 U/ml penicillin, 100 µg/ml streptomycin, 25 ng/ml amphotericin B, and 10% fetal bovine serum (FBS) (Gibco, Grand Island, NY, USA) at 37°C in a humidified incubator with 5% CO₂. For the RNA interference assay, an siRNA for IL-32α was designed to silence IL-32α expression in HCC cell lines, Hu7 and HepG2 (Santa Cruz Biotechnology, Santa Cruz, CA, USA). The cells were transfected with 40 nM of siRNA using Lipofectamine™ LTX (Invitrogen). Silencing efficiency was verified by western blot analysis. Exogenous IL-32α at a similar concentration (500 pg/ml) was added to the culture medium for the rescue assay.

Transfected and control cells were subjected to cell scratch and Transwell invasion assays. For the scratch assay cells were seeded into 6-well plates and cultured until reaching confluence. A wound was created with a sterile pipette tip. The distance was measured by a Nikon DS-5M Camera System mounted on a phase-contrast Leitz microscope. Images of the wound were captured under a phase-contrast microscope at 0, 24 and 48 h. For each experiment, 5 visual fields and 2 repeated wells were measured with 3 replications.

For the Transwell assay a 24-well Transwell chamber (8-mm; Millipore) coated with 30 µl Matrigel was used for the invasion assay. A 100-µl cell suspension was loaded into the upper Matrigel-coated chamber. DMEM (600 ml) with 10% FBS was added to the bottom chamber. Cells were then allowed to migrate or invade for 48 h at 37°C. The cells in the bottom chamber were fixed in paraformaldehyde and permeabilized in methanol, and then stained with crystal violet dye. Cell images were obtained under a light microscope (Leica DM4000 B; Leica Microsystems, Wetzlar, Germany.

The Mann-Whitney U test or Kruskal-Wallis was used for between-group comparisons, where appropriate, and the correlation between the results obtained with the two different analyses was analyzed with the Spearman's test. A paired Student's t-test was used to compare the differences of IL-32-α mRNA and protein expression in tumor tissues and non-cancerous tissues. The correlations of mRNA expression levels were analyzed with Pearson test. P<0.05 was considered statistically significant. All data were analyzed with SPSS 16.0 (SPSS, Inc., Chicago, IL, USA).

In order to investigate the expression pattern of IL-32α in HCC tissue and its prognostic role in HCC patients, we examined the IL-32α expression levels in 100 HCC samples. IL-32α expression was found widely elevated in the HCC tissues, as compared with that in the paired non-cancerous tissues (Fig. 1A). Moreover, we analyzed the correlation between IL-32α serum levels of HCC patients and clinicopathological parameters, including tumor size, virus infection, liver cirrhosis, vascular invasion and metastasis. Statistical results revealed that IL-32α was much higher in the serum samples of patients with distant metastasis than those without distant metastasis and in patients with vascular invasion (Table I; P=0.01). Similarly, IHC staining revealed that high IL-32α expression was often observed in vessel invasion foci (Fig. 1B). Importantly, HCC patients showed a higher IL-32α serum concentration than the controls (571.45±102.28 vs. 144.60±51.172 pg/ml, P=0.007, Fig. 1C). Taken together, these findings suggest that IL-32α overexpression can serve as a predictive indicator for distant metastasis and vascular invasion of HCC patients.

In order to verify the functions of IL-32α in HCC in vitro, an siRNA of IL-32α was designed to silence the IL-32α expression in HCC cell lines, Hu7 and HepG2. The results indicated that IL-32α was significantly downregulated in the si-IL-32α-treated Hu7 and HepG2 cells, as compared with the mock group (Fig. 2A and B). IHC staining assays also confirmed the knockdown of IL-32α in the HepG2 Hu7 cells (Fig. 2C). As a secreted factor, IL-32α was also detected in the culture medium, and its concentration in the supernatant was decreased after siRNA knockdown (Fig. 2D). Cell scratch and Transwell invasion assays were carried out in order to examine the cell migration and invasion abilities, respectively. Cell scratch assays revealed that Hu7-si-IL-32α cells showed sharply reduced migration ability as compared with that of mock cells (Fig. 3A and B; P<0.05). For Transwell invasion assays, Hu7-si-IL-32α cells showed decreased invasive potential than that of Hu7-mock cells (Fig. 3C and D; P<0.05). Similar results were also observed in HepG2 cells both in cell scratching and Transwell invasion assays (Fig. 3E and F). To further confirm the role of IL-32α in regulating cell migration and invasion, exogenous IL-32α at a similar concentration (500 pg/ml) was added to the culture medium. Cell scratch assays showed that Hu7-si-IL-32α cells restored the migration ability with IL-32α treatment (Fig. 3A and B; P<0.05). For Transwell invasion assays, Hu7-si-IL-32α cells also displayed elevated invasive potential after IL-32α treatment (Fig. 3C and D; P<0.05). To sum up, these results indicated that IL-32α could positively regulate the migration and invasion ability of HCC cells.

It has been reported that IL-32α regulates VEGF levels in breast cancer, which is linked to angiogenesis and tumor invasion (14,16). In order to test whether IL-32α also modulates VEGF in HCC cells, we examined the level of VEGF after IL-32α was transiently silenced in the Hu7 and HepG2 cell lines. The present study found that VEGF was significantly reduced after IL-32α knockdown at both the mRNA level and protein level (Fig. 4A and B). Furthermore, a decreased VEGF level in the culture medium was observed after IL-32α knockdown (Fig. 4C; P<0.05). Collectively, these data showed that VEGF was a downstream response factor of IL-32α in HCC cells.

Previous reports have revealed that VEGF-STAT3 signaling is important for vascular invasion in a series of tumors (17,18). In the present study, we provided further proofs for the correlation between IL-32α and VEGF-STAT3 signaling. STAT3 was significantly reduced in the Hu7-si-IL-32α cells than that noted in the mock cells (Fig. 4D). However, exogenous IL-32α treatment increased the level of STAT3 in the Hu7-si-IL-32α cells (Fig. 4D). Taken together, our data revealed that the IL-32α/VEGF/STAT3 signaling pathway plays an essential role in the vascular invasion in HCC.

To further confirm the correlation between IL-32α and VEGF in HCC, VEGF staining was performed on the HCC tissues and corresponding non-cancerous liver tissues. Our result verified that VEGF expression levels were in accordance with IL-32α in the HCC tissues, whereas their expression was low in paired non-cancerous tissues (Fig. 5A). Moreover, western blotting of IL-32α and VEGF were also conducted in 6 cases of HCC and corresponding non-cancerous liver tissues. A significant correlation was also detected between VEGF and IL-32α at the relative protein level (VEGF greyscale/GAPDH greyscale and IL-32α greyscale/GAPDH greyscale) (Fig. 5B). Protein level analysis of the western blotting revealed that IL-32α and VEGF levels were positively correlated (P<0.05; Fig. 5C). Furthermore, we investigated IL-32α and VEGF levels in the HCC patient serum samples. IL-32α and VEGF protein were positively correlated (Fig. 5D; P<0.05). Taken together, these data provide further proof that VEGF may serve as a downstream factor regulated by IL-32α in HCC.