Human liver tissues were obtained from 16 male patients (age, 40–70 years) with liver cancer during surgical resection in The Second and The Third Affiliated Hospital of Hebei Medical University (Shijiazhuang, China) between August 2012 and December 2013. Written informed consent was obtained from each patient prior to resection and experiments were approved by the Ethics Committee of Hebei Medical University (Shijiazhuang, China). Patients with metastatic liver cancer were excluded from the current study. Tumor and adjacent (1 cm from tumor) liver tissues were identified according to pathology results. Tissues were fixed in 10% neutral-buffered formalin overnight at 4°C and embedded in paraffin fro 1 h at 60–62°C.

All animal studies were approved by the Institutional Animal Care and Use Committee of Hebei Medical University (approval no. HebMU20080026; Shijiazhuang, China) and efforts were made to minimize suffering. A total of 16 male Wistar rats (age, 4–5 weeks; weight, 100–120 g) were obtained from the Central Laboratorial Animal Facility at the Faculty of Medicine of Hebei University (Shijiazhuang, China). Rats were housed in cages under controlled environmental conditions at 25°C with 55–65% humidity, a 12-h light/dark cycle and had free access to food and water. Following acclimatization for 1 week, rats were divided into two groups (8/group) as follows: Control group (NC), not receiving DEN, injected with an equal volume of saline; and DEN group (Cancer), receiving a weekly intraperitoneal dose of DEN (70 mg/kg) for 10 weeks. All animals were sacrificed at 20 weeks (17) by exsanguination through cardiac puncture under urethane anesthesia (20%; 1.2 g/kg; intraperitoneally). Livers were isolated and fixed for histopathological analysis.

Human and rat specimens were processed routinely in 10% formalin buffer overnight at 4°C and embedded in paraffin 1 h at 60–62°C. Tissue sections (4-µm) were obtained and stained with hematoxylin and eosin 2–3 min/each at room temperature. Histopathological examinations were performed under a light microscope (magnification, ×200).

Precut rats livers were placed in Total Protein Isolation Buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1% NP-40, 0.1% SDS, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM Na₃VO₄ and 1 mM NaF) containing protease inhibitors (leupeptin and aprotinin, 1 µg/ml each), homogenized using an ice-chilled glass homogenizer with Teflon pestle and lysed for 40 min on ice with vortex mixing every 10 min. Supernatants were collected (15,000 × g; 30 min; 4°C), the protein concentration was determined using the Lowry method and samples were aliquoted and stored at −70°C until further analysis.

Cells were lysed in lysis buffer (50 mM PBS pH 7.5, 200 mM NaCl, 0.5 mM EDTA pH 8.0, 2 mM β-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride and 0.5% Tween-20) for protein extraction. Following incubation on ice for 30 min, cell lysates were centrifuged (12,000 × g; 20 min; 4°C), the protein concentration was determined as indicated above and the supernatants were stored at −20°C until use.

A total of 20 or 40 µg protein was used in the analysis of HSD17B4 overexpressing or knockdown samples, respectively. Proteins were separated on 10% SDS-PAGE gels and transferred to polyvinylidene fluoride membranes. Membranes were blocked with 5% skim milk powder for 1 h at room temperature and incubated overnight at 4°C with the following primary antibodies: Mouse anti-HSD17B4 (cat. no. sc-365167; 1:1,000; Santa Cruz Biotechnology, Inc., Dallas, TX, USA), rabbit anti-cyclin D1 (cat. no. 2922S; 1:1,000; Cell Signaling Technology, Inc., Danvers, MA, USA), rabbit antiproliferating cell nuclear antigen (PCNA; cat. no. 13110; 1:1,000; Cell Signaling Technology, Inc.), mouse anti-β-actin (cat. no. sc-47778; 1:1,000; Santa Cruz Biotechnology, Inc.), rabbit anti-p-STAT3 (cat. no. sc-135649; 1:500; Santa Cruz Biotechnology, Inc.), rabbit anti-STAT3 (cat. no. sc-482; 1:500; Santa Cruz Biotechnology, Inc.), rabbit anti-p-Akt (cat. no. sc-7985; 1:500; Santa Cruz Biotechnology, Inc.), rabbit anti-Akt (cat. no. sc-8312; 1:500; Santa Cruz Biotechnology, Inc.), rabbit anti-p-MEK (cat. no. sc-101733; 1:500; Santa Cruz Biotechnology, Inc.), rabbit anti-MEK (cat. no. sc-9259; 1:50; Santa Cruz Biotechnology, Inc.), rabbit anti-p-ERK (cat. no. sc-23759; 1:500; Santa Cruz Biotechnology, Inc.), rabbit anti-ERK (cat. no. sc-292838; 1:500; Santa Cruz Biotechnology, Inc.), rabbit anti-p-c-Jun N-terminal kinase (JNK; cat. no. sc-135642; 1:500; Santa Cruz Biotechnology, Inc.), rabbit anti-JNK (cat. no. sc-571; 1:500; Santa Cruz Biotechnology, Inc.), rabbit anti-p-p38 (cat. no. sc-101759; 1:500; Santa Cruz Biotechnology, Inc.), and rabbit anti-p38 (cat. no. sc-156091; 1:500; Santa Cruz Biotechnology, Inc.). Following incubation with goat anti-mouse or anti-rabbit IgG horseradish peroxidase (HRP)-conjugated secondary antibody (cat. nos. 610-703-002 and 611-142-002, respectively; 1:5,000; Rockland Immunochemicals Inc., Limerick, Pennsylvania, USA) for 1 h at room temperature, bands were visualized using an enhanced chemiluminescence kit (Fuazon Fx; Vilber Lourmat, Marne-la-Vallée, France). Images were captured and processed by FusionCapt Advance Fx5 (Vilber Lourmat). The relative gray scale indicated the expression of various genes, with β-actin as the protein loading control.

Total RNA was isolated from rat livers using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer's instructions. cDNA was synthesized using a SuperScript Reverse Transcription kit (Invitrogen; Thermo Fisher Scientific, Inc.) with the following protocol: 5 min at 65°C, 2 min at 37°C, 10 min at 25°C, 50 min at 37°C and 15 min at 70°C. cDNA was used as a template for qPCR with a SYBR Green PCR Master Mix kit (Takara Biotechnology Co., Ltd., Dalian, China). qPCR was performed using a Rotor-Gene 3000 Detection System (GeneBio Systems, Inc., Oakville, ON, Canada) using the following protocol: Initial denaturation step at 95°C for 30 sec, followed by 40 cycles of denaturation for 5 sec at 95°C, annealing for 30 sec at 55°C and elongation for 20 sec at 72°C. 18S ribosomal RNA was used as internal control and for normalization. All reactions were performed in triplicate. Relative mRNA expression was determined using the 2^−ΔΔCq method (18). Primer details are presented in Table I.

Immunohistochemistry was performed with paraffin-embedded human or rat tissue sections (4-µm) as previously described (19). Briefly, liver sections were blocked with goat serum (1:1; OriGene Technologies, Inc., Rockville, MD, USA) at 37°C in a wet box for 30 min and incubated with a mouse anti-HSD17B4 antibody (cat. no. PA1727; 1:1,000; Santa Cruz Biotechnology, Inc.) and rabbit anti-STAT3 antibody (cat. no. sc-135649; 1:500; Santa Cruz Biotechnology, Inc.) overnight at 4°C in moist chambers. Following washing with 0.01 mol/l PBS (pH 7.2) three times, slides were incubated with a biotinylated secondary anti-mouse (cat. no. PV-6000; 1:1; OriGene Technologies, Inc.) or anti-rabbit antibodies (cat. no. PV-9000; 1:1; OriGene Technologies, Inc.) for 30 min at 37°C. Samples were then incubated with HRP-avidin D (cat. no. A-2004; Vector Laboratories, Inc.; Maravai LifeSciences, San Diego, CA, USA) at 37°C for 30 min and developed using a diaminobenzidine kit (Vector Laboratories, Inc.; Maravai LifeSciences). Sections were counterstained with hematoxylin for 30 sec at room temperature. Staining intensities were determined by measuring the integrated optical density (IOD) with light microscopy (magnification, ×200) with associated software (SPOT Basic™ image capture software version 3.2.4; cat. no. SPOT53BE; SPOT Imaging; Diagnostic Instruments, Inc., Sterling Heights, MI, USA).

The human liver cancer cell line HepG2 was obtained from the American Type Culture Collection (cat. no. HB-8065; Manassas, VA, USA). HepG2 cells were maintained in RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc.) containing 10% fetal bovine serum (FBS; Abgent Biotech Co., Ltd., Suzhou, China), penicillin and streptomycin (100 U/ml, each) in a humidified environment at 37°C with 5% CO₂. For inhibitor studies, all inhibitors were dissolved in dimethyl sulfoxide (DMSO) and control experiments were performed with equal volumes of DMSO. Cells were treated for 2 h with MEK inhibitor PD98059 (5 µmol/; Calbiochem; Merck KGaA, Darmstadt, Germany) or phosphoinositide 3-kinase (PI3K) inhibitor LY294002 (5 µmol/l; Calbiochem; Merck KGaA) prior to transformation with the HSD17B4 overexpression plasmid.

HepG2 cell proliferation assays were performed using a bromodeoxyuridine (BrdU) Cell Proliferation Assay kit (Millipore; Merck KGaA) according to the manufacturer's instructions. Cells were labeled with BrdU for 6 h at 24 h following HSD17B4 plasmid transfection. Optical density readings were performed at 450 nm to measure the incorporation of BrdU. All groups were evaluated with ≥3 separate wells/experiment.

HSD17B4 cDNA was cloned into the pLL3.7 vector (Addgene, Inc., Cambridge, MA, USA). HepG2 cells were plated at 5×10⁶/well in 6-well culture dishes for 18 h to reach ~70% confluence in Dulbecco's modified Eagle's medium (DMEM; Gibco; Thermo Fisher Scientific, Inc.) containing 10% FBS without antibiotics. Cells were washed twice with serum-free RPMI-1640 and 1 ml serum-free Opti-MEM I (Gibco; Thermo Fisher Scientific, Inc.) was added to each well. A DNA-Lipofectamine 2000 complex was prepared according to the manufacturer's instructions (Invitrogen; Thermo Fisher Scientific, Inc.). A total of 6 µg transfected plasmid and 30 µl Lipofectamine reagent were added to each reaction. The empty vector control is indicated using ‘−’ and the HSD17B4 plasmid is highlighted using ‘+’ in the following. Cells were incubated at 37°C for 6 h. Following transfection, cells were maintained in DMEM containing 1% FBS for 24 h and the transfection efficiency was determined using western blotting. Cells were collected for further analysis.

siRNA targeting HSD17B4 (siHSD17B4; forward, 5′-GUACCUUUGUAUUUGAGGAdTdT-3′ and reverse, 5′-UCCUCAAAUACAAAGGUACdTdT-3′) and non-specific siRNA (siNC; forwards, 5′-UUCUCCGAACGUGUCACGUTT-3′ and reverse, 5′-ACGUGACACGUUCGGAGAATT-3′), were purchased from Sigma-Aldrich (Merck KGaA). Transfections with 10 µmol/l siRNA (siNC as ‘−’ and siHSD17B4 as ‘+’) were performed using the Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) following the manufacturer's instructions, when cells reached 50–60% confluence. Following 24 h of transfection, the transfection efficiency was determined using western blotting. Cells were harvested and lysed as described for RT-qPCR and western blotting.

Data are presented as the mean ± standard deviation of ≥3 independent experiments. All statistical analyses were performed using SPSS 13.0 (SPSS, Inc., Chicago, IL, USA). Statistical differences between groups were assessed using a one-way analysis of variance, followed by Dunnett's or Bonferroni's multiple comparison tests. Correlation analysis was performed using the Pearson's correlation. P<0.05 was considered to indicate a statistically significant difference.

To evaluate HSD17B4 expression during liver cancer development, rats with DEN-induced liver cancer were investigated. All rats in the Cancer group developed liver cancer while the livers of rats in the NC group exhibited a normal lobular architecture with central vein and radiating hepatic cords (Fig. 1A and B). Hepatocytes were polyhedral in shape and the cytoplasm was granulated with small uniform nuclei. The Cancer group exhibited extensive cell swelling and single cell necrosis; necrotic cells were small with basophilic nuclei and dark cytoplasm (Fig. 1B). These results suggested that the rat liver cancer model using DEN was successfully established.

HSD17B4 expression in the liver tissues of the rats was determined. HSD17B4 protein and mRNA levels were significantly increased in the Cancer compared with the NC group (P<0.05; Fig. 1C and D). DEN administration resulted in HSD17B4 accumulation in the cytoplasm of liver tissues from rats with liver cancer (Fig. 1E). Compared with the NC, the IOD was significantly increased in the Cancer group (P<0.01; Fig. 1E). The results indicated that HSD17B4 expression was increased during liver cancer development.

To evaluate an association between HSD17B4 expression and proliferation-associated genes, mRNA levels of cyclin D1 and PCNA were evaluated. It was observed that expression was significantly increased in rats with liver cancer compared with the NC group (P<0.05; Fig. 2A and B). The correlation analysis between HSD17B4 and cyclin D1 or PCNA mRNA expression revealed a positive correlation (P<0.01; Fig. 2C and D).

Protein expression of HSD17B4, cyclin D1 and PCNA was determined in HepG2 HSD17B4 overexpression and knockdown cells. Compared with the empty vector, cyclin D1 and PCNA levels were increased in the HSD17B4 overexpressing cells (Fig. 2E). In the siHSD17B4-treated cells cyclin D1 and PCNA expression was downregulated compared with the siNC (Fig. 2F). The results indicated that HSD17B4 expression affected proliferation-associated genes in HepG2 cells.

To further investigate the role of HSD17B4 in proliferation, BrdU incorporation assays were performed with HSD17B4 overexpression and knockdown HepG2 cells. Proliferation of the HSD17B4 overexpressing cells was significantly increased compared with the empty vector control (P<0.05; Fig. 2G) and BrdU incorporation was significantly decreased in the siHSD17B4-treated cells compared with the siNC (P<0.05; Fig. 2H). The results indicated that HSD17B4 promoted HepG2 proliferation by affecting proliferation-associated genes.

STAT3 phosphorylation contributes to liver cancer progression (5–7). STAT3 phosphorylation was significantly increased in the Cancer compared with the NC group (P<0.05; Fig. 3A). As STAT3 is involved in various signaling pathways (20,21), proteins associated with the Akt and the MEK/ERK signaling pathways were investigated. Western blot analysis revealed that levels of p-Akt, p-ERK and p-MEK were significantly increased in tissues from the Cancer compared with the NC group (P<0.05; Fig. 3B).

A recent study revealed that HSD17B4 is upregulated in patients with liver cancer and HSD17B4 over-expression promotes HepG2 proliferation (16). To determine whether HSD17B4 serves a role in liver cancer progression, HSD17B4 expression in adjacent and tumor tissues from patients with liver cancer were evaluated. IHC images revealed that HSD17B4 expression was increased in tumor tissues compared with adjacent normal tissues obtained from patients with liver cancer and the determined IOD was significantly increased in the cancerous tissues (P<0.01; Fig. 4A).

An association between HSD17B4 expression and activation of STAT3 was evaluated in HSD17B4 overexpression and knockdown HepG2 cells. As presented in Fig. 4B, HSD17B4 overexpression significantly increased STAT3 phosphorylation in HepG2 cells compared with the empty vector group (P<0.05). In contrast, p-STAT3 levels were significantly decreased in the siHSD17B4-treated cells compared to the siNC group (P<0.05; Fig. 4B).

To verify these finding, HSD17B4 and p-STAT3 expression in tumor and adjacent normal tissues of patients liver cancer were evaluated using IHC. IODs determined for HSD17B4 and p-STAT3 suggested a positive correlation between the HSD17B4 expression and STAT3 activation (P<0.01; Fig. 4C). The results suggested that HSD17B4 expression induced STAT3 activation.

To determine the mechanism by which HSD17B4 promotes STAT3 activation in liver cancer, activation of various signaling pathways involved in liver cancer progression were assessed. HSD17B4 overexpression increased the phosphorylation of Akt, MEK and ERK in HepG2 cells compared with the empty vector group (P<0.05; Fig. 5A). In siHSD17B4-treated cells, Akt, MEK and ERK phosphorylation was significantly decreased compared with the siNC group (P<0.05; Fig. 5A). Activation of JNK and p38 were not affected by the induction or knockout of HSD17B4 expression (P>0.05; Fig. 5A). The results suggested that activation of the Akt and the MEK/ERK signaling pathways was associated with HSD17B expression.

To investigate which signaling pathways are involved in STAT3 activation, HepG2 cells were incubated with PD98059 and LY294002, inhibitors of the MEK/ERK and the PI3K/Akt signaling pathway, respectively, prior to transfection with the HSD17B4 overexpression plasmid (22). STAT3 phosphorylation was determined and it was observed that levels significantly decreased in HSD17B4 overexpressing cells in the presence of MEK/ERK and PI3K/Akt signaling pathway inhibitors compared with the DMSO treated cells and a significant decrease was further observed for empty vector-transfected cells in the presence of inhibitors compared with the DMSO-treated cells (P<0.05; Fig. 5B). The results suggested HSD17B4 expression promoted STAT3 activation via the PI3K/Akt and the MEK/ERK signaling pathways.