A total of 69 patients with liver cancer (40 men and 29 women; mean age 58.3±5.9 years) who underwent surgical resection at Affiliated Hospital of Nantong University between October 2010 and May 2012 were recruited for the present study. Tumor and adjacent non-cancerous tissues (>2 cm from the lesion) were collected and stored at −80°C until further use. The patients did not receive chemotherapy, radiotherapy, biological therapy or a similar treatment regimen prior to surgery. The pathological classification and staging of liver cancer were consistent with the 7th edition of the American Joint Committee on Cancer Tumor-Node-Metastasis (TNM) Staging System published in 2010 (14). This study was approved by the Ethics Committee of The Affiliated Hospital of Nantong University and written informed consent was obtained from all patients.

Human liver cancer cell lines HepG2 and Huh7 were purchased from the American Type Culture Collection, and Hep3B and HCCLM3 cell lines were purchased from the type Culture Collection of the Chinese Academy of Sciences. All cells were authenticated by short tandem repeat profiling. Cells were cultured with DMEM (Invitrogen; Thermo Fisher Scientific, Inc.) containing 10% FBS (Invitrogen; Thermo Fisher Scientific, Inc.), 100 U/ml penicillin and 100 µg/ml streptomycin (Sigma-Aldrich; Merck KGaA) and placed at 37°C in a humidified incubator containing 5% CO₂. Vectors expressing AXIN2, AXIN2 small interfering RNA (siRNA) (5′-3′) and scrambled control-sense (5′-3′) were designed and synthesized by Chang Jing Bio-Tech, Ltd. Lipofectamine^® 3000 Transfection Reagent (Invitrogen; Thermo Fisher Scientific, Inc.) was used for cell transfections according to the manufacturer's instructions. The miRNA mimics and inhibitor and AXIN2 siRNA were purchased from Sangon Biotech Co., Ltd. The pcDNA3.1A(−) vector and pcDNA3.1A(−)-AXIN2 overexpression plasmid were obtained from Shanghai GeneChem Co., Ltd. The sequences were as follows: miR-221-3p mimics, 5′-AGCUACAUUGUCUGCUGGGUUUC-3′, 5′-AACCCAGCAGACAAUGUAGCUUU-3′; miR-221-3p inhibitor, 5′-GAAACCCAGCAGACAAUGUAGCU-3; miR-15b-5p inhibitor, 5′-UGUAAACCAUGAUGUGCUGCUA-3′; miR-15b-5p mimics: 5′-UAGCAGCACAUCAUGGUUUACA-3′, 5′-UAAACCAUGAUGUGCUGCUAUU-3′; and negative control, 5′-UUCUCCGAACGUGUCACGUTT-3′ and 5′-ACGUGACACGUUCGGAGAATT-3′. The sequences of siRNA and its negative control were as follows: Axin2, 5′-GCAGAGGGACAGGAATCAT-3′, and the negative control, 5′-GCAGGGACAAGGTAGACAT-3′.

Transfection of liver cancer cells was performed using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.). Transfection efficiency was verified by western blotting 48 h post-transfection. For functional analysis of miR-221-3p and miR-15b-5p, liver cancer cells were transfected with miR-221-3p and miR-15b-5p mimics and the control (miR-con) or miR-221-3p and miR-15b-5p inhibitor and negative control (scramble).

According to manufacturer's instructions, total RNA was extracted by TRIzol^® (Invitrogen; Thermo Fisher Scientific, Inc.) from tissues or liver cancer cells, including Hep3B, HepG2, HCCLM3 and Huh7 cell lines. Reverse transcription reaction was performed using SYBR Premix Ex Taq (Takara Bio, Inc.) at 16°C for 30 min, 42°C for 30 min and 85°C for 5 min. qPCR was performed using SYBR Premix Ex Taq with 1 µl cDNA from the RT reaction on an ABI 7500 system (Applied Biosystems; Thermo Fisher Scientific, Inc.). The PCR thermocycling conditions were as follows: 94°C for 2 min followed by 30 cycles of 94°C for 30 sec, 58°C for 30 sec, and 72°C for 30 sec. U6 small RNA was used as an internal reference. Primers were synthesized by Sangon Biotech Co., Ltd. as follows: miR-221-3p, forward, 5′-CGGCTACATTGTCTGCCTG-3′ and reverse, 5′-CAGTGCGTGTCGTGGAGT-3′; miR-15b-5p forward, 5′-ATGAACTTTCTCTGTCTTGG-3′ and reverse, 5′-CAGTGCGTGTCGTGGAGT-3′; and U6 forward, 5′-CGCTTCGGCAGCACATATAC-3′ and reverse, 5′-AACGCTTCACGAATTTGCGT-3′. The reverse universal miR qPCR primers were included in the PrimeScript™ miRNA RT-PCR kit (cat. no. RR716; Takara Biotechnology Co., Ltd.). The relative expression levels of miR-221-3p and miR-15b-5p were calculated using the 2^−ΔΔCq method as previously described (15). All experiments were conducted in triplicate.

Cell viability assay was performed using the 3-(4,5-dimethylthiazole-2-yl)-2,5-biphenyl tetrazolium bromide (MTT) method with DMSO as solvent (3,000 cells per well in 96-well plates). Absorbance was detected at 570 nm in three parallel samples, and each sample was tested three times.

At 24 h post-transfection, cells were seeded into 6-well plates and cultured for two weeks in DMEM containing 12% FBS at 37°C in a humiliated atmosphere of 5% CO₂ (600 cells per well in 6-well plates). The colonies were fixed and stained with 1 mg/ml crystal violet, and colonies containing >50 cells were counted using Leica microscopy (Leica DMR 3000; Leica Microsystems GmbH).

Cell invasive ability was determined using a Transwell assay with Matrigel (BD Biosciences). To assess cell invasion, Transwell chambers were coated with 30 µl Matrigel, incubated at 37°C for 40 min and placed in 24-well plates. At 24 h post-transfection, HepG2 and HCCLM3 cells were seeded in the upper chamber at a density of 5×10⁴ cells/well in DMEM with 2% FBS. A total of 500 µl DMEM containing 10% FBS was added to the lower chamber. After 24 h at 37°C, invaded cells were fixed with 4% paraformaldehyde for 10 min at 37°C. Non-invaded cells were removed with a cotton swab. Invaded cells were stained with 1 mg/ml crystal violet at room temperature for 10 min and counted in five randomly selected fields using Leica microscope (Leica DMR 3000; Leica Microsystems GmbH).

The 3′ UTR of Axin2 was cloned into a pGL3 control vector (Promega Corporation) containing luciferase genes to generate pGL3-AXIN2-wild-type plasmids. The primers are as follows: Axin2 forward, 5′-GCTCTAGAGCCCTGGGGTCTGGCTTTG-3′ and reverse, 5′-GCTCTAGATTTTGAAAAATATAAAATT-3′. The pGL3-AXIN2-mutant plasmids were constructed using a Takara MutanBEST Kit (Takara Biotechnology Co., Ltd.). HCCLM3 and HepG2 cells were seeded into 24-well plates (1×10⁵ cells/well), cultured for 24 h at 37°C and transfected with the miRNA mimic or inhibitor and the corresponding luciferase reporter plasmids using Lipofectamine^® 3000. pRL-TK Renilla plasmids were co-transfected into the cells as an internal reference. At 48 h, the reporter activity was measured using a Dual Luciferase Reporter Assay Kit (Promega Corporation), and relative to Renilla luciferase activities were determined. The online target gene prediction software packages TargetScan 7.2 (http://www.targetscan.org) and miRanda 3.2 (http://www.mirdb.org) were used.

Total protein of HepG2 cells was extracted according to the manufacturer's instructions using RIPA lysis buffer containing protease inhibitors (Promega Corporation). Protein concentration was determined using bicinchoninic acid assay kit (Bio-Rad Laboratories, Inc.). Proteins (30 µg per lane) were separated using 12% SDS-PAGE and transferred onto a polyvinylidene fluoride membrane. Following blocking with 5% skimmed milk in TBS at room temperature for 2 h, the membrane was incubated with rabbit anti-Axin2 polyclonal antibody (1:1,000; cat. no. ab32197; Abcam), rabbit anti-β-actin polyclonal antibody (1:5,000; cat. no. ab8227; Abcam) and rabbit anti-GAPDH polyclonal antibody (1:5,000; cat. no. ab37168; Abcam) for 12 h at 4°C. The membrane was washed three times in TBS + Tween-20 (0.1% V/V) and incubated with a horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (1:5,000; Sigma-Aldrich; Merck KGaA) at room temperature for 2 h bands were detected using enhanced chemiluminescence substrate, and relative expression of proteins was normalized to GAPDH using Scion Image v. 4.0.2 software (Scion Corporation).

SPSS 17.0 (SPSS, Inc.) was used for statistical analysis. Data are expressed as the mean ± standard deviation. Multiple groups were compared using one-way ANOVA followed by the Student-Newman-Keuls test. The Kaplan-Meier method was used to evaluate the survival rate. Spearman's rank analysis was used to identify the correlation between two groups. P<0.05 was considered to indicate a statistically significant difference.

RT-qPCR was used to determine the relative expression levels of miR-221-3p and miR-15b-5p in liver cancer and adjacent non-cancerous tissues. The results demonstrated that the relative expression level of miR-221-3p was 1.85±1.19 in liver cancer tissues, which was significantly higher compared with that in adjacent non-cancerous tissues, in which an expression level of 1.02±0.39 was observed (P<0.01; Fig. 1A). miR-15b-5p levels in liver cancer tissues was also significantly higher compared within the adjacent tissues (P<0.01; Fig. 1B). The expression levels of miR-221-3p and miR-15b-5p were also determined in liver cancer cell lines Hep3B, HepG2, HCCLM3 and Huh7 using RT-qPCR. The results indicated that miR-221-3p and miR-15b-5p were expressed at a higher level in HepG2 cells and at a lower level in HCCLM3 cells compared with Huh7 cells (P<0.05; Fig. 1C). Spearman's rank analysis was used to identify the correlation between miR-221-3p and miR-15b-5p expression levels in liver cancer tissues; the results demonstrated that their expression in tumor tissues exhibited a weak correlation (r=0.267, P=0.026; Fig. 1D).

The clinicopathological characteristics of 69 patients with liver cancer are presented in Table I. The results of statistical Student-Newman-Keuls test demonstrated that miR-221-3p and miR-15b-5p levels were associated with the TNM stage and tumor capsular infiltration (P<0.05), but not with sex, age, tumor size, AFP, hepatitis B virus surface antigen, cirrhosis or tumor differentiation (P>0.05). Kaplan-Meier analysis revealed that patients with liver cancer with high expression levels of miR-221-3p or miR-15b-5p exhibited lower overall survival rates compared with those with low levels of expression of the respective miRNA (P<0.001 and P=0.035, respectively; Fig. 1E and F).

Based on the analysis of the patient clinicopathological data, it was hypothesized that miR-221-3p and miR-15b-5p may promote liver cancer cell proliferation and invasion. miR-221-3p and miR-15b-5p expression levels were the lowest in HCCLM3 cells, which were selected for overexpression experiments by transfection with miR-221-3p and miR-15b-5p mimics (P<0.05; Fig. 2A). Overexpression of miR-221-3p and/or miR-15b-5p significantly promoted the proliferation of HCCLM3 cells, as indicated by the results of MTT (P<0.05 and P<0.01; Fig. 2B) and colony formation (P<0.05; Fig. 2C) assays. Overexpression of miR-221-3p and/or miR-15b-5p also stimulated cell invasion (P<0.05; Fig. 2D).

Compared with the other liver cancer cell lines, the relative levels of miR-221-3p and miR-15b-5p were the highest in HepG2 cells, which were selected for transfection with miR-221-3p and miR-15b-5p inhibitors to knock down endogenous miR-221-3p and miR-15b-5p expression (P<0.05, Fig. 3A); knockdown of miR-221-3p and/or miR-15b-5p significantly inhibited cell proliferation, colony formation and invasion (P<0.05; Fig. 3B-D).

To establish the specific mechanism by which miR-221-3p and miR-15b-5p may promote the proliferation of liver cancer cells, the common target gene of miR-221-3p and miR-15b-5p was predicted through bioinformatic analysis using online target gene prediction software packages TargetScan 7.2 and miRanda 3.2 to determine whether the candidate target gene 3′UTR contained miR-221-3p and/or miR-15b-5p binding sites (Fig. 4A and B). The results of a dual luciferase activity assay revealed that overexpression of miR-221-3p or miR-15b-5p inhibited the luciferase activity of the pGL3-Axin2-3′-UTR reporter, but not the pGL3-mut-Axin2-3′-UTR reporter (P<0.05; Fig. 4C and D). Western blotting results demonstrated that Axin2 protein levels were significantly lower in liver cancer cells transfected with miR-221-3p or miR-15b-5p mimics (P<0.01; Fig. 4E). Western blotting experiments were also performed to assess Axin2 expression in liver cancer tissues and corresponding adjacent non-cancerous tissues, and the results revealed that the protein expression level of Axin2 in HCC tissues was significantly lower compared with that in adjacent non-cancerous tissues (P<0.05; Fig. 4F).

To establish the roles of miR-221-3p and miR-15b-5p in the regulation of Axin2, HepG2 cells were transfected with si-Axin2 to knock down the endogenous expression of Axin2. The reduction and overexpression in Axin2 levels was determined by western blotting (Fig. 5A and B). Axin2 knockdown had a significant promotive effect on the proliferation and invasion of HepG2 cells, similar to the promotion of liver cancer cells by overexpression of miR-221-3p and miR-15b-5p (Fig. 5C and D). In HepG2 cells stably overexpressing miR-221-3p and miR-15b-5p, the restoration of Axin2 level reversed the promotive effects of miR-221-3p and miR-15b-5p in liver cancer cells (Fig. 5E and F). In addition, the results revealed that when knockdown of Axin2 was followed by transfection with miR-221-3p and miR-15b-5p inhibitor, Axin2 levels and invasive and colony formation abilities in liver cancer cells were partially restored (Fig. 6).