This study was approved by the Ethics Committee of the 4th Affiliated Hospital of Baotou Medical College, Baotou, China. A total of 23 primary HCC tissues, as well as their matched normal adjacent specimens were collected from the 4th Affiliated Hospital of Baotou Medical College from December 2013 to June 2014. Written informed consent was obtained from all participants. Among these 23 cases of HCC, 15 were male, and 8 were female; the youngest patient was 38 years of age, and the oldest was 67 years of age, with the average age being 52.4 years. In addition, 4 patients with HCC were at T1 stage, 6 were at T2 stage, 8 were at T3 stage, and 5 were at T4 stage of the disease. None of the patients had received any pre-operative chemotherapy, radiotherapy, or embolization. The histomorphology of all the samples was confirmed by the Department of Pathology of the 4th Affiliated Hospital of Baotou Medical College). HCC tissues were immediately snap-frozen in liquid nitrogen after surgical removal.

The human HCC cell lines, HepG2, LH86, LMH and PLHC-1, the normal liver cell line, THLE-3, as well as the 293 cells were obtained from the Cell Bank of Baotou Medical College. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS) (both from Life Technologies, Carlsbad, CA, USA) at 37°C in a humidified incubator containing 5% CO₂.

Total RNA was extracted from the cells and tissues using TRIzol reagent (Life Technologies), according to the manufacturer's instructions. RNA was then converted into cDNA using the MiRNA Reverse Transcription kit (Life Technologies). Reverse transcription was performed at 16°C for 30 min, followed by an incubation at 42°C for 30 min and enzyme inactivation at 85°C for 5 min. Quantitative (qPCR) was then performed to examine miR-33b expression using the miRNA qPCR Detection kit (GeneCopoeia, Rockville, MD, USA) on an ABI 7500 thermocycler (Life Technologies). The U6 gene was used as an internal reference. mRNA expression was determined using the SYBR-Green I Real-Time PCR kit (Biomics, Nantong, China) on an ABI 7500 thermocycler. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal reference. The sequences of the primers used for PCR were as follows: SALL4 forward, 5′-CCCGGCAGTAAGGACTGTC-3′ and reverse, 5′-TCTCTGTCTTTAGGTACACCACA-3′; and GAPDH forward, 5′- ACAACTTTGGTATCGTGGAAGG-3′ and reverse, 5′-GCCATCACGCCACAGTTTC-3′. The reaction conditions were 95°C for 5 min, followed by 40 cycles of denaturation at 95°C for 15 sec and annealing/elongation step at 60°C for 30 sec. The relative expression was analyzed using the 2^−ΔΔCt method, as previously described (21).

Cell proliferation was determined by MTT assay. The HepG2 and LH86 cells were cultured in 96-well plate, each well with 100 µl of fresh serum-free medium with 0.5 gl MTT. Following incubation at 37°C for 12, 24, 48 and 72 h, the medium was removed by aspiration and 50 µl of dimethyl sulfoxide (DMSO) was added to each well. Following incubation at 37°C for a further 10 min, the absorbance at 570 nm (A570) of each sample was measured using a plate reader (Tecan Infinite M200; Tecan Group Ltd., Männedorf, Switzerland).

Wound healing assay was used to examine the migratory capacity of the HCC cells. The HepG2 and LH86 cells were cultured to full confluence. Wounds of approximately 1 mm in width were created using a plastic scriber. The cells were washed and then cultured in DMEM containing 10% FBS for 48 h. The cells were then observed and photographed under a microscope (CX23; Olympus, Tokyo, Japan).

The invasive capacity of the HepG2 and LH86 cells was determined in 24-well Transwell chambers (BD Biosciences, Franklin Lake, NJ, USA), which has a layer of Matrigel. A cell suspension containing 5,000 cells was added to the upper chamber, and DMEM containing 10% FBS was added to the lower chamber. Following incubation for 24 h, non-invading cells, as well as the matrix gel on the interior of the inserts was removed using a cotton-tipped swab. The invasive cells on the lower surface of the membrane were stained with gentian violet (Beyotime Biotechnology, Haimen, China), and then rinsed with water and dried in air. The invasive cells were photographed and the cell number was determined under a microscope (CX23; Olympus).

Lipofectamine 2000 (Life Technologies) was used to perform transfection according to the manufacturer's instructions. Briefly, the cells were cultured to 70% confluence, and resuspended in serum-free medium. Scramble miR (miR-NC), miR-33b mimics, miR-33b inhibitor, the pc-DNA3.1-SALL4 plasmid (all generated by Amspring, Changsha, China) and Lipofectamine 2000 were diluted with serum-free medium. The diluted Lipofectamine 2000 was added to the diluted miR or plasmid, and incubated for 20 min at room temperature, and then added to the cell suspension. Following incubation at 37°C, 5% CO₂ for 6 h, the medium was replaced with normal serum-containing medium. The cells were then cultured for 24 h for use in the following assays. Untransfected cells were used as controls.

The cells were solubilized in cold RIPA lysis buffer (Life Technologies). Proteins were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred onto a polyvinylidene difluoride membrane (Life Technologies). The membrane was incubated with phosphate-buffered saline (PBS) containing 5% milk (Yili, Beijing, China) overnight at 4°C, which was then incubated with rabbit anti-SALL4 polyclonal antibody (1:50; ab31968), rabbit anti-GAPDH polyclonal antibody (1:50; ab9485), at room temperature for 3 h. After being washed with PBS 3 times, the membrane was incubated with mouse anti-rabbit secondary antibody (1:10,000; ab99697) at room temperature for 1 h. An ECL kit (Pierce Chemical, Rockford, IL, USA) was then used to perform chemiluminence detection according to the manufacturer's instructions. The relative protein expression was represented as the density ratio versus GAPDH.

Bioinformatics analysis was performed to predict the putative target genes of miR-33b using Targetscan 3.2 software (www.targetscan.org), according to the manufacturer's instructions.

The directed Mutagenesis kit (Stratagene, La Jolla, CA, USA) was used to construct the mutant type (MT) of SALL4 3′UTR lacking complimentarity with the miR-33b seed sequence, in accordance with the manufacturer's instructions. Subsequently, the wild-type (WT) SALL4 3′UTR or MT SALL4 3′UTR was cloned into the psiCHECK-2 vector (Promega, Madison, WI, USA) downstream of the Renilla luciferase gene, respectively. Subsequently, 293 cells were seeded into 48-well plates (50,000 cells/well) and cultured for 24 h. The cells were then co-transfected with the reporter vectors and the miR-33b mimic or miR-NC using Lipofectamine 2000. Luciferase activity was measured at 48 h post-transfection using the Dual-Luciferase Reporter assay system (Promega) on an Lmax multiwell luminometer (Molecular Devices, Sunnyvale, CA, USA), in accordance with the manufacturer's instructions.

Data are expressed as the means ± SD. Statistical analysis was performed using SPSS 17.0 software (SPSS, Armonk, NY, USA). The statistical correlation of data between groups was analyzed by one-way analysis of variance (ANOVA). A value of P<0.05 was considered to indicate a statistically significant difference.

To elucidate the role of miR-33b in HCC, we first conducted RT-qPCR to determine its expression levels in HCC tissues and cell lines. The expression levels of miR-33b were markedly decreased in the HCC tissues compared to their matched adjacent tissues (P<0.01; Fig. 1A and B). Moreover, our data demonstrated that the miR-33b level was markedly lower in the advanced-stage HCC tissues (stages (T3-T4 stages) compared to the early-stage HCC tissues (stages (T1-T2) (P<0.01), suggesting that the decreased expression of miR-33b was significantly associated with the malignant progression of HCC (Fig. 1C). We further demonstrated that miR-33b was significantly downregulated in the HCC cell lines, HepG2, LH86, LMH and PLHC-1, when compared to the normal liver cell line, THLE-3 (Fig. 1D; P<0.01). Therefore, miR-33b is downregulated in HCC.

As the HepG2 and LH86 cell lines showed the most significant decrease in miR-33b expression, we used them in our in vitro experiments to determine the role of miR-33b in regulating HCC cell proliferation, migration and invasion. As miR-33b was downregulated in HCC, a miR-33b mimic was used to transfect the HepG2 and LH86 cells in order to upregulate the expression of miR-33b. As shown in Fig. 2A, the expression level of miR-33b was decreased in the HepG2 and LH86 HCC cells following transfection with the miR-33b mimic, when compared to the control group (P<0.01). However, transfection with miR-NC induced no change in miR-33b levels in the HCC cells (Fig. 2A). MTT assay, wound healing assay and Transwell assay were further conducted to examine cell proliferation, migration and invasion, respectively. The overexpression of miR-33b led to a significant decrease in the proliferation, migration and invasion of the HepG2 and LH86 cells, compared to the control group (P<0.01; Fig. 2B–D). These data indicate that miR-33b plays a suppressive role in HCC growth and metastasis.

We then focused on the putative targets of miR-33b in HCC cells. Bioinformatics analysis was performed to predict the targets of miR-33b. As shown in Fig. 3A, SALL4 was found to be a putative target gene of miR-33b. To further clarify this prediction, we generated the luciferase vectors containing WT or MT of SALL4 3′UTR (Fig. 3B and C). Subsequently, luciferase reporter assay was conducted using the 293 cells. As shown in Fig. 3D, luciferase activity was significantly decreased in the 293 cells co-transfected with the WT SALL4 vector and miR-33b mimics, but was unaltered in the cells co-transfected with the MT SALL4 vector and miR-33b mimic, compared to the control group (P<0.01). These data indicate that miR-33b directly binds to the seed sequences within the SALL4 3′UTR.

As miRNAs generally inhibit the protein expression of their target genes, we then examined the effects of miR-33b overexpression or knockdown on the protein expression of SALL4 in HCC cells. The HepG2 and LH86 cells were transfected with the miR-33b mimic or miR-33b inhibitor. Transfection with the miR-33b mimic led to an upregulated miR-33b level, while transfection with miR-33b inhibitor resulted in a decrease in the miR-33b level (P<0.01; Fig. 3E). We then found that the upregulation of miR-33b significantly inhibited the SALL4 protein level, while the knockdown of miR-33b expression upregulated the protein expression of SALL4 in the HepG2 and LH86 cells (P<0.01; Fig. 3F). Therefore, we suggest that miR-33b negatively regulates the protein expression of SALL4 in HCC cells by directly binding to the 3′UTR of SALL4 mRNA.

We further wished to determine whether SALL4 is involved in the miR-33b-mediated proliferation, migration and invasion of HCC cells. The HepG2 and LH86 cells were transfected with the miR-33b mimic with or without the SALL4 plasmid. Following transfection, MTT assay, wound healing assay and Transwell assay were further conducted to examine cell proliferation, migration and invasion, respectively. As shown in Fig. 4A–C, the proliferative, migratory and invasive capacities of the HepG2 and LH86 cells were significantly enhanced following co-transfection with the miR-33b mimic and SALL4 plasmid, when compared to the cells transfected only with the miR-33b mimic (P<0.01), indicating that the overexpression of SALL4 reversed the suppressive effects of miR-33b overexpression on HCC cell proliferation, migration and invasion. To further verify these findings, we examined the protein levels of SALL4 in each group. The protein level of SALL4 was higher in the HepG2 and LH86 cells co-trasnfected with the miR-33b mimic and SALL4 plasmid, when compared to the HepG2 and LH86 cells transfected only with the miR-33b mimic (P<0.01; Fig. 4D). These data further indicate that SALL4 is involved in the miR-33b-mediated malignant phenotypes of HCC cells.

Finally, we determined the expression levels of SALL4 in HCC tissues, as well as in their matched adjacent non-tumor tissues. The results of RT-qPCR revealed that the mRNA expression of SALL4 was significantly upregulated in the HCC tissues compared to their matched adjacent non-tumor tissues (P<0.01; Fig. 5A and B). Moreover, the SALL4 mRNA level was markedly higher in the advanced-stage HCC tissues (stages T3-T4) compared to the early-stage HCC tissues (stages T1-T2) (P<0.01), suggesting that the increased expression of SALL4 was significantly associated with the malignant progression of HCC (Fig. 5C). Furthermore, the western blot analysis data also indicated that the protein expression of SALL4 was upregulated in the HCC tissues compared to their matched adjacent non-tumor tissues (P<0.01; Fig. 5D). Subsequently, we further determined the expression of SALL4 in HCC cell lines. As shown in Fig. 5E and F, the mRNA and protein expression of SALL4 was also increased in the 4 HCC cell lines compared to the normal liver cells (P<0.01). In addition, we found that the SALL4 level inversely correlated with the miR-33b level in both HCC tissues and HCC cell lines (Fig. 5G and H). Based on these data, we suggest that miR-33b may act as a tumor suppressor in HCC through the direct inhibition of SALL4 expression.