A total of 116 specimens, including 58 tumor-adjacent tissues and 58 tumour tissues (female, 11 and male, 47; mean age, 49.16±13.42 years), were obtained from patients with HCC, who underwent surgical resection from August 2010 to October 2016 at The First College of Clinical Medical Science, China Three Gorges University (Yichang, China). Patients who received chemotherapy or radiotherapy prior to surgery were excluded. The final diagnosis was confirmed by pathological analysis. All the specimens were collected immediately following liver resection and stored in liquid nitrogen at −80°C until analysis. Written consent was obtained from every patient and the research protocol was approved by the Ethics Committee of The First College of Clinical Medical Science, China Three Gorges University.

HCC cell lines, Hep3B and Huh7, and normal hepatocyte LO2 were all purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). In addition, all the cell lines were cultured with their specified basic culture medium [Dulbecco's modified Eagle's medium (DMEM); Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA] supplemented with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.), 100 U/ml penicillin and 100 µg/ml streptomycin sulfate and maintained at 37°C in a humidified atmosphere containing 5% CO₂.

FEZF1-AS1 small interfering (si)RNA (5′-GAAAGUGUUGUGUCAAUAACG-3′) and non-targeting siRNA [si negative control (siNC, 5′-AATTCTCCGAACGTGTCACGT-3′)], miR-4443 mimics (5′-UUGGAGGCGUGGGUUUU-3′), inhibitors (5′-AAAACCCACGCCUCCAA-3′) and controls (5′-ACAUCUGCGUAAGAUUCGAGUCUA-3′) were purchased from Shanghai Integrated Biotech Solutions Co., Ltd. (Shanghai, China). Transfection was performed with Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA), according to the manufacturer's protocol, with plasmids or siRNAs transfected at a concentration of 50 nM. A total of 48 h post-transfection, efficiency was validated by reverse transcription-quantitative polymerase chain reaction (RT-qPCR).

Total RNA was extracted from tissues or cells using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturers' protocol, and subsequently converted into complementary DNA (cDNA) using a PrimeScript RT reagent kit (Takara Biotechnology Co., Ltd., Dalian, China), according to the manufacturer's protocol. The RNA expression levels were examined by real-time PCR using a SYBR Premix Dimmer Eraser kit (Takara Biotechnology Co., Ltd.). The thermocycling conditions of qPCR were as follows: 94°C for 15 min, followed by 45 cycles at 94°C for 10 sec, 60°C for 30 sec and 72°C for 30 sec. Gene expression in each sample was normalized to U6. The expression of miR-4443 was quantified using TaqMan MicroRNA Assay kit (Applied Biosystems; Thermo Fisher Scientific, Inc.), and human U6 RNA, which was amplified as a control. Data are presented as the mean ± standard deviation from three independent experiments. The relative expression fold-change of mRNA was calculated using the 2^−ΔΔCq method (15). The primer sequences were as follows: FEZF1-AS1, forward 5′-TTAGGAGGCTTGTTCTGTGT-3′, reverse 5′-GCGCAGGTACTTAAGAAAGA-3′; miR-4443, forward 5′-GTTGGAGGCGTGGGT-3′, reverse 5′-GGTCCAGTTTTTTTTTTTTTTTAAAACC-3′; and U6, forward 5′-CTCGCTTCGGCAGCACA-3′ and reverse 5′-AACGCTTCACGAATTTGCGT-3′.

Fluorescence-conjugated FEZF1-AS1 probes for RNA-FISH were generated, according to protocols of LGC Biosearch Technologies (Petaluma, CA, USA). HCC cells were treated in a non-denaturing condition, followed by hybridization with DNA probe sets, as previously described (9). DAPI (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) was used to stain the nuclei at room temperature for 10 min. Treated samples were visualized by confocal microscopy at ×600 magnification (FV1000; Olympus Corporation, Tokyo, Japan).

Cells were seeded into 96-well plates at a density of 1×10³ cells/well. Following 24, 48, 72 and 96 h of incubation at 37°C in DMEM (Gibco; Thermo Fisher Scientific, Inc.), cellular viability was evaluated using a Cell Counting Kit-8 (CCK-8) assay (Dojindo Molecular Technologies, Inc., Kumamoto, Japan), according to the manufacturer's protocol. The absorbance was measured at a wavelength of 450 nm using a Multiskan^™ GO Microplate Spectrophotometer (Thermo Fisher Scientific, Inc.).

A colony formation assay was also performed. A total of 2×10³ cells/well were seeded into 6-well plates with DMEM supplemented with 10% FBS and cultured at 37°C with 5% CO₂ for 14 days. Colonies were fixed with 100% methanol at room temperature for 20 min and stained with 0.1% crystal violet (Sigma-Aldrich; Merck KGaA) at 25°C for 30 min. The total number of visible colonies was determined under an optical light microscope at ×40 magnification (Olympus Corporation). All experiments were repeated three times.

Transwell migration and matrigel invasion assays were assessed using transwell membranes (8 µm pore size, 6.5 mm diameter; Corning Incorporated, Corning, NY, USA) in 24-well plates of 8 µm (BD Biosciences, Franklin Lakes, NY, USA) according to the manufacturer's protocol. For the transwell migration assay, cells were resuspended in serum-free 200 µl DMEM medium at a density of 5×10⁴ cells/well. The bottom chamber of transwell plates was supplemented with 600 µl DMEM medium, containing 20% FBS as a chemoattractant. Cells were subsequently seeded into the upper chamber and incubated for another 20 h at 37°C with 5% CO₂. For the invasion assay, matrigel matrix (1:8; 50 µl/well, BD Biosciences) was polymerized in the transwell membrane, according to the manufacturer's protocol. Cells (1×10⁵ cells/well) were used and incubated for another 48 h at 37°C with 5% CO₂. Following incubation, cells on the upper surface of the membrane were scraped off with cotton swabs. Cells that migrated to the lower surface were fixed with polyoxymethylene at 25°C for 30 min, stained with 0.1% Crystal Violet staining solution at 25°C for 30 min. The cells on the bottom of the membrane were calculated from five random light microscopic fields at ×40 magnification.

The potential targets of FEZF1-AS1 were predicted using the miRDB tool (http://mirdb.org/miRDB/index.html). Subsequently, FEZF1-AS1 sequences containing the wild-type (WT) binding site or mutated-type (Mut) binding site for miR-4443 were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China) and cloned into the pmiR-GLO vector (Promega Corporation, Madison, WI, USA). Luciferase reporter gene assays were performed using the Dual-Luciferase Reporter Assay System (Promega Corporation) according to the manufacturer's protocol. Prior to transfection, cells were seeded in 24-well plates (5×10³ cells/well) and cultured for 24 h. Subsequently, the WT or Mut of FEZF1-AS1 was transiently co-transfected with miR-4443 mimics or miR-NC (Guangzhou RiboBio Co., Ltd., Guangzhou, China) into logarithmic phase cells using Lipofectamine^® 2000 transfection reagent. Luciferase assays were performed using the Dual-Luciferase Reporter Assay system (Promega Corporation), according to the manufacturer's protocol, after a further 48 h culture. The value of relative luciferase activity indicates the firefly luciferase activity normalized to that of Renilla for each assay.

The statistical significance of the differences between groups was assessed using Student's t-test for pair-wise comparisons or one-way analysis of variance followed by Fisher's least significant difference post hoc test for multiple comparisons. Pearson's correlation coefficient analysis was used to determine the correlations. Kaplan-Meier survival curve analysis and log-rank test were used for survival analysis. P<0.05 was considered to indicate a statistically significant difference. Data are presented as the mean ± standard deviation of three independent experiments. Statistical analysis was performedv using the SPSS software version 20.0 (IBM Corp., Armonk, NY, USA).

To examine the expression patterns of FEZF1-AS1, RT-qPCR analysis was conducted using HCC cell lines and LO2 cells. It was identified that FEZF1-AS1 expression was significantly upregulated in HCC cell lines (Hep3B and Huh7) compared with normal LO2 hepatocytes (P<0.01; Fig. 1A). To further verify its overexpression in HCC cells, 58 HCC tissues and 58 adjacent normal tissues were obtained, and FEZF1-AS1 expression was analyzed using RT-qPCR. The results suggested that FEZF1-AS1 expression levels were significantly upregulated in HCC tissues compared with adjacent normal tissues (P<0.001; Fig. 1B). To determine whether FEZF1-AS1 is able to serve as a prognostic biomarker of patients with HCC, a Kaplan-Meier survival curve analysis was conducted. It was identified that patients with HCC with higher expression of FEZF1-AS1 exhibited a significantly lower survival rate (P<0.05; Fig. 1C). The location of FEZF1-AS1 in HCC cells was analyzed and it was identified that FEZF1-AS1 was located in the cytoplasm (Fig. 1D). FEZF1-AS1 expression was subsequently knocked down in Hep3B and Huh7 cells by transfection with specific RNAs. As presented, the expression levels of FEZF1-AS1 were significantly downregulated in Hep3B and Huh7 cells following transfection with the two siRNAs (P<0.05; Fig. 1E).

To examine the function of FEZF1-AS1 in HCC cells, CCK-8, colony formation and transwell assays were performed. Through CCK-8 assays, it was observed that knockdown of FEZF1-AS1 significantly inhibited the proliferation of Hep3B and Huh7 cells (Fig. 2A and B) at 96 h after culture. The colony formation assays suggested that the downregulation of FEZF1-AS1 significantly decreased the number of colonies formed by Hep3B and Huh7 cells (P<0.05; Fig. 2C). Furthermore, it was identified that FEZF1-AS1 depletion significantly decreased the numbers of migrated and invaded Hep3B and Huh7 cells (P<0.05; Fig. 2D and E). Collectively, these data demonstrated that FEZF1-AS1 is important for the proliferation, migration and invasion of HCC cells.

lncRNAs serve as competing endogenous RNAs (ceRNAs) to bind with miRs and regulate gene expression (16). In order to investigate the downstream mechanism of FEZF1-AS1, bioinformatics analysis was performed, which identified that numerous miRs exhibited potential binding sites for FEZF1-AS1. Among all targeting miRs, miR-4443 ranked top and has been reported to inhibit colon cancer progression (17). Therefore, miR-4443 was selected for further examination. It was identified that there were two potential binding sites of miR-4443 in FEZF1-AS1 (Fig. 3A). Through luciferase reporter assays, it was demonstrated that miR-4443 overexpression inhibited the luciferase activity in Hep3B and Huh7 cells transfected with FEZF1-AS1-WT, while mutation of the binding sites in FEZF1-AS1 abrogated this effect (Fig. 3B). Furthermore, it was identified that overexpression of miR-4443 significantly inhibited the expression levels of FEZF1-AS1 in Hep3B and Huh7 cells (P<0.01; Fig. 3C); whereas, inhibition of miR-4443 significantly increased the expression of FEZF1-AS1 (P<0.05; Fig. 3D). Consistently, knockdown of FEZF1-AS1 additionally significantly upregulated the expression levels of miR-4443 in Hep3B and Huh7 cells (P<0.05; Fig. 3E). Collectively, these data suggested that FEZF1-AS1 directly targets miR-4443 in HCC cells.

The function of miR-4443 remains unknown in HCC. To further determine whether miR-4443 is involved in HCC progression, the expression patterns of miR-4443 in HCC cell lines were first analyzed by RT-qPCR. The results suggested that miR-4443 was downregulated in HCC cell lines compared with LO2 cells (Fig. 4A). Similarly, the expression of miR-4443 was significantly downregulated in HCC tissues compared with adjacent normal tissues (P<0.001; Fig. 4B). Subsequently, the expression correlation between FEZF1-AS1 and miR-4443 was determined in HCC tissues by RT-qPCR. The results demonstrated that FEZF1-AS1 expression was negatively correlated with that of miR-4443 in HCC tissues (Fig. 4C). In summary, these data implied that miR-4443 may serve a function in HCC progression.

To verify whether miR-4443 was involved in FEZF1-AS1-mediated effects on HCC progression, the expression levels of miR-4443 were demonstrated to be downregulated by RT-qPCR in FEZF1-AS1-depleted Hep3B and Huh7 cells. The results suggested that the miR-4443 expression level was markedly downregulated in Hep3B and Huh7 cells transfected siFEZF1-AS1 plus miR-4443 inhibitor compared with the siFEZF1-AS1 group (Fig. 5A). CCK-8 assays were performed to evaluate the proliferation of Hep3B and Huh7 cells. It was identified that FEZF1-AS1 knockdown inhibited cellular proliferation while inhibition of miR-4443 significantly reversed this trend in Hep3B and Huh7 cells (P<0.05; Fig. 5B and C). Furthermore, transwell assays additionally demonstrated that FEZF1-AS1 depletion decreased the migration and invasion of Hep3B and Huh7 cells; whereas, knockdown of miR-4443 at the simultaneously promoted cell migration and invasion (Fig. 5D and E). FEZF1-AS1 promoted HCC cell proliferation, migration and invasion by acting as a sponge to miR-4443.