A total of 50 pairs of primary liver cancer (PLC) and adjacent liver tissue specimens were randomly selected from patients who had undergone hepatic resection at Xiangya Hospital (Changsha, China) between January and December 2017. The detailed clinicopathological data are presented in Table I. All cases were pathologically diagnosed by two independent pathologists. Furthermore, 30 matched fresh PLC tissues and adjacent non-tumor tissues were collected between January and August 2018 for reverse transcription-quantitative polymerase chain reaction (RT-qPCR) and western blot analyses. All patients or their families provided written informed consent regarding the use of their tissues for research purposes. All the patients were followed up as suggested in the REMARK guidelines (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2361579/). All experiments using human materials were approved by the Ethics Committee of Xiangya Hospital of Central South University (Changsha, China).

Total RNA from fresh PLC, adjacent non-tumor tissues and cell lines was extracted with TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer's instructions. RNA quantity and quality were evaluated using a Nanodrop spectrophotometer (Thermo Fisher Scientific, Inc.). RNA was reverse-transcribed into cDNA by BeyoRT™ II First-Strand cDNA Synthesis kit (Beyotime Institute of Biotechnology, Shanghai, China), and qPCR was conducted using SYBR-Green Master Mix on the Applied Biosystems QuantStudio™ 3 and 5 Real-Time PCR System (Thermo Fisher Scientific, Inc.). The primers were as follows: TMOD3 forward, 5′-TTCCGGCAGAAGAACCAGACATC-3′ and reverse, 5′-CAAGAATTGCTGCGAGGTCACAC-3′; GAPDH forward, 5′-GGCACCGTCAAGGCTGAGAAC-3′ and reverse, 5′-GGTGGCAGTGATGGCATGGAC-3′. Each sample was analyzed in triplicate and the data were calculated using the 2^−ΔΔCq method.

Tissues or cells were dissolved using RIPA lysis buffer supplemented with 1% phenylmethanesulfonyl fluoride. Protein concentration was measured using a bicinchoninic acid protein assay kit (Beyotime Institute of Biotechnology). Next, proteins were separated by 1% SDS-PAGE and transferred onto polyvinylidene difluoride membranes. Then, the membranes were blocked with 5% skimmed milk and incubated with specific primary antibodies overnight at 4°C. Following washing, the membranes were incubated with the appropriate horseradish peroxidase-conjugated secondary antibody at room temperature for 30 min and detected using an enhanced chemiluminescence kit (Thermo Fisher Scientific, Inc.). Antibodies against TMOD3 (cat. no. 70-ab4606-050) were obtained from MultiScicences (Hangzhou, China). Antibodies against p-ERK (cat. no. ab126455) and ERK (cat. no. ab17942) were obtained from Abcam (Cambridge, MA, USA), and those against E-cadherin (cat. no. sc-8426), vimentin (cat. no. sc-6260) and cyclin D1 (cat. no. sc-246) were purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). Antibodies against matrix metalloproteinase (MMP)2 (cat. no. AF0577) and MMP9 (cat. no. AF0220) were purchased from Affinity Biosciences (Cincinnati, OH, USA). The β-actin antibody (cat. no. TA-09) and corresponding secondary antibodies (cat. no. ZB-2305; cat. no. ZB-2301) were purchased from Zhongshan Golden Bridge Biotechnology (ZSGB; Beijing, China).

All tissues were cut into 4-µm sections, dewaxed in xylene and rehydrated in a graded ethanol series. Following heating in a microwave for antigen retrieval (12 min in sodium citrate buffer, pH 6), endogenous peroxidase was inactivated with 0.3% H₂O₂ for 30 min and the sections were incubated with 10% normal goat serum for 30 min. The TMOD3 antibody (1:100; MultiSciences) was applied overnight in a moist chamber at 4°C, followed by incubation with the secondary antibody (ZSGB) for 30 min. The antigen-antibody interactions were detected by 3,3′-diaminobenzidine and counterstained with hematoxylin. Tissue sections were dehydrated in graded ethanols and mounted.

The immunostained sections were independently evaluated by two pathologists who were blinded to all patient clinical data. The staining intensity and the percentage of protein expression were assessed. The staining intensity of TMOD3 was graded between 0 and 3 as follows: 0, negative; 1, weak; 2, moderate; and 3, strong. The percentage of positive cells was classified as 1 (0–25%), 2 (26–50%), 3 (51–75%) or 4 (>75%). The final score was calculated by multiplying these two scores, and the protein expression of TMOD3 in liver cancer specimens was divided into high-expression (≥4) and low-expression (<4) groups for further analysis.

The human liver cancer cell lines Hep3B, HepG2 and PLC/PRF5 were purchased from the American Type Culture Collection (Manassas, VA, USA). The MHCC97-H, MHCC97-L, HCCLM3 and Huh7 liver cancer cell lines and the normal liver cell line L02 were obtained from the Chinese Academy of Sciences (Shanghai, China). Cell culture was conducted according to the manufacturer's instructions and all the cell lines were cultured at 37°C in a humidified atmosphere of 5% CO₂.

A human TMOD3 overexpression clone lentivirus, three short hairpin RNA (shRNA) lentiviruses of TMOD3 and their control vectors were purchased from Shanghai GenePharma Co., Ltd. (Shanghai, China). The cells were cultured in 6-well plates prior to transfection until reaching 80–90% confluence within 24 h. Then, transfection was performed according to standard procedures. Puromycin (2 µg/ml) was used to select stable clones. The three candidate hairpin sequences for TMOD3 were as follows: 5′-CCTTGGGAATCTGTCAGAAACAG-3′ (shRNA-1); 5′-AAAGAAGCATTGGAGCATAAAGA-3′ (shRNA-2); 5′-CCTCGCAGCAATTCTTGGGAGC-3′ (shRNA-3); The efficiency of TMOD3 overexpression and knockdown were assessed by RT-qPCR and western blot analysis.

For the MTT assay, 5×10³ cells were seeded into each well of 96-well plates (6 wells/group). The cells were incubated for 0–7 days, then stained with MTT (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany), and the absorbance was measured at 570 nm. For colony formation, the cells were grown in 6-well plates at a density of 5×10² cells/well and cultured for 14 days. Then, the number of colonies was counted following staining with 1% crystal violet solution. All studies were conducted with 3 replicates.

The cells were seeded into 6-well plates at a density of 1×10⁵ cells/well. When grown to 90% confluence, the cells were incubated with mitomycin C (10 µg/ml) for 1 h at 37°C to suppress cell proliferation, and the cells were then starved for 24 h in serum-free medium. A 10-µl pipette tip was used to create an artificial wound. The results were observed and photographed every 12 h.

The cell invasion assay was performed in a 24-well Transwell plate. Cells were incubated with mitomycin C (10 µg/ml) for 1 h at 37°C to suppress cell proliferation, then 1×10⁵ cells in 500 µl of serum-free medium were placed into the upper chamber with Matrigel-coated membranes (BD Biosciences, Franklin Lakes, NJ, USA). The lower chamber was filled with 500 µl medium supplemented with 10% fetal bovine serum. Following a 48-h incubation at 37°C, the cells that remained in the upper chambers were removed and the cells that adhered to the lower membranes were stained with 0.1% crystal violet solution. The invading cells were counted in 5 random fields per well.

The cells were seeded into 6-well plates with glass coverslips for 24 h. Then, the cells were fixed in 4% paraformaldehyde, permeabilized with 0.5% Triton X-100 and incubated with phalloidin (Sigma-Aldrich; Merck KGaA) according to the manufacturer's protocol. The coverslips were counterstained with DAPI and the results were photographed under an inverted microscope.

Statistical analysis was conducted using SPSS 18.0 (SPSS Inc., Chicago, IL, USA). All measurement data were expressed as the mean ± standard deviation. Student's t-test or one-way analysis of variance were used to test the statistical significance of the differences between the groups, while proportional comparisons were conducted via a Chi-squared test. P<0.05 was considered to indicate a statistically significant difference.

Initially, the Oncomine Database (www.oncomine.org) was utilized to investigate TMOD3 expression in liver cancer. The results revealed that the TMOD3 mRNA levels were higher compared with those observed in normal liver tissues. The P-values recorded in the Wurmbach, Chen, Mas, Roessler 1 and 2 liver datasets were 8.27×10⁻⁵, 1.34×10⁻⁷, 4.81×10⁻⁴, 9.57×10⁻⁴ and 2.06×10⁻¹⁷, respectively (Fig. 1A). Subsequently, RT-qPCR and western blotting were performed in 30 pairs of PLC samples and matched normal liver tissues. Consistently, the mRNA and protein levels of TMOD3 were significantly higher compared with those observed in the normal liver tissues (P<0.001; Fig. 1B-D). IHC was performed to further analyze TMOD3 expression in 50 paired PLC and adjacent liver specimens. As shown in Fig. 1E, TMOD3 was mainly localized in the cytoplasm. It was also positively expressed in the 50 PLC specimens, of which 32 (64%) exhibited a high expression and 18 (36%) exhibited a low expression (Fig. 1F). Compared with the matched non-tumor tissues, TMOD3 expression was significantly higher in liver cancer tissues (P<0.001). These results indicate that TMOD3 is upregulated in liver cancer tissues and may contribute to liver cancer progression.

In order to study the biological functional role of TMOD3 in liver cancer, the present study conducted knockdown and overexpression experiments. First, the expression of the TMOD3 protein in the normal liver cell line L02 and in 7 liver cancer cell lines (MHCC97-H, MHCC97-L, Huh7, HepG2, PLC/PRF5, HCCLM3 and Hep3B) was evaluated. The results revealed that TMOD3 exhibited the greatest expression in Hep3B cells and the lowest in HepG2 cells, which had high and low metastatic potential, respectively (Fig. 2A and B). Stable TMOD3-overexpressing HepG2-TMOD3 cells and TMOD3-knockdown Hep3B-shTMOD3 cells were established via lentivirus transfection. The cell transfection efficiency in each cell type was confirmed by RT-qPCR and western blot analysis. Transfection of TMOD3-expressing lentivirus plasmids increased the expression of TMOD3 in HepG2 cells (P<0.001; Fig. 2C and D). Three shRNAs (shRNA1, shRNA2 and shRNA3) were constructed to silence TMOD3 expression in Hep3B cells. The expression level of TMOD3 was determined by RT-qPCR and western blot analysis, and shRNA1 was found to be the most effective; shRNA1 was consequently selected for further experiments (Fig. 2E and F). The MTT assay demonstrated that the proliferation rate was markedly increased in HepG2-TMOD3 cells, whereas Hep3B-shTMOD3 cells exhibited the opposite effect (Fig. 3A). Consistently, in the colony formation assay, HepG2-TMOD3 cells formed more colonies, while Hep3B-shTMOD3 cells exhibited decreased clonogenic ability (P<0.01; Fig. 3B). These results suggested that TMOD3 promotes the proliferation of liver cancer cells.

Wound healing and Transwell assays were performed to determine the impact of TMOD3 on the migration and invasion capacities of these cells. The results presented in Fig. 3C and D suggested that overexpression of TMOD3 in HepG2 cells significantly enhanced the wound healing ability and promoted cell invasion, while Hep3B-shTMOD3 cells displayed a slow wound closure rate and weak invasive abilities. Therefore, the present study demonstrated that TMOD3 promotes the migration and invasion of liver cancer cells.

As TMOD3 is associated with actin binding and is involved in cell migration and invasion, it was hypothesized that TMOD3 may be associated with the EMT process. IF analysis revealed that ectopic expression of TMOD3 in HepG2 cells displayed fibroblast-like spindled morphology. However, TMOD3 silencing in Hep3B cells produced a cobblestone-like appearance (Fig. 4A). Western blotting was performed to determine the expression of EMT biomarkers in liver cancer cells. The results demonstrated that overexpression of TMOD3 decreased the expression of E-cadherin (epithelial marker), and resulted in the upregulation of vimentin and Snail (mesenchymal markers) levels, while the opposite trend in the expression of these markers was observed in Hep3B-shTMOD3 cells (Fig. 4B). These results indicate that TMOD3 may induce EMT in liver cancer.

To evaluate the potential regulatory mechanism of TMOD3 in promoting liver cancer development, the Gene Set Enrichment Analysis (GSEA) analysis was used to identify the pathways regulated by TMOD3. High TMOD3 levels were positively associated with Kirsten rat sarcoma viral proto-oncogene (KRAS; Fig. 4C), which has previously been defined as a key component of the MAPK/ERK signaling pathway for modulating ERK activity, suggesting that TMOD3 may regulate MAPK/ERK signaling. The MAPK/ERK pathway plays an important role in cell proliferation, migration, differentiation and apoptosis, and it is one of the most important molecular pathways in cancer growth and metastasis (16). Western blot analysis revealed that TMOD3 overexpression increased the phosphorylation of ERK in HepG2 cells, whereas TMOD3 knockdown decreased the levels of p-ERK in Hep3B cells; however, the total level of ERK remained unchanged (Fig. 4D). In addition, the present study detected the expression of MMP2, MMP9 and cyclin D1, which are controlled by the MAPK/ERK signaling pathway and are associated with cancer cell proliferation and invasion. The results revealed that the expression of MMP2, MMP9 and cyclin D1 were significantly increased in HepG2-TMOD3 cells and decreased in Hep3B-shTMOD3 cells (Fig. 4D). These results indicated that TMOD3 may promote liver cancer progression by activating the MAPK/ERK signaling pathway.