Expression of tensin 4 in hepatocellular carcinoma and its role in epithelial-mesenchymal transformation

Abstract

BACKGROUND

Tensin (TNS) 4, also known as C-terminal TNS-like, is a unique oncogenic member of the TNS family that lacks the N-terminal actin-binding domain and acts as a critical signal amplifier rather than a structural scaffold.

AIM

To investigate the expression of TNS4 in hepatocellular carcinoma (HCC) tissues and HepG2 and PLC cells, and its role in epithelial-mesenchymal transition (EMT).

METHODS

HCC and paired paracancerous tissues were collected from April 2015 to October 2017. TNS4 expression was assessed via immunohistochemistry and western blotting. Its correlation with clinicopathological features was analyzed, and Kaplan-Meier survival curves were constructed. Adeno-associated virus 8 vectors carrying TNS4- or Slug-specific shRNA were generated and used to transfect HepG2 and PLC cells. Western blotting was performed to evaluate TNS4, Slug, EMT markers, and transforming growth factor-β1/Smad pathway proteins. Cell proliferation was measured by Cell counting kit-8 assay, while Transwell and wound healing assays were used to assess invasion and migration.

RESULTS

Expression of TNS4 in HCC tissues was significantly higher than that in adjacent tissues (P < 0.05). The tumor-free and overall survival rates of patients with high TNS4 expression were significantly lower than those of patients with low TNS4 expression (P < 0.05). Overexpression of TNS4 promoted EMT in HepG2 and PLC cells. Silencing TNS4 downregulated expression of Slug, p-smad2 and p-smad3; reversed EMT; and inhibited proliferation, invasiveness and metastasis of hepatoma cells (P < 0.05).

CONCLUSION

High expression of TNS4 is closely related to poor prognosis in patients with HCC. TNS4 may affect Slug expression and regulate EMT through the transforming growth factor-β1/Smad pathway, and promote HCC proliferation and migration.

Key Words: Tensin4; C-terminal tensin-like; Slug; Hepatocellular carcinoma; HepG2 cells; PLC cells; Transforming growth factor-β1; Epithelial-mesenchymal transformation

Core Tip: This study showed that tensin (TNS) 4 was highly expressed in hepatocellular carcinoma (HCC) tissues and correlated with poor prognosis. TNS4 may regulate Slug expression to mediate epithelial-mesenchymal transition via the transforming growth factor-β1/Smad signaling pathway, promoting proliferation, migration and invasion of HCC cells. TNS4 is expected to be a novel therapeutic target for HCC and merits further in-depth research.

INTRODUCTION

Hepatocellular carcinoma (HCC) is one of the most common malignant tumors encountered in general surgery. Currently, surgical resection is the main treatment, but relapse and metastasis are common, and the latter is an important cause of death in patients with HCC. Epithelial-mesenchymal transition (EMT) plays an important role in tumor invasiveness and metastasis[1-3]. EMT refers to the process of tumor cells transforming from more polar epithelial cells to more aggressive and metastatic mesenchymal cells in a specific tumor microenvironment, which can promote tumor cell invasiveness and distant metastasis[4]. Transforming growth factor (TGF)-β1 is an effective EMT-inducing factor, and promotes tumor invasiveness and metastasis[5]. Slug (encoded by SNAI2) was first extensively characterized for its regulatory role in neural crest cell development during embryogenesis. It is now well established as a core transcriptional regulator of EMT in many malignancies, and represents a canonical downstream target of the TGF-β1/Smad signaling cascade, whose expression level reliably mirrors EMT progression and the activation status of this pathway[6,7]. The above research evidence has validated its utility as a robust reference marker for assessing EMT phenotypes in HCC and other epithelial tumors. The tensin (TNS) family is a group of focal adhesion-associated proteins involved in cell adhesion, cytoskeletal organization, migration and intracellular signal transduction[8-10]. The family consists of four members (TNS1-TNS4) with context-dependent functional duality. TNS1-TNS3 often act as tumor suppressors by stabilizing cell-matrix adhesion and restricting aberrant cell motility. In sharp contrast, TNS4, also known as C-terminal TNS-like, is an atypical oncogenic member of the TNS family that lacks the conserved N-terminal actin-binding domain. Rather than functioning as a passive structural scaffold, TNS4 acts as a powerful signal amplifier for receptor tyrosine kinase pathways such as epidermal growth factor receptor (EGFR) and MET, and represents a predominant oncogenic driver in nearly all epithelial malignancies[11]. TNS4 is closely related to the interstitial transformation of non-small cell lung cancer[12,13], colorectal cancer[14-18], invasive breast cancer, melanoma, esophagogastric junction cancer and other tumors[19-22]. In HCC specifically, TNS4 has been identified as a downstream effector of the Ras/mitogen-activated protein kinase signaling pathway, and its expression is induced by EGF-activated extracellular signa-regulated kinase 1/2 to promote cancer cell proliferation and migration[23]. Nevertheless, the regulatory mechanism by which TNS4 modulates EMT in HCC, particularly its crosstalk with the canonical TGF-β1/Smad signaling axis, has not yet been fully elucidated. This study aimed to investigate the relationship between TNS4 expression in HCC tissues, clinicopathological features and prognosis, and study the role and mechanism of TNS4 in EMT, to provide new therapeutic targets for HCC.

MATERIALS AND METHODS

Collect and organize clinical patient data

This study included 102 patients diagnosed with primary liver cancer in the Department of Hepatobiliary Surgery, Affiliated Hospital of Chengde Medical College from April 2015 to October 2017; all of whom underwent surgical resection. HCC was diagnosed by postoperative pathological analysis. There were 54 males and 48 females, with an age range of 30-76 years, with an average of 55.17 ± 8.58 years. This study was approved by the Ethics Committee of our hospital, and all patients gave signed informed consent.

Postoperative follow-up

Follow-up data were obtained from the medical records and examination reports of the Department of Hepatobiliary Surgery. Some patients were followed up by telephone. All patients were reviewed every 3-6 months for 2 years after surgery, every 6-12 months thereafter, and every 12 months after 5 years. The follow-up included: (1) Patients’ general condition; (2) Routine blood tests; (3) Liver and kidney function; (4) Serum alpha-fetoprotein (AFP); (5) Viral load (hepatitis B virus DNA or hepatitis C virus RNA) and other serological detection indicators; and (6) Liver plain scan enhanced computed tomography or liver magnetic resonance plain scan enhanced diffusion-weighted imaging; and lung computed tomography plain scan. The primary endpoints of the study were disease-free survival (DFS) from the date of surgery to the date of recurrence and metastasis, and overall survival (OS) from the time of diagnosis of primary liver cancer to the time of death or end of follow-up. One hundred and two patients were followed up for 8.6-60.0 months (Table 1).

Table 1 Clinicopathological data of hepatocellular carcinoma patients (n = 102).

Clinicopathological index		Number of patients	%
Sex	Male	54	52.94
	Female	48	47.06
Age (years)	< 60	67	65.69
	> 60	35	34.31
Hepatitis B/C surface antigen	Positive	68	66.67
	Negative	34	33.33
Liver cirrhosis	Yes	79	77.45
	No	23	22.55
Liver function classification	Child-Pugh A	74	72.55
	Child-Pugh B	28	27.45
Tumor size (cm)	> 5	41	40.20
	< 5	61	59.80
Tumor number	Single	84	82.35
	Multiple	18	17.65
AFP level (ng/mL)	> 400	64	62.75
	< 400	38	37.25
Vascular invasion	Yes	39	38.24
	No	63	61.76
Cancer embolism	Yes	38	37.25
	No	64	62.75
Pathological grade	I/II	59	57.84
	III/IV	43	42.16
TNM staging	I	61	59.80
	II/III	41	40.20

AFP: Alpha-fetoprotein.

Reagents

The normal liver cell line LO2 and human hepatoma cell lines HepG2, PLC, Hep3B and HuH-7 were purchased from Wuhan Pricella Biotechnology Co. Ltd. Main reagents and instruments included: (1) Trypsin and fetal bovine serum (Vivacell, Shanghai, China); (2) Trizol kit [Thermo Fisher (Shanghai) Co. Ltd.]; (3) FastKing RT Kit (With gDNase) (KR116) [Tiangen Biochemical Technology (Beijing) Co. Ltd.]; (4) Immunohistochemical kit (Zhongshan Golden Bridge Biotechnology Co., Ltd., Beijing, China); (5) Transwell chamber (Corning Corporation, Corning, NY, United States); (6) 4′,6-diamidino-2-phenylindole nuclear staining solution, RIPA protein lysis solution, and BCA protein concentration determination kit (Beijing Solarbio Technology Co. Ltd.); (7) TGF-β1, N-cadherin, E-cadherin, vimentin (Abclonal, Wuhan, Hubei Province, China); (8) GAPDH (HUABIO, Hangzhou, Zhejiang Province, China); (9) The p-Smad2, p-Smad3, Smad2 and Smad3 (Affinity, Hubei Province, China); (10) TNS4, Slug Antibody, Goat anti-Rabbit IgG (H+L), HRP-conjugated, Goat anti-Mouse IgG (H+L), HRP-conjugated (Proteintech, Wuhan, Hubei Province, China); (11) Small-molecule inhibitor of Smad3 (SIS3) (Selleck Company); (12) Polyvinylidene difluoride membrane and ECL chemiluminescent liquid [Zhongshi Gene Technology (Tianjin) Co. Ltd.]; (13) The shRNA control, adeno-associated virus (AAV) 8-shRNA virus and TNS4 overexpression (OE) vectors (Shanghai GenePharma Co. Ltd.); (14) M450 [Thermo Fisher (Shanghai) Co. Ltd.]; and (15) Cell incubator (Sanyo Company, Japan).

Cell culture

Normal liver LO2 cell lines and human hepatoma HepG2 and PLC cell lines were cultured in RPMI 1640 containing 10% fetal bovine serum, 100 U/mL penicillin and 0.1 g/L streptomycin. Human hepatoma Hep3B and HuH-7 cell lines were cultured in Dulbecco’s modified Eagle’s medium at 37 °C and 5% CO₂ in an incubator. HepG2 and PLC cell lines selected were grouped into the following groups: (1) Control; (2) TGF-β1; (3) AAV8-control; (4) AAV8-shRNA-TNS4; (5) TGF-β1 + shRNA-normal control (NC); (6) TGF-β1 + shRNA-TNS4; (7) TGF-β1 + control; (8) TGF-β1 + OE-TNS4; (9) TGF-β1 + OE-TNS4 + shRNA-Slug; (10) TGF-β1 + OE-TNS4 + Vehicle; and (11) TGF-β1 + OE-TNS4 + SIS3. Each experiment was repeated three times with sample size n = 3.

Immunohistochemistry

Hematoxylin and eosin (HE) staining was performed on human HCC and paracancer tissue sections (4 μm). Sections were baked at 80 °C for 15 minutes, dewaxed in xylene, rehydrated through graded ethanol, and rinsed with distilled water and phosphate-buffered saline (PBS). Nuclei were stained with hematoxylin for 1 minute, followed by differentiation and tap water washing. Sections were counterstained with eosin for 2 minutes, dehydrated through graded ethanol, cleared in xylene, and mounted with neutral balsam. The other paraffin sections were repaired with citric acid antigen repair solution by microwave after conventional dewaxing and soaking in water. Reagent 1 (endogenous peroxidase blocker) was added to the sections, which were incubated at room temperature for 10 minutes and cleaned twice with PBS. Primary antibody (catalog No. 11580-1-AP; TNS4 Antibody; Proteintech) was applied at a dilution of 1:50, and the sections were stored in a refrigerator at 4 °C overnight. Goat anti-rabbit horseradish-peroxidase-conjugated secondary antibody (catalog No. ZB-2301; Zhongshan Golden Bridge Biotechnology) was applied at a dilution of 1:200. Enzyme-labeled goat anti-rabbit IgG polymer was added at 37 °C for 40 minutes, color was developed with 3,3-diaminodbenzidine working solution, and the reaction was terminated by rinsing with tap water. Sections were redyed with hematoxylin, dehydrated, cleared and sealed. PBS was used as negative control instead of primary antibody. A positive reaction was indicated by the presence of brown-yellow particles in the cell membrane or plasma. In the negative controls, there were no brown-yellow particles in the cell membrane or plasma, although the nuclei were stained blue. A semiquantitative score was calculated according to the staining intensity of tumor cells and the proportion of positive cells. The TNS4 immunohistochemical staining score was obtained by multiplying the tumor positive cell percentage score by the tumor cell staining intensity score, with the lowest score being 0 and the highest score being 12. We defined ≤ 4 as low expression of TNS4 and > 4 as high expression.

The shRNA construction and packaging

Based on TNS4 and Slug gene sequences, shRNA sequences of several targets were designed online by Life Technology, according to shRNA design principles. The target fragment was constructed on the vector plasmid AAV8-TNS4-shRNA and AAV8-Slug-shRNA, transformed, identified and sequenced. Target gene TNS4 primer sequence: (1) Upstream: 5’-GGACAGCAATGACCTCATCCGA-3’; and (2) Downstream: 3’-AATGCTGGCACACGAAGGCAGA-5’. Target gene Slug primer sequence: (1) Upstream: 5’-CCATGCCTGTCATACCACAA-3’; and (2) Downstream: 3’-ACAGTGATGGGGCTGTATGC-5’. In the control group, Scramble sequences with no gene homology were selected. Total 293T cells were transfected with AAV carrier plasmid, packaging plasmid and helper plasmid (type 8). Cells and supernatants were collected, impurities were discarded by centrifugation, bacteria were removed by filtration, and cells and supernatants were concentrated and purified. Virus titer was determined by quantitative reverse transcription polymerase reaction (qRT-PCR), and cells and supernatants were stored at -80 °C for later use.

The qRT-PCR

Total RNA was extracted using Trizol reagent. The concentration and purity were determined by NanoDrop and RNA was stored at -80 °C. Total RNA (1 μg) and 1 μL DNA Eraser were added to a 200-μL RNase-free EP tube, and the final volume was adjusted to 10 μL with RNase-free water. All operations were carried out on ice. The mixture was incubated at 42 °C for 5 minutes. Subsequently, 10 μL of 2 × SPARKscript II RT Plus Master Mix was added to the tube. After gentle mixing and brief centrifugation, the reaction was incubated at 50 °C for 15 minutes, followed by heating at 85 °C for 5 minutes to inactivate the enzyme. The resulting cDNA was immediately used for qPCR analysis. The target gene TNS4 and Slug primer sequences are shown above. The PCR amplification procedure was: (1) Pre-denaturation at 95 °C for 15 minutes; (2) Denaturation at 95 °C for 15 seconds; (3) Annealing at 60 °C; and (4) Extension at 72 °C for 34 seconds, for a total of 40 cycles. The amplification curve and fusion curve were confirmed. Each sample was tested in triplicate and repeated three times. The fold change for each group was calculated using the 2^-DDct method, compared with those of the control group.

Western blotting

Tissues and cells were collected, total protein was extracted and the concentration was determined. Sample solutions containing 30 μg total protein from each group were prepared, separated by gel electrophoresis, transferred onto polyvinylidene difluoride membranes and blocked with 5% nonfat milk. The primary antibody was incubated at 4 °C overnight, and the secondary antibody was incubated at room temperature for 1 hour. Image Pro Plus software was used to analyze the OD of the protein bands and detect the protein expression level.

Cell counting kit-8 assay

HepG2 and PLC cells from TGF-β1 + shRNA1-NC and TGF-β1 + shRNA1-TNS4 groups were inoculated in 96-well plates at 5 × 10⁴/mL, and incubated at 37 °C in 5% CO₂. Cell counting kit (CCK) 8 solution (10 μL) was added to each well after 24 hours, 48 hours, 72 hours and 96 hours, and OD₄₅₀ was determined after 2 hours. The OD value was directly proportional to the number of viable cells, calculation formula: (OD_drug - OD_blank)/(OD_control - OD_blank) × 100%.

Cell scratch test

HepG2 and PLC cells in the TGF-β1 + shRNA1-NC and TGF-β1 + shRNA1-TNS4 groups, which showed exponential growth during routine digestion, were prepared with 4 × 10⁵/mL cell suspension in serum-free medium, inoculated into six-well plates and cultured in a 5% CO₂ incubator at 37 °C. When the cell confluence reached approximately 90%, a 200-mL pipette head was used to scratch the bottom of the plate according to the template. On the back of the six-well plate, six horizontal lines were drawn with a spacing of 0.5 cm between adjacent lines, which served as a background reference. The plate was washed twice with PBS, and culture was continued in the incubator for 24 hours and 72 hours. Cell migration at 0 hour, 24 hours and 72 hours in each group was observed and photographed.

Transwell assay

Fibronectin (1 ×) was diluted 15 times with buffer C solution (0.5 ×). Diluted fibronectin (0.1 mL) was added to the upper Transwell chamber, and incubated at 4 °C in the refrigerator for 3 hours. Culture medium containing 10% serum (0.8 mL) was added to the lower Transwell chamber. HepG2 and PLC cells in the TGF-β1 + shRNA1-NC and TGF-β1 + shRNA1-TNS4 groups were digested and centrifuged by trypsin. A suspension of 7.5 × 10⁵ cells/mL was prepared with serum-free medium, and 0.1 mL was placed into the upper Transwell chamber, which was cultured in a 5% CO₂ incubator at 37 °C for 48 hours. The chamber was removed, blotted dry, rinsed in sterile PBS, and the cells on the membrane surface at the bottom of the upper chamber were wiped with a wet cotton swab. The cells were fixed in methanol for 30 minutes, stained with Giemsa dye for 20 minutes at room temperature, washed and dried. Nine visual fields were randomly selected under the microscope and counted.

Statistical analysis

SPSS version 26.0 was used for statistical analysis. Measurement data were presented as mean ± SD. Pairwise comparison was performed by t test. One-way analysis of variance was used for comparison among multiple groups, and numerical data were expressed as %. The difference between the two groups was analyzed by χ² test. The DFS and OS curves were plotted by Kaplan-Meier method, and the difference in survival rate between the two groups was calculated by log-rank test. P < 0.05 was considered statistically significant. All experiments were repeated three times.

RESULTS

HE staining of HCC and paracancer tissue

HCC and paracancer tissue samples were differentiated by HE staining. Compared with paracancer tissue, HCC tissue showed destruction of hepatic lobular structure, and the cells were arranged into nests or cords with dense distribution. The blood sinuses were abundant, the cancer cells were polygonal, the cytoplasm was abundant and atypia was obvious. All samples were identified as HCC tissue and matched paracancer tissue, after which, follow-up experiments were performed (Figure 1).

Figure 1 Hematoxylin and eosin staining of human hepatocellular carcinoma and paracancer tissue.

Expression of TNS4 protein in HCC detected by immunohistochemistry

Most staining for TNS4 protein was in the cytoplasm, although some cell membranes were also stained. Positive expression was indicated by brown-yellow or brown granule distribution. Specific scoring criteria were: (1) Positive cell percentage score: Positive cell rate ≤ 5%, 0 points, 6%-25%, 1 point, 26%-50%, 2 points, 51%-75%, 3 points, and > 75%, 4 points; and (2) Cell staining intensity score: 0 point for no staining, 1 point for light yellow coloring, 2 points for brown yellow coloring, and 3 points for dark brown coloring. The product of positive cell percentage score and cell staining intensity score yielded the TNS4 immunohistochemical staining score: (1) Lowest score was 0; and (2) Highest score was 12. We set ≤ 4 as low expression of TNS4, and > 4 as high expression. Immunohistochemical results showed that in 102 pairs of HCC and paracancer tissues, 87 cases (85.3%) had high expression of TNS4 protein and 15 had low expression. In 102 pairs of paracancer tissues, 11 had high expression of TNS4 protein (10.8%) and 91 had low expression. Intergroup comparison suggested that expression of TNS4 protein in HCC tissues was significantly higher than that in adjacent tissues (P < 0.05; Figure 2).

Figure 2 Expression of tensin 4 protein in human hepatocellular carcinoma and paracancer tissues was detected by immunohistochemistry.

Expression of TNS4 in HCC detected by western blotting

Compared with expression of TNS4 protein in adjacent tissues, expression in HCC tissues was significantly upregulated (P < 0.05; Figure 3).

Figure 3 Expression of TNS4 in hepatocellular carcinoma detected by western blotting. A: Tensin 4 protein expression in hepatocellular carcinoma and adjacent tissues determined by western blotting; B-D: Tensin 4 protein expression in hepatocellular carcinoma and adjacent tissues of Samples 1-3 by western blotting. ^dP < 0.0001.

Relationship between TNS4 expression and clinicopathological features

The relationship between expression of TNS4 and the clinicopathological features of HCC was studied, including: (1) Gender; (2) Age; (3) Hepatitis B/C surface antigen; (4) Whether the patients were complicated with cirrhosis; (5) Liver function grade; (6) Tumor size and number; (7) AFP level; (8) Vascular invasion; (9) Cancer thrombus formation; (10) Pathological grade; and (11) Postoperative TNM stage of the tumor. Expression of TNS4 was not significantly associated with gender; (1) Age; (2) Hepatitis B/C surface antigen; (3) Cirrhosis; (4) Liver function grade; (5) Tumor size and number; (6) AFP level; and (7) Postoperative tumor TNM stage (P > 0.05). Expression level of TNS4 was significantly correlated with tumor vascular invasion, tumor thrombus formation and pathological grade (P < 0.05; Table 2).

Table 2 Relationship between tensin 4 protein expression and clinicopathological features in hepatocellular carcinoma, n (%).

Clinicopathological feature		n	Negative expression	Positive expression	P value
Sex	Male	54	12 (22.22)	42 (77.78)	0.569
	Female	48	13 (27.08)	35 (72.92)
Age (years)	< 60	67	17 (25.37)	50 (74.63)	0.779
	> 60	35	8 (21.05)	27 (78.95)
Hepatitis B/C surface antigen	Positive	68	19 (27.94)	49 (72.06)	0.255
	Negative	34	6 (17.65)	28 (82.35)
Liver cirrhosis	Yes	79	18 (22.78)	61 (77.22)	0.453
	No	23	7 (30.43)	16 (69.57)
Liver function classification	Child-Pugh A	74	15 (20.27)	59 (79.73)	0.106
	Child-Pugh B	28	10 (35.71)	18 (64.29)
Tumor size, cm	> 5	41	11 (26.83)	30 (73.17)	0.655
	< 5	61	14 (22.95)	47 (77.05)
Tumor number	Single	84	18 (21.43)	66 (78.57)	0.118
	Multiple	18	7 (38.89)	11 (61.11)
AFP level (ng/mL)	> 400	64	13 (20.31)	51 (79.69)	0.201
	< 400	38	12 (31.58)	26 (68.42)
Vascular invasion	Yes	39	5 (12.82)	34 (87.18)	0.031
	No	63	20 (31.75)	43 (68.25)
Cancer embolism	Yes	38	5 (13.16)	33 (86.84)	0.04
	No	64	20 (31.25)	44 (68.75)
Pathological grade	I/II	59	19 (32.20)	40 (67.80)	0.034
	III/IV	43	6 (13.95)	37 (86.05)
TNM staging	I	61	17 (27.87)	44 (72.13)	0.336
	II/III	41	8 (19.51)	33 (80.49)

AFP: Alpha-fetoprotein.

Correlation between TNS4 expression and prognosis of patients with HCC

We used a Kaplan-Meier survival curve to analyze 102 patients with HCC. Median TNS4 expression level was used as the boundary between the high and low expression groups, and the relationship between TNS4 expression and postoperative DFS and OS was analyzed. In HCC tissues, DFS and OS of patients with high expression of TNS4 were significantly lower than those of patients with low expression of TNS4 (P < 0.05). This indicated that high expression of TNS4 was closely related to poor postoperative prognosis of patients (Figure 4).

Figure 4 Correlation between tensin 4 expression and prognosis of patients with hepatocellular carcinoma. A: Relationship between tensin 4 expression level and postoperative tumor-free survival in patients with hepatocellular carcinoma; B: Relationship between tensin 4 expression level and overall survival rate of patients with hepatocellular carcinoma. ^aP < 0.05. TNS4: Tensin 4; HR: Hazard ratio.

Expression of TNS4 in different hepatoma cell lines

Compared with normal human liver LO2 cell lines, TNS4 protein expression was significantly increased in HuH-7, HepG2, PLC and Hep3B cell lines (P < 0.05). HepG2 and PLC cell lines with high TNS4 expression level were selected for follow-up studies (Figure 5).

Figure 5 Expression of tensin 4 in different hepatoma cell lines. A: Expression of tensin 4 protein in normal liver LO2 cell lines and hepatoma cell lines (PLC, HepG2, HuH-7 andHep3B); B: Expression of tensin 4 in each cell line was detected by western blotting. ^cP < 0.001, ^dP < 0.0001. TNS4: Tensin 4.

Influence of TNS4-shRNA on TNS4 mRNA expression in HepG2 and PLC cells

AAV8-TNS4-shRNA1, AAV8-TNS4-shRNA2 and AAV8-TNS4-shRNA3 were transfected into HepG2 and PLC cells, and qRT-PCR was performed after 48 hours. AAV8-TNS4-shRNA1, AAV8-TNS4-shRNA2 and AAV8-TNS4-shRNA3 all silenced expression of TNS4 mRNA in HepG2 and PLC cells. The silencing efficiency of AAV8-TNS4-shRNA1 was the most significant (P < 0.05; Supplementary Figure 1).

Effect of Slug-shRNA on Slug mRNA expression in HepG2 and PLC cells

AAV8-Slug-shRNA1, AAV8-Slug-shRNA2 and AAV8-Slug-shRNA3 were transfected into HepG2 and PLC cells, and qRT-PCR was performed after 48 hours. AAV8-Slug-shRNA1, AAV8-Slug-shRNA2 and AAV8-Slug-shRNA3 all silenced expression of Slug mRNA in HepG2 and PLC cells. The silencing efficiency of AAV8-Slug-shRNA1 was the most significant (P < 0.05; Supplementary Figure 2).

Influence of TNS4 OE vector on TNS4 mRNA expression in HepG2 and PLC cells

Recombinant AAV8-TNS4 OE vector was constructed and transfected into HepG2 and PLC cells for 48 hours. The qRT-PCR showed that expression of TNS4 mRNA in the OE group was significantly higher than that in the control and NC groups (P < 0.05; Supplementary Figure 3).

Effects of TNS4-shRNA transfection on EMT-related protein expression in HepG2 and PLC cells

HepG2 and PLC cells were transfected with AAV8 control and AAV8-shRNA1-TNS4 for 72 hours, respectively. Western blotting showed that expression of E-cadherin in the AAV8-shRNA1-TNS4 group was significantly increased compared with the AAV8 control group. Expression of N-cadherin, vimentin and Slug were significantly decreased (P < 0.05). AAV8-shRNA1-TNS4 reversed EMT in HepG2 and PLC cells, and TNS4 was involved in the process of EMT in HepG2 and PLC cells (Figure 6).

Figure 6 Effects of tensin 4-shRNA transfection on epithelial-mesenchymal transition-related protein expression in HepG2 and PLC cells. A: Influence of tensin (TNS)4-shRNA transfection on epithelial-mesenchymal transition-related protein expression in HepG2 cells, measured by western blotting; B-E: TNS4-shRNA1 transfection significantly downregulated expression of N-cadherin, vimentin and Slug, and upregulated E-cadherin expression in HepG2 cells; F: Influence of TNS4-shRNA transfection on epithelial-mesenchymal transition-related protein expression in PLC cells, measured by western blotting; G-J: TNS4-shRNA1 transfection exerted consistent regulatory effects in PLC cells, with downregulation of N-cadherin, vimentin and Slug and upregulation of E-cadherin. ^bP < 0.01, ^cP < 0.001, ^dP < 0.0001. TNS4: Tensin 4; AAV: Adeno-associated virus.

Effects of TGF-β1 on EMT-related protein expression in HepG2 and PLC cells

After 72 hours induction of TGF-β1, western blotting showed that, compared with the control group, expression of E-cadherin in the TGF-β1 group was significantly decreased, while expression of N-cadherin and vimentin was significantly increased (P < 0.05). TGF-β1 induced EMT in HepG2 and PLC cells (Figure 7).

Figure 7 Effects of transforming growth factor-β1 on epithelial-mesenchymal transition-related protein expression in HepG2 and PLC cells. A: Influence of transforming growth factor (TGF)-β1 on epithelial-mesenchymal transition (EMT)-related protein expression in HepG2 cells detected by western blotting; B-D: TGF-β1 significantly upregulated expression of N-cadherin and vimentin, and downregulated E-cadherin expression in HepG2 cells; E: Influence of TGF-β1 on EMT-related protein expression in PLC cells detected by western blotting; F-H: TGF-β1 induced consistent changes in expression of the aforementioned EMT-related proteins in PLC cells. ^dP < 0.0001. TGF-β1: Transforming growth factor-β1.

Effects of TGF-β1 + shRNA-TNS4 on EMT-related protein expression in HepG2 and PLC cells

HepG2 and PLC cells were induced by TGF-β1, and transfected with shRNA1-NC and shRNA1-TNS4 for 72 hours. Western blotting showed that, compared with TGF-β1 + shRNA1-NC, expression of E-cadherin in the TGF-β1 + shRNA1-TNS4 group was significantly increased, while expression of N-cadherin, vimentin and Slug was significantly decreased (P < 0.05). TGF-β1 induced EMT in HepG2 and PLC cells, and shRNA-TNS4 partially reversed EMT (Figure 8).

Figure 8 Effects of transforming growth factor-β1 + shRNA-tensin 4 on epithelial-mesenchymal transition-related protein expression in HepG2 and PLC cells. A: Influence of transforming growth factor (TGF)-β1 + shRNA-tensin (TNS) 4 on epithelial-mesenchymal transition (EMT)-related protein expression in HepG2 cells, measured by western blotting; B-E: TGF-β1 + shRNA1-TNS4 significantly downregulated expression of N-cadherin, vimentin and Slug, and upregulated E-cadherin expression in HepG2 cells; F: Influence of TGF-β1 + shRNA-TNS4 on EMT-related protein expression in PLC cells, measured by western blotting; G-J: TGF-β1 + shRNA1-TNS4 exerted consistent regulatory effects on the aforementioned EMT-related proteins in PLC cells. ^cP < 0.001, ^dP < 0.0001. TGF-β1: Transforming growth factor-β1; NC: Normal control; TNS4: Tensin 4.

Effects of TGF-β1 + OE of TNS4 on EMT-related protein expression in HepG2 and PLC cells

Under TGF-β1 induction, HepG2 and PLC cell lines were transfected with control and TNS4 OE vectors. After 72 hours, western blotting showed that, compared with the TGF-β1 + control group, expression of E-cadherin in the TGF-β1 + OE-TNS4 group was significantly decreased, while expression of N-cadherin, vimentin and Slug was significantly increased (P < 0.05). TNS4 promoted EMT in HepG2 and PLC cells (Figure 9).

Figure 9 Effects of transforming growth factor-β1+ overexpression of tensin 4 on epithelial-mesenchymal transition-related protein expression in HepG2 and PLC cells. A: Effect of transforming growth factor (TGF)-β1 + overexpression (OE) of tensin (TNS) 4 on epithelial-mesenchymal transition (EMT)-related protein expression in HepG2 cells, measured by western blotting; B-E: TGF-β1 + OE significantly upregulated expression of N-cadherin, vimentin and Slug, and downregulated E-cadherin expression in HepG2 cells; F: Effect of TGF-β1 + OE of TNS4 on EMT-related protein expression in PLC cells, measured by western blotting; G-J: TGF-β1 + OE exerted consistent regulatory effects on the aforementioned EMT-related proteins in PLC cells. ^cP < 0.001, ^dP < 0.0001. TGF-β1: Transforming growth factor-β1; OE: Overexpression; TNS4: Tensin 4.

Relationship between Slug knockdown after OE of TNS4 and EMT of HepG2 and PLC cell lines

Under TGF-β1 induction, HepG2 and PLC cell lines transfected with TNS4 OE vector were transfected with shRNA1-Slug. Western blotting showed that expression of E-cadherin was significantly upregulated, while expression of N-cadherin and vimentin was significantly decreased (P < 0.05). TNS4 regulated EMT through Slug (Figure 10).

Figure 10  Relationship between Slug knockdown after overexpression of tensin 4 and epithelial-mesenchymal transition of HepG2 and PLC cell lines. A: Relationship between Slug knockdown after tensin (TNS) 4 overexpression (OE) and epithelial-mesenchymal transition in HepG2 cells; B-D: After OE of TNS4, Slug knockdown significantly upregulated expression of E-cadherin, and significantly downregulated expression of N-cadherin and vimentin in HepG2 cells; E: Relationship between Slug knockdown after TNS4 OE and epithelial-mesenchymal transition in PLC cells line; F-H: After OE of TNS4, Slug knockdown significantly upregulated expression of E-cadherin, and significantly downregulated expression of N-cadherin and vimentin in PLC cells. ^bP < 0.01, ^dP < 0.0001. TGF-β1: Transforming growth factor-β1; OE: Overexpression; TNS4: Tensin 4.

Effects of shRNA-TNS4 on TGF-β1/Smad signaling pathway in HepG2 and PLC cells

After inducing HepG2 and PLC cells with TGF-β1, the cells were transfected with AAV8-shRNA1-TNS4 for 72 hours, and expression of Smad2, p-Smad2, Smad3 and p-Smad3 proteins was detected by western blotting. Compared with the TGF-β1 + control group, expression of p-Smad2 and p-Smad3 in the TGF-β1 + AAV8-shRNA1-TNS4 group was significantly decreased (P < 0.05), while expression of t-Smad2 and t-Smad3 was not significantly changed (Figure 11).

Figure 11  Effects of shRNA-tensin 4 on transforming growth factor-β1/Smad signaling pathway in HepG2 and PLC cells. A: Transforming growth factor (TGF)-β1/Smad signaling pathway-related protein expression in HepG2 cells, measured by western blotting; B and C: TGF-β1 + adeno-associated virus-shRNA1-TNS4 significantly downregulated expression of p-Smad2/Smad2 and p-Smad3/Smad3 in HepG2 cells; D: TGF-β1/Smad signaling pathway-related protein expression in PLC cells, measured by western blotting; E and F: TGF-β1 + adeno-associated virus-shRNA1-TNS4 significantly downregulated expression of p-Smad2/Smad2 and p-Smad3/Smad3 in PLC cells. ^dP < 0.0001. TGF-β1: Transforming growth factor-β1; TNS4: Tensin 4; AAV8: Adeno-associated virus 8.

Effects of TGF-β1/Smad pathway inhibitor SIS3 on EMT-related protein expression in HepG2 and PLC cells

HepG2 and PLC cell lines were transfected with TNS4 OE vector under TGF-β1 induction, and SIS3, a Smad3 pathway inhibitor, was added. Compared with the TGF-β1 + OE-TNS4 + Vehicle group, expression of Slug protein in the TGF-β1 + OE-TNS4 + SIS3 group was significantly decreased, expression of E-cadherin was significantly increased, and expression of vimentin and N-cadherin was significantly decreased. The TGF-β1/Smad pathway is a key link in the influence of TNS4 on Slug expression and regulation of EMT (Figure 12).

Figure 12  Effects of transforming growth factor-β1/Smad pathway inhibitor small-molecule inhibitor of Smad3 on epithelial-mesenchymal transition-related protein expression in HepG2 and PLC cells. A: Effects of transforming growth factor (TGF)-β1/Smad pathway inhibitor small-molecule inhibitor of Smad3 (SIS3) on epithelial-mesenchymal transition-related protein expression in HepG2 cells, measured by western blotting; B-E: SIS3 significantly downregulated expression of N-cadherin, vimentin and Slug, and significantly upregulated expression of E-cadherin in HepG2 cells; F: Effects of transforming growth factor-β1/Smad pathway inhibitor SIS3 on epithelial-mesenchymal transition-related protein expression in PLC cells, measured by western blotting; G-J: SIS3 significantly downregulated expression of N-cadherin, vimentin and Slug, and significantly upregulated expression of E-cadherin in PLC cells. ^cP < 0.001, ^dP < 0.0001. TGF-β1: Transforming growth factor-β1; TNS4: Tensin 4; OE: Overexpression; AAV8: Adeno-associated virus 8; SIS3: Small-molecule inhibitor of Smad3.

Effects of TNS4-shRNA transfection on proliferation of HepG2 and PLC cells induced by TGF-β1 detected by CCK-8 assay

CCK-8 assay showed that there was no significant difference in OD between the TGF-β1 + shRNA1-TNS4 and TGF-β1 + shRNA1-NC groups at 24 hours. When the culture time point was 48 hours, 72 hours or 96 hours, the OD of the TGF-β1 + shRNA1-TNS4 group was significantly lower than that of the TGF-β1 + shRNA1-NC group, and the difference in OD increased with time (P < 0.05). This indicated that downregulation of TNS4 expression reduced the activity of human hepatoma cells and inhibited their proliferation (Figure 13).

Figure 13  Tensin 4-shRNA1 significantly inhibited proliferation of HepG2 and PLC cells. ^aP < 0.05, ^bP < 0.01, ^dP < 0.0001. TGF-β1: Transforming growth factor-β1; TNS4: Tensin 4; NC: Normal control.

Effects of TNS4-shRNA on migration and invasiveness of HepG2 and PLC cells induced by TGF-β1

The scratch and Transwell assays showed that migration and invasiveness of HepG2 and PLC cells in the TGF-β1 + shRNA1-TNS4 group were significantly decreased compared with in the TGF-β1 + shRNA1-NC group (P < 0.05). shRNA1-TNS4 downregulated the migration and invasiveness of HCC cells (Figure 14).

Figure 14  Effects of tensin 4-shRNA on migration and invasiveness of HepG2 and PLC cells induced by transforming growth factor-β1. A-C: Effect of tensin (TNS)4-shRNA on the migration ability of HepG2 cells induced by transforming growth factor (TGF)-β1 detected by scratch assay. TNS4-shRNA1 significantly inhibited migration of HepG2 cells after 24 hours and 72 hours; D-F: Effect of TNS4-shRNA on migration of PLC cells induced by TGF-β1 detected by scratch assay. TNS4-shRNA1 significantly inhibited migration of PLC cells after 24 hours and 72 hours; G-J: Effect of TNS4-shRNA on the invasiveness of HepG2 and PLC cells induced by TGF-β1 detected by Transwell. TNS4-shRNA1 significantly inhibited invasiveness of HepG2 and PLC cells. ^aP < 0.05, ^cP < 0.001, ^dP < 0.0001. TGF-β1: Transforming growth factor-β1; TNS4: Tensin 4; NC: Normal control.

DISCUSSION

Primary liver cancer is one of the most common cancers worldwide, ranking fifth in global incidence and third as a cause of cancer death[24]. Improvement of surgical techniques and emergence of targeted and immunotherapy have improved the prognosis of liver cancer; however, high recurrence and metastasis rates still affect prognosis[25]. Early metastasis of liver cancer is mainly spread through the portal vein and its branches in the liver, and hematogenous, lymphatic and implantation metastasis may occur in the late stage. Mesenchymal transformation of cancer cells plays an important role in the invasiveness and metastasis of liver cancer[26,27]. TNS4 is highly expressed in non-small cell lung cancer[12,13], colorectal cancer[14-18], invasive breast cancer, melanoma, esophagogastric junction cancer and other tumor tissues[19-22], and is closely related to tumor mesenchymal transformation. However, there are no reports on the mechanism of TNS4 in liver cancer mesenchymal transformation. This study investigated the relationship between TNS4 expression and prognosis in liver cancer, as well as the role and mechanism of TNS4 in interstitial transformation of liver cancer cells. The results show that TNS4 is highly expressed in liver cancer tissues, and is related to tumor vascular invasion, cancer thrombus formation and pathological grade of patients, which indicates poor prognosis.

In this study, TGF-β1 induced HCC cells to construct a model of interstitial transformation. Expression of EMT marker E-cadherin was significantly increased after TNS4 knockdown in HepG2 and PLC cells, while expression of N-cadherin and vimentin was significantly decreased. After OE of TNS4, expression of E-cadherin decreased significantly, while expression of N-cadherin and vimentin increased significantly. The process of EMT was partially reversed after TNS4 knockdown, and OE of TNS4 promoted EMT.

In this study, CCK-8, cell scratch and Transwell assays were performed after TNS4 knockdown in HepG2 and PLC cells. TNS4 knockdown reduced the proliferation, migration and invasiveness of HCC cells, indicating that TNS4 played an important role in promoting liver cancer metastasis. According to the study of Lu et al[28], in non-small cell lung cancer, OE of TNS4 significantly promoted the proliferation and colony formation of PC9 and PC9/GR cells, while they were inhibited by silencing of TNS4. Studies have shown that TNS4 regulates interstitial transformation induced by TGF-β1 in colorectal cancer cells, and upregulation of TNS4 expression promotes migration and invasiveness of colorectal cancer cells[29]. Di-Luoffo et al[30] showed that TNS4 promoted the invasiveness and migration of intrahepatic cholangiocarcinoma. All these results are consistent with our study.

EMT is closely related to tumor invasiveness and metastasis, and Slug is one of the nuclear transcription factors closely related to EMT[31]. Slug is a key regulator that initiates EMT in malignant cells and promotes distant metastasis of cancer cells[32,33]. We found that Slug expression decreased and EMT was inhibited after TNS4 knockdown in HCC cells. The expression of Slug increased after OE of TNS4, which promoted EMT of HCC cells. After OE of TNS4, Slug knockdown inhibited EMT of HCC cells. It has been shown that TNS4 affects EMT by regulating the expression of Slug, a key transcription factor.

In this study, after TGF-β1 induced interstitial transformation of HCC cells, expression of p-Smad2 and p-Smad3 was significantly downregulated after transfection with TNS4-shRNA, while expression of t-Smad2 and t-Smad3 was unaffected. Under induction of TGF-β1, HCC cells were transfected with TNS4 OE vector and treated with SIS3; an inhibitor of the TGF-β1/Smad pathway. Expression of Slug protein was significantly reduced and EMT of HCC cells was inhibited (E-cadherin expression was significantly increased, N-cadherin and vimentin expression was significantly decreased). This study suggests that TNS4 influences Slug expression and regulates EMT through the TGF-β1/Smad pathway. Consistent with our findings, Pan et al[34] verified that activation of the TGF-β1/Smad cascade drives EMT in biliary epithelial cells during primary biliary cholangitis, indicating the conserved pro-EMT role of this pathway across different liver diseases. Wu et al[35] revealed that the adherens junction scaffold protein p120-catenin amplifies TGF-β signaling by promoting Smad2/3 nuclear translocation. This mechanistic model supports our observation that TNS4, a focal adhesion scaffold protein, is regulated by the TGF-b1/Smad signaling axis in HCC cells. Collectively, our findings align with established regulatory mechanisms and confirm that TNS4 promotes EMT in HCC cells via the TGF-b1/Smad signaling pathway.

From a mechanistic standpoint, the protumorigenic effects of TNS4 extend beyond its regulatory role in the TGF-β1/Smad signaling axis. An increasing number of studies have confirmed that TNS4 directly interacts with the E3 ubiquitin ligase c-Cbl, suppressing ubiquitination and proteasome-dependent degradation of activated receptor tyrosine kinases, including EGFR and MET, and maintaining persistent activation of downstream pro-oncogenic signaling cascades[11,36]. This receptor-stabilizing effect works in concert with the TGF-β1/Smad–Slug pathway revealed in our work to form a robust feed-forward regulatory loop, which collectively amplifies EMT-related signaling and promotes acquisition of invasive and metastatic phenotypes in HCC cells. Such a dual regulatory mode provides a molecular explanation for the potent prometastatic activity of TNS4, and enhances our understanding of the mechanistic network underlying TNS4-driven HCC progression.

CONCLUSION

TNS4 is abnormally highly expressed in HCC tissues and is closely related to the clinicopathological features and prognosis of patients with HCC. TNS4 may affect Slug expression and regulate EMT through the TGF-β1/Smad pathway, play a role in promoting the proliferation and migration and invasiveness of HCC, and is expected to become a new therapeutic target for patients with HCC.