TM4SF1 is a potential target for anti-invasion and metastasis in ovarian cancer

Fifty-five cases of patients with primary epithelial ovarian cancer (EOC), 30 cases pair-matched lymph node metastasis specimens. 23 cases of benign ovarian tumor tissues, 11 cases of normal ovarian epithelial tissues. The protocols used in our study were approved by the Ethics Committee of Affiliated Tumor Hospital of Guangxi Medical University. The EOC specimens were classified according to the FIGO stage (2014), and the clinicopathological characteristics of patients with EOC, benign ovarian tumor tissues, normal ovarian epithelial tissues are presented in Table 1. Human ovarian cancer cell lines (SKOV3, A2780, HO8910, and H08910PM) and human kidney epithelial 293 T cells were bought from cell bank of Chinese Academy of Sciences stored at Key laboratory of High-Incidence-Tumor Prevention &Treatment (Guangxi Medical University), Ministry of Education. The cells were cultured in a humidified incubator with 5% CO2 at 37 °C in RPMI1640 medium containing 10% or 20% fetal bovine serum (FBS). 5-w-old female nude mice with a body weight of 17–20 g were purchased from and fed in the experimental animal center of Guangxi Medical University.

Primers were designed using the Primer 5 primer design software according to the coding sequence (CDS) of human TM4SF1 (NM_014220) in GenBank. Primers sequences are presented in Table 2.

The lentiviral plasmid containing human TM4SF1 shRNA (including the GFP reporter gene) was constructed and confirmed by Shanghai KeyGEN (China). Sequences of siRNA of TM4SF1 are presented in Table 3.

Pre-selected paraffin sections were deparaffinized and dehydrated. Antigen retrieval was performed using the high-pressure method in 0.01 mol/L citrate buffer for 20 min. Sections were blocked in goat serum for 15 min and incubated with 1:2000 rabbit anti-human TM4SF1 (Abcam, USA) at 4 °C overnight. Sections were then incubated with goat anti-rabbit secondary antibody (CST, USA) at 37 °C for 1 h and developed using DAB (Beijing ZSGB-BIO, China) for 1 min. Sections were air- dried and mounted. Breast cancer tissues were used as the positive control, and as the negative control, the primary antibody was replaced with PBS.

The presence of yellow granules in the cell membrane or cytoplasm was considered positive TM4SF1 protein staining. Five locations in ovarian tissues without necrosis were randomly selected, and 100 cells in each field were observed under a microscope at 400X magnification. The score of each field was comprehensively determined according to the staining intensity and the number of positive cells. The scores were recorded and averaged to obtain the immunostaining score of each section. The standards of the staining intensity score were as follows: no color (−) was 0 point, light yellow color (+) was 1 point, yellow-brown color (++) was 2 points, and brown color (+++) was 3 points. The standards of the score of the number of positive cells were as follows: ≤5% was 0 points, 6%~ 25% was 1 point, 26%~ 50% was 2 points, 51%~ 75% was 3 points, and > 75% was 4 points. The product of these two scores > 3 was defined as positive TM4SF1 protein expression. All sections were independently reviewed by two experienced pathologists. The average value was obtained after double-blind scoring.

Cells at the logarithmic growth phase were treatment with pancreatic enzyme and inoculated onto 96-well plates at a density of 2 × 10³ (HO8910PM cells) or 6 × 10³ (HUVECs) cells/well. After cell attachment, MTT or CCK-8 reagent (DOJINDO, Japan) was added after 0 h, 24 h, 48 h, 72 h, and 96 h and incubated with cells at 37 °C for 1 h. The optical density (OD) values at 450 nm were detected using a microplate reader (Thermo, USA).

Cells at the logarithmic growth phase were collected and fixed in 75% ice-cold ethanol at 4 °C overnight. According to the manual of the cell cycle detection reagent kit (Shanghai KeyGEN, China), cells were incubated with RNase A in a 37 °C water bath for 30 min, stained with PI at 4 °C in the dark for 30 min, and detected using a flow cytometer (EPICS XL, COULTER, USA).

After single cell suspension was prepared, cells were inoculated onto 6-well plates at densities of 100, 200, 400, and 600 cells/well. Each density was applied in 3 replicate wells. Cells were then incubated in a 37 °C and 5% CO₂ incubator for continuous culture for 5–7 d. The cell culture was terminated according to the conditions of cell colony formation. Cells were fixed in methanol and stained with Giemsa stain. The cell colonies were counted under a microscope.

Single cell suspension under starvation conditions was prepared using serum-free culture medium. Cells were inoculated into the top chamber of Transwell (8.0 μm, Costar, USA) at a density of 1 × 10⁶ cells. The bottom chamber of the Transwell was filled with 600 μl of culture medium containing 20% fetal bovine serum (FBS). Cells were cultured in a 5% CO₂ incubator at 37 °C for 6 h, fixed in 4% paraformaldehyde (30 min at 25 °C), and stained with Giemsa staining solution for 30 min. Three random fields were photographed under an inverted microscope, and the cells that passed through the bottom chamber surface were counted. For each group, three replicate wells were prepared, and the experiment was repeated three times. The mean values were statistically analyzed (\( \overline{x} \)±SD).

Single cell suspension under starvation conditions was prepared using serum-free culture medium. Cells were inoculated into the top chamber of Transwell coated with Matrigel (BD, USA) at a density of 1 × 10⁶ cells/well. The bottom chamber of the Transwell was filled with 600 μl of culture medium containing 20% FBS. Cells were cultured in a 5% CO₂ incubator at 37 °C for 6 h, fixed in 4% paraformaldehyde (30 min at 25 °C), and stained with Giemsa staining solution for 30 min. Three random fields were photographed under an inverted microscope, and the cells that passed through the bottom chamber surface were counted. For each group, three replicate wells were prepared, and the experiment was repeated three times. The mean values were calculated (\( \overline{x} \)±SD).

Twelve nude mice were randomly divided into 2 groups with 6 animals in each group. Mice were labeled; group A was LV-CON-RNAi-Luc/HO8910PM, and group B was LV-TM4SF1-RNAi-Luc/HO8910PM. All animals were used in accordance with institutional guidelines, and the current experiments were approved by the Use Committee for Animal Care.

Cell suspension (100 μl) was subcutaneously inoculated into the right groin at a concentration of 1 × 10⁸ cells/ml. Tumor formation was observed every week, and the experiment was completed 6 w after the inoculation. Every week, the long (L) and short (S) diameters of the tumors were measured and the tumor volume was calculated according to the formula: V = L × S²/2.

Live imaging of xenograft tumors in nude mice was performed every 2 w for 6 w. The luciferase substrate Luciferin (Promega, Beijing, China) was intraperitoneally injected using a microinjector at 150 mg/kg. The growth of xenograft tumors in the mice were observed via In-Vivo Imaging System Fx Pro (BRUKER, USA). The luminescence value at each time point was quantitatively analyzed.

Nude mice were euthanized with inhalation of carbon dioxide, and killing was completed by cervical dislocation. and the subcutaneous xenograft tumors were dissected and collected. TM4SF1 mRNA expression in xenograft tumors from the two groups was detected using RT-qPCR. The method was the same as that described in 2.2.2.

TM4SF1 expression in xenograft tumors from the two groups was detected using western blotting. The method was the same as that described in 2.2.3.

All quantitative data were expressed as mean ± standard deviation (\( \overline{x} \)±SD). The Student’s t-test was used to compare quantitative variables. The Chi-squared test and Fisher’s exact test were used to compare categorical variables. Multivariate analysis was performed using COX proportional hazards regression model. SPSS v.16.0 software was used to statistical analysis. GraphPad Prism 5 Software was used to present graph. A value of P < 0.05 was considered statistically significant.

After RNAi, TM4SF1 mRNA expression (2-△△Ct) in xenograft tumors was significantly decreased compared to that in the control group [(0.08 ± 0.02) vs (1.00 ± 0.13), P < 0.05] (Fig. 6d).

After RNAi, TM4SF1 protein expression (relative gray density value TM4SF1/GAPDH) in xenograft tumors was significantly decreased compared to that in the control group [(0.11 ± 0.05) vs (0.87 ± 0.05), P < 0.05] (Fig. 6e).

At least 34 TM4SF family members have been discovered in the human body. They play important roles in embryonic development, cell migration, adhesion, metastasis, and fusion. Different TM4SF genes are located on different chromosomes and encode 20–30 kD transmembrane proteins. To date, many tumor metastasis- and invasion-related studies have focused on CD9, CD81, CD82/KAI1, CD151, and TM4SF1.

It has been shown that an increase in TM4SF1 expression can promote the migration ability of epidermal keratinocytes [13]. Increased TM4SF1 protein expression can also increase the invasion ability of lung cancer cells, and an anti-TM4SF1 monoclonal antibody can significantly reduce the invasion ability of lung cancer cells [14]. Down-regulation of TM4SF1 gene expression using RNA interference technology can reduce the migration of prostate cancer cells and the metastasis ability of cervical cancer cells [15]. A recent study showed that hsa-miR-141 can effectively interfere with TM4SF1 expression in prostate cancer to inhibit prostate cancer cell invasion and metastasis, which describes a new carcinogenesis mechanism of TM4SF1 and suggests the potential of targeting TM4SF1 in gene therapy [16].

Our group performed immunohistochemistry and showed that the positive expression rate of TM4SF1 protein in epithelial ovarian cancer tissues was higher than that in benign ovarian tumor tissues and normal ovarian tissues. Moreover, the positive expression rate in benign ovarian tumor tissues was higher than that in normal ovarian tissues. The level of TM4SF1 protein expression also increased with an increase in the FIGO stage of epithelial ovarian cancer and a reduction in the differentiation level, indicating that the expression of TM4SF1 increased with abnormal hyperplasia and malignant transformation of ovarian epithelial cells and suggesting that TM4SF1 might participate in the occurrence and development of ovarian cancer. In addition, the positive expression of TM4SF1 protein in tissues in all metastatic lymph node foci and the matched primary loci indicated that TM4SF1 protein might be closely associated with tumor invasion and metastasis. Inhibition of TM4SF1 expression significantly inhibited the migration and invasion ability of HO8910PM and SKOV3 cells.

Otsuka et al. [17] applied internal cDNA microarray technology to detect genetic changes in colon cancer cell lines (SW 480 and SW 620) and 58 cases of clinically stratified colorectal cancer tissues, and the results showed that TM4SF1 gene expression was closely associated with the progression of tumor stage and might be a potential marker of metastasis.

Gordon et al. [18] reported that TM4SF1 was a negative regulatory gene in apoptosis of malignant pleural mesothelioma cells and could promote tumor cell growth. The results in this study showed that inhibition of TM4SF1 expression did not have a significant influence on cell cycle or cell growth and proliferation ability.

Chang et al. [19] showed that TM4SF1 and aminopeptidase N (CD13) co-localized in cells and formed a complex on the surface of lung cancer cells to affect cell migration ability, which further mediated lung cancer cell invasion and migration.

This study showed that TM4SF1 was localized in the cell membrane or the membrane of cells near the basement membrane in normal ovarian epithelial tissues and benign ovarian tumor tissues. However, its expression was concentrated in the cytoplasm in cancer tissues. These results were consistent with the results of Allioli et al. [15], who showed that TM4SF1 protein expression was mainly concentrated in the cytoplasm in prostate cancer and in the apical membrane in benign tissues. This localization change of TM4SF1 protein in different tissues might be associated with cancer cell invasion and metastasis.

The relevant mechanisms underlying the involvement of TM4SF1 in cancer cell invasion and metastasis are still not clear. Huang et al. [5] showed that TM4SF1 mediated tumor metastasis through regulation of the migration-associated genes MMP-2, MMP-9, and VEGF. A recent study showed that TM4SF1 could interact with discoidin domain receptor 1 (DDR1) promoted by collagen I. In addition, TM4SF1 could recruit syntenin2 and PKCa and activate PKCa to induce the noncanonical DDR1 signaling pathway to regulate metastasis of dormant breast cancer cells to multiple distant tissues [20].

Some researchers have also performed studies on the association between TM4SF1 and clinical prognosis. Kao et al. [14] showed that high TM4SF1 protein expression was significantly associated with postoperative early recurrence and a short survival period in lung squamous cell carcinoma and that increased TM4SF1 protein expression shortened the survival period in a tumor-bearing animal model. The results in this study showed that TM4SF1 protein expression in epithelial ovarian cancer at the late stage was higher than that at the early stage. However, the results of multivariate analyses showed that the positive expression rate of TM4SF1 protein was not an independent prognostic factor for epithelial ovarian cancer patients.

At the cell level, we showed that the TM4SF1 expression level was not significantly associated with cell growth and proliferation. However, the nude mice tumor formation experiment showed that inhibition of TM4SF1 expression in HO8910PM cells significantly inhibited the growth of xenograft tumors. The possible reason is that RNAi does not completely inhibit TM4SF1 expression.

Furthermore, it has also been shown that TM4SF1 is a target gene of androgen receptor [15]. Immunization with the cytotoxic T lymphocyte (CTL) epitope of TM4SF1 in HLA-A2 transgenic mice induced CTL immune responses to effectively inhibit the growth of tumor cells that expressed TM4SF1 [21].