TSN (purity ≥99%) was purchased from Shanghai Yuanye Biotechnology Co., Ltd. (Shanghai, China). 3-(4,5)-dimethylthiazol(-2-yl)-2,5-diphenyltetrazolium bromide (MTT), Annexin V/propidium iodide (PI) apoptosis detection kit, cell cycle analysis kit and enhanced chemiluminescence (ECL)-Plus/kit were all from KeyGen Biotech Co., Ltd. (Nanjing, China). TRIzol reagent were from Invitrogen (Carlsbad, CA, USA). SYBR Green Master Mixture was obtained from Takara (Otsu, Japan). miScript Reverse Transcription kit and miScript SYBP Green PCR kit were all from Qiagen (Japan). β-catenin, c-Myc, cyclin D1, Bcl-2, MMP-9, AKT, p-AKT(Ser⁴⁷³), GSK3β, p-GSK3β (Ser⁹) and β-actin polyclonal antibodies were all from ABclonal Biotech Co. (Wuhan, China). Histone H3, E-cadherin, N-cadherin and vimentin monoclonal antibodies were all from Cell Signaling Technology (Beverly, MA, USA). ERK1/2 and p-ERK1/2 (Thr²⁰²/Tyr²⁰⁴) polyclonal antibodies were both from Hangzhou HuaAn Biotechnology Co., Ltd. (Hangzhou, China).

Human GC cell lines MGC-803, BGC-823, HGC-27, AGS, SGC-7901 and MKN-45 and normal human gastric epithelial cells GES-1 were obtained from the Type Culture Collection, Chinese Academy of Sciences (Shanghai, China), cultured in RPMI-1640 medium supplemented with 10% heat-inactivated fetal bovine serum (FBS), 100 U/ml penicillin and 100 µg/ml streptomycin and in a humidified atmosphere containing 5% CO₂ at 37°C.

Lentivirus-mediated β-catenin overexpression vector (LV-CTNNB1), negative control vector (CON238), lentivirus-mediated hsa-miR-200a-3p knockdown vector (LV-hsa-miR-200a-3p-inhibition) and negative control vector (CON137) were all from GeneChem Co., Ltd. (Shanghai, China). The correct sequences and insertions were confirmed by DNA sequencing. Transfection was performed according to the manufacturer's instructions. On the day of vector transduction, SGC-7901 cells were cultured at 5×10⁴ cells/well in 24-well plates containing serum-free medium with polybrene (5 mg/ml). At 50% confluence, cells were transfected with recombinant experimental virus or control virus at the optimal MOI (multiplicity of infection) of 50, and cultured at 37°C and 5% CO₂ for 4 h. Then supernatant was discarded and medium containing serum was added. Positive and stable transfectants were selected and expanded for further study.

Cell proliferation was analyzed using the MTT assay. Briefly, cells were cultured in 96-well plates at a density of 5×10³ cells/well. After adhering to the plate surface, cells were treated with TSN of different doses for various times, followed by 20 µl MTT (5 mg/ml) incubation for further 4 h and subsequent 150 µl DMSO dissolution for 5 min. The optical densities (ODs) were measured at 570 nm using an enzyme immunoassay analyzer (Bio-Rad, Hercules, CA, USA). The cell proliferation inhibition rates were calculated using the following formula: 1 − OD (experiment) / OD (control). The experiment was performed in triplicate.

Annexin V/PI apoptosis detection kit was used to detect cell apoptosis. Briefly, cells (1×10⁶) treated with or without TSN for 48 h were harvested and suspended in binding buffer according to the instruction of the apoptosis kit. Approximately 5 µl Annexin V and 5 µl PI were then added to the fixed cells for 20 min in darkness at room temperature. Then, Annexin V binding buffer was added to the mixture before analysis by fluorescent activated cell sorting (fACS) on a flow cytometer. Three separate experiments were performed for each clone.

For cell cycle distribution analysis, cells were incubated in serum-free medium for 24 h followed by treatment with TSN for another 48 h. Then cells were trypsinized, washed with PBS, fixed with 75% cold ethanol overnight at 4°C and washed with PBS again. Afterwards, the fixed cells were stained with PI in the presence of RNase A for 30 min in the dark. The samples were then analyzed on FACsort flow cytometer (Becton-Dickinson, Mountain View, CA, USA). The DNA content in cells could be read according to PI fluorescence. ModFit3.0 software (Verity Software House, Topsham, ME, USA) was used for cell cycle analysis. The experiment was performed in triplicate.

SGC-7901 cells were suspended in serum-free medium with or without TSN after starvation in serum-free medium for 24 h. The suspension were then added to the upper chamber coated with Matrigel (3.9 µg/µl, 80 µl), while the lower chamber was filled with RPMI-1640 medium containing 10% bovine serum. After 24-h incubation, the non-invaded cells on the upper side of the chamber were washed away and cells attached to the bottom were fixed with 75% alcohol and stained with crystal violet. The number of cells invading through the Matrigel were counted using microscope in three random fields of each group.

Cell migration was assessed by wound-healing assay. Briefly, the adherent cells in a 6-well plate (1×10⁶/well) were wounded with a standard 200-µl pipette tip to form a 1-mm wide strip across the well. Then the wounded monolayers were washed twice with PBS before incubation in medium with or without TSN. After 48-h incubation, the wound closure was observed and photographed with a microscope.

Total RNA was extracted using TRIzol reagent (Invitrogen) according to the manufacturer's protocol. Total RNA (1 µg) was used to synthesize cDNA by reverse transcription using M-MLV reverse transcriptase and cDNA amplification was performed using the Power SYBR Green Master Mix kit (for AKT, ERK, GSK3β, β-catenin, c-Myc, cyclin D1, Bcl-2 and MMP-9). MiScript Reverse Transcription kit and miScript SYBP Green PCR kit were for miR-200a according to the manufacturer's instructions. The primer sequences are shown in Table I. The expression of human β-actin or U6 was used as internal control. Cycle threshold (Ct) values were obtained graphically for the target genes, miR-200a, U6 and β-actin. ΔCt = Ct (target genes or miR-200a) − Ct (β-actin or U6). ΔΔCt = ΔCt (experimental group) − ΔCt (control group). Data were analyzed using the comparative Ct method (2^−ΔΔCt). Each sample was tested in triplicate.

Cytoplasmic proteins were extracted with RIPA buffer and nuclear protein extraction was conducted using Nuclear Extraction kit (Life Technologies, Carlsbad, CA, USA). Protein concentrations were determined using Bradford assay and cell extracts were boiled for 5 min in loading buffer. Equal amount of proteins were separated on 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis before translocation onto polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). Then the membranes were blocked with 5% skim milk for 1 h. The primary antibodies against AKT, p-AKT, ERK, p-ERK, GSK3β, p-GSK3β, β-catenin, histone H3, c-Myc, cyclin D1, Bcl-2 and MMP-9 were diluted according to the manufacturer's instructions and the membranes were incubated overnight at 4°C. Afterwards, the membranes were washed three times followed by incubation in HRP-labeled goat anti-rabbit IgG secondary antibodies at a dilution ratio of 1:4,000 at room temperature for 1 h. Immunodetection was performed with an ECL-Plus/kit.

All animals were purchased from Shanghai Laboratory Animal Center of Chinese Academy Sciences (Shanghai, China) and were adapted to cages with 12-h light/dark cycle in a temperature-controlled environment for study. SGC-7901 cells with or without transduction were subcutaneously injected and maintained by passage in the hypodermis of female BALB/c/nu/nu mice (4–5 weeks of age, 18–20 g). The female BALB/c/nu/nu mice (4–5 weeks of age, 18–20 g) were randomly divided into 4 groups (n=5/group). Mice in groups 1 and 2 were orthotopically implanted in the stomach with subcutaneous GC tissue pieces formed by SGC-7901 cells without transduction while SGC-7901 cells transfected with LV-CTNNB1 or LV-hsa-miR-200a-3p inhibition belonged to groups 3 and 4, respectively. Before sugery, the female mice were administered with rhabarber and glauber to further reduce the immunity and improve the rate of tumor formation. The surgical procedure of gastric orthotopic transplantation was performed as previously reported (41). Briefly, subcutaneous GC tissue pieces (~2.0 cm × 2.0 cm × 1.0 cm) were harvested and minced into small pieces (~1 mm³). The female nude mouse stomach was gently exteriorized via a left-side upper abdominal incision and one small tissue pocket in the middle wall of the greater curvature was cut. One tumor piece was placed into the pocket and fixed with a drop of medical tissue glue (gifts from Department of General Surgery, Shanghai Jiao Tong University Affiliated Shanghai Sixth People's Hospital, Shanghai, China). The stomach was then relocated into the abdominal cavity followed by the abdominal closure with 4-0 absorbable sutures. Seven days after surgery, TSN was dissolved in RPMI-1640 medium and intraperitoneally given to all mice in groups 2, 3 and 4 at a dose of 0.20 mg/kg/day and 0.2 ml once every other day for 35 days. Meanwhile, the equal volume of RPMI-1640 medium was administered to mice in group 1 in the same way. All mice were sacrificed by cervical dislocation and dissected 24 h after the final medicine. The orthotopically implanted tumor tissues were removed from stomach and was measured with a caliper and calculated with the formula volume = (length × width²)/2. The metastatic nodules in each liver were observed and counted. All in vivo animal studies were approved by the Animal Ethics and Research Committee of Shanghai Jiao Tong University and in accordance with the Internal Biosafety and Bioethics Guidelines of School of Medicine, Shanghai Jiao Tong University.

Immunohistochemistry was performed to assay the expression of tissue proteins in β-catenin pathway. Briefly, the gastric tumor tissue slides were deparaffinized, rehydrated, antigen-retrieved (performed with 10 mM sodium citrate buffer, pH 6.0, at 90°C for 30 min), blocked and antibody-incubated in turn. The slides were preincubated with 0.04% bovine serum albumin to block non-specific binding. Subsequently, the slides were incubated with primary polyclonal antibodies (ABclonal Biotech Co., Ltd.) at a dilution of 1:200 overnight at 4°C and a secondary antibody (KeyGen Biotech Co.) at room temperature for 1 h. Finally, the sections were stained with DAB (3,3-diaminoben-zidine) and counterstained with hematoxylin. Images were visualized under a microscope (Olympus, Tokyo, Japan).

SPSS 20.0 was used for the statistical analysis. All of the values were recorded as the mean ± standard deviation (SD). Independent-sample t-test and one-way analysis of variance (ANOVA) were used to analyze statistical differences of experimental data between two groups and more than two groups, respectively. Chi-square test (χ²) test was used to compare the differences of proportions between two groups. Significance was defined as p<0.05 or p<0.01.

We examined the effects of TSN on the viability of various GC cells (MGC-803, BGC-823, HGC-27, AGS, SGC-7901 and MKN-45) and normal human gastric epithelial cells (GES-1) using MTT assay. As shown in Fig. 1B–H, TSN produced marked proliferation inhibition of cells in a dose- and time-dependent manner. The calculated IC₅₀ values on the basis of inhibition rates were 7.95±0.65 to 60.74±6.73 µM (24 h), 0.70±0.09 to 8.15±0.97 µM (48 h) and 0.11±0.04 to 4.10±0.45 µM (72 h) (Fig. 1I–K). Among these cell lines, SGC-7901 cells were most sensitive to the killing effect of TSN, whereas GES-1 was the least sensitive. The results indicate that at a certain dosage range, TSN might selectively kill human GC cells rather than normal human gastric epithelial cells.

Since TSN exerted the most powerful cytotoxicity on SGC-7901 cells, we next used PI staining and flow cytometric analysis to determine if TSN influenced distribution of the cell cycle. As shown in Fig. 2A, the proportion of cells at G0/G1 phase increased while that at S phase decreased after treatment with TSN (0.5 and 1 µM) for 48 h in a dose-dependent manner. Then Annexin V/PI double staining and flow cytometric analysis were performed to assess the rate of apoptosis. As indicated in Fig. 2B, after treating for 48 h, TSN induced a dose-dependent increase of cells undergoing early apoptosis. Furthermore, invasion potential was determined by Transwell assay and the results showed that TSN suppressed cell invasive capacity also in a dose-dependent manner (Fig. 2C). The wound-healing assay confirmed the slower migration of SGC-7901 cells by TSN treatment (Fig. 2D). In addition, western blot analysis showed that expression of E-cadherin, which was an epithelial marker and reduced by TGF-β1 (10 ng/ml), was increased by TSN. TSN also led to significant downregulation of mesenchymal markers (vimentin and N-cadherin) expression elevated by TGF-β1 (Fig. 2E).

We hypothesized effects of TSN on SGC-7901 cells could be mediated by β-catenin signaling. β-catenin and its downstream target genes were assessed by qRT-PCR and western blotting. The β-catenin protein expression (including that in the nucleus) in SGC-7901 cells transfected by LV-CTNNB1 significantly increased comparing with those without transfection (Fig. 3A). TSN could significantly decrease the mRNA and protein levels of genes in β-catenin pathway (β-catenin, c-Myc, cyclin D1, Bcl-2 and MMP-9) in SGC-7901 cells in a dose-dependent manner and exogenous β-catenin overexpression abolished the effects (Fig. 3B and C). functionally, increasing the levels of β-catenin reversed the effects of TSN on biological behavior of SGC-7901 cells, including proliferation, cell cycle progression, apoptosis, invasion, migration and TGF-β1-induced EMT (Fig. 3D–I). These findings suggest that TSN might possess antitumor effects on SGC-7901 cells through repressing β-catenin signaling in vitro.

To further verify whether miR-200a mediates the inhibitory effects of TSN on β-catenin pathway, we established stable SGC-7901 cell lines transfected with LV-hsa-miR-200a-3p-inhibition in which the level of miR-200a was lower compared with that in SGC-7901 cells without transfection (Fig. 4A). After treating for 48 h, TSN dramatically elevated miR-200a expression which was abrogated by miR-200a silencing (Fig. 4A). More importantly, knockdown of miR-200a attenuated suppressive effects of TSN on mRNA and protein expression of β-catenin, c-Myc, cyclin D1, Bcl-2 and MMP-9 (Fig. 4B and C). Nevertheless, it is worth noting that miR-200a silencing did not influence protein level of upstream regulators of β-catenin (p-AKT, p-ERK and p-GSK3β) which were downregulated by TSN (Fig. 4C). Results also showed that the mRNA and protein level of AKT and ERK were not influenced by any of the above treatments. Knockdown of miR-200a did not affect protein level of GSK3β which raised after TSN treatment due to the reduction of p-GSK3β protein level (Fig. 4C). Altogether, these data strongly indicated that TSN suppresses β-catenin pathway by activating AKT and ERK signaling as well as facilitating miR-200a which might only target β-catenin and then influence its downstream genes.

To investigate the role of miR-200a in inhibitory effects of TSN on SGC-7901 cells, MTT, flow cytometry, Transwell, wound healing, and western blot analysis were used. Our results displayed that when cells transfected with CON137 (negative control for LV-hsa-miR-200a-3p-inhibition) were exposed to TSN (1 µM) for 48 h, significant cell cycle arrest at G0/G1 phase and apoptosis were found as well as decline of proliferation, invasion, migration and TGF-β1-induced EMT. Moreover, miR-200a silencing enabled SGC-7901 cells to overcome the above effects of TSN (Fig. 4D–I). Taken together, these findings supported that miR-200a mediated suppressive effect of TSN on β-catenin pathway and biological behaviors in SGC-7901 cells.

Given that TSN has antitumor effects on GC in vitro, potential effects of TSN on GC growth and liver metastasis in vivo were further investigated by establishing orthotopic transplantation GC model in nude mice. The results showed that mice administered by TSN had lower average volume and weight of GC tissues, less number of metastatic nodules in liver and lower liver weight than those of animals in control group. More importantly, in groups of models established with SGC-7901 cells transfected with LV-CTNNB1 or LV-hsa-miR-200a-3p inhibition, the above curative effect of TSN was significantly weakened (Fig. 5A–E). In addition, as indicated in Fig. 5f, IHC assay showed that TSN reduced the expression of p-AKT, p-ERK, p-GSK3β, β-catenin, c-Myc, cyclin D1, Bcl-2 and MMP-9 in GC tissues. Consistent with observations in vitro, miR-200a depletion abolished suppressive effects of TSN on the above proteins excluding p-AKT, p-ERK and p-GSK3β in tissues. These data provided evidence that TSN inhibits GC growth and liver metastasis via miR-200a/β-catenin axis in vivo.