The interaction between terphenyllin and STAT3 was investigated by molecular docking in this study. The structure of terphenyllin was drawn by ChemDraw 20.0. The crystal structure of STAT3 (PDB ID: 6TLC) was obtained from the RCSB Protein Data Bank (PDB, https://www.rcsb.org/). The binding sites between terphenyllin and STAT3 were predicted by AutoDock 4.2 software and the docking result was visualized by PyMol 2.5.

Human gastric cancer cell lines MKN1 and BGC823 were purchased from the American-type culture collection (ATCC). Both cell lines were cultured in RPMI 1640 medium at 37°C in an incubator containing 5% CO₂. 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin were added to cell culture media. The isolation and purification of terphenyllin were performed as previously (Wang et al., 2017). The structure of terphenylin was tested and verified by NMR, MS, UV, and IR spectroscopy. The purity of this compound was determined to be more than 98%. All antibodies were purchased from Cell Signaling Technology: Anti-STAT3 (Cat. No. 12640), Anti-pSTAT3 (Cat. No. 9145), Anti-c-Myc (Cat. No. 18583), Anti-Cyclin D1 (Cat. No. 55506), and Anti-GAPDH (Cat. No. 5174).

The methodology used here was described elsewhere (Wang et al., 2014a; Wang et al., 2018b). To test the cytotoxic effect of terphenyllin, we seeded 3,000 cells in each well of 96-well plates. The compound was given at the indicated concentrations for 24, 48, and 72 h, followed by a CCK8 assay (Biosharp, Anhui, China). For colony formation assay, we seeded 800 cells in each well of 6-well plates and treated cells with the compound at the specific concentrations for 24 h. The cells were kept in the plate for 7–10 days, followed by fixation and crystal violet staining. In the wound healing assay, we passaged enough cells and formed a confluent monolayer of cells. The cells were scratched using a pipette tip and then exposed to the indicated concentrations of terphenyllin, thus assessing the effects of terphenyllin on cell migration. Considering the effect of terphenyllin on cell invasion, 5 × 10⁴ cells were transferred to every upper well of a Boyden chamber (Corning, United States) and treated with terphenyllin for 24 h. The invaded cells were then stained with Crystal Violet Stain solution 2.5% (Solarbio, China) and the stained cells were photographed and counted.

Cell cycle assay and apoptosis assay were performed as described previously (Wang et al., 2014b; Wang et al., 2019). For cell cycle assay, the 3×10⁵ cells were passaged to each well of 6-well plates (3×10⁵ cells/well) and exposed to indicated concentrations of terphenyllin (0, 20, 50, and 200 μM) for 24 h. The cells were harvested and fixed in 95% ethanol at 4 °C for more than 2 h. The fixed cells were incubated and stained with propidium iodide/RNase staining buffer (BD Pharmingen, United States). Subsequently, the cells in different cell cycle phases were analyzed by LSRFortessa (BD Bioscience, United States). The data were analyzed by the ModFit LT software (Verity Software House, Switzerland). For apoptosis assay, 3 × 10⁵ cells were seeded into 6-well plates and then treated with specific concentrations of terphenyllin for 48 h. The treated cells were harvested, washed with PBS, and then re-suspended in binding buffer and Annexin V-FITC and propidium iodide from FITC Annexin V Apoptosis Detection Kit I (BD Pharmingen, United States). The inhibitory effects of terphenyllin on cell apoptosis were processed on a CytoFLEX LX (Beckman Coulter, United States).

Western blot was performed as described previously (Xue et al., 2017; J. J.; Qin et al., 2018). In brief, 3 × 10⁵ cells were seeded into 6-cm dishes and then treated with indicated concentrations of terphenyllin (0, 20, 50, and 200 μM) for 24 h. Subsequently, the cells were scraped and lysed in RIPA lysis buffer (Absin Bioscience Inc., Shanghai, China) containing protease inhibitor mixture (PMSF). After centrifugation for 20 min at 12,000 rpm, 4°C, the supernatants were collected. The concentration of the samples was then quantified by the BCA Protein Assay Kit (Absin, Shanghai, China). Equal amounts of protein were separated by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred onto nitrocellulose membranes (Bio-Rad, Hercules, CA, United States). The membranes were blocked with 5% milk for 2 h at room temperature. The membranes were washed three times with TBST (Tris-buffered saline, 0.1% Tween 20), followed by incubation with required primary antibodies overnight at 4°C. After washing three times with TBST, the membranes were incubated with a secondary antibody for 2 h at room temperature. The working dilution of each antibody involved in Western blot experiments is 1:1000. The membranes were scanned and visualized using the ImageQuant 800 (Amersham, United Kingdom).

The orthotopic gastric cancer mouse model was formed as described previously (Kang et al., 2021). Female four or 5 weeks nude mice were purchased from Shanghai Laboratory Animal Center. In brief, 100 µL of MKN1-Luc cell solution (5 × 10⁶ cells in a 1:1 mixture of Matrigel and serum-free medium) was slowly injected into the left axillary. When the diameter of the MKN1 subcutaneous tumor grew to 1 cm, we took out the tumor and removed the fat and necrotic tissue of the tumor. Then we cut the fresh tumor tissue into a side length of 1-2 mm that were placed on the side of the great curvature of the stomach. Fifteen seconds later, we put the stomach back into the abdominal cavity, put a little powder antibiotic into the abdominal cavity, and suture the abdominal muscle layer and skin layer in turn. Terphenyllin was dissolved in PEG400:ethanol:saline (4:1:2, v/v/v) and administered to mice by intraperitoneal injection at a dose of 10 mg/kg/day, 7 days/week for 5 weeks. For in vivo imaging, mice were administered intraperitoneally with fluorescein substrate (150 mg/kg) and anesthetized with isoflurane using an anesthesia machine (Summit Anesthesia, United States). The images for orthotopic tumor growth and metastasis were detected from a Xenogen IVIS 200 imaging system (Caliper Life Sciences, United States). All the animal data was processed using LT Living Image 4.3 Software. At last, we tested the tumor metastasis to other organs in every experimental rice.

The H&E staining followed the previous paper (Voruganti et al., 2015). At the end of the animal experiments, various major organs including liver, lungs, kidneys, spleen, heart, and brain, were removed from the tumor-bearing mice, fixed in 10% formalin, and embedded in paraffin. These tissue blocks were handled and sectioned at a thickness of 5 µm. These tissue sections were deparaffinized with xylenes, rehydrated, washed using PBS, stained in Mayer’s Hematoxylin for 10 min, and then stained with eosin for less than 1 min. Following staining, the slides were dehydrated, mounted, and detected using an inverted microscope (Axio Observer A1, Zeiss, Germany).

Data were presented as the mean ± SD from three independent experiments and processed with the Prism software version 5 (Graph Pad Software Inc., San Diego, CA, United States). The significant differences between the two groups were compared by t-test. ∗ denotes p < 0.05, ∗∗ denotes p < 0.01, ∗∗∗ denotes p < 0.001.

To identify small molecule inhibitors for STAT3, we performed a molecular docking and structure-based virtual screening assay against a library of natural products using STAT3 as a substrate, and terphenyllin was found to bind to STAT3. As shown in Figures 1B, C, the hydrogen atom of the phenolic hydroxyl group of terphenyllin forms a hydrogen bond with the oxygen atom of Glu612 side-chain amide. To investigate the effect of terphenyllin on the STAT3 signaling pathway, we test the inhibitory effect of terphenyllin on STAT3 protein in gastric cancer cells. We found that terphenyllin concentration-dependently suppressed the protein level of p-STAT3 (Tyr705) but had no obvious influence on the level of total STAT3 (Figures 1D, E). Also, terphenyllin inhibits the expression levels of Cyclin D1 and c-Myc, the downstream target genes of STAT3 (Figures 1D, E). In conclusion, the results above revealed that terphenyllin blocked the STAT3 signaling pathway in both MKN1 and BGC823 gastric cancer cell lines.

CCK8 assay was performed to examine the effect of terphenyllin on the viability of gastric cancer cells. The results demonstrated that terphenyllin showed obvious inhibitory effects on the viability of gastric cancer cells especially when the treatment concentration was higher than 100 μM (Figures 2A, B). When the treatment time reached 72 h, the IC₅₀ of terphenyllin against MKN1 and BGC823 cell lines reduced to 35.5 and 39.9 μM, respectively. We also explored the anticancer activity of terphenyllin using colony formation assay. Terphenyllin decreased the colony numbers of MKN1 and BGC823 cell lines in a concentration-dependent manner (Figures 2C, D). When the treatment concentration came to 20 μM, already inhibited by terphenyllin. We could hardly see any colony formation in both 2 cell lines at the highest concentration.

To test the effect of terphenyllin on the cell cycle progression, the cell cycle distribution of both MKN1 and BGC823 cell lines was measured by flow cytometric assay. It was seen that the proportions of G1 and G2 cells did not show remarkable changes among these four groups but the proportions of terphenyllin-treated groups in the S phase significantly increased, compared with the DMSO groups (Figures 3A, B). Terphenyllin treatment-induced cell cycle arrest was probably related to the down-regulation of cell cycle regulation proteins (such as Cyclin D1 and c-Myc), which was consistent with our western blot results. As for the assessment of the pro-apoptotic influence of terphenyllin on both cell lines, the apoptotic rates were detected by the FITC Annexin V apoptosis detection kit. As shown in Figure 3C, the proportion of apoptotic cells significantly increased in a concentration-dependent manner. In the highest concentration group, the percentage of apoptotic cells in both 2 cell lines was around 50%. Furthermore, it could be discovered that the apoptotic rate of BGC823 is much higher than that of MKN1, which implied that BGC823 cells might be more sensitive to terphenyllin than MKN1 cells.

To determine the effects of terphenyllin on the migration and invasion of MKN1 and BGC823 cell lines, the wound-healing assay and transwell invasion assay were performed. As shown in Figures 4A, B, MKN1 and BGC823 cells in the control group both migrated into the entire wounded area by 48 h, whereas treatment with terphenyllin at specific concentrations (5 and 10 μM) prominently inhibited the cell migration. It was observed that the wound of the MKN1 cells was much easier to heal than of the BGC823 cells at a 10 μM treatment concentration. The results of the invasion assay in Figure 4C showed that compared with the control group, the invasive ability of MKN1 and BGC823 cells in the terphenyllin-treated groups was significantly decreased in a dose-dependent manner. Compared with BGC823 cells, MKN1 cells showed stronger invasive ability according to the number of cells invaded.

Owing to the inhibition of proliferation of gastric cancer cells after treatment with terphenyllin, we next tried to examine whether it could slow the tumor growth in vivo. We established an orthotopic gastric cancer mouse model and the tumor-bearing nude mice were randomly divided into two groups: vehicle group and terphenyllin treatment group. As expected, intraperitoneal injection of terphenyllin caused significant inhibition of gastric tumor growth in the orthotopic model (Figures 5A,B). The rate of tumor growth inhibition is 67.2% compared with the vehicle group. There was no significant difference in body weight between these two groups, revealing that terphenyllin had no obvious host toxicity (Figure 5C). The pathological changes in the vital organs were also evaluated using H&E staining. Compared the vehicle group with the treatment group, no pathological changes were detected, indicating that terphenyllin did not cause damage to the organs of mice (Figure 5D). Collectively, terphenyllin suppressed the growth of tumors and did not show significant side effects in vivo.

To test the effect of terphenyllin on tumor metastasis, vital organs were taken out separately for fluorescence imaging. Terphenyllin showed an obvious inhibitory effect on the metastatic lesions in the liver, spleen, and peritoneum while comparing with the vehicle group. The statistical results indicated that 6, 2, and six out of eight vehicle-treated mice showed metastatic lesions in the liver, spleen and peritoneum, respectively, while the incidence of liver, spleen and peritoneal metastasis in terphenyllin-treated mice was reduced to 2/8, 1/8, and 2/8, respectively (Figures 6A–C). Moreover, there were far more mice without metastasis in the terphenyllin-treated group than in the vehicle group (Figure 6C).