Investigation of Inhibition Effect of Gossypol-Acetic Acid on Gastric Cancer Cells Based on a Network Pharmacology Approach and Experimental Validation.

Gastric cancer (GC) is one of the most common malignant tumors worldwide. 1 In 2018, approximately one million people were diagnosed with GC and the number of deaths due to GC was about 800,000. 2 Although many treatment options are available at present, such as fluorouracil, oxaliplatin, and paclitaxel, 1 the prognosis of GC is still not optimistic. 3 GC represents a serious threat to human health. Therefore, it is critical to develop new and effective treatments for GC patients. In recent years, “conventional drug in new use” has become a research hotspot, such as aspirin, 4 sildenafil, 5 , 6 gossypol 7 and zidovudine. 8 GAA is a commonly-used male contraceptive agent 9 that exhibits significant anti-tumor effects. Previous studies showed that GAA inhibited tumor growth primarily by regulating cell cycle progression and promoting apoptosis. 10 – 12 However, it was unclear whether GAA would exhibit an inhibitory effect on GC cells. A network pharmacology approach is an emerging method based on the “disease-genes-drug” network. It has been widely used to improve the efficiency of drug discovery. Zheng et al uncovered the mechanism underlying the therapeutic effects of herbs on endometriosis by using the network pharmacology approach. 13 Yang et al reported the mechanism of Physalis against acute lung injury by using the same method. 14 The mechanism of geraniol was also unrevealed via network pharmacology. 15 Compared with traditional experimental methods, network pharmacology has comprehensive and systematic characteristics. In this study, we predicted potential binding targets of GAA, identified GC-related genes, constructed the “GC-targets-GAA” network, and explored the mechanism underlying the regulatory role of GAA in GC by using the network pharmacology approach. Then biological experiments were performed to verify the results obtained from the network pharmacology approach. Our findings showed that GAA inhibited the proliferation of GC cells, which may provide a theoretical basis for the development of novel drugs for GC in the future.

We obtained the 3D structure of GAA from the PubChem database ( https://pubchem.ncbi.nlm.nih.gov/ ) ( Figure 1A ). Then we predicted potential binding genes of GAA using the PharmMapper platform. Finally, 282 genes were predicted to be GAA-binding genes ( Table 1S ). A total of 11,352 GC-related genes were obtained from the GeneCard database ( https://www.genecards.org ) ( Table 2S ). As shown in Figure 1B , there were 247 intersection genes ( Table 3S ), through which GAA might regulate the progression of GC. To further investigate the role of GAA in GC, we constructed a “GC-targets-GAA” network using the Cytoscape software ( Figure 1C ). Figure 1 Network pharmacology of GAA and gastric cancer. ( A ) The 3D structure of GAA. ( B ) The Venn map of gastric cancer - related genes and GAA - target genes. ( C ) The “gastric cancer - targets - GAA” network. Network pharmacology of GAA and gastric cancer. ( A ) The 3D structure of GAA. ( B ) The Venn map of gastric cancer - related genes and GAA - target genes. ( C ) The “gastric cancer - targets - GAA” network. To explore the relationship between GAA and GC, we constructed a PPI network of 247 intersection genes by using the String database ( https://string-db.org/ ). The whole PPI network contained 247 nodes and 427 interaction relationships (interaction score > 0.96) ( Figure 2A ) ( Table 4S ). Then we imported this network into the Cytoscape software and identified 10 hub genes: GRB2 , HRAS , AKT1 , PTPN11 , SRC , HSP90AA1 , EGFR , MAPK1 , PTPN1 , and RXRA ( Figure 2B ). These genes may play important roles in the whole PPI network. GRB2 , HRAS , AKT1 , and EGFR have always been of great interest to researchers. Figure 2 The construction of PPI network. ( A ) The PPI network of intersection genes from Figure 1B . ( B ) The hub genes of PPI network. The darker the red color, the higher the connection degree. The construction of PPI network. ( A ) The PPI network of intersection genes from Figure 1B . ( B ) The hub genes of PPI network. The darker the red color, the higher the connection degree. To elucidate the potential mechanism underlying the regulatory effect of GAA on GC, we performed GO and KEGG pathway analyses on 247 intersection genes by using the DAVID software. The GO functional enrichment results showed that GAA was related to cytosol, extracellular exosome, lysosome, mitochondrion, spindle microtubule, cytoskeleton, vesicle, and nuclear chromatin ( Figure 3B ) ( Table 5S ). Moreover, GAA may play a regulatory role in receptor binding, enzyme binding, oxidoreductase activity, and MAP kinase activity ( Figure 4B ) ( Table 6S ). GAA may also participate in protein phosphorylation, apoptosis, drug response, immunity, cell proliferation, cell migration, and DNA synthesis ( Figure 3A ) ( Table 7S ). The results of KEGG pathway analysis showed that these intersection genes were mainly enriched in the Ras signaling pathway, FoxO signaling pathway, PI3K-Akt signaling pathway, T cell receptor signaling pathway, and ErbB signaling pathway ( Figure 4A ) ( Table 8S ). The findings suggested that GAA might regulate GC progression through multiple cancer-related pathways, such as the Ras signaling pathway, PI3K-Akt signaling pathway, VEGF signaling pathway, HIF-1A signaling pathway, and metabolic pathway. Among them, the PI3K-Akt signaling pathway was the highest-ranked pathway ( Figure 5 ). Figure 3 GO analysis of intersection genes. ( A ) Biological processes. ( B ) Cellular components. Figure 4 GO and KEGG pathway analysis of intersection genes. ( A ) KEGG pathway analysis. ( B ) Molecular functions. Figure 5 The PI3K-AKT signing pathway. The nodes marked in red were related to intersection genes. GO analysis of intersection genes. ( A ) Biological processes. ( B ) Cellular components. GO and KEGG pathway analysis of intersection genes. ( A ) KEGG pathway analysis. ( B ) Molecular functions. The PI3K-AKT signing pathway. The nodes marked in red were related to intersection genes. Next, we performed biological experiments to verify the anti-tumor effect of GAA on GC. The MTS proliferation assay showed that GAA suppressed the proliferation of GC cells in a concentration-dependent manner ( Figure 6A and B ). The plate cloning test showed that GAA dose-dependently inhibited the colony formation of GC cells ( Figure 6C and D ). These results confirmed that GAA inhibited the proliferation of GC cells. Figure 6 The effect of GAA on the proliferation of gastric cancer cells. ( A ) Survival rate of gastric cancer cells treated for 48 hours by GAA of gradient concentrations. ( B ) Survival rate of gastric cancer cells treated for 72 hours by GAA of gradient concentrations. ( C ) Colony formation experiments of SGC-7901 treated for 24 hours by GAA of gradient concentrations. ( D ) Colony formation experiments of MGC-803 treated for 24 hours by GAA of gradient concentrations. ** P < 0.01, *** P < 0.001. The effect of GAA on the proliferation of gastric cancer cells. ( A ) Survival rate of gastric cancer cells treated for 48 hours by GAA of gradient concentrations. ( B ) Survival rate of gastric cancer cells treated for 72 hours by GAA of gradient concentrations. ( C ) Colony formation experiments of SGC-7901 treated for 24 hours by GAA of gradient concentrations. ( D ) Colony formation experiments of MGC-803 treated for 24 hours by GAA of gradient concentrations. ** P < 0.01, *** P < 0.001. We hypothesized that GAA inhibited the proliferation of GC cells by promoting apoptotic cell death or regulating cell cycle progression. The above results showed that GAA at 10 μM efficiently inhibited GC cell proliferation. Therefore, a dosage of 10 μM was used for subsequent experiments. The apoptosis rate was also increased in GAA-treated cells compared to the SGC-7901 group (13.7% vs. 7.6%) ( Figure 7A and B ). A similar result was observed in cells treated with MGC-803 ( Figure 7A and B ). As shown in Figure 8A , GAA treatment increased the proportion of cells in G1 phase compared to the SGC-7901 control group (51.3% vs. 42.6%). A similar result was achieved by exposing cells to MGC-803 ( Figure 8B ). Taken together, these data showed that GAA induced apoptotic cell death arrest and G1 phase, thus affecting GC cell proliferation. Figure 7 The effect of GAA on the apoptosis of gastric cancer cells. ( A ) Statistical analysis on chart B. ( B ) GAA induced the apoptosis of gastric cancer cells. *** P < 0.001. Figure 8 The effect of GAA on the cell cycle of gastric cancer cells. ( A ) GAA induced G1 phase arrest of SGC-7901. ( B ) GAA induced G1 phase arrest of MGC-803. *** P < 0.001. The effect of GAA on the apoptosis of gastric cancer cells. ( A ) Statistical analysis on chart B. ( B ) GAA induced the apoptosis of gastric cancer cells. *** P < 0.001. The effect of GAA on the cell cycle of gastric cancer cells. ( A ) GAA induced G1 phase arrest of SGC-7901. ( B ) GAA induced G1 phase arrest of MGC-803. *** P < 0.001.

SGC-7901 and MGC-803 were purchased from ATCC (American Type Culture Collection). The cell lines were complemented in DMEM supplemented with 10% fetal bovine serum (HyClone, USA). Cell lines were cultured at 37°C in 5% CO2. The 3D structure of GAA was obtained from the PubChem database ( https://pubchem.ncbi.nlm.nih.gov/ ). The SDF file of the 3D structure was imported into the Pharmmapper software for target prediction. GC-related genes were obtained from the GeneCard database ( https://www.genecards.org/ ). The predicted drug targets and GC-related genes were imported into the jvenn database ( http://jvenn.toulouse.inra.fr/app/example.html ) to obtain the intersection genes. The “GC-targets-GAA” network was constructed using the Cytoscape software. The PPI analysis of these intersection genes was performed using the string software. In this study, data with a high confidence level (>0.96) were selected to construct the network. The Cytoscape software was used to visualize and analyze the network. The GO and KEGG pathway enrichment analyses of the intersection genes were performed by using the DAVID database ( https://david.ncifcrf.gov/ ). P < 0.05 was set as the threshold to screen biological processes. The 50 top-ranked biological processes were shown using the R software. In MTS proliferation test, GC cells at a density of 8×10 3 cells per well were seeded into 96-well plates. After adhering to the wall, cells were treated with GAA at different concentrations for 0, 24, 48, 72, and 96 h. Then the OD values were detected by MTS. In plate cloning test, GC cells at a density of 2×10 4 cells per well were inoculated into 12-well plates and then treated with GAA at different doses for 24 h. Then cells were fixed with 75% ethanol, stained with crystal violet, and photographed. In cell cycle assay, cells were plated into 6-well plates. After adhering to the wall, cells were treated with GAA at various doses for 12 h. Then cells were digested with trypsin, centrifuged, and fixed with precooled 75% ethanol overnight. On the next day, cells were washed three times with precooled PBS. Then cells were incubated with PI in the dark for 30 minutes. The cell cycle was detected by flow cytometry (BD Biosciences, USA). In apoptosis assay, apoptotic cells were inoculated in 6-well plates and then treated with GAA at different concentrations for 20 h. After trypsin digestion and centrifugation, cells were treated with the reagents in the apoptosis kit (Neobioscience, Shenzhen, China). The apoptosis rate was determined by flow cytometry. The experimental results were analyzed by Student’s t -test (unpaired, two tailed). P <0.05 were considered to be significant. All statistical analyses were performed using Prism5 (GraphPad Software Inc., La Jolla, CA).

The development of conventional drugs as therapeutic strategies for cancers has become a hotspot in research. Conventional drugs are safer than new ones because they have been used for years. Moreover, the use of conventional drugs significantly reduces the cost and time of drug discovery. 16 GAA is a male contraceptive agent that has been used for the treatment of gynecological diseases, such as uterine leiomyoma and endometriosis. It was reported to have antimalarial 17 and anti-tumor activities. 18 – 20 However, the role of GAA in GC had not been reported. The network pharmacology approach is an effective method to explore the relationships of drug, targets, and disease, which may reveal drug mechanisms of action to some extent. In view of the poor prognosis of GC, we explored whether GAA could inhibit GC cell proliferation using the network pharmacology method. First, the “GC-targets-GAA” network was constructed. Then the PPI network analysis of these targets was performed and 10 hub genes that may be involved in the regulation of GC by GAA were identified. The GO and KEGG pathway analyses showed that GAA regulated cell cycle progression and apoptosis in GC cells and might also mediate many cancer-related signaling pathways. As the bioinformatics analysis may not reflect the effect of GAA on GC cells, we then performed biological experiments to verify these results. The MTS proliferation test, plate cloning test, cell cycle and apoptosis assays confirmed that GAA inhibited the proliferation of GC cells by inducing G1 phase arrest and apoptosis. The limitation of the current study was that we did not verify the mechanisms by which GAA inhibited GC in vivo. Further studies using animal models are needed to investigate the effects of GAA on GC progression. In conclusion, our study showed the inhibitory effect of GAA on GC cell growth by using the network pharmacological method and biological experiments. These results provided a theoretical basis for further investigation of GAA in clinical trials and suggested the potential use of GAA as a novel therapeutic agent for the treatment of GC.