Network pharmacology- and molecular docking-based exploration of the molecular mechanism underlying Jianpi Yiwei Recipe treatment of gastric cancer

Abstract

BACKGROUND

Traditional Chinese medicine (TCM) is widely used as an important complementary and alternative healthcare system for cancer treatment in Asian countries. Network pharmacology, which utilizes various database platforms and computer software to study the interactions between complex drug components in vivo, is particularly useful for studying the pharmacodynamic mechanisms of multi-pathway and multi-target Chinese medicines.

AIM

To explore the potential targets and function of Jianpi Yiwei Recipe treatment of gastric cancer (GC) through network pharmacology and molecular docking.

METHODS

Data on the components of Jianpi Yiwei Recipe (Radix Astragali, Radix Codonopsis, Agrimonia eupatoria, Atractylodes macrocephala Koidz., Poria cocos, stir-baked rhizoma dioscoreae, Amomum villosum Lour., fried Fructus Aurantii, pericarpium citri reticulatae, Rhizoma Pinelliae Preparata, and Radix Glycyrrhizae Preparata) were collected and screened by using the TCM systems pharmacology database and analysis platform (TCMSP). Then the targets of these compounds were predicted. GC-related targets were screened using the GeneCards database. Venn diagram was used to identify common targets. An active ingredient-core target interaction network and a protein-protein interaction (PPI) network were built. Moreover, we performed gene ontology (GO) functional annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses on the core targets and validated them by molecular docking.

RESULTS

TCMSP screening revealed 11 active components and 184 targets, whereas GeneCards found 10118 disease-related targets, with 180 shared targets between them. Topology analysis of the PPI network identified 38 targets, including ATK1, TP53, and tumor necrosis factor, as key targets for the treatment of GC by Jianpi Yiwei Recipe. Quercetin, naringenin, luteolin, etc., may be the main active components of Jianpi Yiwei Recipe. GO enrichment analysis identified 2809, 1218, and 553 functions related to biological process, molecular function, and cellular component, respectively. KEGG pathway enrichment analysis revealed 167 related pathways, mainly involved in cancer, endocrine resistance, and AGE-RAGE signaling in diabetic complication. Validation with molecular docking analysis showed docking of key active components with core targets.

CONCLUSION

Jianpi Yiwei Recipe plays a therapeutic role in GC through multiple components, targets, and pathways. These findings form a basis for follow-up exploration of Jianpi Yiwei Recipe in the treatment of GC.

Key Words: Jianpi Yiwei Recipe, Gastric cancer treatment, Network pharmacology, Key target, Molecular docking

Core Tip: Traditional Chinese medicine plays an important complementary and alternative medical system, natural anticancer compounds were shown to have good efficacy, few side effects, and low toxicity, making them effective in anti-tumor therapy. This study revealed that Jianpi Yiwei Recipe plays a therapeutic role in gastric cancer through multiple components, targets, and pathways through network pharmacology and molecular docking.

INTRODUCTION

Gastric cancer (GC), one of the most prevalent gastrointestinal malignancies, is derived from gastric mucosal epithelial cells, with high incidence and poor prognosis[1]. Approximately 70% of the patients with GC are diagnosed as advanced due to atypical early symptoms, which greatly limits the efficacy of surgery and radiotherapy[2]. Therefore, it is necessary to explore treatment strategies that reduce cancer-related mortality and improve patient survival. GC is mostly treated with chemotherapy. In the past few decades, patients with advanced GC have received great clinical benefits with conventional chemotherapy, such as cisplatin and 5-fluorouracil[3]. However, the clinical efficacy of conventional chemotherapy has gradually deteriorated due to drug resistance and cytotoxicity[4]. Natural anticancer compounds were shown to have good efficacy, few side effects, and low toxicity, making them effective in anti-tumor therapy[5]. Therefore, the study of natural anticancer drugs is of great significance for cancer therapy and patient outcomes.

For thousands of years, traditional Chinese medicine (TCM), an important complementary and alternative medical system, has been widely used for cancer treatment in Asian countries (China and Japan in particular)[6]. TCM relieved clinical surgery- and chemotherapy-induced fatigue, pain, vomiting, diarrhea, pancytopenia, etc., and mitigated tumor-associated symptoms, improving immune function, and providing survival benefits[7-10]. Although “gastric cancer” has no specific name in ancient Chinese medicine books, GC was classified as “abdominal mass”, “accumulation”, “epigastric pain”, “regurgitation”, and “dysphagia”, according to clinical manifestations[11]. TCM believes that the disease is located in the spleen and stomach, and pathogenesis is an intermingling of deficiency and excess, with deficiency of the spleen and stomach being the source and phlegm-dampness, heat-toxicity, and blood stasis being the signs. TCM attaches importance to the role of healthy Qi in the treatment of diseases, and emphasizes on the nourishment of the whole body using Qi acquired by the spleen and stomach, to improve the normal physiological functions of the human viscera[12]. Therefore, based on the principle of invigorating Qi and spleen, eliminating dampness, and regulating the stomach, this study chose to examine Jianpi Yiwei Recipe as the therapeutic drug. Radix Astragali, Radix Codonopsis, and Agrimonia eupatoria are monarch drugs in the prescription, which can replenish Qi, nourish the blood, and strengthen resistance to fight cancer, taking both the symptoms and root causes into consideration. A. macrocephala Koidz. and Poria cocos are minister drugs, which invigorate the spleen and improve Qi, in line with the spleen deficiency pathogenesis of GC. Stir-fried Fructus Aurantii, Amomum villosum Lour., pericarpium citri reticulatae, and Rhizoma Pinelliae Preparata invigorate the spleen, promote circulation, and regulate the ascending and descending Qi of the spleen and stomach. Modern pharmacological studies have found that the use of TCMs in invigorating the spleen and benefiting Qi can enhance the body’s immunity, reduce the side effects of chemoradiotherapy, inhibit tumor cell growth, improve anticancer efficacy, and prolong patient survival[13].

Network pharmacology uses various database platforms and computer software to build a network displaying the relationship between drug components, action targets, and diseases, as well as to study the interaction between complex drug components in the body. This is especially useful for studying the pharmacodynamic mechanism of TCM with multiple pathways and targets[14]. When studying the action of TCM on the human body, the introduction of network pharmacology makes it easier and more intuitive to understand the process of drug action, providing a basis for optimizing drug use. Therefore, our research uses network pharmacology and molecular docking to analyze the function of Jianpi Yiwei Recipe in the treatment of GC. We aim to provide a direction for exploration of GC treatment by Jianpi Yiwei Recipe and shedding light on the clinical treatment of GC.

MATERIALS AND METHODS

Active ingredient acquisition and screening

The active ingredients of Jianpi Yiwei Recipe, including Radix Astragali, Radix Codonopsis, A. eupatoria, A. macrocephala Koidz., P. cocos, stir-baked rhizoma dioscoreae, A. villosum Lour., Fried Fructus Aurantii, pericarpium citri reticulatae, Rhizoma Pinelliae Preparata, and Radix Glycyrrhizae Preparata, were searched using TCM systems pharmacology database and analysis platform (TCMSP, https://tcmspw.com). Stir-baked rhizoma dioscoreae, fried Fructus Aurantii, Rhizoma Pinelliae Preparata, and Radix Glycyrrhizae Preparata were excluded if they were not included. Attributes such as absorption, distribution, metabolism, and excretion were screened using oral bioavailability ≥ 30% and drug-likeness ≥ 0.18.

Action target acquisition

After obtaining effective active components from relevant TCMs, we used TCMSP to obtain their corresponding targets, sorted the data, and deleted repetitive and unidentified components. The targets were imported into the UniProt database (https://www.uniprot.org/) one by one. Target protein names (official symbol) and UniProt ID numbers were retrieved, with retrieval condition Organism: Homo sapiens (Human). Unverified target genes were removed during the research.

Prediction of GC-related targets

The targets were searched on the Human Gene Database (GeneCards, http://www.genecards.org/) with the keywords “gastric cancer”, “gastric carcinoma”, or “stomach neoplasms”, where relevance score was greater than the median and the category was protein coding. All targets were combined and GC targets were obtained after deduplication. To clarify the relationship between the related targets of Jianpi Yiwei Recipe and the targets of GC, R software (v4.0.3) was used to match the two groups of targets, and intersecting target genes were obtained to plot a Venn diagram.

Drug-active ingredient-target network building

A multilayer network of relationships between active ingredients and their targets was built after inputting the drug, active ingredients, and targets into Cytoscape v3.8.2 (http://www.Cytoscape.org). Topology analysis was performed with the components ranked by betweenness; the top three components screened were considered key components.

Protein-protein interaction

Common drug targets of the drugs and disease were input into the STRING database (https://string-db.org/) to construct protein-protein interaction (PPI) networks. The organism was set to “Homo sapiens.” Other parameters were left as defaults to obtain a PPI network diagram. The obtained information was imported into Cytoscape for core gene screening by topology analysis of the PPI network using the NetworkAnalyzer plug-in.

Gene ontology-based functional annotation and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis

The core action targets obtained in 1.5 were analyzed using OmicShare Tools (http://www.omicshare.com/tools) for gene ontology (GO)-based functional annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. The organism was set to “human,” and biological process (BP), cellular component (CC), molecular function (MF), and KEGG were enriched and analyzed to identify major biological functions and signal pathways, with corrected P value < 0.05 as the screening threshold.

Molecular docking validation

To validate the accuracy of possible key targets of Jianpi Yiwei Recipe in treating GC, the top-ranked active ingredients sorted in 1.4 and 1.5 were molecularly docked with the drug targets. The 2D structures of the active ingredients’ small molecule ligands were downloaded using TCMSP. The 3D structures of target protein receptors were downloaded from PDB (http://www.rcsb.org). PDB ID screening referred to the literature and selected crystal structures with high resolution. These data were imported into the PyMOL software (https://pymol.org/2/), and the command was input to delete water molecules and primary ligand molecules of the target proteins to complete protein receptor preparation. After processing, such as hydrogenation, using the AutoDock Tools software and ligand separation, the genes were saved in pdbqt format. The key active components of the drugs were saved in pdbqt format after processing. The AutoDock Vina software was utilized for molecular docking, and docking conformation was visually analyzed using the PyMOL software.

RESULTS

Acquisition and screening of active ingredients

A search of the TCMSP database using keywords “Astragalus membranaceus, Codonopsis pilosula, A. eupatoria, Atractylodes macrocephala Koidz., Poria cocos, Amomum villosum Lour. and pericarpium citri reticulatae”, identified 1207 drug targets: 462 for A. membranaceus, 214 for C. pilosula, 305 for A. eupatoria, 23 for A. macrocephala Koidz., 30 for P. cocos, 78 for A. villosum Lour., and 95 for pericarpium citri reticulatae. Of these, 5 validated drugs with 514 targets were identified. A total of 184 targets were obtained after normalized annotation with the UniProt database and deletion of repeated targets (Table 1).

Table 1 Information of “Traditional Chinese medicine-active ingredients-disease targets”.

Serial number	Drug	Active ingredient		Target
1	Astragalus membranaceus	4	MOL000098: Quercetin	127	180
			MOL000354: Isorhamnetin	5
			MOL000392: Formononetin	9
			MOL000422: Kaempferol	39
2	Codonopsis pilosula	2	MOL000006: Luteolin	48	51
			MOL008400: Glycitein	3
3	Agrimonia eupatoria	5	MOL000006: Luteolin	48	230
			MOL000098: Quercetin	127
			MOL000422: Kaempferol	39
			MOL000492: (+)-Catechin	2
			MOL001002: Ellagic acid	14
4	Amomum villosum Lour.	1	MOL000358: Beta-sitosterol	10	10
5	Pericarpium citri reticulatae	2	MOL004328: Naringenin	30	43
			MOL005828: Nobiletin	13

Acquisition of intersecting targets of active ingredients and GC targets

GC targets (13730, 12645, and 21881) were retrieved by searching “gastric cancer”, “gastric carcinoma”, and “stomach neoplasms” in the GeneCards database, respectively. Considering that many targets were retrieved, we sorted targets from high to low according to relevance score, screened targets larger than median, and selected targets in the category of protein coding. The targets were merged and deduplicated, returning 10118 targets. Finally, 180 active ingredient targets of Jianpi Yiwei Recipe and GC targets were obtained by using R language and Venn plug-in. See Figure 1 for the Venn diagram.

Figure 1 Venn diagram of drug targets and disease targets. In total, 180 overlapping genes were identified between gastric cancer-related genes and drug-related genes.

Construction of the active ingredient-core target interaction network

The collected data were sorted and imported into Cytoscape 3.8.2 to map the active ingredient-core target interaction network of Jianpi Yiwei Recipe for GC. The rectangles, rhombuses, and ellipses in the Figure 2 represent the drugs, active ingredients, and target genes, respectively. Active ingredients with the top 3 betweenness values obtained by CytoNCA plug-in were quercetin (QT; MOL000098, betweenness = 27401.44), naringenin (NG; MOL004328, betweenness = 7802.55), and luteolin (LL; MOL000006, betweenness = 6709.474).

Figure 2 Active ingredient-core target interaction network. Green, orange, and blue indicate drug, active ingredient, and targeted gene, respectively.

Target PPI network

The PPI network was constructed by uploading the 180 intersecting target genes obtained from the STRING platform and visualized with Cytoscape (Figure 3). The network had 178 nodes and 3983 edges (mean degree for each node: 44.300, mean local clustering coefficient: 0.667, and PPI enrichment P value < 1.0e-16). The more the adjacent nodes, the more significant the role of these proteins in the network. Data from tsv files were imported into Cytoscape 3.8.2 software, and the NetworkAnalyzer plug-in was used for topology analysis of the PPI network to screen key genes. We obtained 38 targets in the core position, suggesting their key role in PPIs. See Figure 4 for the core genes (in yellow).

Figure 3 Protein-protein interaction network of intersecting genes.

Figure 4 Screening of key genes by topology analysis of the protein-protein interaction network. Pink indicates targeted gene, yellow indicates key genes identified using the NetworkAnalyzer plug-in.

GO functional annotation and KEGG pathway enrichment analysis

The core genes obtained in 2.4 were used for GO functional enrichment analysis using OmicShare tool, mainly including BP, CC, and MF analyses. BPs in which these components played a role mainly involved biological regulation, cellular process, and metabolic process. MFs mainly included binding, MF regulator, and catalytic activity. CCs mainly acted on cellular anatomy and protein-containing complex (Figure 5A). KEGG pathway enrichment analysis revealed that 167 pathways were enriched according to P value and enriched gene number. The top 20 pathways are presented in Figure 5. These include pathways in cancer, endocrine resistance, and AGE-RAGE signaling in diabetic complication. Bubble color represents the significance of enrichment, the redder the color, the higher is the enrichment degree. Bubble size is indicative of the number of genes enriched in this pathway, with a larger bubble suggesting a greater number of genes enriched.

Figure 5 Kyoto Encyclopedia of Genes and Genomes and gene ontology enrichment of key genes. A: Gene ontology enrichment analysis, including biological process, cellular component, and molecular function; B: Top 20 Kyoto Encyclopedia of Genes and Genomes pathway enrichment.

Molecular docking validation

To validate the relationship between the active components of Jianpi Yiwei Recipe and GC targets and explore the mechanism of drug-target action, we mapped using PyMoL, according to the result data of 2.3 and 2.4, and molecularly docked the main active components (QT, NG, and LL) with the top three key targets ranked by degree [AKT1, TP53, and tumor necrosis factor (TNF)]. A lower binding energy between the receptor and the ligand suggests more stable binding conformation and a greater possibility of interaction. We found that the binding energy of the active components to the core targets was negative, indicating that the compounds and receptors had binding activity (Table 2). Figure 6 shows the molecular docking diagram of QT with various targets.

Figure 6 Molecular docking of quercetin with various targets. A: Action mode of quercetin with target AKT1 (PDB ID: 1UNQ); B: Action mode of quercetin with target TP53 (PDB ID: 1A1U); C: Action mode of quercetin with target tumor necrosis factor (PDB ID: 7ATB). TNF: Tumor necrosis factor.

Table 2 Docking of core active ingredients to key targets.

Target	PDB ID	Binding energy to the target/(kJ/mol)
		Quercetin	Naringenin	Luteolin
AKT1	1UNQ	-3.20	-2.92	-2.94
TP53	1A1U	-3.95	-4.05	-3.86
TNF	7ATB	-2.95	-3.77	-3.32

TNF: Tumor necrosis factor.

DISCUSSION

The occurrence of GC is primarily the result of a combination of factors, such as the environment, the host, and Helicobacter pylori (H. pylori) infection[15]. Spleen-invigorating and Qi-invigorating Chinese medicines have been extensively used for GC treatment. In GC, the curative effect of ginseng-containing prescriptions with Western medicine is significantly better than that of Western medicine alone[16]. Through network pharmacology and molecular docking, this study mainly explores the molecular role of Jianpi Yiwei Recipe in treating GC.

In total, 11 active components with 184 targets were screened by network pharmacology, of which QT, NG, and LL were identified as the key active components. QT, a natural occurring flavonol in food[17], is used as a dietary supplement and can serve as an anticancer agent in daily diet[18]. The intake of a tolerable dose of QT may have beneficial biological effects, such as anti-inflammatory, antioxidant, immune regulatory, etc[19]. QT was effective in inhibiting tumor progression through multiple mechanisms of action, while exerting strong anti-proliferation and pro-apoptosis effects in various GC cells. For example, the suppression of uPA/uPAR by QIT may be correlated with the inhibition of the NF-κB, PKC-δ, and ERK1/2 signaling pathways, which are involved in processes critical for GC metastases, including extracellular matrix degradation, cell motility, cytomorphologic alterations, and angiogenesis[20,21]. QT intake was inversely correlated with the likelihood of developing gastric adenocarcinoma[22]. Recent evidence suggests enhanced therapeutic efficacy of irinotecan plus QT in GC[23]. NG is a derivative of naringin or naringin hydrolysis, mainly present in the form of aglycones, but also in the form of glycosylation and neohesperidin[24]. TNG is primarily isolated from grapefruits, lemons, tomatoes, and oranges[25,26]. NG and its derivatives have important pharmacological properties, including estrogen-like activity or anticancer effect. NG effectively suppressed SNU-1 cell growth, inducing G0/G1 phase arrest and apoptosis by inactivating carcinogens and promoting cell cycle arrest; in addition, as a protective nutrient, NG showed chemopreventive and therapeutic effects, confirming its potential in GC treatment[27]. A recent study confirmed that NG inhibits SGC-7901 cell proliferation, migration, and invasiveness and induces apoptosis by downregulating AKT signaling[28]. In a tumor-bearing rat model induced by N-methyl-N’-nitro-N-nitrosoguanidine and promoted by S-NaCl, NG effectively demonstrated anti-tumor and cancer chemoprophylaxis effects by upregulating the redox state and protecting glutathione-metabolizing enzymes and phase I and II exogenous enzymes[29,30]. LL is a natural flavonoid found in fruits, vegetables, natural herbs, etc[31]. In addition to its anti-inflammatory, neuroprotective, and anti-allergic activities, LL is effective against various human malignancies, including GC. LL also has a marked anti-tumor effect in cMet-overexpressing patient-derived human tumor xenograft models of GC, possibly through the cMet/Akt/ERK axis[32,33]. Zang et al[34] demonstrated that LL reversed epithelial-mesenchymal transition by inhibiting the Notch axis, thus suppressing GC progression. The reliability of these components was further validated after molecular docking, which provides a novel basis for follow-up clinical research and treatment of GC.

Topology analysis of the PPI network identified 38 targets of Jianpi Yiwei Recipe for the treatment of GC, with ATK1, TP53, and TNF being the most critical. The AKT1 axis regulates cell proliferation, growth, apoptosis, and glucose metabolism, presenting different expression levels in human breast, colorectal, and ovarian cancers[35-37]. AKT1 overexpression was associated with a lower survival rate in resected gastric adenocarcinoma[38]. The loss of function of TP53 as a tumor suppressor gene plays a significant role in tumorigenesis[39]. During GC, the TP53 mutation is an early event in the formation of gastric adenocarcinoma, which turns intestinal metaplasia into GC[40]. TP53 overexpression is the main event in the transformation of normal gastric mucosa from intestinal metaplasia to GC[41]. TNFs are multi-effect cytokines that are the main mediators of inflammation, virus replication, tumor metastasis, graft rejection, rheumatoid arthritis, and septic shock[42]. The TNF-α-inducing protein gene family, including H. pylori membrane protein 1 and tipα, are tumor promoters involved in the carcinogenicity of H. pylori[43]. Studies have confirmed the close relationship between other target genes and GC, and we will not go into details here. It is speculated that the effective ingredients of TCM in the Jianpi Yiwei formula may exert therapeutic effects on GC through the aforementioned targets.

CONCLUSION

In summary, through network pharmacology and molecular docking analysis, this study found that in GC, active components in Jianpi Yiwei Recipe, such as QT, NG, and LL, may act on multiple targets, such as ATK1, TP53, and TNF; thus, affecting endocrine-related signaling pathways to improve and influence cells and playing an anticancer role.