Mechanism of Bupleurum and Asarum in Traumatic Brain Injury Treatment Based on Network Pharmacology and Molecular Docking Technology

Lianshan GUO, Yanqiu JIANG, Jihua FENG, Huimin ZHAO, Xiaomei TANG, Guang ZENG, Zhengzhao LI

Trauma Surgery, the Second Affiliated Hospital of Guangxi Medical University, Nanning 530000, China

Abstract [Objectives] To explore the active components and mechanism of Bupleurum and Asarum in the treatment of traumatic brain injury (TBI). [Methods] All the active components and potential action targets of Bupleurum and Asarum pairs were collected by online platform Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP) database and literature search. Target genes related to traumatic brain injury were obtained by Online Mendelian Inheritance in Man (OMMI), Therapeutic Target Database (TTD), PharmGKB, Genecards and Drugbank. The "drug-ingredient-target" network diagram was constructed by using Cytoscape software. Venny 2.1.0 was used to integrate the intersection targets of drug targets and disease targets, and the String platform was used to construct a target protein-protein interaction network (PPI). Topological analysis and core target screening of the constructed PPI network were performed using the "CytoNCA" plug-in in Cytoscape software. Gene Ontology (GO) annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed on the intersection targets using the DAVID database. Finally, Autodock and Pymol software were used to simulate the binding activity of key candidate active components and core genes. [Results] 25 active components were screened from Bupleurum-Asarum and 111 potential targets involved in the disease process. GO analysis and KEGG results showed that potential therapeutic targets were mainly enriched in biological processes such as inflammatory response, oxidative stress, cell membrane repair, and cytokine regulation. Network analysis and molecular docking showed that the key compounds of Bupleurum and Asarum were kaempferol and quercetin, which were well docked with the active pockets of four core genes of traumatic brain injury. [Conclusions] Bupleurum and Asarum may be involved in the regulation of inflammatory response, oxidative stress, cell membrane repair through multiple targets and multiple pathways in the treatment of traumatic brain injury.

Key words Bupleurum and Asarum, Traumatic brain injury (TBI), Network Pharmacology, Core target, Molecular docking

1 Introduction

Traumatic brain injury (TBI) generally refers to the transient or permanent structural brain tissue and brain dysfunction caused by external mechanical forces. It is a leading cause of death and disability worldwide. According to findings of modern medical research[1], TBI first causes brain parenchyma damage, cerebral hemorrhage and axonal shear, induces inflammation, oxidative stress, metabolic disorders and central nervous cell death, and finally leads to nerve damage, local blood supply disturbance, blood-brain barrier damage, and swelling of the surrounding brain,etc.Due to the complexity of brain anatomy, there is currently no effective treatment for TBI. At present, commonly and widely used treatment methods are early surgical operation to relieve the compression of brain tissue, supplemented by nutritional nerve, and rehabilitation physiotherapy. Traditional Chinese medicine (TCM) therapy shows unique advantages in the treatment of various diseases, and can act on multiple pathological links of the disease. According to the TCM theory, TBI belongs to the scope of "brain injury", and the syndrome differentiation belongs to qi stagnation and blood stasis, and the treatment principle is to open the orifice. In the combination of Bupleurum and Asarum, Asarum has functions of warming meridians and relieving pain, while Bupleurum has functions of dispersing liver and regulating qi, and combined use of them has functions of relieving stasis and relieving pain. According toLectureNotesonTraditionalChineseMedicineTraumatology, the combination of the two drugs has been successfully used as a classic drug pair in the TBI treatment[2]. However, there is no research to reveal the active components and action mechanism of Bupleurum and Asarum in the TBI treatment.

Network pharmacology is a new method for finding drugs in recent years. From the overall interaction between the components, targets and diseases of traditional Chinese medicine, it collects and sorts out active components, potential targets and disease targets of the discovered drugs through major databases. Then, it studies the targets and pathways of action of the target active components using network pharmacology, Gene Ontology (GO) functional annotation, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. Finally, it uses molecular docking technology to conduct docking analysis between the screened ligand molecules and receptor proteins, and visualize the interaction between ligands and receptors. In this study, we intended to provide a theoretical basis for the molecular mechanism of Bupleurum and Asarum in the treatment of traumatic brain injury through network pharmacology and molecular docking technology, so as to reveal its action mechanism, and provide a reference for subsequent in-depth experimental research.

2 Materials and methods

2.1 Active component screening and target prediction of Bupleurum and AsarumSeparately taking "Bupleurum" and "Asarum" as the keywords, using the TCMSP (http://tcmspw.com/tcmsp.php) database to select oral bioavailability (OB) ≥ 30% and drug-like (DL) property ≥0.18 as screening conditions, we collected and screened the active components of Bupleurum and Asarum[3]. We imported the results into Excel to create Bupleurum and Asarum active component files. Then, we imported the collected active components into the TCMSP database and PharmMapper database supplemented by the Drugbank database to obtain their corresponding targets. Using the Uniprot database (https://www.uniprot.org/), we converted the obtained target names into verified human-derived gene symbols to construct component-target files.

2.2 Collection of TBI-related targetsUsing OMIM (https://www.omim.org), TTD (http://db.idrblab.net/ttd/), PharmGKB (https://www.pharmgkb.org), Genecards (https://www .genecards.org) and Drugbank (https://go.drugbank.com) databases, taking "traumatic brain injury" as the keywords, we searched for TBI-related target genes, and removed duplicates, finally created TBI-related disease target files.

2.3 Construction of drug-component-target networkUsing the component-target file, we generated the gene list, molecule list and type file, and imported Cytoscape 3.9.1 software to construct the "drug-component-target" network diagram.

2.4 Construction of target protein-protein interaction (PPI) networkUsing the Venny 2.1.0 platform, we selected the intersection of drug active components targets with TBI-related disease targets, and plotted a Venn diagram of drug active component targets-disease targets. We imported the intersection targets into the String platform (https://cn.string-db.org/), selected "Multiple Proteins", the study species was "Homo sapiens", and obtained the target interaction relationship (confidence 0.700), deleted proteins with no interaction relationship, and constructed the target protein-protein interaction (PPI) network diagram.

2.5 Screening of core targetsWith the aid of the cytoNAC plug-in in Cytoscape software, we screened the core targets of the PPI network according to Betweenness (BC), Closeness (CC), Degree (DC), Eigenvector (EC),etc.Multiple screenings can be performed according to needs, so as to analyze the key components of Bupleurum and Asarum in the TBI treatment.

2.6 GO enrichment and KEGG pathway analysisWe uploaded the drug-disease intersection targets to the DAVID database (https://david.ncifcrf.gov/) for GO enrichment analysis and KEGG pathway analysis, and selected the top 10 GO enrichment and top 15 pathways of KEGG. With the aid of the website (http://www.bioinformatics.com.cn/), we displayed the results in the form of graphics.

2.7 Molecular dockingWe matched the final screened core targets in the constructed component-target network, and screened out the components corresponding to the core targets. Using the PubChem database, we obtained the 2D structures of the core components, imported the 2D structures into Chemofice software, conducted energy minimization processing, obtained the 3D structures of key compounds, and saved them in Mol2 file format. In the PDB database, we downloaded the 3D structure of the core protein genes, and then used the PyMol software to remove water and ligand molecules from the protein. Using AutoDock 1.5.6 software, we converted compounds and core protein genes into pdbqt format. Then, we searched for the receptor protein active pockets and saved them in gpf format. Next, we used Vina software for molecular docking, and finally used PyMol software to visualize the molecular docking results.

3 Results and analysis

3.1 Screening and target prediction results of active components of Bupleurum and AsarumThrough TCMSP database and literature search supplement, we collected 17 compounds in Bupleurum and 8 compounds in Asarum, and included a total of 25 compounds into this study. Among them, Bupleurum and Asarum have a common compound (Table 1). For 17 compounds in Bupleurum, we collected a total of 196 targets from TCMSP and PharmMapper databases, and for 8 compounds in Asarum, we collected a total of 111 targets. After removing the duplicates, we collected a total of 269 targets from Bupleurum and Asarum. Through the Uniprot database, we converted the obtained target names into the validated gene symbols of human origin.

Table 1 Active components of Bupleurum and Asarum

3.2 Collection results of TBI-related targetsUsing OMIM, TTD, PharmGKB, Genecards, and Drugbank databases, we searched for TBI-related targets and generated Venn diagrams for TBI targets in each disease database (Fig.1). After removing duplicates, we obtained a total of 1 798 targets related to the disease.

Fig.1 Venn diagram for TBI-related targets

3.3 Construction and analysis of drug-component-target networkUsing the component-target files, we generated gene lists, molecular lists and type files, and imported them into Cytoscape software to construct a "drug-component-target" network diagram (Fig.2). The gene distribution was adjusted according to the Degree value, and the larger the degree value, the larger the graph shape of the gene. The drug components are represented by different colors, green indicates the active components of Bupleurum, pink represents the active component of Asarum, and MOL000422 is the common component of Bupleurum and Asarum.

Fig.2 Diagram for drug-component-target network

3.4 Target protein-protein interaction (PPI) networkWe took the intersection of the selected active component targets of Bupleurum and Asarum with disease targets, and plotted a Venn diagram of drug-component-disease targets (Fig.3). We obtained a total of 110 potential target genes. Then, we imported the intersection genes into the String platform to obtain PPI, deleted the proteins without interaction relationship, imported the saved TSV format file into Cytoscape software, and plotted a diagram for the target protein-protein interaction (PPI) network (Fig.4). This network had a total of 102 nodes representing proteins produced by coding genes, and 560 mapping relationships representing predicted functional connections between proteins. The size and color depth of the nodes were set according to the ranking of the Degree value. In other words, the larger the node and the darker the color, the greater the Degree value of the target protein corresponding to the node, and the more important it plays in the network.

Fig.3 Venn diagram for drug-disease target

Fig.4 Diagram for target protein-protein interaction network

3.5 Filtering of core targetsWe screened the PPI network targets. After the first filtering, we obtained a total of 32 nodes and 231 edges. After the second filtering, we obtained a total of 11 nodes and 47 edges, and screened out key targets including RELA, HIF1A, MMP9, IL1B, AKT1, MYC, EGFR, CASP3, P53, TNF, and VEGFA. Finally, from the above candidate targets, we selected TNF, AKT1, TP53, and VEGFA as candidate core targets for subsequent molecular docking verification (Fig.5).

Fig.5 Process of filtering core targets

3.6 GO enrichment and KEGG pathway analysisWe uploaded the intersection targets to the DAVID database (https://david.ncifcrf.gov/) for GO enrichment analysis and KEGG pathway analysis. Among them, there were 2194 GO entries, including 1981 BP entries, 70 CC entries, and 143 MF entries. The results showed that potential therapeutic targets were mainly enriched in biological processes such as inflammatory response, oxidative stress, cell membrane repair, and cytokine regulation. There were a total of 167 KEGG pathways, including lipid and atherosclerosis-related signaling pathways, PI3K-Akt and other signaling pathways closely related to cell metabolism, growth, and proliferation, TNF, IL-17 and other signaling pathways related to cellular inflammation and immunity in the body. We selected the top 10 pathways for GO enrichment and the top 15 pathways for KEGG, and separately displayed the results in the form of histogram and bubble chart (Fig.6 and Fig.7).

Fig.6 GO analysis histogram

Fig.7 Bubble chart for KEGG analysis

3.7 Molecular docking verification of core components and key targetsWe matched the screened core targets in the constructed component-target file, screened the components corresponding to the core targets, and finally kept the two core components Kaempferol and Quercetin and key targets, to verify the molecular docking with key targets TNF, AKT1, TP53 and VEGFA. AutoDock adopts a semi-flexible docking method, that is, the molecular structure of the ligand is changeable, and the protein structure is unchangeable. It is generally believed that when the binding energy is less than zero, it indicates that the molecular docking is successful, and no external energy treatment is required, and if the binding energy is less than -5.0 kcal/mol, it indicates that the two have good binding activity. The larger the absolute value of the binding energy, the better the docking effect between the ligand and the receptor[3]. The molecular docking verification of the screened ligand and receptor proteins showed that the binding energy of all ligand molecules and receptor proteins was <-5.0 kcal/mol. Among them, Kaempferol-AKT1: -8.1 kcal/mol, Kaempferol-TNF: -6.9 kcal/mol, Kaempferol-TP50: -9.3 kcal/mol, Kaempferol-VEGFA: -7.9 kcal/mol, Quercetin-AKT1: -8.2 kcal/mol, Quercetin-TNF: -7.2 kcal/mol, Quercetin-TP50: -7.3 kcal/mol, and Quercetin-VEGFA: -8.0 kcal/mol, indicating that the ligand molecules and receptor proteins can bind well. Finally, we used the PyMol software to visualize the molecular docking results (Fig.8).

Fig.8 Molecular docking of core components and key targets

4 Discussion

The TBI refers to brain damage caused by physical force, and it is characterized by structural damage to the brain tissue and cell death. After TBI, it may lead to inflammation, oxidative stress, excitotoxicity and mitochondrial dysfunction in the brain[4]. Traditional Chinese medicine has been developed for thousands of years, and it has been used in the treatment of various diseases with remarkable curative effect. How to reveal the pharmacodynamic material basis and molecular mechanism of traditional Chinese medicine is one of the core issues in the modern research of traditional Chinese medicine. In 2007, pharmacologist Andrew L. Hopkins first proposed the term "network pharmacology". Network pharmacology technology is a network model that uses nodes and connections to establish connections between individuals, abstracts the interaction of complex biological systems as a network, and realizes the system identification of organisms by analyzing the components and characteristics of complex networks. Designing a new model of drug molecular design based on network pharmacology is an important way to develop innovative drugs[5].

Using the research method of network pharmacology, we screened the core targets of PPI network, and screened out four core targets, namely, TNF, AKT1, TP53 and VEGFA. TNF (tumor necrosis factor) is involved in the process of inflammatory response. Studies have shown that the level of inflammatory cytokines increases significantly after craniocerebral injury, mediates local inflammatory responses, and participates in secondary damage to brain tissue[6].AKT1gene participates in mediating various biological responses, such as inhibiting apoptosis and stimulating cell proliferation, and also plays an important role in inhibiting neuronal cell death[7-9].TP53(Tumor Protein P53) is a protein coding gene, it is a transcription factor that controls biological processes such as apoptosis, cell cycle arrest and DNA repair[10]. As an important tumor suppressor gene in the human body, it can not only prevent tumor cell division, induce tumor cell apoptosis, but also repair normal DNA damage. Sun Jijunetal.[11-12]found that VEGFA (Vascular Endothelial Growth Factor A) could target mediate angiogenesis in a rat middle cerebral artery occlusion model, and VEGFA signaling regulates angiogenesis and attenuates cerebral ischemia after ischemic stroke.

We performed GO and KEGG pathway enrichment analysis on intersection genes. GO enrichment analysis indicated that potential therapeutic targets were mainly enriched in biological processes such as inflammatory response, oxidative stress, cell membrane repair, and cytokine regulation. KEGG enrichment analysis further revealed the potential action mechanism of Bupleurum and Asarum in the TBI treatment. Through analysis, we found that the pathways involved mainly include lipid and atherosclerosis signaling pathway, PI3K-Akt signaling pathway, and cell signaling pathways such as TNF and IL-17. TBI is associated with atherosclerosis and cardiovascular and cerebrovascular mortality in humans[13]. The results of Wang Jintaoetal.[14]showed that the analysis of atherosclerosis in TBI mice compared with sham-treated mice showed that atherosclerosis in TBI mice increased, and after TBI, the atherosclerosis progression was accelerated. PI3K-Akt can increase the accumulation of cyclin D1, and promote the cell cycle and improve cell proliferation through the PI3K-Akt signaling pathway, which is closely related to a variety of enzymatic biological effects and glucose metabolism[15-16]. TNF is a pleiotropic cytokine. Molecular processes involved in TNF synthesis and release by microglia play an important role in regulating important central nervous system functions including neuroinflammatory responses, neuronal circuit formation and synaptic plasticity, and myelin damage and repair after brain injury[17-18]. The pro-inflammatory activity of IL-17 is associated with the pathogenesis of secondary TBI[19]. Findings of Lindsey Carlsenetal.[10]showed that in the central nervous system and serum of TBI rats, the level of IL-17 reached to peak at 3 d after modeling, which was consistent with the severity of secondary brain injury.

In summary, we found that core targets such as TNF, AKT1, TP53, and VEGFA are closely associated with TBI repair. Bupleurum and Asarum may act on these core targets through multiple pathways to regulate inflammatory response, oxidative stress, cell membrane repair and other processes to treat TBI. Finally, we preliminarily verified our conclusion through molecular docking analysis. We are intended to provide a theoretical reference for an in-depth study of the action mechanism of Bupleurum and Asarum.