Conclusion
TCM play critical therapeutic roles through “multi-components, multi-targets and multi-pathways” mechanisms in influenza infection. The antiviral pharmacological mechanism of Xuanbai Chengqi decoction, which was analyzed by network pharmacology and molecular docking, provide a new idea for further exploring the diagnosis and treatment of severe influenza.
Epidemics of influenza typically occur during the winter months and have become a public health event that threatens the world. According to the epidemiological model, influenza-associated respiratory mortality are higher than previously WHO stated, about 291 243–645 832 deaths annually [1]. China government attaches great importance to the prevention and treatment of influenza, and the National Health and Health Commission has issued three versions of the Influenza diagnosis and treatment to promote the prevention and treatment of influenza in recent years [2]. In this scheme, in addition to updating antiviral western drugs, severe influenza patients are recommended for treatment on the basis of Xuanbai Chengqi decoction based on syndrome differentiation. Xuanbai Chengqi decoction (XBCQ) originates from ”differentiation of febrile Diseases”, The prescription is composed of ephedra, raw gypsum, almond, Anemarrhena anemarrhena, Herba Houttuyniae, Cymbals, Radix Scutellariae, Fritillaria thunbergii, rhubarb, Radix Paeoniae Rubra and raw licorice.
TCM has been used in Asia and even all over the world. However, because of its complex composition and changeable prescription, it is relatively difficult to verify its mechanism by traditional experimental methods. Therefore, there is an urgent need for new methods to systematically and comprehensively analyze the mechanism of Chinese herbal medicine. With the rapid development of bioinformatics, network pharmacology has become a new way to effectively reveal the molecular and pharmacological mechanisms of TCM formulae, and it will also become a new paradigm for TCM research [3-7]. The network pharmacological studies on the anti-influenza effects of compound Artemisia annua, Jingyin granule and Reduning injection have been carried out [8-10].
Molecular docking is an effective and intelligent computational technique to estimate the binding affinity of a ligand (such as drug candidate) in the macromolecular target site (receptor). Based on the structure of active components of TCM and disease targets are clearly, molecular docking is a promising way to show the mechanism of drug action [11, 12].
In order to obtain more accurate anti-influenza mechanisms of TCM formulae, we studied the scheme of TCM recommended by the China National Health Commission for severe influenza patients and combined with the methods of network pharmacology and molecular docking to provide a theoretical basis for the prevention and control of influenza with TCM. The findings presented in this thesis add to our understanding of TCM for influenza.
1. Materials and methods
1.1. Active ingredients and ingredient targets collection
In Chinese medicine systematic pharmacology database and analysis platform TCMSP (http://lsp.nwu.edu.cn/tcmsp.php), all of the chemical compositions of herbs were available. In order to obtain the active ingredients of the scheme, we screening active ingredients based on absorption, distribution, metabolism and excretion (ADME) parameters. oral bioavailability (OB)≥30% and drug-likeness (DL)≥ 0.18 are essential for drug screening. Targets corresponding to active components were queried in TCMSP, then, converted target proteins into corresponding Gene names in UniProt database.
1.2. Acquisition for Influenza disease targets
Used Drugbank (https://www.drugbank.ca/) and GeneCard (https://www.genecards.org/), OMIM (https://omim.org/), TTD (http://db.idrblab.net/ttd/), PharmGkb (https://www.pharmgkb.org/) database and entered the keyword ”Influenza” to search for influenza-related targets. The intersection of the active ingredient targets and the influenza related targets were the key antiviral targets of the active ingredients.
1.3. Network construction and analysis of “active ingredients-key targets”
Active ingredients and key targets were imported to software Cytoscape 3.8.0 and constructed the ”active ingredients-key targets” Network, and the ”Network Analyzer” function in Cytoscape was used for Network analysis, and the important antivirus components were screened according to the Degree values.
1.4. Protein-protein interaction (PPI) Network construction and analysis
We imported the key protein into String database, defined species as human, and set confidence value≥0.9. The text data after organizing was imported into Cytoscape3.8.0 software to construct PPI Network, and then applied ”Network Analyzer” function to analyze the index of Network topology, screened important targets whose degree values were higher than average.
1.5. Biological function and pathway analysis
The ClueGO plug-in in Cytoscape3.8.0 software was used to perform GO biological function annotation and KEGG pathway analysis on key targets and set P≤0. 05.
1.6. Molecular docking Verification
AutoDock Vina molecular docking software could be used to predict the possible molecular interaction between a target protein and a small molecule. According to the important components in the “Active ingredients - Key targets” network and the durg target proteins in the PPI network, the docking effect between the active ingredients and the key targets was evaluated by AutoDock Vina molecular docking verification.
2. Results
2.1. Drug and disease-related genes
We obtained 192 active components of 12 herbs (except gypsum) from TCMSP. And then we got 242 targets of active ingredients (Table 1) and 229 influenza-related targets (Figure 1). 31 key targets (figure 2) were obtained after the duplicative terms deleting by Venn diagram.
2.2. Construction and Analysis of ”Active Ingredients - Key Targets” Network
Network diagram (Figure 3) of antiviral “Active ingredients - Key targets” was constructed in Cytoscape 3.8.0 software with 192 active ingredients corresponding to 12 traditional Chinese medicines and 31 targets (Table 2). The network diagram involves 188 nodes and 364 edges, among which the circle represents the active ingredients and the rectangle represents the key target points. In terms of degree evaluation, 10 important ingredients were screened, including quercetin, luteolin, kaempferol, wogonin, aloe-emodin, naringenin and etc. The ingredients with high degree values may be the key antivirus ingredients (Table 3).
2.3. PPI network construction and analysis
31 key target proteins were imported to String database, defined species as human, set confidence value≥0.9, and PPI network (figure 4) was built in Cytoscape software. The network involves 24 nodes which represent targets and 61 sides which represent interaction between proteins and proteins. And the sizes and colors of the nodes express Degree values. Degree values grow from small to large, so do the nodes. Degree values of 11 target proteins are higher than the average, among which, JUN, CASP8, MAPK1, IL1B’s degree values are relatively high, which indicates a strong interaction with other proteins. Therefore, these predictive proteins play an important role in the network.
2.4. Biological function and pathway analysis
In Cytoscape 3.8.0 software, ClueGO plug-in was applied to GO biological function annotation and KEGG pathway analysis on 31 key targets. The bar chart of visual analysis was obtained after setting P≤0.05. As shown in Figure 5, GO analysis revealed that they were significantly enriched in the following biological processes, including response to lipapolysaccharide, molecule of bacterial origin, tumor necrosis factor. Cellular component analysis showed that membrane raft, membrane microdomain and membrane region were common classifications. In terms of molecular functions, they were mainly associated with cytokine receptor binding, cytokine activity, receptor ligand activity, signaling receptor activator. In addition, the pathways enrichment terms were shown by KEGG database. As shown in Figure 6, KEGG pathways included IL-17 signailing pathway, hepatitis B, influenza A, Chagas disease, TNF signaling pathway, toll-like recptor signaling pathway.
2.5. Molecular docking verification
We selected ingredients with rich contents such as quercetin, luteolin, naringin, and the first four key targets JUN, CASP8, MAPK1 and IL-1B in PPI network. Targets crystal structure files were downloaded from PDB database, and water molecules and ligand molecules were removed by Pymol software. From the Protein Data Bank (PDB) database, 1JNM, 1QDU, 5LCK and 9ILB were identified as the protein structures of the four key targets highlighted above for molecular docking experiments. The Grid Box parameters in AutoDockTools were set as following: JUN, grid center 40 40 40, NPTS 10.2 0.5 19.5, spacing 1.0; IL1B, grid center -15.8 13.5 -1.6, number of points in xyz (NPTS) 40 40 40, spacing 1.0. CASP8 and MAPK1 are dimers and trimers, their active pockets are highly conserved. The docking conditions were similar after 20 times docking and binding energy was used as an important criterion for constituents screening (Table 4). Then we uploaded the structure files of target proteins and quercetin, luteolin, naringin respectively in AutoDock Vina online software. The molecular docking results are shown in Table 4. The closer the protein binds to small molecules, the lower the Full Fitness values are. The results show that the three components can dock with target proteins JUN and IL-1b (Figure 7).