Bioinformatics and Network Pharmacology Identify the Therapeutic Role of Guominkang in Allergic Asthma by Inhibiting PI3K/Akt Signaling

Introduction

Allergic asthma (AA) is the result of heterogeneous, complex gene-environment interactions involving multiple cells, especially mast cells, eosinophils and T-lymphocytes, in chronic airway inflammation with different clinical phenotypes, inflammation and remodelling. In susceptible individuals, this inflammation can cause recurrent episodes. The most common symptoms of AA include wheezing, shortness of breath, chest tightness and/or coughing, which are often present at night and/or in the early hours of the morning. Additionally, individuals with AA tend to have increased airway responsiveness to a wide range of irritants. AA affects between 1–18% of the global population,¹ with an estimated 339 million people suffering from the condition, resulting in 1000 deaths per day, as reported in the 2018 Global Asthma Report.² The current clinical treatments for this disease only control the symptoms and reduce inflammation. For example, the most commonly used biologics globally are anti-IgE monoclonal antibodies, which account for 78% of all treatments, followed by anti-interleukin (IL)-5 antibodies (43.9%), anti-IL-13R/IL-4R antibodies (36.7%), and anti-IL-5R antibodies (26.7%).³ The reported adverse reactions include local reactions at the injection site (74%), followed by allergic reactions (6.8%), delayed rashes (5.1%), and serious cutaneous adverse reactions.⁴ This causes a lot of inconvenience to the patient’s life and work. In addition, corticosteroids are one of the main pharmacological treatments for asthma with anti-inflammatory properties, and according to a meta-analysis by Bateman et al (64 RCTs, 42,527 patients), an inhaled corticosteroid (ICS) regimen was found to improve patients’ quality of life, and the inhaled corticosteroid + long-acting β2-agonist fixed-dose combination was well tolerated and safe, and current clinical guidelines recommend prescribing a combination of the two for treatment.⁵

With progressive development over millennia, traditional Chinese medicine (TCM) have gained widespread acceptance in treating numerous illnesses, including allergic diseases. The herbal formulation Guominkang (GMK) formula was found to be markedly effective in regulating allergic diseases, including allergic rhinitis, AA, and atopic dermatitis.⁶ Previously, we have reported that GMK can suppress the recruitment of inflammatory cells, reduce mucus production, decrease airway resistance, attenuate allergic symptoms in asthmatic mouse models, and reestablish the cytoimmune homeostasis of Th1/2 and regulatory T cells (Tregs)/Th17 cells.⁷ Accordingly, it can be speculated that GMK may inhibit AA progression by regulating several targets and pathways. Currently, the effect of TCM on diseases is being explored mainly by overlapping interactions between components and disease targets to unravel the underlying mechanisms of TCM therapy.

PI3K/Akt is a typical signal transduction pathway that regulates a number of cellular processes, including cell growth, proliferation, motility, metabolism and survival and plays a crucial role in the pathogenesis of allergic diseases. The PI3K/AKT pathway is a key contributor to allergic airway inflammation, promoting eosinophil recruitment and Th1/Th2 immune imbalance.⁸ The specific PI3K inhibitor Ly294002 has been demonstrated to effectively attenuate airway inflammation, airway remodelling and eosinophil infiltration in asthma mice.⁹ Consequently, numerous herbs and herbal extracts were observed to alleviate inflammation and lung injury in the OVA-LPS-induced AA mouse model, with the PI3K signaling pathway playing a pivotal role.^10,11

This study employed network pharmacology, molecular docking, experimental methods and bioinformatics to analyze the key components and targets of GMK that exert therapeutic effects on AA, as well as the signaling pathways of action. We carried out preliminary validation through animal experiments, which provided a reference for the precise therapeutic targets of GMK and the multi-targets for the treatment of AA. The workflow is presented as a graphical summary (Figure 1).

Figure 1 Flow chart of this study.

Materials and Methods

Data Source and DEG in Asthma

The Asthma-related microarray datasets GSE104468 were downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/). The microarray GSE104468, detected by GPL21185, Included 24 nasal samples from 12 asthma patients and 12 healthy samples. The PI3K-Akt related genes were obtained from the KEGG database, which contains 359 genes (KEGG PATHWAY: hsa04151). Raw data were batch corrected to remove batch effects for further analysis. Differentially expressed genes (DEGs) were screened by using the “limma” package with P < 0.05 and |log FC| > 0.9 indicating statistically significant differences.¹²,388 genes were analysed further. The R language programs “heatmap” and “ggplot2” were used to create heat maps, volcano maps and box line plots.

Analyses of DEGs Function

To investigate the function of the DEGs in asthma, a consensus clustering algorithm was employed to evaluate the optimal cluster numbers and robustness. The R package “ConsensusClusterPlus” implemented the above steps for 1000 iterations to guarantee the robustness of classification. Gene set enrichment analysis (GSEA) was performed on the gene expression matrix by applying the “clusterProfiler” package. The “ggplot2” packages were used to perform Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses on DEGs. The “ggvenn” package was used to depict the intersecting genes of the dataset differential genes and signaling pathways. The “corrplot” and “paletteer” packages were used to profile the heatmap of the correlation coefficients of the intersecting genes.

Screening of Active Ingredients of GMK

The six GMK components were entered into the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP, https://old.tcmsp-e.com/tcmsp.php) and the HERB (http://herb.ac.cn/) databases to determine the chemical composition of the herbal medicines, with descending screening conditions in the TCMSP website set as follow: oral bioavailability (OB) ≥ 30 and drug-likeness (DL) ≥ 0.18.^13,14 The pharmacochemical components of HERB were downloaded from PubChem (https://pubchem.ncbi.nlm.nih.gov/) with Canonical SMILES for each component. These active ingredients were screened using pharmacokinetic parameters from the swissADME (http://www.swissadme.ch/) database, using the criteria gastrointestinal absorption (GI) = High and the rule of five (Lipiskin, Ghose filter, Veber filter, Egan filter, and Muegge filter) presenting “yes” ≥ 2 to comply with ADME (absorption, distribution, metabolism, and excretion) characteristics of the drug and demonstrate that the orally administered drug can be absorbed.^15,16

Target Prediction of Active Ingredients of GMK

The Canonical SMILES of each component obtained by database screening were uploaded to SwissTargetPrediction (http://www.swisstargetprediction.ch/) to establish a list of targets for each active component, proteins with the probability > 0 were considered more biologically active targets. To obtain uniform gene names, the UniProt (http://www.uniprot.org) website was used to perform ID conversion, download the reviewed files converted to uniform Homo sapiens, and obtain the UniProt IDs of all targets.^17,18

Collection of Disease-Related Targets for AA

The disease name “allergic asthma” was entered into the disease search field of the following five databases: OMIM (https://omim.org/),¹⁹ DisGeNET (https://www.disgenet.org/),²⁰ GeneCards (https://www.genecards.org/),²¹ HERB (http://herb.ac.cn/), Drugbank (https://www.drugbank.ca/),²² the relevant targets of the disease were collected and downloaded from each database. Finally, the UniProt ID and human gene name were obtained from the UniProt database.

Compounds-Disease-Targets Network Construction

The predicted targets of each component (obtained from the above steps in each database) were uploaded to the Venn online tool with AA-related targets. Subsequently, common target sets of drugs and diseases were compared to obtain an intersection set as the GMK target for treating AA. The Cytoscape 3.9.0 software was employed to generate tables for the purpose of a visual analysis of compounds-disease-targets networks.²³

Protein-Protein Interaction, PPI

PPI networks, connected by protein interactions, can systematically analyze massive protein interactions in biosystems and are critical to explore molecular mechanisms of bio-signaling and energetic substance metabolism in physiological and pathological states, as well as to clarify functional linkages among proteins. The target genes in the above intersection were uploaded to the STRING database (https://string-db.org/), and species-selected Homosapiens and PPI network relationships were obtained using the default parameters of the system. To obtain the core target networks, the CytoNCA plugin wasused to calculate topological parameters in PPI networks and perform an in-depth analysis of the properties of nodes in interactive networks. The top 10 key genes were calculated using cytoHubba, with color shades representing the keyness of genes.^24–26

Enrichment Analysis

Enrichment categories and extent of target genes in GO and KEGG were analyzed. GO analysis included molecular function (MF), cellular component (CC), and biological processes (BP). The R (3.6.3) ggplot2 package with the clusterProfiler package and KOBAS (http://kobas.cbi.pku.edu.cn/) website were used with the screening criteria set at P < 0.05.

Molecular Docking

First, candidate proteins were entered into the UniProt website, and structures with a recent update date and high resolution were selected and linked to the PDB website for download. The selected small-molecule ligand 3D structures were downloaded from the PubChem website, and proteins and small molecules were processed using the AutoDock Tools software and output in their respective pdbqt format files.

Animal Model and GMK Preparation

BALB/c mice (6-8-week-old) were purchased from Beijing Vital River Laboratory Animal Technology Co. and housed in a specific pathogen free (SPF) environment. All animal experiments followed the standards of the Animal Care and Use Committee of the Beijing University of Chinese Medicine, AAALAC and IACUC guidelines.

Mice were randomly divided into three groups, with 12 animals assigned per group: the blank control group (control), model group (model), and GMK treatment group (GMK). The AA model was established as described in our previous study.⁷ Sensitization phase: On day 0 and day 14, respectively, each mouse was intraperitoneally administered 0.2 mL of ovalbumin (OVA) antigen solution containing 20 μg OVA (Cat#A5503, Sigma-Aldrich, USA) and 0.2 mg aluminum adjuvant (Cat#77161, Invitrogen, USA), mice in the blank group were intraperitoneally administered the same volume of saline containing 0.2 mg aluminum adjuvant. Stimulation phase: From day 21 to day 25, at least 30 min before completing administration, mice were nebulized with 1% OVA saline solution for 30 min, the blank group was nebulized with the same volume of phosphate-buffered saline (PBS) for 30 min.

The GMK formula was as follows: Wu-Mei (Mume Fructus), Shou Wu Teng (Polygoni Multiflori Caulis), Ling Zhi (Ganoderma), Fang Feng (Saposhnikoviae radix), Chan Tui (Cicadae Periostracum), Tian Ma (Gastrodiae Rhizoma), these components were purchased and immersed in deionized water for 30 minutes, and then decocted them at 100°C. The same volume of deionised water was added again with the same conditions of decoction and repeated once more. 19.5 g/kg of GMK was administered once daily by gavage from day 14 to day 25.

Airway Methacholine (Mch) Excitation Assay in Mice

Acetylcholine powder was dissolved in PBS to achieve gradient concentrations of 0, 6.25, 12.5, 25, and 50 mg/mL, and each mouse was stimulated for 5 min. Enhanced pause (Penh) values of each group of mice stimulated with different acetylcholine concentrations were measured noninvasively. On experiment completion, the Penh values of the mice in each group were counted and calculated, and finally expressed as Penh ratio. The calculation formula is: the average value of Penh in the 5-minute interval of methacholine/the average value of Penh in the 5-minute interval of PBS stimulation.

Pathological Tissue Staining

Lung tissues were fixed in 4% paraformaldehyde solution for 24 h and then subjected to HE staining, PAS staining and Masson staining (Wuhan Servicebio Technology CO.,LTD, China). The sections were dehydrated with 50%, 70%, 80%, 95% ethanol and anhydrous ethanol in turn, and then dipped in xylene for several hours after being transparent, embedded in regular paraffin wax, and stored in slices. HE staining: paraffin sections were dewaxed, stained with hematoxylin-eosin, and dehydrated and sealed, and the nuclei of the cells appeared to be blue to blue-purple and the cytoplasm of the cells appeared to be red under the microscope. PAS staining: the sections were dewaxed and acidified with periodic acid, washed in water, stained with chevron, and sealed with dehydration, and the mucus material appeared to be purple-red. Masson staining: paraffin sections were sequentially deparaffinised, stained with potassium dichromate, stained with ferric hematoxylin, stained with magenta, stained with phosphomolybdic acid, stained with aniline blue and sealed, and the local histopathological changes were observed under the microscope.

Western Blotting

Total protein was extracted from the RIPA lysate and quantified by the BCA method. Subsequently, protein separation was achieved by 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, carefully place the membrane in a film transfer clip, and activate the polyvinylidene difluoride (PVDF) membrane in methanol, followed by transfer to PVDF membranes in TBST buffer. The membranes were blocked using PBST containing 5% skim milk powder at 37°C for 2 h. The primary antibodies (Akt (56 kDa, Abcam, ab238477, 1:1000), P-AKT (56 kDa, Abcam, ab81283, 1:1000), PI3K (123 kDa, Abcam, ab151549, 1:1000), p-PI3K (84 kDa, Abcam, ab182651, 1:1000), β-actin (42 kDa, Abcam, ab8227, 1:1000)) were incubated at 4°C overnight. Incubate the secondary antibody for one hour at room temperature. The bands on PVDF membrane were detected using chemiluminescent substrates and quantified density using the Image Lab 5.2 software. All experiments were performed in triplicate.

Statistical Analysis

SPSS 26.0 software was used to analyze the data. The experimental results were obtained from three independent replications. All data were tested for normality and chi-square. Each group’s results are expressed as mean ± SEM. Differences between the groups were analyzed using unpaired t-tests or chi-square tests, depending on the type of data, while one-way or multi-way ANOVA was utilised for multiple group comparisons. The statistically significant difference is indicated by P value < 0.05. GraphPad Prism 9.5.0 software was used to draw statistical charts.

Results

Active Compounds of GMK

Searching the TCMSP and HERB databases, we identified the following active compounds: Ganoderma, Saposhnikoviae radix, Mume Fructus, Cicadae Periostracum, Caulis Polygoni Multiflori, and Gastrodiae Rhizoma, of these, the pharmacologically active ingredients were screened using the rule of five. After eliminating compounds without relevant targets and duplicate data, 50 compounds of Mume Fructus (Wu-Mei), 4 of Polygoni Multiflori Caulis (Shou Wu Teng), 115 of Saposhnikoviae radix (Fang-Feng), 128 of Ganoderma (Ling Zhi), 17 of Gastrodiae Rhizoma (Tian Ma), and 10 of Cicadae Periostracum (Chan Tui) were identified, among which 27 overlapping ingredients were named using their PubChem IDs in the network. Overall, 267 compounds were obtained, with 1177 related targets predicted. Table 1 presents information regarding the top ten compounds.

Table 1 Top 10 Compounds in the Ingredient-Target Network of GMK

Construction of a Drug-Compound-Target Network

The keyword “allergic asthma” was used to search and download each database separately, and 2109 disease targets were obtained after eliminating duplicate data. The targets of each component and disease were entered into the Venn mapping website to obtain the intersection and plot the Venn diagram (Figure 2a). The plotted Venn diagram revealed 479 genes in the central intersection, and 475 were obtained after converting UniProt ID to the gene name, which showed that one-quarter of the GMK targets were disease-related genes. The rhombus represents the drug, the oval represents the compound, the circle represents the compound with overlapping related drugs, the color corresponds to different compounds, and the round rectangle represents the target. The drug-compound-target network graph includes 775 nodes and 11,138 edges (Figure 2b).

Figure 2 Network pharmacological analysis. (a) Venn diagram of active ingredients and disease targets (b)Network of Compounds-disease-targets (c) PPI network of common targets of GMK and AA (472 nodes,10735 edges) (d) cytoNCA PPI network (164 nodes,5383 edges) (e) Hub gene of PPI network.

PPI Network and Freedom Analysis With Screening of Hub Genes

Files regarding the interaction relationship obtained from the STRING database were imported into Cytoscape 3.9.0 for visualization (Figure 2c). CytoNCA screened core targets based on centrality analysis, with proteins associated with the PI3K-Akt pathway indicated in dark pink (Figure 2d). Among them, the top 10 nodes in terms of degree were TNF, IL6, AKT1, ALB, GAPDH, IL1β, SRC, TP53, VEGFA, and STAT3 (see Table 2 for node details). In the cluster analysis, there were seven modules with MCODE scores ≥ 5 (see Table 3 for module details).

Table 2 Top 10 Targets in Protein-Protein Interaction Network of GMK and AR

Table 3 Modules of the Protein-Protein Interaction Network of Common Targets of GMK and AR

The following ten common target hub genes were identified according to the default MCC algorithm in the cytoHubba plugin, with degree values in descending order and indicated in dark to light colors. The top ten hub genes were GAPDH, JUN, STAT3, TP53, CASP3, TNF, IL-6, HIF1A, CTNNB1, and BCL2L1 (Figure 2e).

GO Functional Enrichment and KEGG Pathway Analysis

The results presented entries with gene enrichment P.adj ≤ 0.05: BP mainly involved processes such as response to molecule of bacterial origin, response to lipopolysaccharide, CC mainly involved processes such as membrane raft, membrane microdomain, and external side of plasma membrane, MF mainly involved processes such as chemokine binding, protein tyrosine kinase activity and drug binding (Figure 3a-c). Significant pathways are shown in the bubble diagram, where predominant signaling pathways include the PI3K/Akt signaling pathway and apoptosis (Figure 3d). Accordingly, the effects of GMK on AA appear to be mediated through multiple compounds, targets, and pathways.

Figure 3 Functional analysis of the target. (a-c) GO functional and (d) KEGG pathway enrichment analysis (e) Network of the main pathways and targets (178 nodes,4719 edges) (f) PI3K-3,7,12,23-Tetrahydroxy-26,28-epoxy-16,23-cyclolanost-8-ene-11,15,26-trione (132113) Dock (g) PI3K-15alpha,26-Dihydroxy-5alpha-lanosta-7,9(11),24-triene-3-one (21635716) Dock (h) PI3K-methyl lucidenateF (21633085) Dock (i) TNF-3,7,12,23-Tetrahydroxy-26,28-epoxy-16,23-cyclolanost-8-ene-11,15,26-trione (132113) Dock (j) PI3K-Dauricine (73400) Dock.

The top 10 signaling pathways with the smallest P values were visualised using Cytoscape 3.9.0, along with their corresponding targets. Consequently, a network graph comprising 178 nodes and 4719 edges was generated (Figure 3e). Using the Network Analysis tool, the average degree of signaling pathway nodes was 53.02, with 74 nodes exceeding this value. The main signaling pathways and target information are listed in Tables 4 and 5, respectively.

Table 4 KEGG Pathways of Main Pathway-Target Network of GMK Treatment for AR (Top 10)

Table 5 Targets of Main Pathway-Target Network of GMK Treatment for AR (Top 10)

Molecular Docking Verification

The top 10 PPI and targets of the main pathway-target network were identified, along with the key targets PI3K. These were then molecularly docked with the top 10 compounds using AutoDock Vina software. The crystal structures of the 7 targets with higher resolution were downloaded: PI3K (PDB ID:5JHB), TNF (PDB ID:4Y6O), IL6 (PDB ID:1ALU), AKT1 (PDB ID:1UNQ), SRC (PDB ID:7NG7), and TP53 (PDB ID:2VUK), and STAT3 (PDB ID:6NJS). The chemical structure SDF files of each compound were downloaded from the PubChem database (Table 6), and the docking fraction between compound-target was less than −5 kcal/mol, indicating that the corresponding compound and target had a higher binding affinity (Table 7), and the binding positions and sites for the five pairs of composite target interactions are shown in results (Figure 3f-j).

Table 6 Target Gene Protein Pocket Coordinates and Grid Box Sizes

Table 7 Affinity Value (kcal/Mol) of the Groups of Main Ligands With Receptors (n = 70)

GMK Inhibits PI3K/Akt Signaling Pathway Activity in AA Mice

The AA animal model was constructed using the classical OVA method (Figure 4a). From the Penh results of each group of mice, GMK can be known to attenuate the airway resistance in AA mice (Figure 4b). Compared with mice in the control group, mice in the model group showed an increasing trend in the Penh ratio when stimulated with 0, 6.25, 12.5, 25, and 50 mg/mL acetylcholine (P < 0.01), thereby corroborating the modelling of AA mice, the GMK group had lower Penh values than the model group (P < 0.001). Based on HE, PAS, and Masson staining analyses (Figure 4c–e), the lung tissues of mice in the model group showed marked inflammatory cell infiltration, goblet cell hyperplasia, a large amount of mucus secretion, and considerable collagen deposition. GMK-treated mice showed reduced numbers of inflammatory cells in the lung tissue, decreased goblet cell metaplasia, and improved collagen deposition.

Figure 4 GMK treatment regulated inflammatory responses in OVA-induced mice. (a) Establishing a timeline of AA mouse model. (b) The Penh values of each group of mice were assessed using a non-invasive method (n=6, **P < 0.01 vs Control group, ***P < 0.001 vs Control group, ΔΔΔP < 0.001 vs Model group, #p < 0.05 vs GMK group, ###p < 0.001 vs GMK group). (c) HE staining and inflammation score of lung tissue (d) Typical PAS staining and percentage of positive cupped cells (e) Masson staining and collagen fibril deposition rate of lung tissue (×200 magnification) (f and g) Comparison of protein expression of p-PI3K and p-Akt in lung tissue of mice in each group (n=6, *P < 0.05, **P < 0.01, ***P < 0.001).

The PI3K/Akt signaling pathway is known to play a key role in AA.^8,27 The pathway enrichment and molecular docking analyses revealed that the PI3K signaling pathway mediates the therapeutic effects of GMK in AA. The results of Western blot also indicated that GMK regulates the PI3K/Akt signaling pathway in the lung tissue of AA mice (Figure 4f and g). Compared with the model group, the GMK group exhibited significantly reduced levels of PI3K and AKT phosphorylation in the lung tissues (P < 0.05). These findings indicate that the PI3K/Akt signaling pathway was suppressed in the lung tissues of GMK-treated mice.

Bioinformatics Analysis of AA and the PI3K/Akt Pathway

After downloading the GSE104468 dataset from the GEO database, 388 DEGs (106 genes up-regulated and 282 down-regulated in AA patients) were screened using the Limma package (P < 0.05, |logFC| > 0.9) (Figure 5a–d). Further GO enrichment analyses of BP, CC and MF showed its role in the positive regulation of phosphatidylinositol 3-kinase activity, T cell receptor signaling pathway and MHC class II receptor activity, KEGG pathway was enriched in Asthma, Antigen processing and presentation and Th1 and Th2 cell differentiation (Figure 5e and f). To further analyse the correlation of these asthma disease genes with the PI3K-Akt pathway, we downloaded the PI3K-Akt pathway genes included in the KEGG pathway, showing that nine genes overlapped, and the PPI network showed the links between them, using pie charts to represent the strength of the positive and negative correlations between these genes (Figure 5g–i).

Figure 5 Bioinformatics analysis of data sets. (a) Box line blots of GSE104468 (b) Volcano map of DEGs (Up-regulated genes were labelled pink and down-regulated genes were labelled blue) (c) Heatmap of the significantly upregulated or downregulated DEGs (d) Heatmap of the matrix of co-occurrence information for AA samples (e and f) GO and KEGG analyses of DEGs between asthma group and healthy group (g) Venn diagram showing 9 overlapping genes associated with PI3K-AKT and DEGs (h) PPI of DEGs and overlapping genes (i) Correlation between 9 intersecting genes.

Discussion

AA, an allergen-induced subtype of asthma, is characterized by eosinophilic airway inflammation, bronchial hyperreactivity, and elevated IgE levels, frequently combined with allergic rhinitis and conjunctivitis. AA has become a serious threat to human health. Our study used network pharmacology, molecular docking, bioinformatics and other experimental methods to analyze the key components and targets of GMK’s therapeutic effect on AA, proving that the PI3K/Akt pathway may play an important role in the treatment of AA.

Previous studies have shown that numerous factors participate in the development of AA, with genetic inheritance, epigenetic inheritance, intestinal flora, and cellular immune balance all influencing the development of AA.²⁸ AA involves a variety of cells, including mast cells, eosinophils, and T lymphocytes. Th2 cells are classical helper T cells that contribute to the development of asthma and produce IL-4, IL-5, and IL-13, important cytokines in AA.^29,30 In response to transforming growth factor (TGF)-β and IL-4 cytokines, CD4⁺ T cells can differentiate into a type of IL-9-secreting helper T lymphocyte, Th9, which is a growth and differentiation factor for mast cells and promotes inflammation by stimulating the secretion of chemokines that recruit other pro-inflammatory cells to the inflammatory site.^31,32 Th17 and Tregs are functionally antagonistic and act together to maintain immune homeostasis. Th17 cells secrete glucocorticoid-resistant IL-17A, IL-17F, and IL-22, cytokines capable of inducing neutrophilic airway inflammation, cupped cell hyperplasia, increased mucus secretion, and airway smooth muscle hyperplasia, thereby resulting in airway remodeling.^33,34

Currently, medications used to treat asthma are classified into two categories: control medications and reliever medications. However, the use of these medications is associated with a number of side effects.³⁵ Epidemiological surveys have revealed that asthma-related morbidity and mortality are increasing annually, and there is a lack of specific drugs for treating asthma in modern medicine, therefore, there is an urgent need to develop alternative therapeutic agents that can effectively control or treat asthma.³⁶ TCM has unique advantages in the treatment of allergic diseases. GMK is a traditional Chinese medicine compound formula composed of six Chinese medicines under the guidance of the diagnostic and therapeutic modes of constitution-disease differentiation, which has shown good efficacy in clinical practice, but its mechanism of action has been less studied. And the holistic nature of network pharmacology is well-suited to Chinese medicine research, with the help of databases and networks to analyze the interrelationships between drugs, targets, and diseases. Here, 267 active ingredients and 475 related genes of GMK were screened for compounds with potent biological activity and oral availability based on the principle of five classes of drugs. Among them, decursin derivatives have been found to considerably inhibit elevated leukocyte and eosinophil levels in the bronchoalveolar lavage fluid, decrease mucus secretion and IgE levels, and induce mucus hypersecretion.³⁷ Imperatorin reportedly exerts good expectorant, analgesic, and anti-inflammatory effects and has been shown to improve mast cell-mediated allergic airway inflammation by inhibiting the MRGPRX2 and CamKII/ERK signaling pathways.³⁸ Furthermore, imperatorin was shown to reduce IL-4, IL-5, and IL-13 production via Th2 cells, promote IFN-γ and IL-12 production via Th1 cells, and enhance IL-10 production in bronchoalveolar lavage fluid.³⁹ However, in the present study, molecular docking revealed that the most important active ingredient was 3,7,12,23-Tetrahydroxy-26,28-epoxy-16,23-cyclolanost-8-ene-11,15,26-trione (PubChem CID: 132113), 15alpha,26-Dihydroxy-5alpha-lanosta-7,9(11),24-triene-3-one (PubChem CID: 21635716), which is the first time that a compound has been identified that can act on the disease target of AA to exert therapeutic effects. Therefore, in future studies, we will continue to investigate the potential mechanisms by which multiple components of network pharmacology work together to treat AA.

We constructed a PPI network using common targets of GMK and AA and analyzed the main gene functions using GO and KEGG, then six targets were identified in the data mining process. Allergen-induced inflammation drives TNF-dependent innate memory, which may maintain and aggravate chronic type 2 airway inflammation, affording a potential target for asthma therapy.⁴⁰ IL-6, a key factor necessary for Th17 cell differentiation, interacts with TGF-β to promote Th17 cell differentiation and inhibit Treg proliferation, resulting in a Th17/Treg imbalance. In a mouse model of asthma, Src kinase mediates epidermal growth factor receptor transactivation, which, in turn, stimulates multiple downstream signaling pathways involving ERK1/2 and PI3Kδ, as well as the transcription factor NF-κB to regulate allergic lung inflammation, with SCR inhibition contributing to an improved phenotype of AA.⁴¹ Inhibition of vascular endothelial growth factor (VEGF) was found to attenuate asthma symptoms in mice, and studies have analyzed polymorphisms in the VEGFA gene that may be associated with the response to asthma treatment.⁴² Elevated levels of phosphorylated STAT3 or STAT3 deficiency reportedly ameliorate Th2-associated allergic lung inflammation.^43,44

In addition, AKT1 is an important target that mediates the PIK3/Akt signaling pathway, while PI3K is a kinase that regulates tumor signaling pathways that affect and regulate inflammatory cells and inflammatory factors closely associated with the development of AA. Phenotypic regulation of AA through PI3K/Akt has been found in this network pharmacology studies. In asthma, PI3K regulates many inflammatory cytokines as well as neutrophil, eosinophil, and macrophage recruitment and activation, and in OVA-sensitized mice, accompanied by enhanced expression of PI3K pathway proteins, enhanced inflammatory cell infiltration and an IL-13 / IFN-γ cytokine imbalance were observed in lung tissue.⁴⁵ Targeting PI3K/Akt for the treatment of asthma has also been reported in previous studies. Inhibition of PI3K/AKT/mTOR signaling by targeted molecules attenuates asthma pathology and plays an important role in airway protection,⁸ eg, the natural compound Bixin acts as a potent antagonist of PI3K/Akt signaling to prevent AA.⁴⁶ This target is therefore central to the treatment of asthma. Using a mouse model of AA, we showed that the PI3K/Akt signaling pathway was suppressed in the GMK-treated group, with mice exhibiting fewer symptoms of airway inflammation than the control group. We proceeded to verify the correlation between allergic asthma and the PI3K-Akt pathway using publicly available data from GEO, employing the R language statistical software. Overall, our findings suggest that PI3K/Akt plays a major role in mediating the effects of GMK against AA. However, the current study has several limitations. Firstly, the mechanisms by which the key active components of GMK modulate other signaling pathways in AA were not investigated in this study. Secondly, in vitro experiments using cells to directly demonstrate the protective effects of GMK on the airway were not performed.

Conclusions

The objective of this study was to elucidate the mechanisms by which GMK exerts its therapeutic effects on AA. To achieve this goal, network pharmacological methods, bioinformatics analyses, and animal experiments were employed. The results demonstrated that the PI3K/AKT signaling pathway plays a pivotal role in AA, and GMK exerts therapeutic effects by regulating the PI3K/AKT signaling pathway. This provides an experimental basis for the clinical application and further development of GMK.