The TCMSP database (https://old.tcmsp-e.com/tcmsp.php) was queried to retrieve information on each of the four herbs mentioned in BYD, namely, Yinyanghuo, Huangqi, Danshen, and Beiwuweizi. The key pharmacokinetic metrics, namely, oral bioavailability (OB) and drug-likeness (DL), conventionally established as screening criteria with thresholds of greater than 30% and 0.18 respectively. Subsequently, the active components and their corresponding protein target names of the four herbs were obtained. Target names were converted to Homo sapiens gene symbols using the Uniprot database (https://www.uniprot.org/), while overlapping target genes and components lacking gene symbols were excluded.

To identify genes associated with asthma, the keyword “asthma” was utilized to query the GeneCards database (https://www.genecards.org/Search) and OMIM databases (https://www.omim.org/). The genes with a relevance score higher than 1.5 were screened, and any duplicated genes were eliminated. The overlapping genes between BYD target genes and asthma-related genes were identified using the R package “VennDiagram”.

The pathway information pertaining to the overlapping genes of BYD-asthma was acquired from the DAVID database (https://david.ncifcrf.gov/tools.jsp). The table file was prepared and imported into Cytoscape 3.7.2 software, along with comprehensive information on active components of herbs, targets related to these components, and overlapping genes in BYD-asthma. Herbs, active components, genes, pathway IDs, and asthma acted as nodes; the interaction between them acted as edges.

The STRING database (https://cn.string-db.org/) was utilized to import and identify 75 overlapping targets of BYD-asthma in Homo sapiens. The network was then constructed based on a minimum required interaction score of 0.9.

Based on the previously published data on overlapping targets between BYD and asthma, the PPI network, and the Bioconductor database (https://bioconductor.org/), the visualization of GO functional enrichment analysis and KEGG pathway enrichment analysis were plotted using the R packages “DOSE”, “clusterProfiler”, and “pathview”.

The 3D structures of selected core components were downloaded from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/), and OpenBabel 2.4.2 was utilized to convert the file format of these structures to MOL2. The appropriate 3D protein structures of selected core genes were collected from the RCSB PDB database (https://www.rcsb.org/) and exported as PDB format files. Small molecular and receptor proteins were selected as ligands and receptors, respectively, using AutoDockTools 1.5.7 software optimization. We used AutoDock 4.2.6 software to set docking parameters and then carried out the molecular docking. The binding energy results with values below −1.2 kcal/mol were conventionally regarded as indicative of favorable binding activities. PyMol software performed the docking results visualization.

BYD required Yinyanguho 21g, Huangqi 30g, Wuweizi 9g, and Danshen 6g, which were purchased from Shandong Provincial Hospital of Traditional Chinese Medicine (Jinan, China). Distilled water was added to the above herbs, and the mixture was boiled twice for 30 min each. Subsequently, the resulting liquid was mixed and filtered to remove impurities, yielding 99 mL of BYD with a decocting concentration of 1.5 g/mL.

Aluminum hydroxide [Al(OH)3, alum] was provided by Waltham, USA. OVA, acetyl-β-methyl choline chloride (methacholine), and phosphate-buffered saline (PBS) were purchased from Sigma, United States. Reagents utilized in lung tissue staining (hydrochloric ethanol, masson bluing solution, masson ponceau s staining solution, phosphomolybdic acid solution, aniline blue, paraformaldehyde, xylene, hematoxylin, eosin, periodic acid, and fuchsin basic) were provided by Solarbio Life Sciences (Beijing, China). RIPA buffer, TBST buffer solution, primary antibody dilutions, secondary antibody dilutions, antibody eluent, and goat anti-rabbit IgG labeled with horseradish peroxidase were supplied by Servicebio Biology (Wuhan, China). Antibodies against phosphatidylinositol3 kinase (PI3K, cat.4691T), phospho-PI3K (P-PI3K, cat.17366S), RAC-α serine/threonine-protein kinase (AKT, cat.4691T), phospho-AKT (P-AKT, cat.4060T), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH, cat. 2118T) were purchased from Cell Signal Technology (Beverly, MA, United States).

A total of 24 pathogen-free female BALB/c mice (aged 7–9 weeks, weight 15 ± 2 g) were purchased from the Experimental Center of Shandong University (Jinan, China). The mice were housed in a pathogen-free standard environment for 1 week, during which they received adjustable feeding. The environmental conditions included a 12-h light/dark cycle, a temperature of 20°C, and a relative humidity of 60%. All animal experiments are performed in accordance with the regulations and guidelines from the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (National Research Council Committee for the Update of the Guide for the and Use of Laboratory, 2011). All animal experimental designs and procedures were approved by the ethics committee on animal care of Shandong Provincial Hospital & Shandong First Medical University (Jinan, China).

We determined the method for asthma modeling based on slight improvements of the methods in the previous studies (Huang et al., 2021; Huang et al., 2022), while the timing and concentration of administration were determined by preliminary experiment of this study. 24 mice began to enter the asthma modeling stage after 7 days of adaptive feeding. The mice were randomly divided into three groups with eight mice in each group: normal control group (control), OVA-induced asthma group (model), BYD treatment group (BYD). In asthma groups, mice were sensitized by intraperitoneal injection (i.p.) of a mixed solution consisting of 20 µg OVA, 2 mg alum on day 0 and day 7. From day 14–28, mice were challenged by atomizing inhalation (i.h.) of an OVA solution consisting of 3% OVA for 30 min each day. The sensitizing and challenging materials in control group were replaced by an equal dose of PBS. 14 days before the second sensitization, BYD (40 g/kg) were given once a day by intragastric administration (i.g.) in BYD treatment group, while the same dose of distilled water was given to control group. From day 14–28, 1 h before each challenge, the treatment measures and doses for each group were the same as above. 24 h after the last challenge, lung function was tested, bronchoalveolar lavage fluid (BALF), blood, and lung tissues were taken (Figure 1).

The measurement of airway responsiveness was conducted invasively using the PFT pulmonary maneuvers system (Buxco Electronics, Inc., Troy, NY, United States). On day 29, mice were anesthetized with pentobarbital (50 mg/kg) and subsequently underwent tracheotomy, followed by attachment of the trachea to the pulmonary function test system. Initially, atomization was performed using a PBS solution, followed by the addition of methacholine at concentrations of 3.125, 6.25, and 12.5 mg/mL for subsequent atomization with an interval of 5 min each. After 10s of atomization, pulmonary resistance and pulmonary compliance were recorded. The ratio of lung resistance (RL) and dynamic lung compliance (Cdyn) to the baseline (PBS challenge) level was employed as an indicator for assessing airway responsiveness of mice.

After assessing airway responsiveness, the mice were euthanized and the trachea was carefully dissected. A tracheal catheter was inserted and 1 mL of PBS was used for triple irrigation. The collected BALF was centrifuged at 4°C and 1200 R/min for 5 min, and the supernatant was absorbed and stored in the refrigerator at −80°C. The precipitated cells were re-suspended with 1 mL of PBS. The total number of cells and the number of inflammatory cells in BALF were measured by the IDEXX ProCyte DX blood analyzer (IDEXX Laboratories Inc., Westbrook, ME, United States).

Th2 cytokine expressions in BALF and serum of mice were tested using ELISA to evaluate the airway inflammation in asthma mice. Fresh blood from mice was left at room temperature for 1 h, then centrifuged at 4°C and 2500R/min for 10 min, reserving the supernatant for detection. The BALF supernatant was removed from the −80°C refrigerator and dissolved over ice cubes. The levels of IL-4, IL-5, and IL-13 in serum and BALF were respectively assessed using the ELISA kit (R&D Systems, Inc.) following the standard procedures described in the product instructions.

Fresh mice lung tissues were fixed in 4% paraformaldehyde for 24 h, followed by dehydration using a gradient concentration of alcohol and immersion in a wax solution. Subsequently, the tissues were embedded and sliced into sections of 4–5 μm thickness for HE, PAS, and Masson’s trichrome staining.

HE staining: Paraffin sections were dewaxed using xylene, followed by hydration in a gradient concentration of alcohol and subsequent washing with distilled water. The sections were stained with hematoxylin for a duration of 10 min, followed by differentiation in 1% hydrochloric ethanol for 30 s. Subsequently, the sections were stained with 0.5% eosin for a period of 1 min, and rinsed with distilled water after each step. Finally, dehydrated the sections in gradient concentration alcohol, transparentized with xylene, then sealed using neutral gum.

PAS staining: The sections were dewaxed and hydrated using the same aforementioned steps, followed by impregnation in a 5% periodate alcohol solution for 8 min and subsequent washing in 70% ethanol. Subsequently, the sections were immersed in a colorless fuchsin basic solution for 20 min, rinsed, and finally stained with hematoxylin for 30 s to visualize cellular structures. Dehydrated and transparentized sections as above.

Masson’s trichrome staining: Dewaxing and hydrating the sections as above. Stained cell nucleus with hematoxylin for 3 min, dripped acidic alcohol solution for 5 s for differentiating, added masson bluing solution for 5 min to reverse blue, while rinsing in distilled water after each step. Stained with masson ponceau s staining solution for 5 min, and drip-washed with weak acid working liquid. Differentiated with 1% phosphomolybdic acid solution for 2 min and drip-washed with weak acid working liquid. Stained with aniline blue for 2 min, then dehydrated and transparentized sections as above.

HE staining was used to evaluate the airway inflammation. The inflammation surrounding the bronchus and blood vessels was observed using an optical microscope in each section, with the inflammation score ranging from 0 to 5 points (Zhang et al., 2023). PAS staining was used to assess airway mucus secretion, while masson’s trichrome staining evaluated collagen deposition. The positive-staining airway wall area (Wat) and bronchial basal membrane circumference (Pbm) were quantified using Image J software. The results of PAS and Masson’s trichrome staining were presented as the ratio of Wat to Pbm.

Add protein lysate on ice to dissolve lung tissues for 30 min. Tested the protein concentration according to the instructions of the BCA protein quantification assay kit (G2026, Servicebio Biology, Wuhan, China). After the electrophoresis, the separation glue was immersed in the membrane transfer solution configured using 10% PAGE color (red) gel ultra-fast formulation kit (G2043-50T, Servicebio Biology, Wuhan, China) and transferred to the PVDF membrane for 30 min at a constant current of 300 mA. At room temperature, imprint with 5% skim milk in TBST buffer for 2 h, and then incubate overnight at 4°C with the following primary antibodies: anti-P-PI3K (1:500), anti-PI3K (1:1,000), anti-AKT (1:1,000), anti-P-AKT (1:1,000), anti-GAPDH at 1:5,000). On the second day, washed with TBST for 3 times, then diluted the second antibody (goat anti-rabbit IgG labeled with horseradish peroxidase) with TBST at a ratio of 1:3,000, incubated in a shaker at room temperature for 1 h. Reapply the mixed ECL luminescent liquid onto the PVDF membrane and initiate ECL chemiluminescence by employing the ECL chemiluminescence kit (G2014-50ML, Servicebio Biology, Wuhan, China). Integrated densities were measured by AIWBwell analysis software, and the ratio of indicator integrated density to parameter integrated density (relative content of the sample) was considered as the evaluation index.

The obtained data are presented as a mean ± standard error of the mean (SEM). The systematic analysis was performed with IBM SPSS Statistics 25 software using one-way ANOVA comparing data between multiple groups which were normally distributed. A value of p < 0.05 was considered to have statistically significant differences between groups.

The screening criteria in the TCMSP database were set as OB ≥ 30% and DL ≥ 0.18 to search for the four drugs, resulting in the identification of 116 active components and their corresponding target protein names. The protein names were mapped to 926 gene symbols of Homo sapiens using the Uniprot database. After excluding 767 duplicate genes and 11 active components lacking gene symbols, we identified 166 BYD targets, 21 YYH components, 18 HQ components, 6 BWWZ components, and 60 DS components. We identified 1,485 asthma-related genes through screening in the GeneCards database, eliminating 39 duplicate genes obtained from the OMIM database. The identification of the 75 DBY-asthma overlapping genes and the visualization image were yielded via running the R package “VennDiagram” (Figure 2A).

The interaction network was constructed utilizing the Cytoscape software (Figure 3). The engagement information of the active components in the network was acquired through utilization of a network analysis tool. When degree, betweenness centrality, and closeness centrality were considered as the indicators, 11 potentially key active components in the BYD treatment for asthma were identified (Table 1): quercetin (MOL000098), kaempferol (MOL000422), luteolin (MOL000006), isorhamnetin (MOL000354), tanshinone iia (MOL007154), anhydroicaritin (MOL004373), 7-O-methylisomucronulatol (MOL000378), dan-shexinkum d (MOL007093), NSC 1224219 (MOL007149), α-amyrin (MOL006824), and salviolone (MOL007145).

The PPI network, consisting of 75 nodes and 126 edges, was constructed by importing 75 BYD-asthma overlapping targets into the STRING database (Figure 2B). And the network analysis was performed by Cytoscape software. After deleting targets lower than the mean value of degree (4.271186441), betweenness centrality (0.057581465), and closeness centrality (0.338852949), the remaining 9 targets were considered as the potential key targets in BYD treatment for asthma (Table 2): heat shock protein 90 alpha family class A member 1 (HSP90AA1), estrogen receptor 1 (ESR1), proto-oncogenec-Fos (FOS), epidermal growth factor receptor (EGFR), mitogen-activated protein kinase 8 (MAPK8), B-cell lymphoma-2 (BCL2), proto-oncogene Myc (MYC), IL6, caspase 3 (CASP3), and most of which are involved in the PI3K/AKT signaling pathways (Laboratories, 2024).

GO and KEGG enrichment analysis can facilitate a comprehensive understanding of the underlying molecular mechanisms involved in BYD treatment for asthma, enabling the identification of functional biases and significant pathways throughout its therapeutic process. The barplot and dotplot images of GO enrichment analysis (Figures 2C, D) implied that BYD might exert its therapeutic effect via regulating DNA-binding transcription factor activity and promoting transcription coregulator binding to stimulate the transcription of anti-inflammatory molecules, thereby regulating the body’ state of oxidative stress. Visual images of KEGG pathway enrichment analysis (Figures 2E, F) indicated the BYD-asthma targets were mainly concentrated in the PI3K/AKT signaling, tumor necrosis factor (TNF) signaling, Th17 cell differentiation, hypoxia inducible factor-1 (HIF-1) signaling, and IL-17 signaling pathways, which are closely associated to the immune inflammation in asthma. According to the identified key genes predominantly involved in the PI3K/AKT signaling pathway and the results of KEGG analysis, it can be inferred that the PI3K/AKT signaling pathway plays a pivotal role in BYD’s mechanisms of action for asthma treatment.

Through literature search, we found that the research on IL6, EGFR, and HIF1A to asthma was relatively richer compared with these of other targets. Moreover, we selected them as core target genes based on their significance in both the PPI network and the PI3K/AKT signaling pathway (Figure 4). We selected the top three active components out of the 11 key components, namely, quercetin, kaempferol, and luteolin, based on their extensive literature support and high centrality in the drug-component-target-pathway-disease network. Subsequently, Autodock software was employed to dock the 3 core targets with the 3 core components. The results of nine dockings showed binding energies ranging from −5.26 to −6.76 kcal/mol, which were all much lower than −1.2 kcal/mol (Table 3). The highest binding activity was observed between IL-6 and luteolin, with a binding energy of −6.13 kcal/mol. Similarly, quercetin exhibited the highest binding activity towards EGFR, with a binding energy of −6.76 kcal/mol. Additionally, HIF1A showed the strongest affinity for kaempferol, with a binding energy of −5.41 kcal/mol. Our molecular docking results demonstrated that the three core active components of BYD can directly target the three core targets of asthma, implying that quercetin, kaempferol, and luteolin in BYD may exert therapeutic effects by intervening and modulating the PI3K/AKT signaling pathway related to asthma.

The expressions of RL and Cydn in 3 groups of mice were measured under different concentrations of methacholine by the pulmonary maneuvers system, evaluating the changes compared with the baseline. As shown in (Figure 5A), with the increase in the concentration of methacholine used for challenge, the increase level and rising trend were most significant in the model group compared with control group. BYD could significantly reduce the RL of asthma mice in the face of methacholine stimulation. The Cydn displayed a decreasing trend when the concentration of methacholine increased in 3 groups (Figure 5B). BYD performed a significant effect in resisting the decreasing tendency of Cdyn. In the RL and Cydn line graphs, the line trend and expression level of BYD group were most similar to those of the control group, indicating that the BYD treatment could effectively alleviate AHR in mice with allergic asthma.

By assessing the inflammatory cells in BALF of mice, including total cells, eosinophile percentage and neutrophil percentage, it was found that the inflammation degree of mice in model group was significantly elevated than that in control group, while BYD treatment could significantly reduce the levels of total inflammatory cells, eosinophils and neutrophils in BALF of asthma mice (Figures 5C–E). The levels of Th2 inflammatory cytokines, including IL-4, IL-5, and IL-13, were quantified in both BALF and serum samples from mice via ELISA (Figures 5F–K). Results displayed that BYD could significantly reduce the level of inflammatory cytokines in asthma mice, indicating the important regulatory and inhibitory effects of BYD on airway inflammation in asthma mice.

The HE staining technique was employed to examine the extent of inflammatory infiltration surrounding the trachea and blood vessels in lung tissues of asthmatic mice. As shown in (Figure 6A), the bronchus in lung tissues of control group mice exhibited a normal morphology, with uniform thickness of the tube wall, unobstructed lumens, and clear delineation of alveoli. In contrast, asthmatic mice displayed thickened tube walls, deformed lumens, damaged alveoli and significant infiltration of high-density inflammatory cells around the tube wall and blood vessels. Asthmatic mice with BYD treatment showed significantly alleviated inflammatory infiltration and reduced inflammatory cell number. The trachea walls exhibited a decrease in thickness, and there was an improvement in the shape of alveolar walls. Additionally, a significant reduction in the inflammatory score was observed compared to the model group (Figure 6D). Additionally, PAS staining and Masson’s trichrome staining were employed to further examine collagen deposition and mucus secretion in lung tissues. It was observed that BYD treatment led to a significant reduction in collagen deposition along the endotracheal lining and mucous secretion within the lumen (Figures 6B, C, E, F), indicating notable improvements compared to the model group. Therefore, we considered that BYD may play an important role in reducing airway inflammation, collagen deposition, and mucus secretion in asthma.

The network pharmacology results mentioned above revealed that the PI3K/AKT signaling pathway may potentially contribute to the therapeutic effects of BYD. In order to further validate the involvement of this pathway in BYD-mediated suppression of airway inflammation in asthmatic mice, we used Western blot analysis to determine the protein expression of PI3K, phosphorylated PI3K, AKT, and phosphorylated AKT in lung tissues, with GAPDH serving as an internal control.

In order to better ascertain the influence of BYD to asthmatic mice, we take the left lower lobe of lung tissues from each group of mice and conduct extractions of tissue homogenate and proteins, then repeated the WB detection for three times. The remaining lung tissues were used for conduct different experiments, such as slicing and staining of lung tissues. The expression levels of phosphorylated PI3K and AKT were significantly increased in the model group, suggesting a close association between airway inflammation in asthmatic mice and activation of the PI3K/AKT signaling pathway (Figure 7A). The protein expressions of phosphorylated PI3K and AKT in lung tissues of BYD-treated mice were significantly diminished compared to those in the model group, implying that BYD may modulate airway inflammation in asthma mice by suppressing the phosphorylation of PI3K and AKT and inhibiting the activation of the PI3K/AKT signaling pathway (Figures 7B, C).