Systematic Analysis Strategy Based on Network Pharmacology to Investigate the Potential Mechanism of Fritillaria thunbergii Miq. against Idiopathic Pulmonary Fibrosis

Idiopathic pulmonary fibrosis (IPF) is a long-term, distressing, and age-related interstitial lung disease characterized by a complicated etiology and irreversible progression. Fritillaria thunbergii Miq. (Zhe Beimu, ZBM) is frequently used for its heat-clearing and phlegm-resolving properties in herbal compounds for the treatment of IPF. However, the specific mechanisms underlying the effects of ZBM against IPF have not yet been reported. In this study, we applied a systematic analysis strategy based on network pharmacology to explore the probable core targets and major pathways of ZBM against IPF. In addition, molecular docking simulation and quantitative real-time polymerase chain reaction (qRT-PCR) were performed to preliminarily investigate the possible mechanisms underlying the therapeutic effects of ZBM on IPF. We collected a total of 86 components of ZBM and used network pharmacology analysis to screen nine presumptive targets of ZBM against IPF. The molecular-docking results indicated that the components of ZBM exhibited good binding activity with presumptive targets. The qRT-PCR results also suggested that ZBM may partly alleviate IPF by regulating the expression of presumptive targets. This study laid the foundation for further clinical applications of ZBM and the development of IPF-related therapeutic products.


Introduction
Idiopathic pulmonary fbrosis (IPF) is a long-term, distressing, and age-related interstitial lung disease characterized by the overactivation of lung fbroblasts and overexpression of extracellular matrix components, which are associated with the remodeling of the lung structure and an irreversible decline in lung function [1][2][3]. In Europe alone, more than 40,000 people are diagnosed annually with IPF, and the morbidity and mortality caused by IPF have shown an upward trend [4,5]. With a median survival of only 2-3 years, the quality of survival of IPF patients is discouraging, creating substantial economic and survival pressures for the patients and imposing a huge burden on the society and the medical system [6]. After more than a decade of clinical trials, only two drugs, nintedanib and pirfenidone, have received marketing approval from the Food and Drug Administration for the treatment of IPF. Although these two drugs have shown positive efects in alleviating the decline in lung function in IPF patients, their efects on improving the survival rate or quality of life of patients with IPF are quite limited. Moreover, despite the signifcance of lung transplantation in improving survival in selected IPF patients, its applicability is greatly limited by the physical condition of IPF patients [7,8]. Terefore, further development of novel therapeutic strategies for IPF patients is essential.
China, with thousands of years of history in the practice and development of traditional Chinese medicine (TCM), has established its own unique TCM-based treatment system. With the deepening of TCM research, this medical system has gained increasing recognition among clinicians and researchers due to its promising therapeutic efects and relatively mild side efects [9,10]. Te dry bulbs of Fritillaria thunbergii Miq. (Zhe Beimu, ZBM), which are mainly produced in Zhejiang, China, have shown wide clinical use in TCM [11,12]. In clinical practice, ZBM is considered to have the ability to clear away heat, resolve phlegm, relieve cough, remove toxicity, reduce swelling, and dissipate knots, allowing its application for the treatment of lung abscess, lung atrophy, cough, and dyspnea [13,14]. Modern pharmacological studies have shown that ZBM has antioxidative, anticancer, anti-infammatory, antitussive, expectorant, antithyroid, and neuroprotective activities [15][16][17]. In addition, ZBM is frequently used as a heat-clearing and phlegm-resolving agent in TCM for the treatment of IPF [18][19][20]. However, since previous studies have emphasized the roles of single components and their targeted pathways, the overall mechanism underlying the therapeutic efects of ZBM on IPF requires elucidation [21,22].
Unlike research on modern drugs, which are identifed by targeting specifc proteins, TCM research is based on systematic analysis strategies to elucidate the mechanisms of drugs, thereby accounting for the complex composition and extensive coverage of these drugs. With recent improvements in systems biology and biological network equilibrium theory, network pharmacology has gradually emerged as a new model of drug molecule design characterized by "network target and multicausal analysis" [23,24]. Network pharmacology emphasizes research on the relationships between active ingredients and therapeutic objects from the perspective of the overall connection of biological relationships. Its research philosophy coincides with the holistic theory of TCM [25].
In this study, the main compounds of ZBM were collected using the TCM systems pharmacology (TCMSP) database, TCM Database@Taiwan, and the relevant literature. Secondly, compounds with superior pharmacokinetic properties and druggable properties were screened by the SwissADME analysis tool, and network pharmacology analyses were conducted to predict crucial targets and possible mechanisms of ZBM against IPF. In addition, the afnity between components and targets was analyzed by molecular-docking simulations, which verifed the reliability of the network pharmacology analysis. Finally, the partial mechanism of ZBM against IPF was verifed by experiments. Tis study provides a new direction for research strategies to elucidate the mechanisms of TCM against IPF from a holistic perspective. Te graphic abstract for the study is presented in Figure 1.

Establishment of the Chemical Composition
Library of ZBM. "Fritillariae Tunbrgii Bulbus" was used as a keyword to obtain chemical information from the TCMSP database, an open system pharmacology database for capturing the relationships among drugs, targets, and diseases [26] (https://old.tcmsp-e.com/tcmsp.php; obtained on April 16, 2022). Similarly, "Fritillariae Tunbrgii Bulbus" was used as the search term to collect the relevant chemical composition information in the TCM Database@Taiwan, a noncommercial database for providing downloads of TCM small molecules for virtual screening [27] (https://tcm.cmu.edu. tw/; obtained on April 16, 2022). Additional chemical composition information was obtained from relevant review articles [14,15]. Te chemical composition library of ZBM was established by sorting and classifying the compound information obtained from the databases and literature and eliminating duplicate values and some compounds for which PubChem CIDs were not available. Detailed information regarding the library is provided in the Supporting Information (Supplementary Material S1-Chemical Composition Library of ZBM).

Screening of Compounds with Superior Properties.
Te SwissADME analysis tool was employed to screen ingredients in the chemical composition library of ZBM. SwissADME can predict the absorption, distribution, metabolism, and excretion (ADME) or pharmacokinetic performance of small molecules on the basis of specifc models [28] (https://www.swissadme.ch/; screened on April 26, 2022). Compounds that met all of the following screening conditions were analyzed in the next step: (1) high gastrointestinal (GI) absorption in the pharmacokinetics analysis; (2) 4 or 5 YES conditions in the drug-likeness analysis; and (3) no violations within the YES condition in the drug-likeness analysis. After screening, 17 compounds with superior properties were obtained, and a library of these compounds was established. Detailed information regarding these compounds is presented in the Supporting Information (Supplementary Material S2-Library of Compounds with Superior Properties).

Predictive Targets of Components with Superior Properties
in the Compound Library. Information from the simplifed molecular input line entry system (SMILES) for components with superior properties in the compound library was obtained from the PubChem website (https://pubchem.ncbi. nlm.nih.gov/; obtained on April 27, 2022). Ten, we pasted the SMILES information of the components into the Swiss Target Prediction platform (a platform for predicting possible biological targets of active molecules based on structural similarity algorithms [29] (https://www. swisstargetprediction.ch/; analysis performed on April 27, 2022) to obtain compound targets and retain targets with probability values >0. Compounds for which no targets were provided on the Swiss Target Prediction website were imported into the TCMSP for target collection. After removing duplicate values, the target cluster of compounds with superior properties was constructed. Detailed information is presented in the Supporting Information (Supplementary Material S3-Compounds with Superior Properties and Teir Corresponding Targets).

Collection of Potential Targets against IPF.
Te Comparative Toxicogenomics Database (CTD), GeneCards, Online Mendelian Inheritance in Man Database (OMIM), Terapeutic Target Database (TTD), and DrugBank Database were searched to obtain IPF-related targets. CTD (https://ctdbase.org/) is a publicly available database for studying chemical exposures and their biological efects [30]. GeneCards (https://www.genecards.org) is an integrative database providing functional information on the relationship between human genes and disease [31]. OMIM (https://www.omim.org/) is an online database for describing the association between disease phenotypes and their causative genes [32]. TTD (https://db.idrblab.net/ttd/) is a publicly accessible database for information on targets known to be of therapeutic value for diseases [33]. DrugBank (https://go.drugbank.com/) is a database providing detailed drug data, drug target information, and computerized drug predictions [34]. All information was obtained on April 28, 2022. In each database, "idiopathic pulmonary fbrosis" was used as the key search term. Among the fve databases, due to the large amount of target information obtained from the CTD and GeneCards databases, we screened the top 1000 targets according to the inference score values and relevance score values, sorted from high to low. After removing duplicates and collating the results from the fve databases, the target cluster of disease was constructed. Finally, the intersection of the target cluster of compounds with superior properties and the target cluster of disease was analyzed by drawing a Venn diagram (https://www.bioinformatics.com. cn; obtained on April 29, 2022) to represent potential targets of ZBM against IPF. Specifc information regarding the targets is provided in the Supporting Information (Supplementary Material S4-Details of Intersection Targets).  Evidence-Based Complementary and Alternative Medicine format and imported into Cytoscape software (v3.7.2, Massachusetts, USA) for network analysis. Te medians of the three parameters, "betweenness centrality," "closeness centrality," and "degree," were calculated by the software. Te crucial gene cluster was screened by retaining genes whose corresponding data values were greater than medians. Network analysis and visualization of the crucial gene clusters were performed in Cytoscape software. Detailed information regarding the crucial gene clusters is presented in the Supporting Information (Supplementary Material S5-Information Regarding the Crucial Gene Cluster).

Enrichment Analysis through Gene Ontology and Kyoto
Encyclopedia of Genes and Genome. Te list of crucial genes obtained in Section 2.5 was entered into the Database for Annotation, Visualization, and Integrated Discovery (DA-VID, version 6.8, a database for providing gene function annotation and related biological information [36], https:// david.ncifcrf.gov/tools.jsp; obtained on April 30, 2022) to analyze the biological links among them. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were used to obtain information regarding the biological function and signaling pathways of the crucial gene clusters. Ten, the microbiographic mapping platform (https://www.bioinformatics. com.cn/; employed on May 2, 2022) and Cytoscape software were employed to visualize the results by drawing a dot bubble chart, GO term chart, and the "compound-targetpathway" network. Te correspondence information between active compounds, core targets, and pathways was input into Cytoscape, which performed network analysis among these elements and sorted them according to the "degree" value.

Molecular Docking Simulation.
Te 3D structure information of the proteins was obtained from the RCSB Protein Data Bank (PDB) (https://www.rcsb.org/, obtained on May 10, 2022, PDB IDs are shown in Table 1). After removing water molecules, heteroatoms, and original ligands in the protein structures with PyMOL software (v2.4.0, New York, USA, https://www.pymol.org/), hydrogen bonds were added to protein molecules in MGLTools software (v1.5.6, California, USA, https://ccsb.scripps.edu/ mgltools/). Te proteins were then processed by calculating their Gasteiger charge values and assigning AD4 atom types and saved in the PDBQT format as receptor molecules. By searching the positions of the existing ligands in protein receptors, the docking boxes of the functional pockets of protein receptors were determined to prepare for molecular docking. Subsequently, compound molecules in the SDF 3D format downloaded from PubChem were converted to the PDB 3D format, and the 3D compound molecules were imported into MGLTools software for the detection of roots and selection of torsions. Ten, the processed compound molecules were saved in the PDBQT format as ligand molecules and treated together with the processed receptor protein molecules to perform molecular docking simulation in AutoDock Vina software (v1.1.2, California, USA, https:// vina.scripps.edu/, a program for molecular docking and virtual screening [37]) to obtain docking afnity scores and docking information. Finally, the docking results were visualized with PyMOL software. Where the binding energies were equal, the conformation that contained hydrogen bonds or contained more hydrogen bonds was preferentially displayed.

Drug Preparation.
Te TCM Fritillaria thunbergii Miq. material was purchased from the Chinese herbal medicine market (Anhui, China) and authenticated by Dr. Ding Qi (Shenzhen Research Institute, Beijing University of Chinese Medicine, Beijing, China). A total of 10.03 g of the TCM material was soaked for 2 h with ten times the amount of double distilled water and then refuxed for 1 h. After fltering the extracted liquid, the drug residues were extracted with eight times the amount of water under refux for 1 h. Te two fltrated liquids were combined, concentrated, and dried to fnally obtain 1.68 g of ZBM extract powder (a yield of 16.75%). Te extract powder was dissolved in dimethyl sulfoxide (DMSO) before the experiment.

Cell Culture and Treatment.
Human embryonic lung fbroblast cell line MRC-5 (BNCC353614) was purchased from BeNa Culture Collection (Henan, China). A basal culture of MRC5 cells was established using Dulbecco's Modifed Eagle Medium (DMEM) containing 12% fetal bovine serum, and the cells showed good proliferation at 37°C and 5% CO 2 . When the cells were ready for plate laying, they were digested with 0.25% trypsin and inoculated at a density of 3 × 10 4 cells/well in 96-well plates or a density of 4 × 10 5 cells/well in 6-well plates. After starvation for at least 4 h, the cells in the administration groups were treated with transforming growth factor-β1 (TGF-β1) stimulators (10 ng/ mL; Proteintech, Wuhan, China) for an additional 48 h in the absence or presence of ZBM extracts (0.01-10 μg/mL).

Cell Viability Assay with the Cell Counting Kit-8.
Te cell counting kit-8 (CCK-8, MedChemExpress, Shanghai, China) was used to detect the proliferation of MRC-5 cells after treatment with ZBM extracts and TGF-β1. Briefy, the day after cell inoculation into 96-well plates, cells were starved with serum-free medium for at least 4 h. Te cells were then treated with ZBM (0, 0.01, 0.1, 1, and 10 μg/

Chemical Composition Library of ZBM.
We collected compound information from the TCM Database@Taiwan, TCMSP databases, and literature. After removing duplicates and compounds without structure information, the remaining 86 compounds were used to establish the chemical composition library of ZBM. Te main components of ZBM included alkaloids, diterpenoids, and essential oils, among others. Detailed information regarding the library is presented in the Supporting Information (Supplementary Material S1, Chemical Composition Library of ZBM).

Collection of IPF Disease Targets.
We collected information on targets related to IPF from the CTD (1000 targets), GeneCards (1000 targets), OMIM (339 targets), TTD (30 targets), and DrugBank (30 targets) databases. A total of 2081 IPF-related targets were obtained to construct the target cluster of disease after removing duplicates. Te results were visualized with Venn diagrams (Figure 2).

Construction of the Protein-Protein Interaction Network and Selection of the Crucial Gene Cluster.
A total of 134 cross-targets were identifed from the intersection of the target cluster of compounds with superior properties and the target cluster of disease ( Figure 3). Te STRING database and Cytoscape software were used to construct the protein-protein interaction network. After calculating the median values of the parameters "betweenness centrality" (median value � 0.00192664), "closeness centrality" (median value � 0.47482014), and "degree" (median value � 14), 46 core targets were selected to construct the crucial gene cluster. Te network relationships in the crucial gene clusters were shown by Cytoscape software (Figure 4). Specifc information is presented in the Supporting Information (Supplementary Material S4-Details of Intersection Targets) and the Supporting Information (Supplementary Material S5-Information regarding the Crucial Gene Cluster).

GO and KEGG Pathway Enrichment
Analysis. Te list of 46 core targets was pasted into the DAVID database for KEGG and GO enrichment analysis to further investigate the role of core targets in the development of IPF. Te top 10 items in the biological process (BP), cellular component (CC), and molecular function (MF) analysis results in the GO enrichment analysis were presented in the form of a merged bar graph in Figure 5. Te results indicated that core targets afect the development of IPF mainly by regulating biological processes such as cellular response to reactive oxygen species, regulation of protein phosphorylation, cell proliferation, and the cellular response to hypoxia. In addition, using KEGG enrichment analysis, the top 10 signaling pathways involved in IPF disease related to core targets were selected and visualized with a dot bubble chart ( Figure 6). On the basis of the existing research on IPF, the enriched KEGG pathways were divided into hormone regulation-related signaling pathways (thyroid hormone, relaxin, prolactin, and estrogen), cell growth state regulation-related signaling pathways (VEGF, PI3K/AKT, ErbB, FoxO, and HIF-1), and immune regulation-related signaling pathways (IL-17).

Compound-Target-Pathway Network Construction.
Based on the top 10 signaling pathways obtained above, the compound-target-pathway network, including 17 compounds with superior properties and 31 core targets, was constructed by Cytoscape software to analyze the relationships among them. As shown in Figure 7, the compound-

Molecular Docking.
We simulated the binding of the compounds and the targets through molecular-docking analyses to further analyze the connection between the targets and compounds. After treatment, fve compounds were used as small molecule ligands and nine target proteins   : Te protein-protein interaction network of the crucial gene clusters. As the degree of the target decreased, the circle became smaller and the color became lighter (from dark red to light green). As the combined score decreased, the combined line became thinner.  AKT1  GCTTCTTTGCCGGTATCGTG  TCCACACACTCCATGCTGTC  EGFR  CCATCCAAACTGCACCTACG  ACACGCTGCCATCATTACTTTG  GSK3B  CAAATGGGCGAGACACACCT  GGCATTTGTGGGGGTTGAAG  JAK3  GTCGTACCGGCATCTCGTG  TAGGCCAGCTGTTTGACCAC  MAPK1  GACCTACTGCCAGAGAACCC  TTGCTCGATGGTTGGTGCT  MAPK3  CCCCAAGTCAGACTCCAAAGC  TCCTTAGGTAGGTCATCCAGC  MDM2  GGCAGGGGAGAGTGATACAGA  GAAGCCAATTCTCACGAAGGG  PIK3CA  TAGGCAAGTCGAGGCAATGG  CTGGTCGCCTCATTTGCTCA  SRC  ACAACACAGAGGGAGACTGG  CACGTAGTTGCTGGGGATGT  ACTA2  CTCCGGAGCGCAAATACTCT  CCCGGCTTCATCGTATTCCT  COL1A2  GAGGGCAACAGCAGGTTCACTTA  TCAGCACCACCGATGTCCAA  COL3A1  AAGTCAAGGAGAAAGTGGTCG  CTCGTTCTCCATTCTTACCAGG  GAPDH  CAAATTCCATGGCACCGTCA  GACTCCACGACGTACTCAGC   6 Evidence-Based Complementary and Alternative Medicine were used as docking receptors. Using AutoDock Vina software, semifexible simulated docking was performed between them. Te docking results were presented as afnity scores obtained with AutoDock Vina software (Table 3). Te larger the absolute value of the docking afnity score, the stronger the binding ability. Te results showed that the docking afnity scores of all compounds had scores greater than −5.0 kcal/mol, and most of them had scores greater than −7.0 kcal/mol, indicating good binding among the top nine targets and fve compounds. Te groups with the highest afnity score with target proteins and compounds were visualized with PyMOL software and are presented in Figure 8. Te abovementioned results suggested that the constituents in ZBM may show favorable interactions with the top nine targets and exert antipulmonary fbrosis efects through these targets. However, further experiments are required to elucidate and validate these fndings.

Efects of ZBM on the Viability of MRC-5 Cells.
We conducted in vitro experiments to further verify the correctness of the network pharmacology method and the results of molecular docking. First, we investigated the efect of the ZBM extract on the viability of MRC-5 cells. As shown in Figure 9, in the absence of TGF-β1 stimulators, ZBM at a concentration of 0.01-10 μg/mL showed no signifcant efect on the cell viability of MRC-5 cells, suggesting that ZBM is safe and reliable within a concentration range of 0.01-10 μg/mL. In comparison with the control group (no stimulation or drug), the cell viability in the model group (only TGF-β1 treatment) increased signifcantly after administration of the TGF-β1 stimulator (10 ng/mL). However, in the group that received simultaneous administration of TGF-β1 and drugs, ZBM at concentrations of 0.1-10 μg/ mL signifcantly inhibited the cell proliferation efect induced by TGF-β1.

Efects of ZBM on mRNA Expression of Collagen and α-SMA in MRC-5 Cells.
Lung myofbroblasts are key efector cells of pulmonary fbrosis [39]. Overexpression of collagen and α-smooth muscle actin (α-SMA) is one of the characteristics of lung myofbroblasts in comparison with normal    Evidence-Based Complementary and Alternative Medicine lung fbroblasts [40]. Terefore, we next investigated the efect of ZBM administration on α-SMA and collagen expression in MRC-5 cells to determine the efect of ZBM on the process of fbroblast transdiferentiation into myofbroblasts. As shown in Figure 10, the mRNA expression levels of ACTA2, COL1A2, and COL3A1 were signifcantly increased in the TGF-β1-stimulated group, while administration of 0.1-10 μg/mL ZBM resulted in signifcant improvement, indicating that ZBM can inhibit the transformation of fbroblasts into myofbroblasts.

Exploration of the Possible Mechanisms Underlying the Terapeutic Efects of ZBM on IPF.
Combined with the results of the above network pharmacology, we performed additional experimental analysis of the core targets to investigate the possible mechanisms underlying the therapeutic efects of ZBM on IPF. As shown in Figure 11, the mRNA expression levels of AKT1, EGFR, GSK3B, JAK3, MAPK1, MAPK3, MDM2, PIK3CA, and SRC were increased in the TGF-β1-stimulated group, while administration of 0.1-10 μg/mL ZBM resulted in various degrees of improvement in these expression levels, indicating that ZBM inhibited IPF partially by regulating the expression levels of these targets and that the systematic analysis strategy based on network pharmacology was reliable.

Discussion
Although many risk factors for IPF, including smoking, drug exposure, pollutant exposure, viral infection, and genetic susceptibility, have been identifed after decades of clinical research, therapeutic options based on the complex pathogenesis of this disease have still not been developed [41]. Multiple biological regulatory processes involving complex signaling pathways, including cellular metabolism, regulation of cell survival state, cellular senescence, immune stress, and hormone regulation, have been reported to contribute to the development of fbrosis [42,43]. Given the suboptimal clinical outcomes of most existing single-targeted or single-pathway small molecule drugs such as SAR156597, simtuzumab, and lebrikizumab, the development of drugs with multitargeted efects may be a better approach for the treatment of IPF [8]. Because of the unique features of TCM, the principles and drugs used in this system of medicine are being increasingly employed in the treatment of and research on IPF worldwide, potentially serving as an attractive source for the identifcation of anti-IPF drugs through a multilevel, multitarget, and multipathway approach [10]. ZBM is frequently used in herbal compounds for the treatment of IPF, but its mechanism of action is still unclear and merits an in-depth investigation. Te existing understanding of the pathogenesis of IPF has been enhanced by ongoing research. In general, under the stimulation of multiple risk factors, repeatedly injured alveolar epithelial cells undergo abnormal activation or apoptosis, resulting in the release of multiple cytokines to promote abnormal recruitment and activation of lung fbroblasts [7,44,45]. Tese activated lung fbroblasts secrete large amounts of extracellular matrix components, which is a key aspect of the pulmonary fbrosis process [46]. In our study, we stimulated the activation of lung fbroblasts in vitro with TGF-β1. We found that ZBM signifcantly reduced the mRNA expression levels of ACTA2, COL1A2, and COL3A1 in the TGF-β1-stimulated lung fbroblasts. Tis result suggested that ZBM had an antipulmonary fbrosis efect in vitro.
For complex diseases, most drug studies focused on a single drug or a single target end in failure. Consequently, with the development of the feld of network pharmacology, an increasing number of studies are employing synergistic, multicompound approaches for therapeutic drug development, which can favorably accelerate clinical translation [47]. In this study, by employing an analytical method based on network pharmacology, we predicted nine targets showing the highest association with the therapeutic efects of ZBM against IPF. Since molecular docking is the most used tool in structure-based drug design and optimization studies [48], we performed molecular-docking simulations using AutoDock Vina software for evaluating the binding of the top nine targets and fve compounds. Te results revealed strong binding abilities of each compound with each target, implying that ZBM may exert antipulmonary fbrosis efects through these targets. To gain a comprehensive understanding of the roles of these nine crucial targets in IPF and the interactions among them, we conducted a literature survey and performed in vitro validation experiments with qRT-PCR.

Evidence-Based Complementary and Alternative Medicine 11
Te phosphatidylinositol 3-kinase/protein kinase B (PI3K/AKT) signaling pathway is considered to be a major regulator of IPF, through its involvement in cell growth, diferentiation, and metabolism [49]. Activation of the PI3K/AKT is closely related to the overexpression of α-SMA in pulmonary fbrosis and interactions with the TGF-β signaling pathway, promoting the development of pulmonary fbrosis [50]. Specifcally, the expression of phosphorylated AKT was shown to be signifcant in livecultured, precisely dissected lung sections obtained from IPF patients, while the progression of fbrosis was suppressed in a concentration-dependent manner after administration of GSK2126458 (a potent and highly selective inhibitor of PI3K) [51]. In our study, the mRNA expression levels of PIK3CA and AKT1 in MRC-5 cells increased after TGF-β1 stimulation and reduced after administration of ZBM, suggesting that ZBM has a regulatory efect on the PI3K/AKT pathway. Mouse double minute 2 (MDM2) mediates the ubiquitination of p53 (one of the key regulators of IPF [52]) and is responsible for the inhibition of p53 activity, making it important for the regulation of cellular homeostasis [53]. Interestingly, AKT can promote the activation of MDM2 through phosphorylation, indicating that the regulation of AKT signaling also afects MDM2 [54]. Furthermore, GSK3B is a serine/threonine kinase that can be activated by the AKT signaling cascade to regulate TGFβ and β-catenin signaling pathways to infuence the development of pulmonary fbrosis [55][56][57]. In our study, we observed the same trend for MDM2 and GSK3B in TGF-β1stimulated MRC-5 cells, whereas the administration of ZBM had an inhibitory efect. Taken together, these fndings suggest that ZBM may regulate the development of pulmonary fbrosis partly through the PI3K/AKT/ MDM2 and PI3K/AKT/GSK3B signaling pathways.
Te mitogen-activated protein kinase (MAPK) signaling cascade, which is responsible for the transmission of extracellular signals to intracellular targets, is critical for cell growth, survival, and stress [58]. Te MAPK signaling cascade has been reported to mediate the regulation of the TGF-β1 signaling pathway, which is critical for the fbrotic process [59]. TGF-β-induced epithelial-mesenchymal transition (EMT) is dependent on the activation of the MAPK signaling cascade [60]. Mefunidone was shown to (h), and SRC (i) was detected using the qRT-PCR assay. Data were presented as the mean ± standard deviation (n � 3). ### p < 0.001 and ## p < 0.01 versus the control group; * * * p < 0.001, * * p < 0.01, and * p < 0.05 versus the model group.
reduce the expression of Snail and vimentin by inhibiting the transduction of the TGF-β/Smad2 signaling pathway and the activation of the MAPK pathway, thereby inhibiting the process of EMT and improving pulmonary fbrosis [61]. In our study, MAPK1 (ERK2) and MAPK3 (ERK1) were enriched. We studied the changes in MAPK1 and MAPK3 expression after administration of TGFβ1 and ZBM and found that ZBM could signifcantly reduce the TGF-β1induced elevation of MAPK1 and MAPK3 expression in MRC-5 cells. Terefore, we speculate that ZBM may partially prevent the development of pulmonary fbrosis by inhibiting the MAPK signaling pathway. Janus kinases (JAKs) are a group of intracellular tyrosine kinases that form the JAK/STAT signaling cascade with downstream signal transducers and activators of transcription (STAT), which play a regulatory role in the development of fbrosis [62]. Te JAK/STAT pathway is activated through the binding of multiple cytokines to their specifc plasma membrane receptors, thereby regulating the cellular transformation of fbroblasts to myofbroblasts, EMT, immune regulation, and cell survival [63,64]. Fei Kang, a TCM compound prescription with benefcial clinical efcacy against IPF, has been reported to play an antipulmonary fbrosis efect by inhibiting the expression levels of proteins related to the JAK1/STAT3 signaling pathway [65]. In our study, ZBM signifcantly reduced the expression level of JAK3 after TGF-β1 stimulation in MRC-5 cells, indicating that ZBM regulated the JAK signaling pathway. On the basis of these fndings, we speculate that ZBM can also exert antipulmonary fbrosis efects through the JAK/STAT signaling axis. Interestingly, the epidermal growth factor receptor (EGFR), a member of the ErbB tyrosine kinase receptor family, was shown to participate in wound healing and tissue regeneration by activating downstream PI3K/AKT, MAPK/ERK, and STAT signaling pathways [66]. In our study, we found that ZBM also signifcantly modulated the expression of EGFR and SRC, a nonreceptor tyrosine kinase that activates STAT3 and transactivates EGFR. Terefore, we speculated that EGFR and SRC act as regulators of the abovementioned multiple pathways.
In summary, the use of ZBM in the management of IPF may be based on a multitarget and multipathway approach. Figure 12 presents a general illustration of this approach. Tese pathways are interconnected and synergistically infuence the overall development of IPF. However, our study only tentatively proposed the partial mechanism of ZBM against IPF on the basis of computational predictions from network pharmacology analyses. Other unproven or relevant targets may also regulate IPF after administration of ZBM. Terefore, in vivo experiments, gene silencing, and other experimental methods should be considered in future studies to comprehensively describe the mechanisms of ZBM against IPF. Tis study has laid the foundation for further clinical applications of ZBM and the development of related therapeutic products.

Conclusions
In this study, we used a systematic analysis strategy based on network pharmacology to predict the potential mechanisms underlying the efects of ZBM against IPF. Our results suggested that ZBM improves IPF partly by infuencing the trans-diferentiation of fbroblasts to myofbroblasts through multiple targets such as EGRF, AKT1, MAPK1, SRC, PIK3CA, MAPK3, and JAK3 and multiple pathways such as the PI3K/AKT, MAPK, and JAK/STAT signaling pathways. Tis study provided new insights for elucidating the mechanism of TCM against IPF and will facilitate the further development of drugs related to ZBM.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.