Multitarget and Multipathway Regulation of Zhenqi Fuzheng Granule against Non-Small Cell Lung Cancer Based On Network Pharmacology and Molecular Docking

Background and Objective. The morbidity and mortality rates of non-small cell lung cancer (NSCLC) remain high. Zhenqi Fuzheng (ZQFZ) granule, which consists of Astragali Radix and Ligustri Lucidi Fructus, is commonly used to improve the immunity of cancer patients. However, the mechanism of ZQFZ granule against NSCLC is still unclear. In this study, the network pharmacology and molecular docking approaches were used to investigate the potential mechanism of ZQFZ granule on NSCLC. Methods. The ingredients in the ZQFZ granule were considered in one study based on UPLC, and the potential targets were predicted in the SwissTargetPrediction database. NSCLC targets were gathered from GeneCards, OMIM, and TTD databases. The ingredient-target-NSCLC network was drawn by Cytoscape. The protein–protein interaction was obtained from the STRING database, and the gene function and biological pathways were analyzed by Metascape. AutoDock Vina was used to verify the molecular docking between the key compounds and core targets, and PyMol visualized the results. Results. 244 targets were related to 13 candidate compounds and 1904 targets were related to NSCLC, of which a total of 106 anti-NSCLC targets were predicted. The compound-target-NSCLC network indicated that sinapinic acid, ferulic acid, asiatic acid, pratensein, and glycitein might be the key components for treating NSCLC. The 41 vital targets (out of 106 targets) above the median calculated by PPI degree were selected for bioinformatics analysis. The top 10 targets out of 41 ranked by MCC were IL-6, SRC, CTNNB1, STAT3, CASP3, TNF, EGFR, MAPK8, HSP90AA1, and PTGS2. ZQFZ granule treatment for NSCLC involved many pathways through KEGG analyses, which included pathways in cancer (hsa05200), proteoglycans in cancer (hsa05205), endocrine resistance (hsa01522), microRNAs in cancer (hsa05206), PI3K-Akt signaling pathway (hsa04151), and IL-17 signaling pathway (hsa04657). Molecular docking studies revealed that sinapinic acid, ferulic acid, asiatic acid, pratensein, and glycitein had good infinity with most core targets. Conclusions. This study indicated that ZQFZ granule with multicompounds could treat NSCLC through multitargets and multipathways.


Introduction
GLOBOCAN 2020 showed that there were an estimated 19 million new cancer cases and 10 million deaths worldwide in 2020. Te cancer morbidity and mortality rates remained high, with breast cancer ranking frst among women and lung cancer being the leading cause of cancer death [1]. Surgery, radiotherapy, chemotherapy, targeted therapy, immunotherapy, and adjuvant therapy with Chinese medicine are the common means of cancer treatment. It is well known that Chinese medicine compounding has a long history with multicomponent and multitarget characteristics. In complementary treatment, Chinese medicine compounding has a great efect on increasing chemotherapy drugs' chemosensitivity, reducing adverse drug reactions and toxicity, relieving patients' pain, and improving life quality particularly [2].
As a traditional Chinese medicine (TCM), ZQFZ granule has the ability to improve immunity and protect bone marrow and adrenal cortex functions [3]. Combined with surgery, chemotherapy, and radiotherapy, ZQFZ granule could efectively reduce adverse efects caused by chemotherapy and promote recovery of immune function, subsequently improving the patient's quality of life and ultimately decreasing the recurrence rate. A prospective, open-label, randomized controlled trial confrmed that TCM treatment could promote NSCLC patients' quality of life, relieve symptoms, and reduce adverse events. Te TCM was composed of the immuno-boosting drug called Kangai injection, herbal decoction, and ZQFZ capsule [4].
ZQFZ granule consists of Astragali Radix (AsR, dried root of Astragalus membranaceus, known as HuangQi in Chinese) and Ligustri Lucidi Fructus (LLF, dried ripe fruit of Ligustrum lucidum, known as NvZhenZi in Chinese). Triterpenoids and phenylethanoid glycosides are the two major types of constituents present in AsR [5]. As one of the most commonly used herbal remedies for cancer in China, AsR can activate immune regulation, inhibit cell proliferation, and attenuate adverse efects caused by cytotoxic therapy [6]. Similarly, the chemical constituents of LLF mainly include saponins, favonoids, and nitrogen-containing chemicals [7]. Studies have shown that LLF can enhance the sensitivity of adriamycin-induced apoptosis, improve immunity, and promote apoptosis, so as to inhibit tumor growth [8]. Te classical formulation of ZQFZ is a (2 : 1, w/w) mixture of AsR and LLF. Te mixture was decocted twice with 10-30 times water, and the fltrate was concentrated into a thick paste, mixed evenly with sucrose, and dried into granules [9].
However, due to the complexity of the components of Chinese herbal compounds, the mechanisms and targets of reaction are difcult to elucidate. Te mechanisms are being studied by examining the impacts of monomeric components, which are not exhaustive. Based on the whole perspective, network pharmacology combines network biology and multipharmacology to compensate for the shortage of multicomponent system research of TCM. In this study, we focused primarily on the active components from a recent study based on ultra-performance liquid chromatography (UPLC) [10] and explored the molecular mechanism of ZQFZ granule for the treatment of NSCLC. Te fowchart is shown in Figure 1.

Collection of the Main Ingredients from ZQFZ Granule and ADME Evaluation.
Te ingredients in ZQFZ granule were considered from one study based on UPLC [10]. Te main active compounds were selected according to Lipinski's rule of fve (RO5) [11]. Te criteria of Lipinski's RO5 were as follows: a molecule weight (MW) should be less than 500, the number of hydrogen bond donors (Hdon) should be less than 5, the number of hydrogen bond acceptors (Hacc) should be less than 10, lipid-water partition coefcient (LogP) no more than 5, and the number of rotatable bonds (Rbon) no more than 10. Te chemical structures and canonical SMILES were searched in PubChem (https:// pubchem.ncbi.nlm.nih.gov/) and the ADME of compounds were uploaded into SwissADME (https://www. swissadme.ch/) to evaluate the druggability.

Protein-Protein Interaction (PPI) Analysis and Vital
Target Screening. Te PPI of common genes was analyzed in STRING (https://string-db.org/). Te organism was set to Homo sapiens and the confdence of 0.4 was selected as signifcant. Te active interaction sources for PPI analysis and median calculation included known, predicted, and other interactions. Te PPI network was analyzed and visualized using the Cytoscape software. Te node size represents the degree value and the edge represents the connection between proteins. In this study, the vital genes, which were above the median calculated by degree, were then used for biological function analysis. Moreover, the top 10 proteins ranked by maximal clique centrality (MCC) using the cytoHubba plug-in were selected as core targets.

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and
Genomes (KEGG) Pathways Enrichment Analyses. GO analysis of the vital genes, including the biological process (BP), cellular component (CC), and molecular function (MF), was enriched and analyzed by the Metascape database (https://metascape.org/gp). Te screening thresholds were a minimum of 3 genes, p < 0.01 and an enrichment factor greater than 1.5. Te same method and settings were used for KEGG pathway enrichment. Finally, the results were plotted as bubble charts using the online bioinformatics tool (https://bioinformatics.com.cn/). Furthermore, the Cytoscape plug-in ClueGo was used for analyzing the relevant KEGG pathway.

Molecular Docking.
To verify the binding of the vital ingredients of ZQFZ granule with predicted core genes, compounds in mol2 format were downloaded from TCMSP (https://tcmsp-e.com/) and the core protein structures were acquired from RCSB PDB (https://www.rcsb.org/). Te compounds and proteins were imported to AutoDockTools 1.5.6 software for dehydration, hydrogenation, and other pretreatments. Ten, molecular docking for analyzing the binding activity was performed using AutoDock Vina 1.1.2 [12] software, and some great results were visualized using PyMol 4.6.0 software.

Te Chemical Structure and ADME Properties of the Vital
Ingredients from ZQFZ Granule. A recent study has identifed 95 chemical components from ZQFZ granule including 15 batches from 3 producers based on UPLC [10]. Te chemical structures of the compounds were searched in the PubChem database, and ADME properties were evaluated in SwissADME. A total of 15 compounds met Lipinski's RO5, indicating that these components had good drug-like properties. Moreover, potential target genes were predicted based on the chemical structure via the Swis-sTargetPrediction database, and two of them had no targets. Terefore, we selected 13 compounds ( Figure 2 and Table 1) as candidate compounds.

Construction of a Candidate Ingredients from the ZQFZ Granule Target NSCLC Network.
To study the mechanism of ZQFZ granule in the treatment of NSCLC, 106 common targets and 13 candidate compounds were used for constructing the potential ZQFZ granule target NSCLC network. As shown in Figure 4, all compounds were associated with multiple targets, generating a total of 344 componenttarget connections between the 13 compounds and 106 Evidence-Based Complementary and Alternative Medicine targets. Te network revealed sinapinic acid (degree = 40) had the most targets, followed by ferulic acid (degree = 33), asiatic acid (degree = 30), pratensein (degree = 24), and glycitein (degree = 21), indicating that these compounds from ZQFZ granule were highly likely to become key components for treating NSCLC.

Protein-Protein Interaction (PPI) Analysis.
Te 106 common targets of the predicted ZQFZ granule and NSCLC were explored in the STRING database and protein interaction relationships were gained ( Figure 5(a)). To further study the relationship between these targets, 41 vital genes above the median calculated by degree were selected and the protein interaction relationships were analyzed via Cytoscape software. Table 2 shows the details of the 41 vital targets. Te PPI network generated 41 nodes, which represent proteins, and 552 edges which represent the connections between proteins. Te larger node means the greater degree ( Figure 5(b)). Furthermore, the top 10 targets ranked by MCC using the cytoHubba plug-in were IL-6,     Figure 5(c)) and selected as core targets for interaction with NSCLC.  Supplementary Table S2.

Molecular Docking Simulation.
According to the potential compounds from the ZQFZ granule target NSCLC network (Figure 4), sinapinic acid, ferulic acid, asiatic acid, pratensein, and glycitein had a high number of targets against NSCLC and were the top 5 compounds ranked by degree. Te PDBIDs and resolutions of the core target proteins are listed in Supplementary Table S3. Tus, molecular docking predicted the binding of the fve compounds to core targets related to NSCLC. Te afnity is usually regarded as the binding efect of the compound to the target; the lower the absolute value of afnity is, the more stable the binding is. As shown in Figure 9, sinapinic acid showed a strong binding to TNF, with an afnity value of −7.1 kcal/ mol; ferulic acid showed a strong binding to TNF and   Table 4 shows some representative docking results with the core targets. As shown in Table 4, the structure of asiatic acid could form one hydrogen bond, two hydrogen bonds, and two hydrogen bonds with Tyr103, Tyr32, and Asp56, respectively, in IL-6. Respectively, asiatic acid could interact with Leu273 and Asp404 via one hydrogen bond in SRC. Te structure of pratensein could interact with Asp207 and Lys281 through one hydrogen bond in CTNNB1. Te structure of asiatic acid could form one hydrogen bond and two hydrogen bonds with Glu638, and Gln644, respectively, in STAT3. Te structure of asiatic acid could form one hydrogen bond with Arg207, Phe250, and Phe256, respectively, in CASP3. Sinapinic acid could form one hydrogen bond with Tyr195 and Tyr227 in TNF. Glycitein could interact with Lys745 and Gly796 through one hydrogen bond in EGFR. Te structure of pratensein could interact with Ala36, Gln37, and Asn114 through one hydrogen bond, respectively, and Met111 via two hydrogen bonds in MAPK8. Te structure of ferulic acid was linked to Asn51, Tyr139, and Trp162 through one hydrogen bond,

Discussion
Lung cancer remains the leading cause of cancer death, accounting for 18% of all cancer deaths [1]. To explore the active ingredients and potential mechanisms of ZQFZ granule for the treatment of NSCLC, this study comprehensively investigated the therapeutic efect of ZQFZ granule in NSCLC.
Generally, compounds for network pharmacology studies are obtained from public databases, while the efect of the content may be ignored, resulting in inaccurate predictions. Especially, active compounds were derived from the study that analyzed 15 batches from 3 producers, which signifcantly increased the credibility of data [10]. Based on Lipinski's RO5 and SwissTargetPrediction databases, 13 compounds were selected as candidate compounds and 106 potential targets were screened against NSCLC. Moreover, a potential compound from the ZQFZ granule target NSCLC network implied that sinapinic acid, ferulic acid, asiatic acid, pratensein, and glycitein were probably the critical components for NSCLC treatment. Sinapinic acid, one of the most common hydroxycinnamic acids, has an anticancer efect on prostate cancer cells. Treatment with sinapinic acid (1 mM for 72 h) could eliminate half the number of PC-3 and LNCaP cells [13]. Ferulic acid, as a 4-hydroxy-3-methoxycinnamic acid, can inhibit phorbol-12-myristate-13-acetate (PMA)-stimulated invasion of A549 cells at a concentration  [14]. Asiatic acid, as a natural triterpene, has been proven to enhance the sensitivity of multidrugresistant A549 cells to cisplatin by downregulating Pglycoprotein (MDR1) and its targets [15]. Glycitein had signifcant cytotoxic efects on gastric cancer cells by inducing cell apoptosis and G0/G1 phase cell cycle arrest via the ROS-related MAPK/STAT3/NF-κB signaling pathway [16]. Terefore, these compounds play an important role in the treatment of tumors. Notably, there are few studies on the treatment of NSCLC with sinapinic acid, pratensein, and glycitein. Te 106 targets in common between ZQFZ granule and NSCLC were considered potential targets. Te top 10 core targets shown in Figure 5(c) included IL-6, SRC, CTNNB1, STAT3, CASP3, TNF, EGFR, MAPK8, HSP90AA1, and PTGS2. Tese target genes are associated with the regulation of tumorigenesis. IL-6, STAT3, and TNF are involved in immunoactivation and play critical roles in the development of NSCLC. CTNNB1 and EGFR are highly expressed in carcinomas such as lung and breast cancers. SRC is related to endocrine regulation and CASP3 is commonly regarded as the predominant terminal shear enzyme in cell apoptosis. Furthermore, the relationship between the potential compounds and the core targets could be validated by the literature. For instance, sinapinic acid signifcantly repressed TNF levels and activated CASP3 in acute doxorubicin-induced cardiotoxicity rats [17]. Ferulic acid inhibited UVB-induced TNF and IL-6 protein expression in mice skin tissue, which might be related to the activation of MAPK and NF-kappa B signaling pathways [18]. Asiatic acid induced HT-29 cell apoptosis via CASP3 activation and inhibited the growth and metastasis of breast cancer in mice by downregulating SRC protein expression [19,20].  Figure 7: A compound-target-pathway network. Te green round rectangle represents the chemical compounds, the pink circles represent the targets, and the yellow hexagon represents the signal pathways.
Te core targets of ZQFZ granule for the treatment of NSCLC were applied to obtain an enrichment map of GO and KEGG pathway analyses via the Metascape database. Te results showed BP was involved in oxidative stress, protein phosphorylation, protein-tyrosine kinase signaling pathway, and infammatory response. CC was primarily concerned with various cell bodies such as membrane, complexes, and cytoplasm. Additionally, MF was mainly associated with protein kinase binding. Furthermore, the main signaling pathways were endocrine resistance, PI3K/ Akt, IL-17, and NF-kappa B signaling pathways, while the diseases involved were cancers, atherosclerosis, and measles.
IL-6 is a pleiotropic cytokine with various biological functions in immunity, tissue regeneration, and metabolism.
It is highly expressed in lung cancer and negatively correlated with survival [21]. Blocking IL-6 expression can inhibit lung cancer promotion, cell proliferation, angiogenesis markers, and tumor cell-intrinsic STAT3 activation [22]. STAT3 is mainly expressed in naive CD4+ T cells and activated p-STAT3 can modulate the transcriptional activity of target genes associated with tumor cell migration and invasion [23]. IL-6/STAT3 signaling is activated in lung tumorigenesis and metastasis. IL-6 can activate STAT3 signaling and escape host immunity by upregulating and recruiting granulocyte-likemyeloid-derived suppressor cells and type II macrophage polarization [24]. EGFR, a member of the receptor tyrosine kinase ERBB family, is an important oncogene in NSCLC progression. EGFR-TKIs are used for

14
Evidence-Based Complementary and Alternative Medicine treating NSCLC by modulating the immune microenvironment. However, EGFR-TKI resistance is widely present in NSCLC and promotes immune escape via increased PD-L1 expression and, subsequently, activates the PI3K/ AKT/mTOR pathway aberrantly [25,26]. Te PI3K/Akt signaling pathway, a classical mediator that regulates cell growth and infammation, was found to be activated in this study. Tere are 17 targets identifed in this pathway, and IL-6, TNF, EGFR, MAPK8, and HSP90AA1 are core targets. Tumor necrosis factor (TNF), a cytokine, can combine with extracellular death receptors to activate the apoptosis process and regulate the infammatory microenvironment via the PI3K/Akt pathway in NSCLC [27,28]. Upregulation of HSP90AA1 is related to poor overall survival in cancer patients [29]. Tis could be due to the fact that HSP90AA1 overexpression reduces immune surveillance and resists foreign substances [30]. Tis is a novel way to study the mechanism of ZQFZ granule in the treatment of NSCLC via a network pharmacologic method. Te selected compounds of ZQFZ granule were derived from a recent study based on UPLC. Te results may be useful to better understand the multipathway and multitarget regulation of ZQFZ granule against NSCLC.
In total, 13 compounds in the ZQFZ granule and 106 common targets against NSCLC were screened by network pharmacology analysis. GO and KEGG enrichment analyses implied the 13 compounds in the ZQFZ granule could treat NSCLC via multiple NSCLC pathological processes. However, the theoretical observations still need to be validated by clinical studies in the future. Tis study provided a theoretical basis for the clinical application of ZQFZ granule in NSCLC.

Conclusion
Tis study revealed the pharmacological mechanism of ZQFZ granule in NSCLC based on network pharmacology and molecular docking. GO and KEGG enrichment analyses implied that the 13 compounds in the ZQFZ granule played an important role in the treatment of NSCLC. Tis study provided a theoretical basis for further research on the clinical application of ZQFZ granule in NSCLC.

Supplementary Materials
Table S1. Te 106 common targets of ZQFZ granule against NSCLC. Table S2. Top 5 BP, CC, and MF pathway terms