Based on Network Pharmacology and Molecular Docking to Explore the Underlying Mechanism of Huangqi Gegen Decoction for Treating Diabetic Nephropathy

Background Huangqi Gegen decoction (HGD), a Chinese herb formula, has been widely used to treat diabetic nephropathy in China, while the pharmacological mechanisms are still unclear. Therefore, the present study aims to explore the underlying mechanism of HGD for treating diabetic nephropathy (DN). Materials and Methods Traditional Chinese Medicine Systems Pharmacology Database (TCMSP), UniProt, and SwissTargetPrediction databases were used to search the active ingredients and potential targets of HGD. In addition, multiple disease-related databases were used to collect DN-related targets. Common targets of the protein-protein interaction (PPI) network were established using the STRING database. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed using the DAVID database. At last, AutoDockVina was used to conduct molecular docking verification for the core components and targets. Results A total of 27 active ingredients and 354 putative identified target genes were screened from HGD, of which 99 overlapped with the targets of DN and were considered potential therapeutic targets. Further analysis showed that the HGD activity of quercetin, formononetin, kaempferol, isorhamnetin, and beta-sitosterol ingredients is possible through VEGFA, IL6, TNF, AKT1, and TP53 targets involved in TNF, toll-like receptors, and MAPK-related pathways, which have anti-inflammatory, antiapoptosis, antioxidation, and autophagy effects, relieve renal fibrosis and renal cortex injury, and improve renal function, thus delaying the development of DN. The molecular docking results showed that quercetin, formononetin, kaempferol, isorhamnetin, beta-sitosterol had a good binding activity with VEGFA, IL6, TNF, AKT1, and TP53. Conclusion This study demonstrated that HGD might take part in the treatment of DN through multicomponent, multitarget, and multichannel combined action.


Introduction
Diabetic nephropathy (DN) is one of the most important microvascular complications of diabetes, accounting for about 20%-40% of all diabetic patients [1]. It has become the main cause of chronic kidney disease and end-stage renal disease.
e main clinical manifestations were persistent proteinuria, hyperglycemia, hypertension, edema, and decreased glomerular filtration rate. e etiology and pathogenesis of DN are still unclear. It is believed that DN is caused by multiple factors such as long-term hyperglycemia, glomerular ultrafiltration, extracellular matrix deposition, activation of the polyol pathway, accumulation of advanced glycation end products, protein kinase C activation, oxidative stress, and genetic factors [2]. At present, the clinical treatment for DN is limited, and it is still mainly to control blood sugar, reduce blood pressure, and reduce urinary protein.
Huangqi Gegen decoction (HGD), derived from ancient books, has the effects of invigorating the spleen and replenishing Qi, nourishing Yin and clearing heat, invigorating body fluid, and relieving thirst. It can be used in hypertension and diabetes due to deficiency of both Qi and Yin. e whole prescription is composed of Astragalus membranaceus (Huangqi) and Pueraria (Gegen). It has been found that Astragalus membranaceus has a therapeutic effect on DN characterized by renal insufficiency and pathological changes by reducing blood glucose and antioxidation [3]. Pueraria extract exerts an antidiabetic effect by activating a variety of mechanisms, such as reducing insulin resistance, increasing insulin release, inhibiting glucose absorption and reabsorption, and improving insulin sensitivity, glucose uptake, and metabolism [4]. HGD can reduce fasting blood glucose and improve insulin resistance in diabetic rats [5]. However, the specific mechanism and related pathways of the action of HGD related to DN remain elusive.
Network pharmacology integrates multidisciplinary technology and content to construct a drug-target-disease network to explore the relationship between diseases and drugs [6]. At the same time, molecular docking promotes the pharmacological action to the molecular level and reveals the mechanism of drug action at the molecular level. Network pharmacology combined with molecular docking can make up for one-sided and unsystematic research on the mechanism of drug therapy. erefore, this study explored the mechanism of HGD in the treatment of DN through network pharmacology method and molecular docking technology and provided new ideas for clinical application. Our flowchart is shown in Figure 1. (Table 1)

Collection of Active Ingredients and Potential Targets.
e active components of Astragalus membranaceus and Pueraria were obtained by searching "Huangqi" and "Gegen" in the TCMSP database. According to absorption, distribution, metabolism, and excretion (ADME) parameters, the active components of Astragalus membranaceus and Pueraria were obtained under the conditions of oral bioavailability (OB) ≥ 30% and drug-likeness (DL) ≥ 0.18, combined with related literature reports. e potential target information of active components was retrieved one by one through the UniProt database to obtain the corresponding gene names of drug targets. For other active ingredients that have not been retrieved, first, the 2D structure of their chemical components was downloaded from the PubChem website and saved in SDF format and then submitted to the SwissTargetPrediction database one by one to obtain the corresponding target of compounds.

Construction of Active Components-Target Network.
e active components and drug targets of HGD obtained in 2.2 were used to construct the active ingredients-target network through Cytoscape software.

Collection of DN Disease Targets.
Here, "diabetic nephropathy" was used as the keyword to search DN-related targets in three databases: GeneCards, OMIM, and CTD. In the integration of disease targets, the larger the score value of the target is, the closer the relationship between the target and the disease is. erefore, the target is screened with score ≥10 as the standard in the GeneCards database, and the target database of DN is established by combining the targets in the three databases.

Core Target Screening and PPI Construction.
Using the online tool of Venny, the potential targets of active components and the disease targets of DN were introduced, and the predictive targets of active components of HGD acting on DN were screened, and Venn maps were drawn. Upload the prediction target to the STRING database, limit the species to "Homosapiens", select the medium confidence data with score condition greater than 0.40, hide the isolated proteins in the map, and finally derive the visual PPI network map. At the same time, the network is imported into Cytoscape to further analyze the core targets.

GO and KEGG Pathway Enrichment Analysis.
DAVID database was used to analyze the enrichment of GO and KEGG pathways, and the gene type was "official gene symbol," and the species type was "Homo sapiens." Finally, GO functional annotation and KEGG pathway enrichment analyses were performed using the Weshengxin online platform, and a P value less than 0.05 was employed for further analysis.

Construction of Main Active Components-DN-Core Targets-Signal Pathways Network.
e main active components, DN key targets, core targets, and signal pathways of HGD were introduced into Cytoscape to construct the network of main active components-DN key targets-core targets-signal pathways.

Molecular Docking.
e core components of HGD were docked with the top five targets of the degree in 1.5. Firstly, the 3D structure of the target protein was downloaded from the PDB database, and the unique ligands similar to small molecules were found as protoligands. e target protein is then removed from water molecules and original ligands by PyMOL and introduced into AutoDockTools for hydrogenation; its charge is calculated; then, it is combined with nonpolar hydrogen and and saved into "pdbqt" format. Finally, AutoDock Vina was run for molecular docking, and PyMOL was used to visualize the results.

Active Components and Potential Targets of HGD.
Twenty-seven active components of Astragalus membranaceus and 18 active components of Pueraria were found by the TCMSP database. After screening by ADME parameters, 20 active components of Astragalus membranaceus and four active components of Pueraria were obtained. By consulting the literature, it was found that Astragalus polysaccharides can inhibit the TGF-β/Smad pathway and protect the kidney in DN model rats [17]. According to the literature, it was found that astragaloside IV can inhibit glomerular mesangial overproliferation and renal fibrosis through the TGF-β1/Smad/miR-192 pathway in rat glomerular mesangial cells and DN model rats, thus reducing renal injury [18]. Moreover, it was found that puerarin can inhibit the expression of NOX4 in podocytes, alleviate kidney injury in DN mice, and protect the kidney of DN mice by reducing oxidative stress and inhibiting inflammation [19,20]. erefore, astragalus polysaccharides, astragaloside IV, and puerarin found in the literature were also included in the active components of HGD, a total of 27 active components (Table 2). ere were 1176 targets corresponding to each active component of HGD, and a total of 354 targets were obtained after screening and deleting repetitive values.

Active Ingredient-Target Network.
ree hundred fiftyfour targets and 27 active ingredients were imported into the Cytoscape software to construct the active ingredient-target network, and a network diagram composed of 351 nodes and 717 edges was obtained. e node size was proportional to its degree value, as shown in Figure 2

Collection of DN Disease Targets.
e DN target is obtained using three databases: GeneCards, OMIM, CTD. Among them, 3320 targets were obtained by GeneCards, and 655 targets were screened by GeneCards under the condition of score ≥10. Two hundred fifty-one targets were obtained by OMIM, and 114 targets were obtained by CTD. A total of 845 DN disease targets were obtained by combining and deduplication of the above targets.

Core Target Screening and PPI Construction.
e potential targets of active components of HGD were intersected with DN disease targets by Venny2.1.0 online tool, and finally, 99 main targets were obtained, and the Venn diagram was drawn, as shown in Figure 3. Upload the main targets to the STRING11.0 database to build the PPI network, as shown in Figure 4, which has 99 nodes and 1983 edges. e PPI results were imported into Cytoscape software and analyzed by the network analyzer. e results showed that the average degree of PPI was 40.06061, the average closeness centrality was 0.630402, and the average betweenness centrality was 0.006467. e targets that were all higher than the average were selected as the core targets of the active components of HGD acting on DN. A total of 27 targets were selected, as shown in Table 3. Among them, the top five target proteins were VEGFA, IL6, TNF, AKT1, and TP53, which indicated that the active components of HGD might treat DN through these targets.

GO and KEGG Pathway Enrichment Analysis.
e above 27 core targets were uploaded to the DAVID database for enrichment analysis of GO and KEGG pathways. Setting P < 0.01, a total of 125 biological processes, 10 molecular functions, 13 cell components, and 81 signal pathways were obtained.
According to the size of the P value, the top 10 functional information sets are selected for analysis and mapping ( Figure 5). e main biological processes involved in the core targets are a cellular response to hypoxia, positive regulation of transcription, angiogenesis, positive regulation of nitric oxide biosynthetic process, positive regulation of gene expression, positive regulation of sequence-specific DNA-binding transcription factor activity, positive regulation of transcription from RNA polymerase II promoter, negative regulation of the apoptotic process, positive regulation of protein phosphorylation, and positive regulation of endothelial cell proliferation. ese results suggest that HGD may play an important role in the treatment of DN by regulating these biological processes. In addition, the main molecular functions involved are extracellular region, cytosol, extracellular space, caveola, nucleoplasm, PML body, nucleus, cytoplasm, neuron projection, and mitochondrion. At the same time, the main cellular components involved include enzyme binding, identical protein binding, transcription factor binding, protein binding, core promoter sequence-specific DNA binding, cytokine activity, transcription regulatory region DNA binding, kinase activity, sequence-specific DNA binding, and growth factor activity. ese may play an important role in the occurrence and development of DN.
According to the size of the P value and the literature, the enrichment results of 10 KEGG pathways were selected.
e specific information is shown in Table 4. It includes TNF signaling pathway, toll-like receptor signaling pathway, MAPK signaling pathway, FoxO signaling pathway, HIF-1 signaling pathway, NOD-like receptor signaling pathway, VEGF signaling pathway, PI3K-Akt signaling pathway, insulin resistance, and mTOR signaling pathway. e results showed that HGD alleviated the DN by regulating inflammatory reaction, autophagy, immunization, and antioxidant stress. Visualize the results, as shown in Figures 6 and 7 .

Construction of Main Active Components-DN-Core Targets-Signal Pathways Network.
e network of the active components-DN-core targets-signal pathways was constructed by Cytoscape software, as shown in Figure 8.
e analysis results show that there are 61 nodes and 208 edges in the network. e larger the nodes, the more significant the impact of the node on DN. e quercetin degree, closeness centrality, and betweenness centrality are 23, 0.57142857, and 0.19158668, respectively. It is predicted that quercetin is an important component in the intervention of DN, followed by formononetin, kaempferol, isorhamnetin, and beta-sitosterol. Detailed information is shown in Table 5. Table 1: Involved databases and related analysis platforms in the study.

Name
Website Version TCMSP [7] https://tcmspw.com/tcmsp.php 2.3 UniProt [8] https://www.uniprot.org/ 2020.8.3 PubChem [9] https://pubchem.ncbi.nlm.nih.gov/ 2019.3 SwissTargetPrediction [10] http://www.swisstargetprediction.ch/ 2020 GeneCards [11] https://www.genecards.org/ 5.0 OMIM [12] https://www.omim.org/ 2020.10.16 CTD [13] http://ctdbase.org/ 2020.10.5 STRING [14] https://string-db.org/ 11.0 DAVID [15] https://David.ncifcrf.gov/ 6.8 RCSB PDB [16] http://www.rcsb.org 2020. 11 e core components in HGD were docking with the top five targets with a degree value of 1.5. e results showed that the binding energies were all less than −5 kcal/mol, indicating that they had good binding activity [21]. e specific information is shown in Table 6. Furthermore, the structure matching analysis is carried out using PyMoL software, as shown in Figure 9. e results showed that kaempferol formed π-π conjugation with TYR-59 amino acid residues in TNF, quercetin formed hydrogen bonds with TYR-151, TYR-119, and TYR-119 amino acid residues in TNF, formononetin formed hydrogen bonds with TYR-119 and LEU-120 amino acid residues in TNF, and formononetin formed hydrogen bonds with TYR-229 amino acid residues in AKT1. Quercetin  Evidence-Based Complementary and Alternative Medicine Figure 4: Common target PPI network between HGD and DN. Each bubble node represents a protein, and the 3D structure in the bubble nodes represents that the spatial protein structure is known or predicted. e lines among inner nodes display the relationship between different proteins, and the width of lines was based on the strength of data support. forms a hydrogen bond with three amino acid residues of LYS-276, LEU-295, and GLY-294 in AKT1, and kaempferol forms a hydrogen bond with three amino acid residues of LYS-179, LEU-295, and GLY-294 in AKT1. ese interactions make the structure tend to be stable.

Discussion
In this study, we screened out 27 active components of HGD and 99 targets related to DN by the network pharmacology method. It shows that HGD can play a role in treating DN through multicomponent and multitarget combination, which has research significance for DN treatment. According to the topological analysis results of the component-disease-core target-signal pathway network, the top five core components were quercetin, formononetin, kaempferol, isorhamnetin, beta-sitosterol.
Quercetin is a kind of flavonoid widely existing in various plants. It has anti-inflammatory and antioxidation activities, reduces blood glucose, blood pressure, and lipid, and has other pharmacological effects [22,23]. It has been reported that quercetin can inhibit glomerular mesangial cell proliferation in DN mice by activating the Hippo pathway in DN mice, thereby reducing renal injury [24]. As a nonsteroidal polyphenol phytochemical, formononetin has a wide range of biological characteristics, involving anti-inflammatory, hypoglycemic, and other effects [25,26]. Studies have shown that formononetin can improve glucose and lipid metabolism, reduce oxidative stress, and protect the kidney in diabetic db/ db mice. Moreover, the inhibition of Smad3 protein and related regulatory factors of extracellular matrix could contribute to alleviation of renal fibrosis [27]. Kaempferol has anti-inflammation, antioxidation, antidiabetes, and antiaging effects [28]. In rat mesangial cells induced by advanced glycation end products, kaempferol could increase superoxide dismutase activity and decrease the content of malondialdehyde. It can inhibit the production of reactive oxygen species, the expression of type IV collagen and TGF-β1, and mitochondrial/cytochrome c-mediated apoptosis pathway and improve the recovery of mitochondrial membrane potential [29]. Isorhamnetin protects cardiocerebral vessels, has anti-inflammation and antioxidation effects, and prevents obesity [30]. A recent study showed that isorhamnetin protects the kidney by inhibiting the NF-κB signaling pathway in diabetic rats [31]. Beta-Sitosterol has good antidiabetes and antioxidant effects in diabetic rats [32]. To sum up, most of these core components have anti-inflammatory, antioxidant, and hypoglycemic effects and can protect the kidney. e treatment of DN may be related to anti-inflammatory, antioxidant, and hypoglycemic effects.
In order to predict the mechanism of HGD in the treatment of DN, we carried out the enrichment analysis of the GO and KEGG pathways. e results showed that HGD could regulate the cellular response to hypoxia, positive regulation of transcription, angiogenesis, positive regulation of nitric oxide biosynthetic process, positive regulation of gene expression, and other biological processes, including extracellular region, cytosol, extracellular space, caveola, and nucleoplasm. At the same time, the activity is produced through molecular functions such as enzyme binding, identical protein binding, transcription factor binding, protein binding, and core promoter sequence-specific DNA binding. TNF signal pathway, toll-like receptor signal pathway, MAPK signal pathway, and other signal pathways are important pathways of HGD on DN, which are mainly involved in inflammation, immunity, autophagy, and oxidative stress. e activation of the TNF signaling pathway can lead to glomerular vasoconstriction, a decrease of renal blood flow and glomerular filtration rate, and participation in the development of DN [33]. Moreover, TNF-α can cause inflammation of renal tissue, collagen deposition, and glomerular sclerosis, leading to kidney injury [34]. Toll-like receptor signaling pathway can promote the release of inflammatory factors such as IL-1, TNF-α, and MAPK, aggravate the inflammatory response, and lead to renal injury. It has been proved that toll-like receptor .0 induces the activation of NF-κB, leading to renal cortex inflammation and renal fibrosis [35]. MAPK signaling pathway is related to inflammation, oxidative stress, cell differentiation, cell proliferation, and apoptosis. ere are three major MAPK kinases in mammalian cells, namely, extracellular signalregulated kinases (ERKs), c-Jun-N-terminal kinase, and p38MAPK. It was also found that ERKs and p38MAPK were activated in glomerular mesangial cells induced by high      cytokines to mediate renal injury. erefore, HGD in the treatment of DN is related to a series of inflammatory pathways. HIF-1 signaling pathway mediates hypoxia-induced cell response by regulating genes involved in cell metabolism, glucose utilization, angiogenesis, oxidative stress, apoptosis, and proliferation, inhibits the expression of HIF-1 in a high glucose environment, affects podocyte survival, and leads to glomerular injury [37]. VEGF signaling pathway can stimulate endothelial cell proliferation and play a key role in physiological and pathological angiogenesis in different tissues. In the kidney, VEGF regulates glomerular permeability and maintains renal integrity, while DN is usually associated with impaired angiogenesis and capillary loss [38]. Studies have shown that the level of VEGF in the kidney of DN rats is increased. Inhibition of VEGF can alleviate the problems of glomerular hyperfiltration rate, glomerular hypertrophy, and urinary albumin excretion [39]. A series of studies have shown that activating PI3 K/Akt can not only inhibit the expression of mTOR protein, increase autophagy of renal cells, and improve renal function, but also increase the expression of FoxO protein and reduce the damage of renal podocytes induced by high glucose [40,41]. In addition, molecular docking results showed that    Evidence-Based Complementary and Alternative Medicine the core components of HGD had a strong binding activity with the AKT1 target. erefore, these results also indicate that HGD can play a therapeutic role in DN by regulating oxidative stress and autophagy and counteracting hypoxia and other pathways.

Conclusions
In conclusion, the results of this study show that HGD can initially explore the potential mechanism of HGD in the treatment of DN through multitarget and multipathway combined action of DN. rough network pharmacology analysis, our study revealed that HGD might take part in the treatment of DN through pathways associated with insulin resistance, PI3K-Akt, toll-like receptors, MAPK, and TNF. en, we validated the core targets and key components by molecular docking technology. e results showed that quercetin, formononetin, kaempferol, isorhamnetin, and beta-sitosterol had good binding activity with VEGFA, IL6, TNF, AKT1, and TP53. VEGFA, IL6, TNF, AKT1, and TP53 may be potential targets of HGD in treating DN. ese results suggest that the antidiabetic nephropathy effect of HGD may be related to its direct or indirect regulation of the above potential targets and pathways, which provides a theoretical basis and research ideas for the clinical application of HGD in the treatment of DN. After all, so far, due to the complexity of traditional Chinese medicine compounds, although HGD has been used in the prevention and treatment of DN, its pharmacodynamic material basis and mechanism of action are still unclear. is study reveals the pharmacodynamic material basis and potential mechanism of HGD in the treatment of DN. Of course, some limitations exist in this study due to experimental conditions, time, and other constraints; we are unable to perform in vivo or in vitro experiments to validate the results in a short time. We will definitely conduct in-depth experimental research on the key targets and pathways of this article when conditions permit at a later period.

Data Availability
e data used in the study are available upon request to the corresponding author.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
D-SS performed the work and drafted the manuscript. W-WH, S-XJ, and M-H contributed to interpreting the data and refined the drafted manuscript. All authors read and approved the final manuscript.
Acknowledgments is work was supported by the PhD research startup foundation of Yichun University (212-3360116031).