Mechanism Prediction of Astragalus membranaceus against Cisplatin-Induced Kidney Damage by Network Pharmacology and Molecular Docking

Background Cisplatin is a frequently used and effective chemotherapy drug in clinical practice, but severe side effects limit its use, among which nephrotoxicity is considered the most serious and prolonged damage to the body. Astragalus membranaceus (AM) is a well-known herbal medicine, and modern pharmacological studies have confirmed its antioxidant, immunomodulatory, and antiapoptotic effects. Clinical studies have shown that AM and its active components can attenuate cisplatin-induced kidney damage, but the molecular mechanism has not been fully expounded. Materials and Methods First, the components and targets information of AM were collected from the TCMSP, and the relevant targets of cisplatin-induced kidney damage were accessed from the GeneCards and OMIM databases. Then, the core targets were selected by the Venn diagram and network topology analysis, which was followed by GO and KEGG pathway enrichment analysis. Finally, we construct a component-target-pathway network. Furthermore, molecular docking was carried out to identify the binding activity between active components and key targets. Results A total of 20 active components and 200 targets of AM and 646 targets related to cisplatin-induced kidney damage were obtained. 91 intersection targets were found between AM and cisplatin-induced kidney damage. Then, 16 core targets were identified, such as MAPK1, TNF-α, and p53. Furthermore, GO and KEGG pathway enrichment analysis suggested that MAPK, Toll-like receptor, and PI3K-Akt signaling pathways may be of significance in the treatment of cisplatin-induced kidney damage by AM. Molecular docking indicated that quercetin and kaempferol had high binding affinities with many core targets. Conclusion In summary, the active components, key targets, and signaling pathways of AM in the treatment of cisplatin-induced kidney damage were predicted in this study, which contributed to the development and application of AM.


Introduction
Cisplatin, a heavy metallic compound with powerful anticancer effects, is synergistic with many antitumor drugs without cross-resistance, thus being one of the most frequently used and effective chemotherapy drugs at present [1]. However, it has many serious side effects that greatly limit its therapeutic use, including gastric toxicity, ototoxicity, nephrotoxicity, allergic reactions, and bone marrow suppression [2]. Nephrotoxicity of cisplatin is considered the most serious and prolonged damage to the body [3]. Acute kidney injury (AKI) has been reported in 25-30% of patients roughly, who have received cisplatin chemotherapy [4]. e mechanisms of cisplatin-induced kidney damage are very complicated, involving multiple factors such as oxidative damage, dysfunctional mitochondria, and renal tubular epithelial cell apoptosis [5][6][7][8][9]. Although hydration diuresis and diuretic forced diuresis are commonly used clinically to reduce cisplatininduced kidney damage, the results are still unsatisfactory [10,11]. erefore, there is an immediate need to pursue an effective measure to mitigate cisplatin-induced kidney damage.
Astragalus membranaceus (AM), also known as "Huang Qi" in traditional Chinese medicine, is the dried root of membranous Astragalus or Astragalus mongolicus, which has been widely used in the clinical setting [12]. AM, which has a long history in China, was first recorded in Shennong's Herbal Classic (an ancient classic Chinese herbal medicine book). Modern pharmacological studies have shown that AM contains flavonoids, saponins, and other active ingredients, which have a wide range of pharmacological effects, such as antioxidant, immune regulation, antiapoptosis, and antitumor [13][14][15][16]. A recent study indicated that Astragalus polysaccharide (an active ingredient of AM) could reduce AKI induced by cisplatin through protecting mitochondria, inhibiting oxidative damage, and improving mitochondriamediated apoptosis [17]. Also, according to a meta-analysis, AM injection can alleviate the clinical side effects of platinum-based chemotherapy drugs [18]. However, the targets and specific mechanism of AM against cisplatin-induced kidney damage remain largely unknown.
Recently, network pharmacology has been a potent tool for gaining insights into the mechanisms of drugs [19]. is method explores the drug-component-target-pathway-disease relationship from a systems biology perspective, predicts potential drug treatment mechanisms for diseases, and provides a theoretical basis for the mechanisms of Chinese medicine treatment for diseases [20]. Molecular docking is an efficient computational method for the prediction of the binding affinity between a ligand and a target protein, which can be used to identify potential drug targets [21]. In this paper, the possible targets and the molecular mechanism of AM on the treatment of cisplatin-induced kidney damage were investigated by network pharmacology and molecular docking ( Figure 1).

Screening Main Active Components and Targets of AM.
Information of the active components and targets associated with AM was collected from the TCMSP database (https:// tcmspw.com/, February 20, 2021) [22]. e TCMSP database is a systematized pharmacology resource that provides the ADME (absorption, distribution, metabolism, and excretion) characteristics of Chinese herbal medicines or ingredients including oral bioavailability (OB), drug likeness (DL), blood-brain barrier, Caco-2 permeability, etc. [23]. Among the pharmacokinetic properties mentioned above, OB and DL are the most vital ones for drug delivery, as they help assess the effect of drug distribution in the circulatory system and how drug-like compounds regard factors such as bioavailability [24]. In the TCMSP database, OB is based on the internal model OBioavail 1.1 and DL is evaluated according to molecular descriptors and Tanimoto coefficients [22,25]. In this study, the name of the herb "Huang Qi" was inputted into the searching box to get the active components and related targets. We selected active components with OB ≥ 30% and DL ≥ 0.18 and their corresponding targets. All targets associated with these active ingredients (both validated and predicted targets) were then imported into the UniProt database (https://www.uniprot. org/, February 15, 2021) to get the gene symbol of targets.
e results accessed from GeneCards were screened for the relevance score ≥10.00 as the screening index. We chose Gene Map in "Advanced Search" to search target genes related to disease in the OMIM database. e ultimate disease-related targets were obtained by removing the duplicates.

Screening Core Targets of AM in Cisplatin-Induced Kidney Damage Treatment.
e VENNY 2.1 online platform (https://bioinfogp.cnb.csic.es/tools/venny/) was used for intersection targets between AM and cisplatin-induced kidney damage. e intersection targets were used to constitute the protein-protein interaction (PPI) network using the STRING database (https://string-db.org/, February 28, 2021) [29] with confidence scores larger than 0.7, and the other variables were left at their original values. We then exported the PPI network in TSV format and analyzed its topology properties (degree centrality) using Cytoscape 3.8.0 (https://cytoscape.org/) [30]. Degree centrality (DC) is the most commonly used topological parameter, which is utilized to appraise the central attribute of nodes in the network. e targets with the DC ≥ 2 × median were selected as the core targets.

Constructing and Analyzing PPI Network of Core Targets.
e core targets obtained above were inputted into the STRING database to get the PPI network. All settings were the same as above. en, the PPI network was extracted in TSV format and inputted into Cytoscape 3.8.0 for further analysis.

Functional and Pathway Enrichment Analysis of Core
Targets. To investigate the multiple mechanisms of AM against cisplatin-induced kidney damage at the systematic level further, we used DAVID 6.8 (https://david.ncifcrf.gov/, March 5, 2021) [31] to perform the functional enrichment analysis of core targets, including the Gene Ontology (GO) terms as well as Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways [32,33].

AM Active Component-Target-Pathway Analysis.
To investigate the relationship between active components, targets, and pathways, we constructed the network of active component-target-pathway (C-T-P network). We imported active components, targets, and pathways into Cytoscape 3.8.0 as three types of nodes and connected associated nodes with edges to construct the active components-targetspathways network.

Molecular Docking.
We used molecular docking to predict the binding activity between the active ingredient and the core target. e structural formula of active components was obtained from PubChem (https://pubchem. ncbi.nlm.nih.gov/, March 10, 2021) and stored in SDF format. Afterward, the core target conformations were obtained from the Protein Data Bank database (PDB, https:// www.rcsb.org/, March 10, 2021). e following were the screening conditions: (1) e biological source of protein structure is human; (2) e protein structure is acquired by X-crystal diffraction; (3) e resolution of protein crystals is less than 3Å; and (4) Protein structures with unique ligands are preferred. e structures of the ligands and the protein receptors were prepared by using Discovery Studio (DS, BIOVIA, 2019). en, we used the flexible docking procedure CDOCKER built-in DS for molecular docking with CHARMM force field (Chemistry at HARvard Macromolecular Mechanics).

Database and Software
Summary. e databases and software used in this study are listed in Table 1 for the convenience of readers.

Active Components and Targets of AM.
We gained 86 components from the TCMSP database, mainly including quercetin, isoflavanone, kaempferol, and so forth. A total of 20 active components satisfied the criteria of OB ≥ 30% and DL ≥ 0.18 (Table 2). en, we obtained a total of 200 targets related to these 20 active components from TCMSP. Figure 2 shows the interaction between active components and targets. It could be seen that quercetin had the maximal DC, that is to say, quercetin had the most targets, suggesting that quercetin might be an important active component of AM.

Targets Related to Cisplatin-Induced Kidney Damage.
We found 534 relevant targets of cisplatin-induced kidney damage in GeneCards and 121 targets in OMIM databases. After deleting duplicate targets, 646 disease targets were finally obtained. e top 5 targets for relevance score were cellular tumor antigen p53 (p53), tumor necrosis factoralpha (TNF-α), interleukin-6 (IL-6), angiotensin-converting enzyme, and transforming growth factor-beta 1.

Core Targets of AM in Cisplatin-Induced Kidney Damage
Treatment.
e Venn diagram showed that there were 91 intersection targets of AM and cisplatin-induced kidney   Figure 3). e intersection targets were used to construct the PPI network using the STRING data set. Immediately afterward, the network was introduced into Cytoscape 3.8.0 and the DC of each node was calculated by using the function of "network analyzer." 16 core targets of AM in cisplatin-induced kidney damage treatment were singled out with DC ≥ 33 (2 × 16.5) (Figure 4). Table 3 shows the specific information of the 16 core targets.

Constructing and Analyzing PPI Network.
e 16 core targets were introduced into the STRING database to get the PPI network ( Figure 5(a)). e DC was reckoned with "Network Analyzer" in Cytoscape 3.8.0 to reflect the significance of a target in the network. Significant targets are indicated by darker colors, larger sizes, and higher DC of interaction in the network, such as mitogen-activated protein kinase 1 (MAPK1), TNF-α, IL-6, vascular endothelial growth factor A (VEGFA) (Figure 5(b)).

GO and KEGG Enrichment Analysis.
For the purpose of figuring out the mechanism of AM against cisplatin-induced kidney damage more systematically, the 16 core targets were inputted into DAVID 6.8 to GO and KEGG enrichment analysis.
ere were 186 items related to biological processes, mainly including MAPK cascade, positive regulation of smooth muscle cell proliferation, inflammatory response,  Evidence-Based Complementary and Alternative Medicine negative regulation of the apoptotic process, and so on. In the cell composition, the extracellular region is the main classification of these targets. Additionally, there were 23 terms enriched in the molecular function category, mainly including protein serine/threonine kinase activity, cytokine activity, transcription factor binding, growth factor activity, and so on. We selected the top 10 items according to the gene ratio, which are presented in Figure 6. e KEGG enrichment pathways analysis results revealed that there were 91 pathways in total, mainly include MAPK signaling pathway, PI3K-Akt signaling pathway, TNF signaling pathway, etc. We selected the top 20 pathways of KEGG according to the gene ratio, which are presented in Figure 7.

Active Component-Target-Pathway Analysis.
For the aim at clarifying what the relationships between components, core targets, and pathways are, the significant pathways and biological processes, core targets, and active components of AM were used to build the C-T-P network (Figure 8). e targets related to pathways and biological processes are also listed in Table 4.

Molecular Docking.
To identify potential drug-target interactions, we performed molecular docking to predict the binding affinities between the active components and core targets in the C-T-P network. Docking results are shown in Table 5. Herein, CDOCKER_ENERGY was used to evaluate the binding affinities. Lower binding energy represents a higher affinity between a ligand and a protein [34]. It can be seen that quercetin has a higher affinity with MAPK1, p53, matrix metalloproteinase-9 (MMP-9), epidermal growth factor receptor (EGFR), and caspase-3, suggesting that quercetin may mitigate cisplatin-induced kidney damage by binding to them.

Discussion
Cisplatin-induced kidney damage is a renal impairment caused by cisplatin, which is extremely harmful to human beings. It is reported that about 25-30% of patients developed symptoms of AKI after receiving cisplatin chemotherapy [4]. Besides hydration combined with mannitol diuretic, injection of AM or its active components is often used for treatment in the clinic [35][36][37]. Although the anticisplatin-induced kidney damage efficacy of AM has been demonstrated, the mechanism has not yet been clarified. Consequently, network pharmacology and molecular docking were performed to analyze the active components, targets, and related signaling pathways of AM against cisplatin-induced kidney damage and to figure out the potential mechanism of AM anti-cisplatin-induced kidney. We obtained 20 active components and 200 corresponding targets from the TCMSP database and then constructed a component-target network, which mainly included quercetin, formononetin, isorhamnetin, and kaempferol. Modern pharmacological studies showed that quercetin has an anti-cisplatin-induced kidney injury effect, which is related to its antioxidant activity and the ability to inhibit kidney inflammation and renal tubular cell apoptosis [38]. In a recent study, it was reported that pretreatment with kaempferol reduced cisplatin-mediated oxidative stress, apoptosis, inflammation, kidney injury, as well as improved its function [39]. Formononetin, an O-methylated isoflavone, is one of the main bioactive ingredients in red clover plants. It was reported that formononetin can alleviate AKI induced by cisplatin and has many potential pharmacological effects, like anti-inflammatory, antioxidant, and antiapoptosis [40]. Isorhamnetin, a metabolite of quercetin, not only has anti-inflammatory and antioxidant effects but also enhances the anticancer effect of cisplatin [41,42]  Evidence-Based Complementary and Alternative Medicine addition, it is interesting that multiple AM active components can act on multiple different targets and have common pharmacological effects, which embodies the peculiarities of AM multicomponent, multitarget synergistic treatment. We obtained 91 intersection targets between AM and cisplatin-induced kidney damage and further screened out 16 core targets of AM in the treatment of cisplatin-induced kidney damage. e targets are mainly associated with oxidative stress, apoptosis, inflammation, and cell proliferation. According to quantities of studies, apoptosis induced by TNF-α, p53, and caspase-3 [43,44] and inflammation induced by IL-6, TNF-α, and interleukin-1 beta (IL-1β) played an essential role in cisplatin-induced kidney damage [45,46]. It was reported that cisplatin increased the expression of proinflammatory cytokine TNF-α, which induced an exogenous apoptotic pathway through its tumor necrosis factor receptor 1 (TNFR1) [47], and TNF-α can activate proinflammatory cytokines and chemokines such as NF-κB and trigger oxidative stress, which ultimately aggravates kidney damage [48]. e study by Xu et al. also demonstrated that the expression of TNF-α further helped induce the expression of receptor-interacting protein 1 (RIP1), RIP3, and mixed lineage kinase domain-like protein (MLKL) in proximal tubular cells and enhanced the necrotic signaling pathway through positive feedback [49]. As mentioned in a literature review, the role of p53 in cisplatin-induced cytotoxicity mainly involved activation of the mitochondrial apoptotic pathway [47]. After exposure to cisplatin-induced cellular DNA damage, p53 was phosphorylated, and the proapoptotic protein Bax underwent structural modifications and mitochondrial membrane integrity changes, eventually downregulating the antiapoptotic proteins Bcl-2 and Bcl-xL and triggering the mitochondrial apoptotic pathway [50]. In addition, Yuan et al. proved that p53 promotes cisplatin-induced renal oxidative damage and apoptosis by regulating P66shc and manganese-dependent superoxide dismutase (MnSOD) [51]. A recent study showed that the AMPK-p53-Bax signaling pathway played a crucial  Evidence-Based Complementary and Alternative Medicine role in cisplatin-induced apoptosis of renal tubular epithelial cells [44]. Many studies have shown that IL-6 is closely related to the inflammatory response in cisplatin-induced renal injury and cisplatin-induced inflammatory response can be alleviated by reducing the expression of IL-6 [52][53][54][55].
Another study showed that IL-6 mediates the production and elimination of ROS in cisplatin-induced AKI [56]. All of the above-mentioned core targets and those not mentioned due to word limits are summarized in Table 6. Subsequently, GO and KEGG enrichment analyses were applied to 16 core targets. e result suggested that the PI3K-Akt signaling pathway, MAPK signaling pathway, and Tolllike receptor signaling pathway were suggested to have a significant role in the treatment of cisplatin-induced kidney damage by AM. It has been reported that the MAPK pathway is central to the regulation of inflammation and oxidative stress in cisplatin renal injury and that TNF-α can activate the MAPK pathway [65]. Besides, Ning et al. found the PI3K-Akt signaling pathway plays an essential role in defending renal tubular epithelial cells from apoptosis induced by cisplatin [66]. Meanwhile, many studies have demonstrated that quercetin can suppress the formation of reactive oxygen species, activate the JNK/P38 MAPK signaling pathway, and regulate the PI3K-Akt pathway [67][68][69]. erefore, quercetin may reduce oxidative stress by regulating the MAPK pathway and PI3K-Akt pathway in the treatment of cisplatin-induced kidney. e results of molecular docking suggested that the most likely key targets are MAPK1, TP53, and EGFR. In addition, kaempferol could block the MAPK cascade and reduce the expression of JNK, TNF-α, and ERK1/2, thus improving cisplatin-induced renal injury [39,70]. What is more, the Toll-like receptor signaling pathway can activate the releasing of downstream cytokines such as TNF-α and IL-1β, leading to increased inflammatory cell infiltration and inflammatory responses in renal tissue [71,72]. Meanwhile, a study has shown that kaempferol dramatically suppresses the upregulation of toll-like receptor 4, blocking the Toll-like receptor signaling pathway to reduce the inflammatory response [70]. Molecular docking suggested that quercetin most likely affected some pathways by combining with p53, MAPK, MMP-9, EGFR, and caspase-3 and then finally achieved the effect of reducing cisplatin-induced renal damage. e pathways discussed above are shown in Figure 9.
In the C-T-P network, each pathway corresponds to a plurality of targets, and each target is connected with a plurality of pathways that are connected with each other    Toll-like receptor signaling pathway Figure 9: Key biological pathways.  Evidence-Based Complementary and Alternative Medicine through a common target. It was suggested that AM synergistically treats cisplatin-induced kidney damage through a multicomponent, multitarget, and multipathway mechanism.

Conclusions
In summary, the present research aimed to figure out the molecular mechanism of AM against cisplatin-induced kidney damage by network pharmacology and molecular docking. Screening of active ingredients and molecular docking studies showed that quercetin, kaempferol, and formononetin were significant active ingredients to AM against cisplatin-induced kidney damage. e C-T-P network suggested that the MAPK pathway, PI3K-Akt signaling pathway, and Toll-like receptor signaling pathway had important roles in AM treatment of cisplatin-induced kidney damage. In addition, one of the significant findings is that AM treats cisplatin-induced kidney injury to multicomponent, multitarget, and multipathway synergistic effects. Unfortunately, this study has no corresponding experimental validation that is what we need to do in the future. In a nutshell, this study provided a rationale for further research of cisplatin-induced renal injury and a new orientation of development and application of AM.

Data Availability
Processed data are contained within the article. Raw data are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that there are no conflicts of interest.

Authors' Contributions
Congchao Jia conceived and mainly performed the research; Pengyu Wang collected and collated data on drug and disease; Binyou Wang and Yiwei Wang carried out the molecular docking analysis; Congchao Jia and Rong Chen drafted the manuscript; and Xianchao Pan revised the manuscript. All authors read and approved the final manuscript.