Network Pharmacology Study Reveals the Mechanism of Astragali Radix in Treatment of Diabetic Nephropathy

The same potential target genes from Astragali Radix and DN were analyzed and constructed the protein interaction network. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment-related major targets and signal pathways were performed. The drug-ingredients-target-disease network was visually built using Cytoscape 3.6.1. The benecial pharmacological activities of quercetin from Astragali Radix were conrmed by CCK-8 assay, determination of antioxidant parameters and Western blotting analysis.


Results
There are 12 bioactive components from Astragali Radix and 56 same targets between Astragali Radix and DN. The GO analysis results showed that the biological processes mainly included protein homodimerization activity.
KEGG analysis indicate that the screened targets were most closely linked to the mitogen-activated protein kinase (MAPK) signaling pathway. The drug-ingredients-target-disease network results revealed that the therapeutic effects of Astragali Radix mainly included oxidative stress, in ammatory reaction and apoptosis. During the veri cation process, quercetin from Astragali Radix could attenuate cytotoxicity, enhance catalase (CAT) and superoxide dismutase (SOD) activities and suppress MAPK signaling pathway.

Conclusions
In the current study, network pharmacology with experimental analysis predicted and proved the therapeutic function of Astragali Radix by improving antioxidant capacity and suppressing MAPK signaling pathway, these investigations could provide a new perspective for further exploration of Astragali Radix on DN treatment.

Background
Diabetic nephropathy (DN) is a unique complication of diabetes representing main reason of a signi cant increase in mortality [1]. It is well known that DN is the most prevalent disease resulting in end-stage renal disease (ESRD), in addition its morbidity is signi cantly rising worldwide [2]. Moreover, the patients with ESRD often need hemodialysis or kidney transplantation to restore the normal kidney function [3]. At present, the therapies for DN mainly focus on controlling the blood pressure and glycemic, and regulating the renin-angiotensin system (RAS) to control the further development of DN [4]. Furthermore, DN is an extremely complicated pathophysiological process, in which multiple biological and pathological processes are involved [5]. Therefore, it is imperative to develop an effective therapies and further understand the pathogenesis for DN. HK-2 cells were incubated into 96-well plates (5 × 10 4 cells/well) and divided into control group (5.5 mmol·L − 1 glucose), high glucose group (30 mmol·L − 1 ), high glucose (30 mmol·L − 1 ) + 50 µmol·L − 1 quercetin group and quercetin group (5.5 mmol·L − 1 glucose + 50 µmol·L − 1 quercetin) for 24 h. Subsequently, CCK-8 solution (Biosharp Life Sciences, China) was supplemented at 37 °C for an hour. Finally, the absorbance results were detected at 450 nm with a multi-well plate reader (Biotek, USA).
Afterwards the HK-2 cells were collected to analyze the antioxidant activities of catalase (CAT) and superoxide dismutase (SOD). The indicators were measured according to the available manufacturer's instructions (Nanjing Jiancheng Bioengineering Institute, China).

Western Blotting Analysis
HK-2 cells were treated with high glucose (30 mmol·L -1 ) and quercetin (50 μmol·L -1 ) in 6-well plates for 24 h. HK-2 cells were collected and resuspended with 500 μL cold phosphate buffered solution (PBS) buffer (Gibco, CA, USA). Protein sample concentrations were determined through bicinchoninic acid assay (BCA) protein sample assay kit after cell lysed (Kangwei Century Biotechnology, Beijing, China). Then the experimental protein samples were treated and separated by 6-12% SDS-PAGE electrophoresis and subsequently transferred to activated PVDF membranes. Subsequently, membranes were blocked with 5% skim milk at room temperature (RT) for 2 h and incubated with primer antibodies against P38 MAPK, JNK, phosph-P38 MAPK, phosph-JNK, or β-actin (Cell Signaling Technology, USA) at 4 ℃ overnight. Next, membranes were correspondingly performed uorescent secondary antibodies (Cell Signaling Technology, USA) at RT for 2 h. Chemiluminescent protein bands were analyzed by Odyessey Imager (LI-COR, USA). ImageJ 1.41 software (Bethesda, USA) was used to calculate the optical density.
Statistical analysis SPSS17.0 statistical software was offered for all data analysis. Values reported were expressed as means ± SD. Student's t-test and two-way ANOVA were used to analyze signi cant differences. P<0.05 was set statistically signi cant..

Results
Drug target prediction TCMSP database was performed to predict the candidate ingredients and targets from Astragali Radix. As shown in Table 1, a total of 12 compounds of Astragali Radix were obtained.  Fig. 1, 56 same targets between Astragali Radix and DN in totally are obtained. And speci c information is described in Table 2.

Go Analysis
We entered 56 same target genes, further, "person" was selected as the species, yielding 20 enrichment results to the DAVID database. The ggplot2.R package (3.2.0 Version) was ready for visualization. The biological processes mainly include protein homodimerization activity, proximal promoter sequence-speci c DNA binding, DNA-binding transcription activator activity and cofactor binding. As shown in Fig. 3, the size of the dots represents the number of genes included, and the color changes from blue to red indicates a gradual increase in signi cance. The protein homodimerization activity contains 11 genes and the gene rate is 19.64%. Therefore, this may become a potential research direction of Astragali Radix for treatment of DN.

Kegg Pathway Analysis
Furthermore, we further imported 56 same target genes, "person" was also selected as the species, producing 13 enrichment results. The enrichment of results was shown in Fig. 4. The potential targets network of Astragali Radix for treatment of DN is mainly related to MAPK, HIF, p53 and NF-κB signaling pathway. Results indicated that the target of Astragali Radix for treatment of DN was distributed in different metabolic pathways, and multicomponent and multi-target interaction may be the mechanism for treating DN. As shown in Fig. 5, the pathway map of Astragali Radix for treatment of DN using KEGG Mapper. Those top-ranking signaling pathways from KEGG analysis were integrated and mapped to obtain the nal path map. The 13 targets of Astragali Radix for treatment of DN was marked in red, and the proportion was 23.21%. Moreover, the results showed that the targets of Astragali Radix for treatment of DN was connected with MAPK, HIF, p53, CASP, Elk1, TNF and NF-κB signaling pathway. It demonstrated that the targets of Astragali Radix for treatment of DN was mainly distributed in these signaling pathways, which probably through the interaction of several links to show therapeutic effect.
Drug-ingredients-target-disease Network Cytoscape 3.6.1 was utilised to build the drug-ingredients-target-disease network of Astragali Radix. The results were shown in Fig. 6. This network contains 70 nodes (1 drug, 12 effective ingredients, 56 targets and 1 disease). The drug, diseases, ingredients and targets respectively used V-shaped, V-shaped, elliptical and rectangular nodes, and the relationship was represented by edges. The more action targets were connected, the greater treatment effect of DN, in which quercetin of Astragali Radix may be potential material basis for therapeutic effects. There were 56 targets for Astragali Radix for treatment of DN, among which GSTP1, NQO1, NOS3 [19] were related to oxidative stress [20], and IL-6 was related to in ammation [21], furthermore, MAPK8, CASP3 and BCL2 were related to apoptosis [22]. Therefore, it is presumed that quercetin of Astragali Radix has the most potential to treat DN mainly through inhibiting oxidative stress, in ammatory reaction and apoptosis in the body.

Quercetin attenuates cytotoxicity in high glucose-induced HK-2 cells
Quercetin from Astragali Radix is a potential natural active product for the treatment of DN, which has the chemical structure of ortho phenolic hydroxyl (Fig. 7A). To investigate the effects of quercetin on the cytotoxicity of high glucose-treated renal cells, HK-2 cells were incubated with high glucose (30 mmol·L − 1 ) and/or quercetin (50 µmol·L − 1 ). The experimental result meant that quercetin treatment signi cantly improved cell viability in high glucose-induced HK-2 cells (Fig. 7B).

Quercetin enhances antioxidant capacity in high glucose-induced HK-2 cells
The activities of antioxidant-related enzymes may be used as an indicator to re ect the antioxidant status. Therefore, the antioxidant capacity were determined to certify the protective roles of quercetin on high glucoseinduced oxidative stress. As shown in Fig. 7C-D, antioxidant parameters SOD and CAT were signi cantly decreased in high glucose group. On the contrary, quercetin could signi cantly enhance SOD and CAT activities in high glucose-induced HK-2 cells.

Quercetin regulates MAPK signaling pathway in high glucose-induced HK-2 cells
In order to further study the molecular mechanisms of quercetin in the treatment of DN, we analyzed the effect of quercetin on MAPK signaling pathway. Western blot analysis results showed high glucose could increase phosphorylation of P38 MAPK and JNK. However, treatment with quercetin signi cantly suppressed the phosphorylation of P38 MAPK and JNK in MAPK signaling pathway (Fig. 8).

Discussion
Studies have shown that kidney damage is the most severe in DN, and the damage can aggravate the risk of kidney disease. Importantly, the acute and chronic high glucose state can inhibit the growth of kidney-related cells, further causing cell in ammation, and then induce apoptosis. Network pharmacological method can be useful in providing a global picture of the disease pathogenesis and to identify potential new drug targets for DN. Our results con rmed the evaluation system of DN based on network pharmacology method. Meanwhile, it also revealed that key nodes were selected to get the pathway enrichment by combining GO with KEGG analysis. To explore the correlation between Astragali Radix and DN from a integral direction, network pharmacology can offer a novel strategy for in-depth research of TCM due to its holistic and systematic.
Here, TCMSP database could predict the potential effective ingredients and targets from Astragali Radix. As shown in Table 1, after deleting the components without corresponding targets, a total of 12 compounds of Astragali Radix were found. Upon the analysis data of Gene Cards and OMIM database for DN disease targets, 56 same target genes between Astragali Radix and DN could be obtained. As shown in Fig. 1 and Table 2, the data of Astragali Radix in the treatment of DN revealed that the same target genes included IL6, RELA, CASP3, CASP8, CASP9, BCL2, MAPK8, NQO1, PTGS1, HIF1A, VCAM1, CYP3A4, CYP1A1, GSTM1 and PPARG, which were involved in many biological processes, such as in ammation, oxidative stress [23], apoptosis, aging [24]. As well documented, researches have certi ed that in ammation reaction plays a pivotal role in DN [25]. High blood sugar may cause the accumulation of advanced glycation end products, which leads to in ammation in the human kidney cells, increases cell apoptosis, and accelerates the development of DN [26]. Besides, it is reported that the pathogenesis of DN is closely correlated with oxidative stress and apoptosis [27]. The drug protects rats from DN by reducing the expression of IL-1β, IL-6, and TNF-α in ammatory cytokines and inhibiting the oxidative stress response mediated by hyperglycemia [28]. Based on the results above, they indicate that Astragali Radix could treat DN by ameliorating these mentioned biological processes. Here, protein interaction networks analysis using the String database indicates that the same targets between Astragali Radix and DN are interrelated and play a positive effect in DN through multi-channel and multi-faceted coordination (Fig. 2).
Enrichment analysis showed that the therapeutic effects of Astragali Radix mainly involved the biological process of oxidative stress, in ammatory reaction and apoptosis. The signaling pathways concerning Astragali Radix for treatment of DN largely related to MAPK, HIF, p53 and NF-κB signaling pathway. These pathways may be the future direction of research in the treatment of DN. GO analysis results displayed that the biological processes included protein homodimerization activity and proximal promoter sequence-speci c DNA binding (Fig. 3). The above-mentioned molecular functions were associated with various genes and were signi cant. So the biological processes may become potential research directions of Astragali Radix for treatment of DN. As shown in Figs. 4 and 5, KEGG analysis results were mainly contained p53, MAPK, HIF and NF-κB signaling pathway. More importantly, the potential targets were most closely linked to the MAPK signaling pathway. p53 is a tumor suppressor [29]. It can respond rapidly to genotoxic stresses caused by DNA damage and induce caspase-9 activation and BCL-2 expression, leading to accelerated cell apoptosis [30]. MAPK signaling pathway, including P38 and JNK, mediates cell apoptosis through combining a complex with a proapoptotic factor of p53 [31]. In addition, it can motivate the expression of in ammatory factors mediated through NF-κB signaling pathways and accelerate the kidney pathological changes [32]. HIF-1 was increased in DN mice, which mediates DN-induced kidney brotic disease [33]. NF-κB is a key transcription factor regulating in ammatory response in DN. Besides, in ammatory cytokines could stimulate the protein expression of NF-κB signaling pathway [34]. It indicated that the target of Astragali Radix for DN treatment was distributed in various metabolic pathways. As shown in Fig. 6, the drug-ingredients-target-disease network of Astragali Radix was drawn by Cytoscape software. The results showed that drugs, ingredients, targets and diseases coordinate with each other. In addition, the treatment of Astragali Radix in DN was through the regulation of multi-component, multi-target, and multi-pathway. Meanwhile, its mechanism mainly involved anti-in ammatory, antioxidation and anti-apoptosis.
Although the above results manifested that the molecular pharmacological mechanism of Astragali Radix in DN treatment from network pharmacology, there were still some limitations. Then we conducted an experimental veri cation in the next research. Network pharmacology analysis results prompted quercetin from Astragali Radix had the most effective suppression ability against high glucose-induced HK-2 cells owing to ranking rst among other ingredients. Therefore, quercetin was chosen as the most representative bioactive ingredient of Astragali Radix for further molecular pharmacology experiments. Moreover, KEGG analysis suggested MAPK signaling pathway were the most potential pathway on the treatment of Astragali Radix in DN. In this study, quercetin remarkably reduced high glucose-induced cytotoxicity of HK-2 cells (Fig. 7B). Oxidative stress is an obvious trigger for DN, which is induced by continuously accumulation of advanced glycation end metabolites in the kindey tissue [35]. Additionally, antioxidant enzymes can scavenge mass production of reactive oxygen species (ROS) to maintain the balance in redox system [36]. CAT and SOD are essential antioxidant elements, they form a mutually supportive defense system against ROS [37]. Further analysis certi ed that enhancement of antioxidant ability may be the underlying molecular mechanism by which quercetin inhibited high glucose-induced renal cell apoptosis. In our experiments, quercetin treatment dramatically induced an increase of CAT and SOD activities in high glucose-induced HK-2 cells (Fig. 7C-D). Finally, our results veri ed the ameliorative role of quercetin in high glucose-induced HK-2 cells by regulating MAPK signaling pathway. In our experiments, quercetin inhibited the phosphorylation of P38 MAPK and JNK, due to signi cantly reduction of cell apoptosis (Fig. 8).

Conclusion
Here, we con rmed the positive effect of Astragali Radix on DN via network pharmacology. Results showed that Astragali Radix could act on multi-target and play a potential therapeutic role on DN by multi-pathway. Molecular mechanism analysis showed that it mainly involved anti-in ammatory, antioxidation and anti-apoptosis. Additionally, experimental analysis demonstrated the therapeutic function of quercetin from Astragali Radix by improving antioxidant capacity and suppressing MAPK signaling pathway, which could generate a theoretical foundation for further research of the molecular pharmacological mechanism of Astragali Radix in the process of DN treatment.