Network Pharmacology Approach to Investigate the Mechanism of Danggui-Shaoyao-San against Diabetic Kidney Disease

Background Danggui-Shaoyao-San (DSS) is a traditional Chinese medicine formula that has been widely used to treat a variety of disorders, including renal diseases. Despite being well-established in clinical practice, the mechanisms behind the therapeutic effects of DSS on diabetic nephropathy (DN) remain elusive. Methods To explore the therapeutic mechanism, we explored the action mechanism of DSS on DN using network pharmacology strategies. All ingredients were selected from the relevant databases, and active ingredients were chosen on the basis of their oral bioavailability prediction and drug-likeness evaluation. The putative proteins of DSS were obtained from the Traditional Chinese Medicine Systems Pharmacology (TCMSP) database, whereas the potential genes of DN were obtained from the GeneCards and OMIM databases. Enrichment analysis using gene ontology (GO) and the Kyoto encyclopedia of genes and genomes (KEGG) was performed to discover possible hub targets and gene-related pathways. Afterwards, the underlying molecular mechanisms of DSS against DN were validated experimentally in vivo against db/db mice. Results We identified 91 phytochemicals using the comprehensive network pharmacology technique, 51 of which were chosen as bioactive components. There were 40 proteins and 20 pathways in the target-pathway network. The experimental validation results demonstrated that DSS may reduce the expression of TNF-α, IL-6, and ICAM-1, as well as extracellular matrix deposition, by blocking the JNK pathway activation, which protects against kidney injury. Conclusion This study discovered the putative molecular mechanisms of action of DSS against diabetic kidney damage through a network pharmacology approach and experimental validation.


Introduction
Te increasing prevalence of diabetic nephropathy (DN) parallels the dramatic rise in the prevalence of diabetes globally. Nearly half of all type 2 diabetic (T2D) patients and one-third of type 1 diabetic (T1D) patients eventually sufer from CKD [1]. Every year, around 25 out of every 10,000 diabetic patients are diagnosed with end-stage renal disease [2]. Diabetic nephropathy etiology is complicated, and the treatment methods are limited and only delay disease progression [3]. Currently, the use of conventional approaches, including aldosterone system blockage, blood glucose level management, and bodyweight reduction, does not always produce satisfying results, as they are not efective for preventing diabetes from progressing to end-stage renal disease [4,5]. As a result, it is necessary to develop more efective therapies for DN patients. TCM could be an alternative to western medicine for treating DN. For example, Chinese herbal medicine and acupuncture reportedly have therapeutic benefts for DN [6].
TCM places emphasis on the notion of the integrity of the entire human body. Diferent from the single-target curative efect of chemical drugs, TCM compound ingredients have an overall curative efect, which is usually regulated by "multicomponent" and "multitarget" [7]. Network pharmacology is efcient in building a "compound-protein/gene-disease" network, which is able to describe complexities among biological systems, medicines, and illnesses from a network viewpoint and has a holistic philosophy similar to TCM [8]. By providing the compound-target and target-pathway networks, network pharmacology aids in evaluating the rationale and compatibility of TCM [9]. In conclusion, network pharmacology is a rational application for drug discovery, notably in the area of TCM preparations' studies and development [10,11]. Danggui-Shaoyao-San (DSS) is a formula made up of six Chinese herbs, including Paeoniae Radix Alba (PRA), Angelica Sinensis Radix (ASR), Chuanxiong Rhizoma (CR), Poria cocos (PC), Atractylodis Macrocephalae Rhizoma (AMR), and Alismatis Rhizoma (AR), which has long been employed as a blood-quickeningstasis-transforming formula for gynecological disorders in China, such as algomenorrhea, irregular menstruation, menopausal syndrome, and infertility [12][13][14][15]. Recent studies have mainly focused on therapeutic efcacy for neural dysfunctions, including depression, Alzheimer's disease, poor memory, and other neuropsychiatric symptoms [16][17][18][19]. But emerging evidence has also revealed the therapeutic efcacy of drugs for renal disease, including fbrosis [20], diabetic nephropathy [21,22], nephrotic syndrome [23], and so on. Despite the fact that so many favorable benefts have been demonstrated, the underlying mechanisms of DSS in DN have not been previously explored.
In this current study, we adopted a network pharmacology approach to analyze the underlying mechanisms of DSS in DN. Furthermore, the potential targets of DSS against DN hypothesized by the network pharmacological approach were confrmed through in vivo experiments. Te detailed graphical abstract of this work is illustrated in Figure 1.

Screening of Active Components in DSS.
Active DSS compounds were retrieved from the TCM systems pharmacology database and analysis platform (TCMSP, http:// tcmspw.com), the encyclopedia of TCM (ETCM), and the TCM integrative pharmacology-based research platform (TCMIP), which lists the pharmacokinetic properties of active phytochemicals such as oral bioavailability (OB) and drug-likeness (DL), among others. Te OB index is an important pharmacokinetic parameter that represents the proportion of the medication in the blood circulation. Te similarity between a novel compound and a pre-existing drug can be expressed in terms of the DL index; a high DL indicates that although a compound is not yet used as a drug, it may become one in the future. In this study, the DSS bioactive components meriting further analysis were those with an OB ≥ 30% and DL ≥ 0.18.

Disease-Targets-Compound Network Construction and
Analysis. Potential targets of DSS were predicted using the TCMSP, while target genes for diabetic nephropathy were collected from the GeneCards and OMIM databases. Subsequently, we used the target genes of the active DSS components and DN therapeutic targets obtained from the databases to identify overlapping genes. We applied Cytoscape 3.8.0 software to construct the "disease-targetscompound network" visual network diagram.

PPI Network Construction.
Targets common to both diabetic nephropathy and DSS were imported into the STRING database, with the research species set as human. Protein relationships were then obtained; those with values >0.4 were screened, and the free ends were removed. Finally, a PPI protein interaction network diagram was drawn; more adjacent genes in the PPI map play a more important role.

Gene Ontology and Pathway Enrichment Analysis.
GO and Kyoto encyclopedia of genes and genomes (KEGG) enrichment analyses were performed using the DAVID 6.8 functional annotation tool (https://david.ncifcrf.gov/). A false discovery rate cut-of <0.05 was applied for clustering during the functional annotation.

Experimental Validation
3.1. Preparation of the DSS Extract. DSS comprises ASR, PRA, CR, PC, AMR, and AR (a ratio of 3 : 16 : 8 : 8 : 4 : 4). It was provided by the pharmacy department of the Shenzhen Traditional Chinese Medicine Hospital and prepared in the following manner: the mixture is submerged in distilled water (1 : 8, w/v) at room temperature for 30 minutes with a refux device before boiling for 1 hour and fltering the extract. Te boiling and extraction steps were performed two times. Te extracted fltrate was combined and then concentrated to a fnal extract concentration of 0.64 g/ml, referred to as DSS extract. Te extract was dried to a powder form by freeze-drying and stored at −20°C, and the powder was redissolved with Milli-Q water to obtain DSS extract before the treatment.

Animals and Treatment.
Male diabetic mice (db/db) and mice (db/m) of 10 weeks were purchased from Jiangsu Jicui Yaokang Biotechnology Co., Ltd. (Nanjing, China). Te mice were maintained in the specifc pathogen-free animal room of Shenzhen Top Biotech Co., Ltd. (Shenzhen, China) on a 12-h light/dark cycle with a relative humidity of 40-60% and temperature of 20-25°C. Animals were free to access food and water. Fasting blood glucose greater than 11.1 mmol/L after 30 days was diagnosed as diabetes. Te diabetic mice were randomly assigned to the db/db group and the DSS group (n � 5). Eight db/m mice were used as a control group (n � 8). Te db/db group and DSS group received gavages of 0.9% saline (10 ml/kg/day) and the DSS extract (6.4 g/kg/day), respectively. Te study was approved by the Animal Ethics Committee of Shenzhen Top Biotech Co., Ltd. and conducted in full compliance with the directives of the National Institutes of Health guidelines. All mice were killed through cervical dislocation without consciousness at the end of the study. Blood samples were collected by eye enucleation, and the kidneys were excised and promptly snap-frozen to detect various markers.

Serum Biochemical Analysis.
Samples of serum were taken and centrifuged for 15 minutes at 1,500 rpm in sterile tubes. A serum creatinine (Scr) kit was used to assess the Scr levels according to the manufacturer's instructions.

Histological Examination.
3-μm sections were prepared from the kidneys after dehydration, embedment in wax, and slicing for the periodic-acid-schif (PAS), Masson's trichrome staining (Masson), and hematoxylin and eosin staining (H&E). Images were visualized with a Zeiss optical microscope. Ninety glomeruli from three mice in each group were assessed for the degree of glomerular mesangial expansion and glomerular area employing Image-J (NIH, Bethesda, MD, USA) in a blinded fashion, under ×400 magnifcation by PAS staining. Quantifcation of fbrosis was further performed under ×200 magnifcation by Masson trichrome staining [24].

Statistical Analysis
Te data were expressed as means ± SEM. SPSS (23.0; SPSS Inc., USA) was employed for analyses. One-way analysis of variance was used to detect diferences between groups, followed by Tukey's multiple comparisons test. A P value <0.05 was considered signifcant.

Active Components in DSS.
A total of 752 ingredients were identifed through the TCMSP database, ETCM, and TCMIP v2.0 (up to July 2022). Based on the cut-ofs of OB ≥ 30% and DL ≥ 0.18 detailed above, 51 candidate bioactive components were subjected to further analyses after eliminating overlapping genes. Detailed information is shown in Table 1.

Disease-Targets-Compound Network Construction and
Analysis. Ninety-one potential DSS targets were predicted by the TCMSP database. A total of 1,153 targets related to diabetic nephropathy were collected from the GeneCards and OMIM databases. After eliminating the overlaps, we identifed 126 common (i.e., both drug and disease target genes) (Figure 2(a)), suggesting that these genes potentially play a signifcant role in DSS treatment of DN. To better understand the potential mechanism of DSS in DN, we constructed a disease-targets-compound network using Cytoscape 3.8.0 software (Figure 2(b)). Nodes with diferent colors and shapes in the fgure represent diferent types of information. Green ovals represent common targets, blue rectangular nodes represent active ingredients of TCM, purple hexagons represent DN, and orange prisms represent DSS.

Target Proteins PPI Network Construction and Analysis.
To determine how the overlapping genes interact, the 126 common targets were imported into the STRING database, and a PPI network diagram was drawn (Figure 3(a)). As stated above, more adjacent genes in the PPI map played more important roles. By calculating the number of nodes connected to each gene, the top 30 genes in terms of centrality were identifed as the most important genes of DSS with respect to the treatment of CKD (Figure 3(b)); these included IL-6, AKT1, TNF, CAT, JUN, and PTGS2.

Gene Ontology and KEGG Enrichment Analyses.
We carried out a GO enrichment analysis using the DAVID database to identify the associations of the 126 common genes with DN; 69 GO terms were obtained, of which the top 20 were signifcantly enriched in terms of MFs ( Figure 4). Tese terms mainly involved heme oxygenase, tetrapyrrole binding, cysteine-type endopeptidase activity during apoptosis, cysteine-type endopeptidase activity in the apoptosis signaling pathway, and peroxidase activity. Tese results show that DSS improves diabetic nephropathy by regulating various BPs.
To further explore how DSS afected DN through the 126 common genes, they were upregulated in the DAVID database and a KEGG enrichment analysis was performed. We identifed 140 signaling pathways; the top 20 items are listed in Figure 5 and they include lipid, atherosclerosis, tumor necrosis factor, Posey's sarcoma-related herpes virus infection, Epstein-Barr virus infection, and advanced glycation end product-receptor for AGE (AGE-RAGE) signaling pathways, indicating that the active ingredients in DSS exert their efects through multiple pathways.

DSS Ameliorated Diabetic Kidney Injury in the db/db
Mice. Renal function was assessed by the Scr level. Te Scr levels in the db/db group increased signifcantly compared to those in the control group (P < 0.01). Administrating DSS signifcantly reduced the Scr (P < 0.01) level in db/db mice ( Figure 6(a)). Tese data indicate that DSS prevented renal functional decline in db/db mice.
Te histological pattern of DN includes thickening of the glomerular basement membrane (GBM), mesangial matrix enlargement, nodular glomerulosclerosis, and arteriolar hyalinosis [2]. On PAS staining, db/db mice kidneys exhibited obvious signs of DN. When compared to db/m mice, db/db mice displayed signifcant glomerular hypertrophy and mesangial matrix enlargement, but these histological alterations were signifcantly alleviated in DSStreated mice. Total glomerular area was likewise increased in db/db mice compared to db/m mice, and DSS dramatically lowered this parameter (Figures 6(b)-6(d)). Masson's trichrome staining revealed patchy collagen deposition in the tubular interstitium of the db/db mice, while it was attenuated in DSS-treated db/db mice (Figures 6(b) and 6(f )). Infammation was severe in these mice. H&E-stained histological lesions were nearly absent in the kidneys of DSStreated mice (Figure 6(b)). Tese results indicate that DSS ameliorated renal injury in the db/db mice.

DSS Might Depress Infammatory Response and Trigger the Infammatory-Mediated JNK Pathway in the Kidneys of the db/db Mice.
Te western blot analysis has demonstrated that JNK and c-Jun were upregulated in the db/db mice expression compared with the db/m mice while these proteins were signifcantly decreased in the DSS group. We also found that the treatment with DSS signifcantly reduced the phosphorylation levels of JNK and c-Jun in db/db mice (Figures 7(a) and 7(b)). Meanwhile, the infammatory cytokines, IL-6, ICAM-1, and TNF-α were increased in the db/db mice but decreased after the DSS treatment (Figure 7(c)). Tese data indicated that DSS inhibited the infammatory response and triggered the JNK pathway in the db/db mice.

Discussion
Diabetic nephropathy is a multifactorial, complex disease process caused mainly by hyperglycemia, oxidative stress,

Evidence-Based Complementary and Alternative Medicine
AGEs, and angiotensin II. All of these factors are linked to various proteins or pathways throughout the development and progression [25]. In particular, Chinese medicine has promised to be the main or alternative treatment for DN due to its multiple targets and functions. Research has focused on discovering the bioactive components and molecular mechanisms for the renoprotective efects of Chinese medicines [26]. TCM formulae are extensively used as the   CYP3A4  CYP1A2  NOS2  CYP2C9  CYP1A1  GSTP1  CASP8  PPARG  NOS3  NFKB1  ICAM1  NFKBIA  MAPK8  CYP2E1  VEGFA  CAT  MAPK14  INS  CASP3  TLR4  ALB  PTGS2  IL1B  HSP90AA1  JUN  TNF  IL6  AKT1   20  22  22  22   30 [27]. Based on the system biology and multipharmacology, network pharmacology provides a new network model of "multiple targets, multiple efects, and complex diseases," which is suitable for mechanistic investigation of complex TCM formulae [28]. He et al. found that the AGE-RAGE signaling pathway, the TNF signaling pathway, and the NF-kappa B signaling pathway were critical nodes for the LiuWei DiHuang Pill against type 2 diabetes mellitus (T2DM) throughout the network analysis [29]. Tis technique was applied in the current investigation to determine the pharmacological mechanism by which DSS alleviates DN.
In this study, the compounds in DSS with an OB of 30% and DL > 0.18 were considered pharmacokinetically active and may be largely responsible for the therapeutic efects of DSS in DN. Paeoniforgenone was the most abundant compound, followed by paeoniforin, lactiforin, paeoniforin qt, albiforin qt, benzoyl, paeoniforin, beta-sitosterol,  Evidence-Based Complementary and Alternative Medicine and stigmasterol, which is consistent with the results of Luo et al. [18]. We demonstrated that DSS acted on several targets and signaling pathways. Hub targets of the pharmacokinetically active compounds included IL-6, AKT1, TNF, CAT, JUN, and PTGS2. Infammation, cell proliferation, apoptosis, and fbrosis were all linked to these genes. Tus, DSS may exert its antirenal damaging efects in DN by preventing fbrosis, lowering infammation, and modulating mitochondrial homeostasis, cell proliferation, and apoptosis, which are key mechanisms in the development of DN [30]. Among the pathophysiological mechanisms responsible for the development of DN and the pathogenesis of diabetic neuropathy, infammation is crucial and has become an important target in DN therapy [31]. DSS may exert therapeutic efects in DN by inhibiting fbrosis and reducing infammation via the ((c)-(e)) Te relative glomerular area, mesangial index, and interstitial fbrosis score were quantifed in the three groups of mice, respectively. n � 3 mice per group ( * * * P < 0.001, * * * * P < 0.0001 vs. the db/m group; ## P < 0.01, ### P < 0.001 vs. the db/ db group). 8 Evidence-Based Complementary and Alternative Medicine JUN, IL-6, and TNF-signaling pathways. We studied the therapeutic efects of DSS on db/db mice in vivo to further support this hypothesis. We explored the curative efects of DSS in db/db mice with the goal of providing further support for this hypothesis. Te in vivo data showed that DSS decreased Scr levels and considerably reduced pathological damage in db/ db mice. DSS substantially decreased JNK, P-JNK, c-Jun, and P-c-Jun expression in vivo, according to the western blotting. We also found that DSS reduced the expression of ICAM-1, IL-6, and TNF-α; this fnding was validated by H&E staining, which revealed infammatory infltration. Infammation is the initial reaction of the body to kidney damage. JNK is a mitogen-activated protein kinase subfamily protein that regulates critical biological functions, such as cell proliferation, diferentiation, and apoptosis [32]. Te JNK pathway is activated by a variety of stimuli associated with kidney damage, including proinfammatory cytokines, danger-associated molecular pattern ligands (alarmins), oxidative stress, profbrotic agents, and nephrotoxins [33]. Seen in the most types of human diabetic nephropathy, the activation of the JNK pathway can worsen symptoms and inhibiting the JNK pathway is a therapeutic target for diabetic nephropathy [34]. JNKs not only have a role in the synthesis of a variety of infammatory cytokines but also the expression of proinfammatory cytokines, such  as TNF-α and IL-6, is linked to the continued activation of JNK1 [35,36]. Activated by multiple pathogens and other infammatory diseases, JNK is able to phosphorylate c-Jun, triggering a series of phosphorylation cascade events and regulating critical downstream efector molecules such as TNF-α, IL-6, IL-1, ICAM-1 to exert its biological impact [37][38][39]. Intercellular adhesion molecule 1 (ICAM-1) induces a proinfammatory and proatherogenic response in atherosis, insulin resistance, and the development of coronary disease [40]. Recently, clinical studies have revealed that increased serum/plasma ICAM-1 levels are signifcantly connected with albuminuria in T1D and T2D patients [41,42]. As a pleiotropic cytokine, IL-6 signaling is critical in the regulation of infammation and immunological responses [43]. Numerous studies have shown that IL-6 is essential to the etiology and progression of DN via gp130-STAT3 dependent processes and also acts locally in tissue remodeling and immune cell infltration [44]. TNF-α mediates infammatory processes and is also implicated in the progression of diabetic complications [45]. Almost all that reside in the kidney can generate TNF-α [46]. Tere is growing evidence that elevated TNF-α concentrations enhance sodium absorption, resulting in renal hypertrophy and salt retention, which are characteristic abnormalities during the early stages of diabetic neuropathy [47]. We demonstrated that DSS may suppress infammatory and fbrotic responses in db/db mice through network pharmacology analysis.

Conclusion
Te active pharmacological components and mechanisms by which DSS suppresses infammation in DN were demonstrated using an integrated approach, including network pharmacology and experimental validation. We found that DSS suppresses infammation primarily by regulating the JNK signaling pathway. In conclusion, the combination of network pharmacology and experimental validation was useful in defning the mechanism of the action of DSS.

Data Availability
Te datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.