Potential Mechanism of Guizhi Decoction in Hypertension Treatment Based on Network Pharmacology and Dahl Salt-Sensitive Rat Model

Introduction: Guizhi decoction (GZD), a classical Chinese herbal formula, has been widely used to treat hypertension, but its underlying mechanisms remain elusive. The present study aimed to explore its therapeutic effects and potential mechanisms in the treatment of hypertension using network pharmacology and experimental validation. Methods: The active ingredients and corresponding targets were collected from Traditional Chinese Medicine Systems Pharmacology database and Analysis Platform (TCMSP). The targets related to hypertension were identied from multiple databases, and multiple networks were constructed to identify key compounds, hub targets, and main biological processes and pathways of GZD against hypertension. The Surex-Dock software was used to validate the binding anity between key targets and their corresponding active compounds. The Dahl salt-sensitive rat model was used to evaluate the therapeutic effects of GZD on hypertension. Results: A total of 112 active ingredients, 222 targets of GZD and 341 hypertension- related targets were obtained. Furthermore, 56 overlapping targets were identied, ve of which were determined as the hub targets to perform experimental verication, including interleukin 6 (IL-6), C-C motif chemokine 2 (CCL2), IL-1β, matrix metalloproteinase 2(MMP-2), and MMP9. Pathway enrichment results indicated that 56 overlapping targets mainly enriched in several inammation pathways such as the tumor necrosis factor (TNF) signaling pathway, toll-like receptor (TLR) signaling pathway and nuclear factor kappa-B (NF-κB) signaling pathway. Molecular docking conrmed that most active compounds of GZD showed tight binding ability with the key targets. Experimental results demonstrated that compared with the group fed a high-salt diet in this study, the GZD improved blood pressure, reduced the area of cardiac brosis, and inhibited the expression of IL6, CCL2, IL1β, MMP2 and MMP9 in rats. Conclusions: The potential mechanism of the therapeutic effect of GZD on hypertension may be attributed to the regulation of cardiac inammation and brosis. AKT-serine/threonine-protein kinase; ALB-albumin; CCL2-C-C motif chemokine 2; VEGFA-Vascular endothelial growth factor A; NF-κB:Nuclear factor kappa-B; HIF-Hypoxia-inducible factor; TLR:Toll-like receptor.


Introduction
Hypertension is one of the most common cardiovascular diseases worldwide. [1] Although substantial progress has been made in recent years in the diagnosis and therapy of hypertension, it remains one of the most severe public health issues in the world. It is predicted that more than 1.5 billion people worldwide will suffer from hypertension by 2025, and this situation is expected to worsen over the next decade with the increasing global and ageing populations [2,3]. Furthermore, long-lasting hypertension usually leads to myocardial infarction, stroke, chronic kidney dysfunction, heart failure, and other complications, which are the major causes of disability and premature death in humans. [4] Generally, the pathological mechanisms of hypertension are mainly related to the overactivation of the reninangiotensin-aldosterone system (RAAS) and sympathetic nervous system (SNS), and sodium and water retention. [2,4] Hence, the rst-line medications for hypertension treatment primarily include diuretics, β-blockers, calcium channel blockers, angiotensin-converting enzyme inhibitors, and angiotensin II receptor blockers. [5] However, some antihypertensive medications are reportedly associated with numerous side effects including peripheral edema, anemia, persistent cough, and decreased sexual function. [2,4,5] Therefore, it is necessary to develop more feasible and safer therapeutic strategies for the management of hypertension.
Cumulative valuable research showed that Chinese medicinal formulas, such as Xiao Yao San, Banxia Baizhu Tianma decoction, and Niuhuang Jiangya preparation, demonstrated satisfactory e cacy and minimal side effects in the treatment of hypertension. [6][7][8] Therefore, several scholars are paying attention to the scienti c research value of traditional Chinese medicine (TCM) in the prevention and management of hypertension. [6] According to the TCM therapeutic theory of "regulating Ying-Wei in case of heart damage," we have discovered an association between the dysfunction of Ying-Wei and hypertension-related clinical symptoms. Guizhi decoction (GZD) is a classical TCM prescription for the treatment of Ying-Wei disharmony. [9] GZD is composed of ve Chinese medicinal herbs: Cinnamomi ramulus (Guizhi, GZ), Paeoniae radix alba (Baishao, BS), licorice (Gancao, GC), Zingiber o cinale roscoe (Shengjiang, SJ), and Jujubae fructus (Dazao, DZ). Our previous study showed that GZD not only effectively reduced blood pressure, but also prevented myocardial brosis, restored balance in the autonomic nervous system, and inhibited the expression of pro-in ammatory cytokines in the Dahl saltsensitive rats. [10] Additionally, we found that GZD could signi cantly improve the heart rate variability and the activity of the vagus nerve associated with the progression of hypertension. [9] However, the underlying mechanisms and targets of GZD against hypertension are not fully understood, which have impeded its clinical practice. Further research should be conducted to provide scienti c evidences regarding the clinical usage of GZD in hypertension treatment.
TCM prescriptions are a complicated system that usually interact with multiple targets and multiple channels to exert their pharmacological effects; hence, it is relatively di cult to elucidate their active ingredients and therapeutic mechanisms. [11] Network pharmacology is an innovative approach based on computational systems pharmacology that can accurately decipher the association among the drugs, targets, and diseases at a systematic and comprehensive level. [12] The systematic and holistic characteristics of network pharmacology correspond to the concepts of holistic view and dialectical treatment theory of TCM, providing a scienti c basis for the in-depth study of TCM prescriptions. [13] Currently, there is growing evidence about the reliability of the network pharmacology method, which may be an effective way to study the pharmacological mechanisms of TCM prescriptions. [11][12][13][14] In this study, network pharmacology analysis was used to visualize and decipher the complex relationships among the key compounds, targets, main biological processes and pathways, and disease. Moreover, molecular docking experiment was used to validate the binding a nity between key targets and their corresponding active compounds. Finally, the therapeutic effect of GZD on hypertension was validated by an experimental model. The detailed owchart is illustrated in Figure 1.
Methods And Materials 2.1 Identi cation of active compounds and prediction of corresponding targets of Guizhi decoction. The active constituents of the ve herbal medicines in GZD were acquired from the Traditional Chinese Medicine Systems Pharmacology database and Analysis Platform (TCMSP) (http://lsp.nwu.edu.cn/ tcmsp.php). This database provides comprehensive and accurate compound information, including the herbal ingredients' chemical structural data, drug half-life, oral bioavailability (OB), intestinal epithelial permeability, and drug-likeness (DL). [15] According to the recommended drug screening criteria of the TCMSP database, chemical constituents with OB ≥30% and DL ≥0.18 may present ideal pharmacological activities, and they were selected as the active ingredients for further analysis.
Subsequently, we screened the targets of active ingredients in GZD through the TCMSP database. The target names were imported into the UniProt database (http:// www.uniprot.org/) with the species selected as "Homo sapiens," and the gene names of the targets were obtained from the UniProt database.
2.2 Identi cation of hypertension-related targets. "Hypertension" was the keyword used to extract the hypertension-related targets from the Online Mendelian Inheritance in Man (OMIM, https://omim.org/), Comparative Toxicogenomics Database (CTD, http://ctdbase.org/), Genecards (http://www.genecards.org/), and Drugbank (https://www.drugbank.ca/) databases. The disease-related targets obtained were standardized as gene names from the UniProt database with the species selected as "Homo sapiens." A Venn diagram was drawn using an online website (http://bioinformatics.psb.ugent.be/Webtools/Venn/) to obtain the overlapping targets between the hypertension-related targets and active compound-related targets, which could be the potential targets of GZD in hypertension treatment.

Construction of protein-protein interaction network and screening of hub targets.
The Search Tool for the Retrieval of Interacting Genes (STRING) database (https://string-db.org/) can explore and analyze direct and indirect interactions between proteins. [16] Based on the overlapping targets of GZD and hypertension, we constructed the protein-protein interaction (PPI) network by using the STRING 11.0 database with the species limited to "Homo sapiens" and con dence score >0.7. The degree value was applied to choose the putative targets for molecular docking and experimental veri cation.
2.4 Gene ontology and pathway enrichment analyses. For better clari cation of the potential biological processes and pathways of GZD in the treatment of hypertension in this study, we utilized the Database for Annotation Visualization and Integrated Discovery (DAVID, https://david.ncifcrf.gov/) to conduct gene ontology (GO) function enrichment analysis and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. The GO terms and pathway terms with p-value ≤0.05 were considered signi cant enrichment entries.
2.5 Network construction and analysis. In this study, several networks were established to visualize and analyze the complicated interconnection of compounds, targets, and disease using Cytoscape 3.7.1 software (https://cytoscape.org/; version 3.7.1). Based on the results, the compound-target (C-T) network, target-disease (T-D) network, compound-target-disease (C-T-D) network, and target-pathway (T-P) network were constructed using Cytoscape software. In these networks, nodes of different colors and shapes represented different active compounds, potential targets, or signal pathways, and the edges represented the connections between the nodes.
2.6 Molecular docking. The Sur ex-Dock program in SYBYL 2.1 software (Certara Inc., USA) was applied to verify the network pharmacology screening results by docking the key targets with their active compounds and positive anti-hypertension drugs. The Sur ex-Dock program is one of the most e cient ligand-receptor docking techniques and virtual screening programs with favorable features of high precision, high true-positive rate, and fast speed. [17] Common antihypertensive agents, valsartan, candesartan, captopril, enalapril, furosemide, metoprolol, nifedipine, amlodipine, bisoprolol, and hydrochlorothiazide were used as positive drugs for molecular docking. Seventeen key targets with degrees >30 in the PPI network were selected for docking simulation, and these targets were mapped to 80 active compounds for molecular docking. The three-dimensional structures of the active compounds were downloaded from the PubChem database (https://pubchem.ncbi.nlm. nih.gov/). The structures of the key target proteins were downloaded from the Protein Data Bank database (http://www.pdb. org/) and modi ed through the Sur ex-Dock software. [18] The complex ligand and water molecules in protein receptors were removed, hydrogen atoms were added to the receptor, and amino acids were optimized and patched. [19] After molecular docking with the default parameters, the docking score values were generated for each compound docking with key targets. The docking score could be used to estimate the binding capacity between the targets and their active compounds.
2.7 Experimental animals and protocol. Six-week-old male, speci c pathogen-free (SPF) grade Dahl saltsensitive rats (body weight, 160-180 g) were provided by the Charles River Animal Laboratory (Beijing, China, Certi cate No. 2016-0006). The rats were reared in the SPF room at a temperature of 20 ± 2℃ and 50 ± 10% humidity on a 12 h light/dark cycle. After acclimatization for 1 week, systolic blood pressure of all rats was measured weekly using a 12-channel tail-cuff blood pressure system (MRBP, IITC Life Science Instruments, USA). All experimental protocols used in this study were performed in accordance with the Institutional Animal Care and Use Committee of Shandong University of Traditional Chinese Medicine (Permit Number: SDVTCM2018071501).

Preparation of Guizhi decoction and intervention. The herbal medicines were supplied by the A liated Hospital of Shandong University of Traditional Chinese Medicine (Jinan, China) and veri ed by Prof. Feng
Li. Cinnamomi ramulus, Paeoniae radix alba, and licorice were mixed in the standard ratio of 3:2:2, with re ux extraction of 10 times the volume of distilled water twice for 1 h each. The extracts were then mixed thoroughly, and concentrated to a relative density of 1.20-1.25 (70-80℃). The solution with 1.5 g/mL of the initial herb was applied in further experiments. After 1 week of acclimatization, the rats were randomly allocated to three groups (n = 8, each): NS group (normal-salt diet), HS group (high-salt diet), and GZD group. The NS group was fed low-salt (0.3% NaCl) diet throughout the experimental period. At 8 weeks of age, the HS group and GZD group fed with high-salt (8% NaCl) diet progressively developed hypertension. At 12 weeks of age, the NS and HS groups were fed with physiological saline at 2 mL/day, while 2 mL/4.0 g of GZD crude drug/kg/day was administered to the GZD group according to the results of a previous study. [9] 2.9 Collection of left ventricle tissue. At the end of the experiment, all the rats were anesthetized by intraperitoneal injection of 20 mg/kg pentobarbital sodium. The left ventricle of each rat was carefully isolated and cut into three parts. One part was xed in 6% paraformaldehyde solution for morphological examination. The second part was xed in 2.5% glutaraldehyde solution and observed under transmission electron microscope. The last part was placed in liquid nitrogen for western blot and quantitative real-time polymerase chain reaction (qRT-PCR).
2.10 Histological examination and transmission electron microscopy. The left ventricles were cut into 4µm sized sections, and they were subjected to hematoxylin-eosin (H&E) staining or Masson staining to evaluate in ammatory cell in ltration or cardiac brosis, respectively. [10] Histological images of myocardium in rats were detected using ZEN 1.01.0 Imaging analysis software (Carl Zeiss Microscopy GmbH, German). Additionally, transmission electron microscopy was used to reveal the extent of myocardial damage.
2.11 Quantitative real-time polymerase chain reaction. In order to verify the reliability of network pharmacology analysis, qRT-PCR experiments were used to examine the mRNA expression levels of hub targets in the PPI network. Total RNA of the left ventricle was isolated using Trizol (Invitrogen, USA), and reverse transcription was performed with the PrimeScript RT reagent kit with gDNA Eraser (Takara, Japan) for 5 min at 85℃ according to the manufacturer's instructions. The Light Cycler 480 SYBR Premix Ex Taq II (Roche, Germany) was used to perform qRT-PCR. The reaction conditions were 94℃ for 2 min, 94℃ for 30 s, and 60℃ for 30 s, and 40 cycles were performed in total. Each RNA sample was performed in triplicate, and the results were normalized with β-actin. The relative quanti cation analysis was performed by the 2 -ΔΔCT method.
3.2 Compound-target network construction and analysis. In order to re ect the interactions intuitively between the active compounds of GZD and their potential targets from a systematic and holistic view, the C-T network was constructed by mapping 112 active compounds to their 222 corresponding potential targets. As shown in Figure 2(a), the network consisted of 339 nodes (5 herbal medicines nodes, 112 active compound nodes, and 222 compound-associated target nodes) and 1932 interaction edges. In the network, quercetin (degree 272), β-sitosterol (degree 104), and kaempferol (degree 102) showed the maximum interactions with potential targets, indicating that these active compounds with high degree values could play an important role in the potential pharmacological effects of GZD. The C-T network showed intimate communications between active compounds and related targets, which provided reference to investigate the pharmacological mechanisms of GZD further.
3.3 Screening of hypertension-related targets and construction of disease-target network. A total of 477 targets associated with hypertension were identi ed from the CTD, GeneCards, OMIM, and Drugbank databases. We chose the top 200 results related to hypertension with the highest relevance from GeneCards as major targets of hypertension. After duplicate targets were deleted, 341 hypertensionrelated targets were nally obtained. These targets were then uploaded to Cytoscape 3.7.1 for visualization. The D-T network consisted of 342 nodes (1 hypertension node and 341 hypertension-target nodes) and 477 edges. Figure 2(b) shows the top 14 targets with the highest degree values (degree ≥100) were recognized as the most crucial targets in the D-T network, namely INS, ALB, AKT1, IL-6, VEGFA, TNF, NOS3, REN, EDN1, AGT, TP53, ACE, FN1, and IGF1. Therefore, these targets could be the potential targets for the treatment of hypertension.
3.4 Construction of PPI network. PPI network was established to better interpret the mechanisms of GZD in hypertension treatment by using STRING software. As shown in Figure 2(c), we obtained 56 overlapping targets after merging hypertension-related targets and active compound-related targets. A PPI network was then established by importing the overlapping targets to the STRING database. As shown in Figure 2 3.5 Construction of compound-target-disease network. In order to illustrate the complex relationship among the active compounds, overlapping targets, and hypertension, we constructed the C-T-D network. The network comprised of 165 nodes (107 active compounds, 56 potential targets, 1 formula, and 1 disease) and 762 edges. In this network, the key active compounds of GZD against hypertension mainly included quercetin (degree = 84), kaempferol (degree = 40), and β-sitosterol (degree = 24), which exhibited degree values higher than that of other compounds. As shown in Figure 3, the same target could interact with different active compounds, and the same compound could act on multiple targets. These results suggested that GZD exhibits multi-component and multi-target characteristics in the treatment of hypertension.  Figure 4 (a-c). The BP results mainly comprised of response to hypoxia, response to drug, aging, in ammatory response, and angiogenesis. The CC analysis indicated that the overlapping targets were mainly related to extracellular space, caveola, membrane raft, extracellular region, and plasma membrane. The MF results mainly included protein binding, protein homodimerization activity, protein heterodimerization activity, enzyme binding, and transcription factor binding.
The results of the KEGG enrichment analysis showed that 63 pathways meeting the threshold value of p <0.05 were considered as signi cant enrichment pathways. After sorting according to the p-value, the top 20 pathways and overlapping targets were used to construct the T-P network using Cytoscape 3.7.1 software. As depicted in Figure 4(d), the KEGG pathways of GZD against hypertension were mainly related to the tumor necrosis factor (TNF) signaling pathway, hypoxia-inducible factor 1 (HIF-1) signaling pathway, toll-like receptor (TLR) signaling pathway, insulin resistance, PI3K-AKT signaling pathway, and nuclear factor kappa-B (NF-ĸB) signaling pathway. 3.7 Docking results analysis. The docking scores of the 17 key targets are shown in Figure 5, ranging from 0 to 10, and representing the binding ability from weak to strong. The docking scores and the spatial structure showed that multiple active compounds of GZD show tight binding ability with the key targets, which further indicates that the speci c therapeutic effect of GZD in the treatment of hypertension, especially for β-sitosterol, kaempferol, and quercetin, were very close to the Vina scores of furosemide and nifedipine. As shown in Figure 6, we observed that many active compounds could form hydrogen bonds with the key targets.
3.8 Effect of Guizhi decoction on the blood pressure in rats. As shown in Figure 7, no obvious differences were observed in the systolic blood pressure between the HS group and GZD group before administration of GZD. The systolic blood pressure in these two groups was signi cantly higher than that of the NS group at 8 weeks of age (p < 0.01). After 4 weeks of GZD administration, systolic blood pressure in the GZD group reduced as compared to the HS group (NS: 126 ± 5.18 mmHg, HS: 220 ± 3.64 mmHg, and GZD: 209 ± 4.17 mmHg at 15 weeks; p < 0.01).
3.9 Effect of Guizhi decoction on in ammation in rats. As presented in Figure 8(a), H&E staining showed that the ratio of in ammatory cell in ltration in the HS group was remarkably higher than that of the NS group (p < 0.01), but this pathological change markedly reduced after administration of GZD (p < 0.01).
As shown in Figures 8(b) and 9, the mRNA and protein expressions of IL-6, CCL2, and IL-1β were signi cantly higher in the HS group than in the NS group (p < 0.01). After administration of GZD for 4 weeks, the mRNA and protein expressions of IL-6, CCL2, and IL-1β were obviously lowered as compared to the HS group (p < 0.01).
3.10 Effect of Guizhi decoction on the myocardial brosis in rats. In the results of transmission electron microscopy, the HS group presented disordered arrangement of the cardiomyocytes, apparent myocardial injury, and excessive collagen bers, which were markedly attenuated with GZD treatment (Figure 10(a)). Similarly, the results of Masson staining showed that the degree of interstitial brosis decreased in the GZD group as compared to the HS group ( Figure 10(b), p < 0.01). As shown in Figures 9 and 10c, the mRNA and protein expressions of MMP-2 and MMP-9 were signi cantly higher in the HS group as compared to the NS group (p < 0.01). After 4 weeks of GZD treatment, the mRNA and protein expression of MMP-2 and MMP-9 in the GZD group were obviously decreased as compared to the HS group (p < 0.05 and p < 0.01, respectively), suggesting that GZD could decrease mRNA and protein expression of MMP-2 and MMP-9 to improve the collagen deposition in Dahl salt-sensitive rats.

Discussion
Hypertension is a multifactorial disease affected by the complex interactions between genetic predisposition and environmental factors. [4] There have been several successful experiences in the treatment of cardiovascular diseases such as hypertension with GZD. [9,10,20] However, its material basis and potential mechanism have not been fully elucidated. In the present study, we initially identi ed key compounds, hub targets, and main biological processes and pathways of GZD against hypertension by network pharmacology analysis, and then examined the therapeutic effects of GZD on hypertension in the Dahl salt-sensitive rat model. Our integrative approach uncovered the potential mechanisms of GZD against hypertension from the systematic network perspective and demonstrated that GZD can effectively attenuate elevated blood pressure, improve both in ammatory cell in ltration and myocardial brosis, and inhibit the expression levels of IL-6, CCL2, IL-1β, MMP-2, and MMP-9 in the Dahl saltsensitive rats.
In this study, 112 active compounds of GZD were identi ed from the TCMSP database, and we screened out three key compounds by the network analysis that might have potential anti-hypertensive function. The three key active compounds were β-sitosterol, kaempferol, and quercetin. Moreover, these three key compounds exhibited optimal binding a nity in molecular docking, indicating that they might play a crucial role in the anti-hypertensive effects of GZD. β-sitosterol is a type of phytosterol that exerts protective cardiovascular effects mainly by increasing the intracellular antioxidant defense, improving endothelial function, and inhibiting serum cholesterol levels. [21][22][23] Kaempferol has shown to be effective in maintaining blood pressure by effectively repressing the generation of in ammatory cytokines and apoptosis, stimulating the release of NO from the vascular endothelium, and decreasing myocardial brosis by inhibiting proliferation of cardiac broblasts. [24][25][26] Quercetin exerts remarkable effects in the treatment of hypertension by improving the endothelial function by increasing NO bioavailability and NO production, inhibiting the overactivation of RAAS, and reducing the generation of the adhesion molecules and other in ammatory factors. [27,28] Overall, these observations suggest that the key compounds of GZD could be crucial in the treatment of hypertension.
In the PPI network, we observed that IL-6, CCL2, IL-1β, MMP-2, and MMP-9 were the hub targets of the GZD for improving hypertension. Notably, these hub targets were closely related to various compounds and biological processes and pathways based on the analysis of multi-layered networks, suggesting that these targets may be important in the role of GZD against hypertension. Several studies showed that a long-term in ammatory response can trigger the sympathetic activation and result in myocardial brosis and endothelial dysfunction. [29] Thus, the inhibition of in ammatory response can effectively delay or control the development of hypertension and severe complications. [30] The in ammatory process of hypertension is characterized by increased levels of local in ammatory cytokines such as IL-6, IL-1β, TNFα, and ICAM-1, which are highly correlated with increased risk of hypertension, and they could be useful diagnostic tools for hypertension in the future. [4] Particularly, IL-6, a well-known pro-in ammatory cytokine, participates in the pathological process of hypertension by promoting endothelial dysfunction and in ammatory cell recruitment. [31] The inhibition of IL-1β could inhibit the overactivation of RAAS and decrease overproduction of other pro-in ammatory cytokines, thereby improving hypertension and cardiac brosis. [32,33] CCL2 is a chemokine that contributes to progression of hypertension by recruiting the circulating monocytes to the blood vessel walls and promoting macrophage in ltration. [34] Myocardial brosis is a crucial pathological feature in the progression of hypertension, and it is predominantly related to an excessive accumulation of extracellular matrix proteins that contribute to increased ventricular wall stiffness and impaired diastolic function. [32,35] MMPs participate in the development of myocardial brosis by regulating the degradation and production of collagen, and higher levels of MMP-2 and MMP-9 were considered as markers of cardiovascular risk and aberrant accumulation of collagen. [36,37] As expected, our experimental results showed that GZD can reduce the degree of in ammatory in ltration and area of interstitial brosis, as well as downregulate the protein and mRNA levels of IL-6, IL-1β, CCL2, MMP-9, and MMP-2 in Dahl salt-sensitive rats. These results were in line with earlier reports that showed that the improvement of cardiac in ammation and brosis was associated with decrease in blood pressure. [10,38] Based on the results of the network analysis and molecular docking, several targets such as AKT1, VEGFA, eNOS, ICAM-1, PTGS2, and ALB could be associated with the potential effect of GZD on hypertension. For example, AKT1 directly participates in the phosphorylation of eNOS at serine 1177, which could increase enzyme activity, NO production, and angiogenesis. [39] Vascular endothelium is implicated in the regulation of vascular tone and structure, and the abnormal vascular endothelial function could be a major contributor to the adverse outcomes of hypertension. [4,40] VEGFA serves as a homologue of the VEGF family and regulates cell migration, division, and angiogenesis regulation in normal microvascular endothelial cells. [41] Several studies have provided evidence that eNOS is a signi cant contributor in the maintenance of vascular function and cardiovascular homeostasis. [42] NO, produced by eNOS, mediates control of the in ammatory process as well as regulation of neoangiogenesis and vasodilatation. The decrease in bioavailability of NO has been implicated as a major cause of endothelial dysfunction in hypertension. [4] ICAM-1 promotes the adhesion of leukocytes and vascular endothelial cells, and subsequently leukocyte activation, which probably triggers the endothelial dysfunction, in ammatory response, and blood-vessel remodeling. [32,43] PTGS2-derived products have been proven to be extensively linked with the regulation of uid balance, endothelial function, and ROS production. [44]As a major protein in human serum, the low level of serum ALB is associated with hypertension, increased risk of cardiovascular disease, and carotid atherosclerosis. [45] These targets need to be veri ed in follow-up experiments.
In the KEGG pathway analysis, the overlapping targets were related to multiple pathways, such as TNF signaling pathway, HIF-1 signaling pathway, TLR signaling pathway, insulin resistance, PI3K-AKT signaling pathway, and NF-ĸB signaling pathway. These signaling pathways play a signi cant role in the pathogenesis and management of hypertension. The activation of the TNF signaling pathway is an important contributor to in ammatory processes, which plays an essential part in modulating the gene expression of many cytokines and chemokines involved in vascular in ammation and remodeling. [46] In the in ammatory process associated with hypertension, the inappropriate activation of the NF-kB pathway facilitates disease progression by inducing in ammatory cytokine release, vascular dysfunction, and generation of reactive oxygen species. [32,47] The TLR signaling pathway, a critical upstream mechanism activating in ammatory signaling, regulates the in ammatory response by directly promoting the release of a variety of in ammatory mediators, inducing the migration of immune cells to in ammatory sites, and increasing the adhesion and in ltration ability of in ammatory cells. [29,48] The HIF-1 pathway is highly correlated with energy metabolism and angiogenesis, and participates in the pathophysiology of in ammation and ischemia. HIF-1 is the chief hypoxia-regulated transcription factor that regulates cellular responses in hypoxic and ischemic conditions. [49,50] The PI3K/AKT pathway plays a seminal role in regulating multiple biological effects, including cell growth and proliferation, apoptosis, and angiogenesis. [51] The PI3K/AKT pathway could activate the phosphorylation and activation of eNOS, and maintain blood pressure homeostasis, endothelium function, and vascular integrity. [52] Insulin resistance is a risk factor in hypertensive patients that closely correlates with the activation of RAAS and SNS, resulting in increased peripheral vascular resistance and circulating plasma volume. [53] Therefore, the improvement of insulin resistance would be signi cant in the management of hypertension and its complications. [54] However, the effects of these signaling pathways in the mechanisms of GZD against hypertension need to be validated through rigorous investigations.
Because of the limitations regarding screening conditions of database and statistical software, several ingredients and targets of GZD against hypertension may have been missed during the screening process. Despite the limitations of this study, our results revealed the potential mechanism of GZD against hypertension, and provided scienti c basis and valuable enlightenment for guiding future indepth research and clinical applications.

Conclusion
In this study, the results of network pharmacology indicated that quercetin, β-sitosterol, kaempferol, and other effective compounds of GZD showed therapeutic e cacy in hypertension treatment by means of multiple targets and multi-pathways. Furthermore, the experimental results suggested that GZD could downregulate the expression levels of IL-6, IL-1β, CCL2, MMP-2, and MMP-9, thus inhibiting the in ammatory response and myocardial brosis in Dahl salt-sensitive rats. In conclusion, this study holistically illuminated the potential mechanisms of GZD in improving hypertension, and provided scienti c basis for further pharmacological study and clinical applications.

Declarations Data Availability
All data used to support the ndings of this study are included within the article.

Disclosure
Ji-ye Chen, Yong-jian Zhang and Yong-cheng Wang are co-rst authors of the article.