Exploring the mechanisms of Gui Zhi Fu Ling Wan on varicocele via network pharmacology and molecular docking.

Conception and design: Ruipeng Wang, Xiaoye Qiao; Administrative support: Xiaobin Wang; Provision of study materials or patients: Ruipeng Wang; Collection and assembly of data: Xiaobin Wang; Data analysis and interpretation: Xiaobin Wang; Manuscript writing: All authors; Final approval of manuscript: All authors.

The key active components and core targets of GFW in treating VC were demonstrated using an integrated network pharmacology method and molecular docking. We began by exploring databases for GFW and VC‐related targets. To acquire the major active components, we built the GFW component‐target network and the GFW‐VC common‐target network. To forecast the primary possible targets, a protein–protein interaction (PPI) network was created. On this basis, significant signalling pathways were identified using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. Finally, molecular docking was used to evaluate the interaction of the key active components with their core targets. In Traditional Chinese Medicine Systems Pharmacology (TCMSP; https://tcmspw.com/tcmsp.php ), active ingredients of five components (Cinnamomi Ramulus, Poria Cocos, Cortex Moutan, Radix Paeoniae Rubra, and Persicae Semen) of GFW and corresponding targets were retrieved until 15 February 2022. After oral administration, the drug must play its role in the body through pharmacokinetic processes such as absorption, distribution, metabolism, and excretion (ADME). The screening conditions were determined as oral bioavailability (OB) ≥30% and drug‐likeness (DL) ≥0.18 (Huang et al.,  2014 ). The targets corresponding to the active ingredients were derived, and the targets were standardised and sorted out through the Uniprot database ( https://www.uniprot.org ). Using ‘Varicocele’ as a key word, the Gene Cards ( https://www.genecards.org/ ), Online Mendelian Inheritance in Man (OMIM; https://omim.org/ ) and Disease Gene Network (DisGeNET; https://www.disgenet.org/ ) were used to retrieve target genes, and final targets for VC were obtained after integration and duplicates removal. We used the R programme ( https://www.r-project.org/ ) to identify the intersection of VC and GFW targets and construct a Venn diagram. A common target PPI was constructed using Search Tool for the Retrieval of Interacting Genes/Proteins (STRING 11.5; https://string-db.org/ ) platform, and the minimum interaction threshold was set as ‘medium confidence (>0.4)‘. Different targets are represented by ‘nodes’ and the link between them is shown by ‘edges’. We also used R for visual processing, the frequency of occurrence was counted, and a histogram was drawn. The PPI network map was imported into Cytoscape ( https://cytoscape.org/ ) to find the main targets using the CytoNCA plug‐in and draw the main target network map. The Cytoscape 3.8.2 software was used to construct a ‘component‐target‐disease’ network. In the network diagram, ‘Node’ represented drugs, active ingredients, diseases, and targets, and ‘edge’ represented the relationship between nodes. The CytoNCA plug‐in was used to find the main active ingredients. The Metascape online platform ( https://metascape.org/ ) was used to find common targets for GO and KEGG pathway enrichment analysis, and R was used to draw a histogram and a bubble chart to visualise the information. We downloaded the mol2 format file of GFW from the TCMSP database, found the ID and 3D structure of target protein in the Protein Data Bank (PDB) database ( https://www.rcsb.org/ ), and exported it in PDB format. To complete molecular docking, the Autodock Tools software ( https://autodock.scripps.edu/ ) was used to remove water and small molecule ligands, followed by hydrogenation and charge computation. Finally, the docking effect was evaluated preliminarily based on the binding energy size, and visual processing was performed using Pymol software ( https://pymol.org/2/ ).

A total of 76 active components were obtained by ADME screening. The number of active components in GZ, FL, MDP, CS, and TR were 7, 15, 11, 29, and 23, respectively. The corresponding targets of each component were searched, including 62 targets of GZ, 24 targets of FL, 218 targets of MDP, 142 targets of CS, and 123 targets of TR. A total of 199 targets were obtained after removal of duplicates, as shown in Figure  1 . Common active ingredients of all Chinese medicines in GFW were listed, as shown in Table  1 . The details are available in Excel 1. Distribution of GFW active ingredients and corresponding targets. GFW, Gui Zhi Fu Ling Wan Common active ingredients of all Chinese medicines in GFW Abbreviations: CS, Chi Shao; FL, Fu Ling; GFW, Gui Zhi Fu Ling Wan; GZ, Gui Zhi; MDP, Mu Dan Pi; TCM, traditional Chinese medicine; TR, Tao Ren. The VC target genes were retrieved from GeneCards, OMIM, and DisGeNET with ‘Varicocele’ as the key word, and a total of 267 disease target genes were obtained after integration and duplication removal. Part of targets of TCM in GFW was listed, as shown in Table  2 . Targets of TCM in GFW (only 20 items are listed) Abbreviations: CS, Chi Shao; FL, Fu Ling; GFW, Gui Zhi Fu Ling Wan; GZ, Gui Zhi; MDP, Mu Dan Pi; TCM, traditional Chinese medicine; TR, Tao Ren. The intersection of the targets of the main active components of GFW and the targets of VC disease was taken, and the Venn diagram was drawn, as shown in Figure  2 . It can be seen that there are 44 common targets. The STRING platform was used to construct a PPI network for common targets. Species ‘Homo sapiens’ was selected. The confidence of target association was set at 0.40 and free nodes were hidden to obtain 43 nodes and 503 links, as shown in Figure  3 . In order to highlight important nodes, the obtained nodes were visualised according to degree value, as shown in Figure  4 . The top 10 targets were AKT serine/threonine kinase 1 (AKT1), tumour protein p53 (TP53), tumour necrosis factor (TNF), hypoxia inducible factor 1 subunit alpha (HIF1A), interleukin‐6 (IL‐6), caspase 3 (CASP3), catalase (CAT), prostaglandin‐ ‐endoperoxide synthase 2 (PTGS2), vascular endothelial growth factor A (VEGFA), and interleukin 1 beta (IL1B), as shown in Figure  5 , which involve apoptosis, inflammatory factors, vasoactive factors, hypoxia regulatory factors, oxidative stress factors, and so on. These results indicated that GFW may act on VC through multi‐target and multi‐biological processes (BPs). Venn diagram of disease and drug targets PPI network of common targets. PPI, protein–protein interaction Visual processing of PPI network nodes by degree value. PPI, protein–protein interaction Core targets in PPI network. PPI, protein–protein interaction Cytoscape 3.8.2 software was used to construct the network diagram of ‘component‐target‐disease’, as shown in Figure  6 . The CytoNCA plug‐in was used to find the top 10 main active ingredients, including quercetin, kaempferol, baicalein, beta‐sitosterol, ellagic acid and so on, as shown in Table  3 . ‘Component‐target‐disease’ network Main active components of VC treated by GFW Abbreviations: GFW, Gui Zhi Fu Ling Wan; VC, varicocele. The common targets of GFW and VC were uploaded to Metascape platform for GO and KEGG analysis. Species was set as ‘homo sapiens’; min overlap was set as 3; min enrichment was set as 1.5; and p ‐value cutoff was set as 0.01. The biological process (BP), cell component (CC), molecular function (MF), and KEGG were analysed respectively, and the top 10 genes were selected. The R software was used to draw a bar chart for visual processing of the results, as shown in Figures  7 and 8 . GO analysis of the targets. GO, gene ontology KEGG analysis of the targets. KEGG, Kyoto encyclopedia of genes and genomes The results showed that GFW was involved in the BPs, including apoptosis signalling pathway, response to inorganic substance, cellular response to organic cyclic compound, response to oxidative stress, negative regulation of cell population proliferation, response to metal ion, regulation of cellular response to stress, blood vessel morphogenesis, blood vessel development, and so on. The CCs involved membrane raft, membrane microdomain, mitochondrial envelope, organelle outer membrane, outer membrane, and so on. Related MFs mainly included protein homodimerisation activity, cytokine receptor binding, oxidoreductase activity, transcription factor binding, protein kinase binding, and so on. The main pathways included Fluid shear stress and atherosclerosis, AGE‐RAGE signalling pathway, PI3K‐Akt signalling pathway, HIF‐1 signalling pathway, apoptosis, IL‐17 signalling pathway, and so on. The Autodock software was used to verify the molecular docking between the five most important active ingredients in GFW (quercetin, kaempferol, baicalein, beta‐sitosterol, and ellagic acid) and the top five targets according to degree value (AKT1, TP53, TNF, HIF1A, and IL‐6). The results are shown in Table  4 . The molecular docking effects of HIF1A and AKT1 with the five main active ingredients mentioned above are shown in Figure  9 . It is generally believed that those with binding energy <−4.25 kcal mol −1 have certain binding activity, while those with binding energy <−5.0 kcal mol −1 have better binding activity (Hsin et al.,  2013 ). In this study, it was found that the binding energies of AKT1, TP53, HIF1A, and IL‐6 with 5 main active ingredients were all less than −5.0 kcal mol −1 , showing good binding activities, as shown in Table  4 . Studies on binding sites of AKT1 and HIF1A found that hydrogen bonds were formed with the ends of multiple residues of active ingredients, such as GLN‐320, GLN‐447, and THR‐445 of baicalein, as shown in Figure  9 . Binding energy of main active ingredients and key targets Abbreviations: AKT1, AKT serine/threonine kinase 1; HIF1A, hypoxia inducible factor 1 subunit alpha; IL‐6, interleukin‐6; TNF, tumour necrosis factor; TP53, tumour protein p53. Molecular docking effect of the main active ingredients of GFW with AKT1 and HIF1A. GFW, Gui Zhi Fu Ling Wan

VC is a common cause affecting male fertility. Oxidative stress injury, increased scrotal temperature, hypoxia, renal and adrenal metabolite reflux, hormone imbalance, anti‐sperm antibody generation, and other pathophysiological mechanisms may be involved in infertility induced by VC. Those VC patients with reduced antioxidant capacity and increased oxidation products will experience further damage to vascular endothelial function. Increased testicular oxidative stress can cause cell death by directly or indirectly affecting spermatogenic cells and the basal membrane of seminiferous tubules. In patients with VC, increased scrotal temperature affects testicular function. Because of the left spermatic vein draining into the left renal vein, when venous regurgitation occurs, renal and adrenal metabolite reflux happen, such as epinephrine, which causes local vasoconstriction and endangers spermatogenesis. When venous blood pressure exceeds the micro‐circulation pressure of the testicular artery, testicular tissue will be in relative hypoxia, and local ischemia results. Elevated temperature, ischemia, and hypoxia can lead to heightened levels of inflammatory factors and metabolic disorders in blood vessels (Jensen et al.,  2017 ). In 2007, Hopkins firstly proposed the concept of ‘network pharmacology’ describing it as a branch of pharmacology that employed network methodology to investigate the synergistic relationship between medications, diseases and targets with ‘multiple components, multiple targets, and multiple pathways’ (Hopkins,  2008 ). In brief, the basic components of network pharmacology research are network construction, network analysis, and experimental verification. At first, by searching public databases and literatures, combining high‐throughput and bioinformatics techniques, drug active ingredients and disease‐related targets are found. Second, visualised tools are used to construct ‘component‐target‐disease’ network and PPI network to predict core active ingredients and hub targets. And analyse the potential signalling pathways and pharmacological mechanisms. Finally, to ensure the reliability of the predicted results, in vivo and in vitro experiments should be further verified (Hopkins,  2008 ). In clinical practice, TCM treatment usually uses compound formulations, following the basic principles such as ‘sovereign‐minister‐assistant‐envoy (Jun‐Chen‐Zuo‐Shi)’ and ‘Yin‐Yang’, to enhance the curative effect or reduce drug toxicity through the deployment of various Chinese medicines, which could achieve comprehensive therapeutic effect. The characteristics of multi‐compound, multi‐target, and multi‐pathway of TCM are congruent with the notion of comprehensiveness, systematism, and integrity of network pharmacology, which conforms to the understanding of TCM's disease essence and gives a new vision for TCM research (Luo et al.,  2020 ). In addition to targets prediction, network pharmacology has been utilised to investigate TCM's compatibility and network toxicology. Although network pharmacology has broad prospects in the field of TCM, there are still some limitations. First, current databases may be incomplete and they are dynamically updated. Second, the use of network pharmacology in the research of TCM prescriptions is mostly qualitative, despite the fact that there is a dose–response relationship between medications and diseases, which is difficult to be explained by network pharmacology. On the other hand, network pharmacology is a static network analysis. However, physical functioning is a continuous and dynamic process, and the same is true of diseases, drug development and pharmacodynamics. In the clinical practice of TCM, the same TCM formula will add or subtract drugs according to different conditions to achieve personalised treatment. Therefore, a considerable number of in vivo or in vitro experimental verifications are necessary (Luo et al.,  2020 ). The main components in GFW include quercetin, kaempferol, baicalein, and ellagic acid. After literature retrieval, it was found that there were few studies about the treatment of VC by the above effective ingredients alone. Quercetin, kaempferol, and baicalein are all flavonoids, which are the main components of many TCM formulae and have significant antioxidant and anti‐inflammatory properties (Dinda et al.,  2017 ; Xu et al.,  2019 ). Animal experiments have shown that adding quercetin to spermatozoa freezing fluid can effectively protect spermatozoa motility and fertilisation ability (Kawasaki et al.,  2020 ). Kaempferol can play an antioxidant role by chelating with heavy metals and reducing malondialdehyde content (Jamalan et al.,  2016 ). Ellagic acid is an organic heterotetracyclic compound. Ellagic acid and its metabolites can effectively scavenge oxygen free radicals and exert antioxidant properties. In addition, ellagic acid effectively down‐regulates the production of inflammatory mediators, and cell experiments have shown that even low concentrations of ellagic acid significantly inhibit TNF‐α and IL‐6 production (Ríos et al.,  2018 ). In this study, the main targets of GFW on VC were screened out through network pharmacology including AKT1, TP53, TNF, HIF1A, IL‐6, CASP3, CAT, PTGS2, and VEGFA. Both IL‐6 and TNF are common inflammatory indicators. Studies have shown that TNF‐α and IL‐6 levels in semen of VC patients are significantly increased, and increased inflammation may lead to increased spermatogenic cell apoptosis (Micheli et al.,  2019 ). CAT is an important antioxidant enzyme that converts hydrogen peroxide into water and oxygen. Ammar et al. found a significant decrease of CAT in semen of VC patients (Ammar et al.,  2021 ). Compared with hernia (no VC) in the control group, Paper et al. found the expression of CAT was lower in the venous specimen of varicocelectomy, and found of MDA, SOD, and CAT were significantly increased in infertility subgroup, which may indicate that intravascular oxidation levels increase and antioxidant levels compensatory rise as the disease progresses (Altunoluk et al.,  2012 ). Also known as COX2, PTGS2 is highly expressed in semen of infertile men with VC, which is positively correlated with VC grade, indicating that VC may lead to increased oxidative stress level and weakened antioxidant capacity of semen (Mostafa et al.,  2016 ). Hypoxia‐inducible factors (HIF) are induced by hypoxia, including HIF1 and HIF2, and HIF‐1α is a common marker of the hypoxia pathway. Lee et al. found that HIF‐1α expression was significantly elevated in venous specimens from VC patients (Lee et al.,  2012 ). In addition, studies have shown that the expression of P53 and HIF‐1α in semen of infertile men with VC were significantly increased and there was a significant positive correlation between P53 and HIF‐1α, while some inflammatory markers such as TNF‐α were not significantly increased (Ghandehari‐Alavijeh et al.,  2019 ). Zhu et al. found that VC leads to testicular hypoxia, which activates the HIF‐1α/BNIP3/Beclin‐1 signalling pathway and induces spermatogenic epithelial cell autophagy (Zhu et al.,  2017 ). Zhao et al. silenced the HIF‐1α gene using CRISPR/Cas9 technique and found that testicular spermatogenic cell apoptosis was reduced and testicular spermatogenic function was significantly improved in VC mice (Zhao et al.,  2019 ). All of these findings imply that the hypoxia pathway is involved in the VC‐induced sperm hypogenesis process. The diseased blood vessels dilate and the blood vessel wall thickens, but the phenomenon remains unclear. As a downstream regulator of HIF‐1α, VEGF is a major regulator of endothelial cell proliferation, angiogenesis, and vascular permeability. VEGF has been found highly expressed in the muscularis and endodermis of human VC venous specimens (Wang et al.,  2017 ). It was shown that VC may cause testicular hypoxia and increased testicular VEGFA expression, which is negatively correlated with total motile sperm count (TMSC) and testicular volume (Shiraishi & Naito,  2008 ). The expression levels of HIF‐1α and VEGFA in testis were significantly decreased after VC repair in mice (Goren et al.,  2017 ). Hypoxia promotes HIF‐1 levels, which stimulates the VEGF‐mediated PI3K/Akt signalling pathway and activates downstream translational regulators 4 E‐BP1 and p70S6K expression. Interestingly, the regulators increase HIF‐1α synthesis, which further increases VEGF expression and promotes spermatogenic cell apoptosis (Wang et al.,  2021 ). The CASP family member CASP3 mainly regulates programmed cell death. Lee et al. found that CASP3 expression decreased in the analysis of varicose veins removed during surgery, and speculated that this dysregulation of apoptosis might lead to dilation and thickening of vascular walls (Lee et al.,  2010 ). An experiment showed that the expression of CASP3 in testis of VC mice was increased and sperm motility was significantly decreased, suggesting that dysregulation of the testicular apoptosis pathway is one of the causes of VC‐induced infertility (Zhu et al.,  2020 ). The primary pathways identified by KEGG enrichment analysis were PI3K‐Akt signalling pathway, HIF‐1 signalling pathway and apoptosis. Based on the above analysis, we propose that GFW has an effect on VC treatment via the following pathways. (1) HIF‐1 signalling pathway and PI3K‐Akt signalling pathway: Both the signalling pathways are involved in hypoxia. PI3K‐Akt signalling pathway could potentially regulates HIF‐1 α via mTOR or FRAP (Zhang et al.,  2018 ). The HIF‐1 signalling pathway, which may also trigger cell autophagy, was a key mechanism in hypoxia response. The principal biological processes identified by GO enrichment analysis involved blood vessel morphogenesis and blood vessel development. VEGF expression increased as a result of VC‐induced hypoxia, which led vascular endothelial cells to proliferate and migrate, and was critical for both healthy and pathological angiogenesis (Wang et al.,  2021 ). The core active components and AKT/HIF1A had strong binding potential which may verify the important role of hypoxia‐related signalling pathway in treating VC, according to the results of molecular docking.(2) Apoptosis signalling pathway: Research has indicated that the apoptotic signalling pathway is engaged in VC‐induced infertility. Response to metal ion and regulation of cellular response to stress were identified as primary biological processes by GO enrichment analysis. Metal ion such as cadmium induces apoptosis. In certain individuals with varicocele, Benoff et al. showed that the percentage of apoptotic nuclei and cadmium levels were high (Benoff et al.,  2004 ). Some inflammatory cytokin like TNF‐α and IL‐6 were increased in VC because of heat stress (Micheli et al.,  2019 ). Besides, the level of oxidative stress increased and antioxidant capacity decreased. The main components of GFW such as quercetin, kaempferol, and baicalein have significant antioxidant and anti‐inflammatory properties (Dinda et al.,  2017 ; Xu et al.,  2019 ). Heat stress is associated with increased levels of inflammation and oxidative stress, which can finally induce apoptosis (Hassanin et al.,  2018 ). In conclusion, we performed a systematic integrated network pharmacology study and molecular docking to illustrate the key active components and primary targets of GFW on VC treatment. GFW was discovered to be able to prevent apoptosis, participate in venous vessel morphogenesis, and reduce oxidative stress in the treatment of VC. The main active components are quercetin, kaempferol, baicalein, beta‐sitosterol, ellagic acid. The core targets are AKT1, TP53, TNF, HIF1A, IL‐6, CASP3, CAT, PTGS2, and VEGFA. The PI3K/Akt signalling pathway, HIF‐1 signalling pathway, and apoptosis are the most representative pathways. Further, the substantial binding interaction between the key active components and core targets was confirmed using molecular docking. This research can be utilised as a foundation for future clinical and scientific research, which can lead to the development of novel medications and therapeutic techniques to treat VC. Although all relevant active ingredients have been included as far as possible in this study, changes of active ingredients caused by decocting or special preparation of TCM have not been considered, which represents a limitation of this study. These results need to be verified by further experiments in the future.

The aberrant extension, elongation, and tortuosity of the pampiniform venous plexus within the spermatic cord is known as varicocele (VC). Testicular pain, impaired sperm quality, and even testicular atrophy are all possible side effects of VC. In the general male population, the prevalence of VC is around 10%–15%, but it can be as high as 30%–40% in primary male infertility (Jensen et al.,  2017 ). Methods of treatment for VC are mainly divided into three kinds: (I) general therapy, such as scrotallift, lifestyle adjustment; (II) drug therapy, such as antioxidants, escin (Aescuven Forte), vasoactive traditional Chinese medicine (TCM); (III) operation treatment, including microscopic/laparoscopic spermatic vein ligation, spermatic vein embolisation (Fang et al.,  2010 ; Jensen et al.,  2017 ; Su et al.,  2021 ). Antioxidants (such as vitamins E, selenium, l ‐carnitine) are widely used in the treatment of infertility caused by varicocele in clinical practice, but there are used empirically and lack of controlled, randomised prospective trials to test their effectiveness and safety (Su et al.,  2021 ). Escin is an extract of the Aesculus hippocastanum seed that has been shown to be useful in the treatment of chronic venous disorders like VC, haemorrhoids, and venous varicocele in the lower limbs. However, studies have shown that escin is less effective in treating severe VC (Fang et al.,  2010 ). There are complications to surgical treatment, including scrotal edema, vas deferens and artery injury, and there is a certain probability of recurrence. Therefore, the development of new effective drugs for the treatment of VC is necessary and important. The use of TCM in China has a history spanning over a 1000 years, and TCM is an attractive option for VC treatment. In East Asian countries such as China, Japan, and Korea, GFW is commonly used. There are five components in GFW, Cinnamomi Ramulus (Gui Zhi, GZ), Poria Cocos (Fu Ling, FL), Cortex Moutan (Mu Dan Pi, MDP), Radix Paeoniae Rubra (Chi Shao, CS), and Persicae Semen (Tao Ren, TR). According to TCM theory, GFW promotes blood circulation to remove blood stasis, and is used for the treatment of various diseases such as menopause syndrome, uterine fibroids, endometriosis, and systemic sclerosis (Kim et al.,  2017 ; Wang et al.,  2019 ). The main efficacy of GFW is removing blood stasis (anti‐Oketsu) via through increasing microhemodynamic measures such as erythrocyte congestion resolution and the cell‐free layer, both of which may be related to endothelial NO generation (Tomita et al.,  2017 ). In addition, GFW also has the properties like anti‐coagulation, anti‐fibrosis and anti‐inflammatory (Tomita et al.,  2017 ). VC is mainly caused by impaired spermatic vein function, resulting in testicular hypoxia and increased ROS, which further harms testicular spermatogenesis (Jensen et al.,  2017 ).TCM is known for the theory ‘Treating different diseases with the same treatment (Yi Bing Tong Zhi)’, and the treatment of VC using GFW is the innovative application of this theory in the field of andrology. Ishikawa et al. showed that GFW could not only effectively treat VC, but also improve semen quality (Ishikawa et al.,  1996 ). However, the molecular mechanism of the pharmacological action of GFW in treating VC is still unclear and further experimental studies are needed. Network pharmacology is a research method integrating medicine, computer technology, and bioinformatics. By constructing a ‘drug‐disease‐target’ network and analysing the biological functions and action pathways of the main targets, the molecular mechanism of drugs acting on diseases can be deeply understood and ideas for drug research can be provided. The system of TCM is characterised by multiple components, targets, and pathways. Based on network pharmacology, Chen et al. found that Wuzi Yanzong Pill could treat spermatogenesis disorders by regulating the apoptosis pathway (Chen et al.,  2020 ). Through network pharmacology, Zhao et al. found that Zuogui Yin can treat male infertility by acting on vascular endothelial growth factor A (VEGFA), caspase 3 (CASP3), and tumour necrosis factor (TNF), through the TNF signalling pathway, PI3K‐Akt signalling pathway, and other pathways (Zhao et al.,  2021 ). In this study, the main targets of GFW treating VC were identified, the related biological functions and signal pathways were analysed, and the molecular mechanism of GFW was further understood using the network pharmacology method, which is conducive to the clinical application of GFW and its worldwide promotion.

The authors declare no conflict of interest.

Appendix S1: Supporting Information Click here for additional data file.