Eucommia ulmoides Oliver's Multitarget Mechanism for Treatment of Ankylosing Spondylitis: A Study Based on Network Pharmacology and Molecular Docking

Background Eucommia ulmoides Oliver (EU) is a plant used in Chinese medicine as a medicinal herb to treat autoimmune and inflammatory conditions. We used network pharmacology to examine the active ingredients and estimate the main targets and pathways affected by EU when it is used to treat ankylosing spondylitis (AS). Materials and Methods The Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform was used to search for active ingredients in EU and their target proteins. The GeneCards Database was used to find AS-related targets. The targets from the EU and AS searches that coincided were selected by constructing a Venn diagram. Then, a STRING network platform and Cytoscape software were used to analyse the protein-protein interaction (PPI) network and key targets. The strong affinity between EU and its targets was confirmed using molecular docking techniques. The Gene Ontology and the Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathway enrichment analysis of overlapping targets was performed using the database for annotation, visualization, and integrated discovery online tool. Results The number of active ingredients against AS in EU was discovered to be 28. Major targets against AS in the PPI network and core targets analyses were identified as IL-1B, PTGS2, IL-8, nMMP-9, CCL2, MYC, and IL-2. Furthermore, molecular docking studies showed the strong affinity between EU's bioactive molecules and their AS targets. Enrichment analysis revealed that active ingredients from EU were involved in a variety of biological processes, including the response to molecules derived from bacteria, extracellular stimuli, nutrient levels, and the regulation of reactive oxygen species, all of which are mediated by interleukin-17, TNF-α, and other signalling pathways. Conclusion The therapy for AS using EU involves a multitarget, multipathway, and multiselection mechanism that includes anti-inflammatory and analgesic effects. This study provides a theoretical basis for future research into targeted molecular therapies for AS.


Introduction
Ankylosing spondylitis (AS) affects the spine and sacroiliac joints and is a chronic inflammatory illness. Due to the inflammatory damage induced by AS, patients have persistent back pain and morning stiffness, which limits their spinal mobility [1]. Furthermore, AS commonly aggravates extraarticular conditions such as uveitis [2]. Multiple environmental and genetic factors are known to have a role in determining the likelihood of acquiring this condition, even though the underlying pathophysiology is still unclear. More than 90% of the genetic causes have been determined so far. Research has revealed that AS is associated with the HLA-B27 gene [3]. Patients suffering from AS are frequently treated with nonsteroidal anti-inflammatory medicines, disease-modifying antirheumatic drugs, and tumour necrosis factor-alpha (TNF-α)-specific therapies, with surgery recommended in cases where there is severe pain or joint damage. However, long-term medication has been linked to several adverse effects, including cardiovascular, gastrointestinal, and renal concerns [4,5]. Another significant disadvantage of this treatment is the high cost of these medications [6].
Traditional Chinese medicine (TCM) divides AS into three categories: lower back pain, spinal stiffness, and kyphosis. Many researchers have studied the function of TCM in the treatment of AS [7]. Eucommia ulmoides Oliver (EU), known as 'du zhong' in Chinese and native to China, is the only plant in the Eucommia genus of the Eucommiaceae family.
e chemical constituents of EU leaves include phenylpropanoid compounds such as sitosterol (SIT), chlorogenic acid, and ursolic acid; iridoids such as eucommiol, geniposide, geniposidic acid, and aucubin; and flavonoids such as quercetin and kaempferol [8]. As a natural herb, EU may help to strengthen bones and muscles, improve liver and kidney function, and relieve pain in the waist region and the knees [9]. It is most often used to treat renal failure and lumbago, muscular and bone weakness, and haemorrhage during pregnancy [10]. e extract of EU also lowers the activity of the PI3K/Akt pathway, which helps to reduce the production of inflammatory mediators [11]. One of the main ingredients of Jitongning tablets is EU, and these are regularly used to treat AS and rheumatoid arthritis. In a clinical trial, Jitongning tablets were shown to be an effective and safe therapy for AS [12,13]. In addition, EU contains various phytochemicals including lignin, phenolics, terpenoids, and flavonoids [14]. However, the chemical component of EU that is effective in the treatment of AS may be unknown, so this needs further study.
Regarding biological networks, network pharmacology is a uniquely promising and cost-effective strategy for discovering bioactive components, predicting drug targets, and analysing mechanisms of drug action [15]. Furthermore, network pharmacology, in contrast to experimental pharmacology approaches, stresses multichannel control of signalling pathways, making it particularly well suited to explaining the mechanism of TCM, as this involves many chemical components and molecular targets [16].
In this research, we used network pharmacology to examine the active constituents of EU for the treatment of AS and suggest the main targets and pathways involved for the first time. e targets of EU active ingredients in the treatment of AS were analysed and their enrichment analysis was carried out. e targets were then preliminarily confirmed using molecular docking, which might provide a theoretical basis for later research into targeted molecular therapies for AS. Figure 1 depicts the general design of this study.

Searching for Active Ingredients.
e TCM Systems Pharmacology Database and Analysis Platform (TCMSP) (https://tcmspw.com/tcmsp) [17] was used to look for active ingredients in EU. e rate and degree to which a medicine is absorbed into the body's circulation is referred to as its oral bioavailability (OB). Drug-like features (DL) describe the functional groups or physical characteristics of a chemical compound that are the same or comparable to those of known drugs. e Caco-2 cell line from human intestinal epithelial cells is often used in studies of absorption and transportation in intestinal epithelial cells. e half-life of a drug is the critical statistic for estimating the dosing interval, dosage delivered, and drug accumulation since it indicates the drug's concentration in the blood or tissues [16]. As previously reported [18], the compounds with greater activity, OB > 30% and DL > 0.18 were further screened.

Searching for Drug Targets and Construction of 'Active
Drug-Target Component' Network. Drug target proteins were discovered when the active ingredients identified as described in Section 2.2 were entered into the TCMSP database for retrieval. After that, UniProt numbers were acquired. e database (https://www.uniprot.org/) was used to connect the target proteins and gene information. After entering the active ingredients and target proteins into the Cytoscape 3.8.0 program, an 'active ingredient-target component' network was created. To select important active ingredients, a significance analysis was performed.

Acquisition of Potential Targets for the Treatment of AS by EU.
e OMIM (https://omim.org/), PharmGKB (https:// www.pharmgkb.org/), TTD (https://db.idrblab.net/ttd/), DrugBank (https://go.drugbank.com/), and GeneCards (https://www.genecards.org/) databases were used to find AS-related targets. To specify the AS condition, the terms 'ankylosing spondylitis' or 'AS' were used; the term 'Homo sapiens' was used to search for human disease targets related to AS. Using a Venn diagram, the pharmacological targets that were identified in Section 2.2 were mapped to relevant disease targets. To discover the possible specific targets of EU in the treatment of AS, common nonspecific targets were filtered out.

Protein-Protein Interaction Network and Core Target
Analysis.
e chosen targets were entered into the STRING network platform (https://string-db.org/) [19] to create a protein-protein interaction (PPI) network to determine the probable direct targets of EU and the interactions between them. e confidence level was set to medium, and the protein type was set to Homo sapiens (0.400). e PPI network was built using Cytoscape 3.8.0 software, and the key targets were examined and selected.

Gene Ontology (GO) Annotation.
e key mechanisms of the action of the targets were investigated using GO enrichment. e GO enrichment of EU's targets in treating AS was investigated using the ClueGO plug-in in Cytoscape. Only relevant GO keywords with P < 0.05 were presented [20], and the protein type was assigned to Homo sapiens. We used the 'Analysis' function to create a graphic depiction of the top 10 elements in biological process (BP), molecular function (MF), and cellular component (CC) based on their relevance. Evidence-Based Complementary and Alternative Medicine

Kyoto Encyclopaedia of Genes and Genomes (KEGG)
Pathway. e KEGG pathway was used to further explain the function of the targets in metabolism, signal transduction, and other activities by monitoring their distribution throughout the network. For pathway enrichment analysis, the targets were loaded into the database for annotation, visualization, and integrated discovery (DAVID) database (https://david.ncifcrf.gov/). e official gene symbols were used as the identifiers. GeneList was chosen as the list type. Homo sapiens was the only species available. en, we looked for KEGG pathway information that differed significantly. We used R software, R 3.6.0 for Windows to create a bubble plot of the signalling pathways. We used the DAVID database to download the most enriched pathway, and the enriched genes were indicated with red stars.

Verification of Compound-Target Interaction Using Molecular Docking.
To represent the relationship between small molecule ligands and protein receptors, molecular docking technology developed based on the lock and key concept. In addition, based on the conformation change and energy matching procedure, the optimal binding mode and location were found. e SYBYL-X 2.1.1 for Windows was used in this work to confirm the compound-protein target interaction.
e Research Collaboratory for Structural Bioinformatics Protein Data Bank database (https://www.rcsb. org/) was used to get the crystal structures of the protein targets. e imported crystal structures were used to build the receptor protein grid in the molecular operating environment (MOE), which was related to protonation, water removal, repeated structure building, structure preparation, and energy reduction. e ligand binding site was chosen using the receptor grid design. Finally, to acquire the graphical results of molecular docking, the threedimensional BioData Mining (2020) 13 : 12 Page 5 of 18 chemical structure was altered using ChemBioDraw software, imported into the MOE, and docked with the protein construct. response to molecule of bacterial origin response to extracellular stimulus response to nutrient levels reactive oxygen species metabolic process multi-multicellular organism process cellular response to biotic stimulus regulation of reactive oxygen species metabolic process reactive oxygen species biosynthetic process female pregnancy external side of plasma membrane membrane ra membrane microdomain ficolin-1-rich granule lumen secretory granule lumen dendrite cytoplasm cytoplasmic vesicle lumen vesicle lumen myelin sheath ficolin-1-rich granule  Evidence-Based Complementary and Alternative Medicine

Evidence-Based Complementary and Alternative Medicine
Evidence-Based Complementary and Alternative Medicine

Searching for Bioactive Compounds in EU and eir Related Targets for the Treatment of AS.
e TCMSP was used to identify 28 active compounds from EU and 1474 associated targets based on the following criteria: OB > 30% and DL > 0.18 (Table 1).

Active Drug-Target Component Network Construction and Analysis.
e network had 784 edges and 73 nodes, including 21 active ingredient nodes and 73 target nodes ( Figure 2). e network topology characteristics were analysed using the network analysis tool. We looked at the nodes that had been selected the most. ese nodes were shown to have an important function in the network and might represent important chemicals or targets. e network's active components had an average selection value of 21.5. We conducted a study to determine the association between the active ingredient and AS target ( Table 2).

Potential Targets of EU in the Treatment of AS.
e OMIM, PharmGKB, TTD, DrugBank, and GeneCards databases were used to find 2,266 AS targets. After mapping EU-related targets and AS-related targets in a Venn diagram, 73 targets overlapped with matching target-related bioactive compounds. Finally, Cytoscape was used to show 21 bioactive substances with their corresponding 73 overlapping targets. is highlights the targets of EU's active ingredients for the treatment of AS (Figure 3). Figure 4 the network included 73 nodes and 784 edges, with an average node degree of 21.5. e size, color, and degree value of the nodes in the diagram were all positively associated. e greater the size and darker the hue of nodes, the higher their degree value. A greater degree suggested stronger interactions with other proteins, which might play a key role in the biological activity. e proteins IL-1B, PTGS2, IL-8, nMMP-9, CCL2, MYC, and IL2 were the top seven targets in terms of degree value (Figure 4), and they may be important targets for EU's effectiveness in the treatment of AS. We undertook a review of the literature to see whether there was a link between these active components and AS or inflammation (Table 3).

Enrichment Analysis of the GO Annotation and the KEGG Pathway.
e FunRich 3.1.3 program screened a total of 1734 GO items (P < 0.05). ere were 1604 BP, 105 MF, and 25 CC protein targets in all. Responses to lipopolysaccharide (LPS), bacterial molecules, extracellular stimuli, nutrient  e exterior side of the plasma membrane, membrane raft, membrane microdomain, ficolin-1-rich granule lumen, and secretory granule lumen are all part of the CC. All the top ten protein targets are shown in Figure 5.
Response to molecules of bacterial origin, LPS, response to extracellular stimuli, response to nutrient levels, reactive oxygen species, multi-cellular organism processes, cellular response to biotic stimuli, and regulation of reactive oxygen species were among the 139 KEGG-enriched signalling pathways discovered by searching the DAVID database (P < 0.05). Figure 6 depicts the top 30 signalling pathways with the greatest enrichment significance. Simultaneously, we conducted a KEGG analysis on the interleukin-17 (IL-17) signalling pathway as an example. Multiple components of the IL-17 signalling cascade, including AP1 and Hsp90, were linked to AS, according to these findings (Figure 7).

Verification of Compound-Target Interactions.
e top three target proteins in the PPI network with intermediate value (IL-1B, PTGS2, and IL-8) were chosen, and the active ingredient quercetin was rated first in the median of the 'active drug-target component' network. For molecular docking, SYBYL-X 2.1.1 was employed (Table 4). Higher absolute docking scores imply a greater binding force between the molecule and the target, and a more stable molecular conformation [46,47]. A docking score absolute value > 4.25 suggests average binding activity, >5.0 shows good binding activity, and >7.0 indicates excellent binding activity. Molecular docking experiments revealed that PTGS2 had the greatest affinity and binding energy for quercetin. Each protein was effectively docked with the bioactive compound. Quercetin, extracted from the compound-target network, and a stable docking model with a particular binding location, distance, and structure was obtained. e binding site where the drug's active component interacts with the target protein may be as shown in Figures 8(a)-8(c).

Discussion
e rational method of network pharmacology is used to investigate the possible biological mechanisms of TCM in the treatment and prevention of many illnesses, particularly chronic and recurrent disorders [48]. Previous research has shown that EU has a key role in the treatment of AS [49], although the mechanism of the action is still unknown. For the first time, pharmacology was employed to comprehensively reveal the mechanism of the action of EU against AS as a network of effects. Furthermore, the compoundtarget stable molecular docking model demonstrated successful binding between representative compounds and key AS targets, confirming the interactions between EU active ingredients and protein targets affecting AS. ese examples of computer models/figures demonstrated EU's direct influence on AS and possible methods including 'multiple chemicals, multiple targets, and multiple mechanisms. ' Previous studies have found that many TCM components play important roles in many human diseases. For example, Averrhoa carambola leaf extract depresses the central nervous system in the thiopental-sodium model of Swiss albino mice [50]. Also, the leaf extract of Lannea coromandelica is potentially antidiabetic [51]. In addition, polyphenol nanoformulations have been used to treat diabetes mellitus [52]. As reported by Saikat Mitra et al., the leaf extract of Avicennia alba has antidiabetic, antiinflammatory, analgesic, and antidiarrheal activity [53]. Furthermore, zinc oxide nanoparticles and natural small molecules can act as important regulatory factors in human cancer [54,55]. Bioactive compounds such as resveratrol also have therapeutic effects on Alzheimer's disease [56,57]. In this study, quercetin, kaempferol, ent-epicatechin, catechin, and beta-SIT were discovered as the most significant active ingredients of EU in the treatment of AS after searching for all the active ingredients and analysing the compound-target network. Previous research has shown that quercetin stimulates chondrocyte proliferation and acts on inflammatory damage to speed up cartilage repair [58]. Kaempferol is increasingly being linked to the inflammatory process. Kaempferol's anti-inflammatory properties include that it regulates cytokines and inhibits the MAPK pathways [59,60]. Also, by downregulating ICAM-1 and VCAM-1, kaempferol can ameliorate endothelial dysfunction and rheumatic disease symptoms. Epicatechin has antioxidant properties, may efficiently scavenge free radicals, and inhibits the cell's oxidative potential [61]. e activity of cells involved in the immunological response is modulated by ent-epicatechin [62]. e catechin group's most significant actions include anti-inflammatory, antioxidant, and chemopreventive activity [63]. e compound SIT is a bioactive phytosterol found in plant cell membranes that has a chemical structure like cholesterol found in human cells [64].  Evidence-Based Complementary and Alternative Medicine    Names e role of key targets in antiosteoarthritis or anti-inflammatory and related mechanisms References IL1B IL-1B belongs to the IL-1 family and is a proinflammatory cytokine mainly produced by activated macrophages. It is positively correlated with inflammation [29,30] PTGS2 Inflammation regulates the expression of PTGS2. Numerous investigations have demonstrated that inflammation can express PTGS2 opportunistically.
CXCL8 is a chemokine that binds to the CXCR1 and CXCR2 receptors. Extensive experimental evidence indicates that CXCL8 and its receptors not only aid in pathogen elimination, but also significantly contribute to disease-related processes such as tissue damage, fibrosis, angiogenesis, and carcinogenesis. [32]

MMP9
MMP9 and other MMP protein may be involved in the interaction between inflammatory and progenitor cells of the skeletal system. MMP9 inhibits the inflammatory response between bone cells. [33,34] CCL2 CCL2 is implicated in malignant tumour metastasis, which promotes malignant tumours. In addition, CCL2 can exacerbate stomach inflammatory symptoms. [35][36][37][38] c-MYC c-MYC can inhibit inflammation by activating the expression of inflammatory cytokines. [39][40][41][42][43] IL2 IL2 may indirectly promote resistance to inflammation [44,45]  response to lipopolysaccharide response to molecule of bacterial origin response to extracellular stimulus response to nutrient levels reactive oxygen species metabolic process multi-multicellular organism process cellular response to biotic stimulus regulation of reactive oxygen species metabolic process reactive oxygen species biosynthetic process female pregnancy external side of plasma membrane membrane raft membrane microdomain ficolin-1-rich granule lumen secretory granule lumen dendrite cytoplasm cytoplasmic vesicle lumen vesicle lumen myelin sheath ficolin-1-rich granule DNA-binding transcription factor binding nuclear receptor activity RNA polymerase II-specific DNA-binding transcription factor binding transcription coactivator binding cytokine receptor binding repressing transcription factor binding cytokine activity transcription coregulator binding receptor ligand activity ligand-activated transcription factor activity      Notably, these EU active ingredients work together to produce anti-inflammatory, anti-apoptotic, and cartilage homeostatic effects, all of which might be considered viable treatment options for AS. e PPI network analysis revealed that IL-1B, PTGS2, IL-8, and nMMP-9 may be significant targets for EU's effectiveness in the treatment of AS. ese targets had the most contact with others in EU's overall active ingredient-AS target network, and they were most likely to cause a cascade of antagonistic effects on AS. Furthermore, gene targets working on several pathways were discovered in the CTP network, indicating the relevance of certain targets in the broader AS bioinformatics network.
is also supported EU's 'multiple targets-multiple routes' regulatory mechanism.
e cytokine IL-1B is a proinflammatory cytokine generated primarily by activated macrophages and belongs to the IL-1 family. Monocytes and macrophages and B and T lymphocytes are all activated by IL-1B, which causes fibroblast proliferation and synovial pannus development. In addition, IL-1B has a role in chronic inflammation. Furthermore, the intracellular actions of IL-1A and IL-1B may cause inflammation in AS [29].
ere are currently few studies linking IL-1 and AS. Chromosome location 2q13 is associated with AS. e interleukin 1 (IL-1) gene cluster is located 132 cm from the telomere of chromosome 2p, in a region linked to AS risk [30]. Prostaglandins implicated in inflammation and/or cell proliferation are produced by PTGS2. e elevation of PTGS2 by inflammatory agents and during inflammatory responses in cells that mediate inflammation, and the decrease of PTGS2 expression in the presence of anti-inflammatory drugs, provide significant evidence that this enzyme plays a role in inflammation [31]. Interleukin IL-8 is generated by fibroblasts and capillary endothelial cells in response to a variety of stimuli, including exogenous microbial products and proinflammatory cytokines including IL-1 and TNF-α. e mechanism of IL-8 action in AS, like that of IL-1, is unclear. e levels of IL-8 are higher in patients with AS, according to existing research [32]. In osteoarthritis, nMMP-9-mediated syndecan-4 shedding correlates with severity [33]. In addition, nMMP-9 protects against LPS-induced inflammation in osteoblasts [34]. While the above research found several EU active ingredients acting on AS targets, how do these targets affect AS? More research was needed.
Our KEGG study demonstrated that lipids and atherosclerosis, the IL-17 signalling pathway, and the TNF signalling pathway are all linked to EU's therapeutic mechanism for the treatment of AS. As a chronic inflammatory disease, AS affects the spine and sacroiliac joints. It has the potential to harm the skeletal system's structural and functional integrity. Several other organs, including the eyes, skin, gastrointestinal tract, and cardiovascular system (CVS) may also be damaged [65]. Among CVS disorders, atherosclerosis is one of the leading causes of morbidity and mortality [66]. e major risk factor for atherosclerosis is dyslipidaemia [67]. e abnormalities in lipid profiles in people with AS are assumed to be caused by systemic inflammation [68]. Systemic inflammation triggers the generation of free oxygen radicals via immune-mediated pathways, which have been linked to the pathophysiology of inflammatory diseases. e formation and progression of atherosclerosis are influenced by systemic inflammation, endothelial inflammation, and increased oxidative stress [69]. Several clinical investigations [70] have shown that TNF-α, coupled with IL-17 and interleukin-23 (IL-23), plays a crucial role. e cytokine IL-17 has been demonstrated to improve T cell priming and induce the production of proinflammatory mediators (such as IL-1, IL-6, TNF-α, and chemokines) by a variety of cell types, including fibroblasts, endothelial cells, macrophages, and epithelial cells [71]. e cytokine IL-17 has also been found to play a key role in the effector phase of the inflammatory response. e expression of IL-17 is elevated in AS patients [72]. e signalling mechanisms involving TNF-α are complicated and not completely understood. When TNF-α binds to the TNFR1, it activates proinflammatory and programmed cell death pathways, which are linked to tissue damage, according to results from in vitro and in vivo studies. e TNFR2 has been demonstrated to modulate tissue healing and angiogenesis signals [73]. e signalling molecule TNF-α has been found in the sacroiliac joints of individuals suffering from AS, especially in the early active stages of the condition [74,75], and increased blood TNF-α levels are linked to the disease activity.
Molecular docking experiments also revealed that a representative selection of AS targets, IL-1B, PTGS2, and IL-8 had a high affinity for the active ingredients of EU. In this research, network pharmacology was employed to build on EU's possible anti-AS actions, which were then visualised using molecular docking. However, the findings from the prediction need additional experimental verification due to the limits of network pharmacology and molecular docking technologies.

Conclusion
According to this research, treatment of AS using EU might involve a complex multitarget, multipathway, and multiselection mechanism. e proteins IL-1B, PTGS2, IL-8, nMMP-9, CCL2, MYC, and IL2 were identified as major targets against AS in the PPI network and core target analyses. e strong affinity between EU's bioactive molecules and their AS targets was confirmed. As a medicinal herb, EU may be used as an analgesic, immunomodulator, antimicrobial, and anti-inflammatory. is study might provide a theoretical basis for research into targeted molecular therapies for AS.

Data Availability
All data generated or analysed during this study are included in this article.

Disclosure
Sheng-Nan Zhang is Co-first author.