Potential Mechanisms of Acupuncture Treatment for Rheumatoid Arthritis: A Study Based on Network Topology and Machine Learning

Potential Mechanisms of Acupuncture Treatment for Rheumatoid Arthritis: A Study Based on Network Topology and Machine Learning | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Potential Mechanisms of Acupuncture Treatment for Rheumatoid Arthritis: A Study Based on Network Topology and Machine Learning Feiyang Li, Zhen Liu, Yuan Xu, Yi Guo, Zhifang Xu, Gongming Yuan, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6535408/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 Oct, 2025 Read the published version in Chinese Medicine → Version 1 posted 9 You are reading this latest preprint version Abstract Objective: Rheumatoid arthritis (RA) is a systemic autoimmune disease requiring multi-target therapeutic strategies. Acupuncture, a holistic therapy in traditional Chinese medicine (TCM), has shown clinical efficacy in RA, yet its molecular mechanisms remain elusive. By integrating network topology and machine learning methods, decode the systemic regulatory effects of acupuncture on RA of acupuncture on RA. Methods: Data on the interactions between acupuncture-affected endogenous compounds and RA-related targets were extracted from databases, built a multi-dimensional interaction network to map the interactions between acupuncture and RA. screened RA differentially expressed genes (DEGs) from GEO, intersecting with acupuncture-responsive genes. Utilize clusterProfiler for KEGG/GO enrichment analysis of these DEGs, Analyze the immune microenvironment using CIBERSORTx and xCell algorithms. Utilize ConsensusClusterPlus (R package) for unsupervised clustering to obtain DEGs. Subsequently, identify key genes using an ensemble machine learning model (GLM/SVM/XGB/RF) and create nomograms. Apply TwoSample MR and colocalization analysis to validate the causal relationship between core acupuncture-affected DEGs and RA risk. Results: This study identified 10 acupuncture-regulated metabolites and 49 RA-related DEGs. KEGG analysis showed DEGs enriched in immune pathways including the JAK/STAT pathway mediating inflammatory responses, the T-cell receptor signaling pathway involved in T cell differentiation and the TNF signaling pathway. Immunome profiling based on the CIBERSORT algorithm indicated that DEGs were primarily enriched in key immune cell subpopulations such as M1 macrophages, activated CD4⁺ T cells, Tregs, and B lymphocytes. Machine learning identified five key genes associated with immune infiltration (STAT1, GAPDH, JAK2, PTGS2, MDM2). MR/colocalization confirmed acupuncture-regulated STAT1 expression positively correlated with RA genetic susceptibility, highlighting STAT1-mediated JAK/STAT pathway in immune remodeling. Conclusion: STAT1, GAPDH, JAK2, PTGS2, and MDM2 may be potential targets for acupuncture treatment of RA. Acupuncture may achieve systemic immune regulation by synergistically targeting multiple pathways (JAK/STAT, TNF) and immune cells (M1 macrophages, CD4 + T cells), This initiative integrates the holistic philosophy of TCM with the precision of AI-driven medical science. Acupuncture Rheumatoid Arthritis Machine Learning Network Topology Mendelian Randomization Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 1. Introduction Rheumatoid arthritis (RA), a prevalent chronic inflammatory disorder, profoundly impacts patients' quality of life through persistent joint damage and systemic complications [1]. Characterized by progressive musculoskeletal deterioration, the disease leads to functional impairment, reduced productivity, and substantial healthcare burdens [2,3]. Epidemiological studies indicate a global prevalence of 0.5-1.0%, with peak onset between 30-50 years and a 3:1 female predominance [4]. Current pharmacological management employs disease-modifying antirheumatic drugs (DMARDs), biologics, and anti-inflammatory agents. While these therapies demonstrate clinical efficacy, they present limitations including adverse effects (gastrointestinal, hepatotoxicity) [5], drug resistance [6], and increased infection risk with prolonged biologic use [7]. Notably, monoclonal antibodies targeting specific cytokines (TNF-α, IL-6) show variable response rates, with some patients developing treatment resistance [8]. This therapeutic challenge stems from RA's complex pathogenesis involving dysregulated immune networks rather than isolated cytokine dysfunction [9]. These clinical observations have prompted a paradigm shift in therapeutic development - from single-target approaches to multi-target strategies addressing interconnected inflammatory pathways. This transition reflects growing recognition of RA as a systemic immune network disorder requiring comprehensive intervention strategies. Acupuncture, rooted in TCM's holistic philosophy of “regulating the body as an organic whole”, has been empirically proven to alleviate RA symptoms [10-12]. However, its multi-target mechanisms—particularly in immune-microenvironment modulation—are poorly understood. Traditional reductionist approaches struggle to capture the complexity of acupuncture's systemic effects, necessitating advanced computational strategies. Bioinformatics methods play a vital role in the study of acupuncture transformation in academia. The amount of data generated by each stage of acupuncture discovery is increasing, and the use of these data for calculation can solve the key challenges in this process [13]. Network topology analysis has proven particularly valuable for mechanistic studies and target prediction in traditional medicine [14,15]. Using machine learning algorithms such as clustering and support vector machine (SVM) to mine useful information from a large number of TCM data, optimize TCM research design, reduce clinical research costs, and improve research quality and efficiency. Mendelian randomization (MR) research is a method that uses genetic variation as a tool variable to explore the causal relationship between risk factors and diseases. In the field of TCM, MR can effectively overcome the bias caused by confounding factors and reverse causality, and provide a new causal inference method for TCM research. Recent studies have emphasized the overall and multi-target effects of acupuncture on various body systems. Recent advances in Network topology and artificial intelligence (AI) offer unprecedented opportunities to decode acupuncture's “multi-component, multi-target” nature. By integrating topological analysis of compound-target networks with ensemble machine learning models, we can systematically identify hub genes that serve as convergence points of acupuncture's systemic regulation. Therefore, it is of great clinical value and research significance to explore the potential mechanism of acupuncture in the treatment of RA with a new research perspective [16]. In this study, database retrieval is used to identify the active components of acupuncture in the treatment of RA. Based on the previous study by Hanet al. [17], we employed network topology and machine learning to investigate the potential mechanism of acupuncture in treating RA and analyze its effectiveness. By exploring the interaction between acupuncture and disease, the therapeutic targets of acupuncture were clarified, and gene expression pattern were confirmed by bioinformatics analysis of the Gene Expression Omnibus (GEO) data set. Through GO/KEGG functional enrichment analysis, we systematically analyzed the key biological processes and signaling pathways of acupuncture intervention in RA, revealing its potential molecular mechanisms for restoring immune homeostasis through pathways such as JAK-STAT/TNF. We also identified the characteristics of immune cell infiltration in RA, providing valuable insights for future research. Additionally, machine learning algorithms were used to screen core therapeutic targets for RA. MR analysis and colocalization analysis established causal relationships between key targets and RA, validating the impact of the key gene STAT1 on RA risk. This study developed an AI-guided comprehensive framework that not only elucidated molecular targets for acupuncture treatment of RA (such as STAT1) but also clarified the overall logic of acupuncture therapy, offering promising prospects for current research and clinical interventions. 2. Material and methods 2.1 Collection of potentially effective active compounds produced in the human body after acupuncture treatment for RA In this analysis, we systematically searched four databases: Web of Science, PubMed, CNKI, and Wan Fang Database (last updated on June 30, 2024), using the following terms: “acupuncture”, “body acupuncture”, “electroacupuncture”, “warm needle”, “fire needle”,” blood-letting puncture”, “RA”, and “arthritis”. Acupuncture is defined here as the stimulation of acupuncture points on the skin with or without electrical stimulation. Studies involving other forms of stimulation, such as acupressure, transcutaneous electrical nerve stimulation, and laser acupuncture, were excluded. We included randomized controlled trials in humans or animals, as well as non-randomized comparative trials (prospective and retrospective), and excluded single-group observational studies that evaluated outcomes before and after interventions. Control interventions could involve placebo acupuncture, sham acupuncture, no treatment, another form of active treatment, or medication. Studies comparing only different forms of acupuncture were excluded. The reported outcomes of interest were also analyzed. Two researchers independently searched the databases using predefined inclusion and exclusion criteria and selected appropriate full-text articles. Disagreements among the researchers were resolved through discussion, with a third researcher resolving disputes in cases of non-consensus. 2.2 Identification of protein targets for active components produced in the body after acupuncture and acquisition of genes related to RA PharmMapper is an integrated platform for pharmacophore matching that utilizes statistical methods to identify potential targets [18]. Swiss Target Prediction, a web server based on 2D and 3D similarity metrics and known ligand binding, accurately predicts the targets of bioactive molecules [19]. Target proteins with high binding affinity () to Endogenous compounds affected by acupuncture. 5-hydroxyindole-3-acetic acid, cell cholecystokinin octapeptide (CCK-8), corticosterone, dinoprostone, dopamine, epinephrine, histamine, methionine-enkephalin (MENK), norepinephrine, and serotonin were retrieved from the PharmMapper and Swiss Target Prediction databases. The targets of the active compounds were then identified by merging the search results from the two databases and removing duplicates. Subsequently, using “RA” as a keyword, relevant genes associated with RA were retrieved from two disease gene databases: GeneCards (https://www.genecards.org/) and Comparative Toxicogenomics Database (CTD) (https://ctdbase.org/) [20,21]. The data obtained from both databases were sorted and organized, finally entered as disease genes. 2.3 Construction of the acupuncture-component-gene-disease network and protein-protein interaction network The acupuncture-related genes and RA-related genes were obtained using the online software Venn Diagrams (https://bioinfogp.cnb.csic.es/tools/venny/). The intersection of these two components was used to identify the effective treatment genes for acupuncture intervention in RA. Subsequently, a network was constructed using Cytoscape 3.9.0, with disease, acupuncture, components, and related genes as nodes and their relationships as edges, for topological analysis to determine the core components within the network. STRING (Search Tool for the Retrieval of Interacting Genes/Proteins, https://cn.string-db.org/) is an online bioinformatics database aimed at providing information on gene and protein interactions [22]. The intersection genes were imported into the STRING database with a filtering criterion of “minimum required interaction score ≥0.4,” and the Protein-Protein Interaction (PPI) network graph and Tab-Separated Values (TSV) format file were downloaded and saved. Subsequently, Cytoscape software (version 3.9.0) was used for the visualization and multidimensional network construction of the PPI network for acupuncture-RA. 2.4 Collection and preprocessing of GEO samples The keyword “RA” was utilized to filter RA-related samples in the GEO database (https://www.ncbi.nlm.nih.gov/geo/). The data type was set as gene expression profiles, limited to human samples. Gene expression matrices and clinical grouping information were collected from these samples, followed by gene symbol annotation and data correction using Perl code. This process aimed to determine the intersecting gene expression levels between acupuncture treatment for RA in the normal and RA groups in the field of network pharmacology. 2.5 The expression difference of intersection genes, chromosome location, and the expression correlation of core genes The R software packages “limma”, “heatmap”, and “ggpubr” were utilized to analyze the expression of overlapping genes in individuals with RA and in healthy individuals. Genes with a significance level of were considered as core genes, and their differential expression was visualized using box plots and heat maps. Perl scripting was used to identify the core genes, which were then represented on a circular map. The correlation coefficient of each core gene was visually assessed using the “Rcircos” package in the R programming language. 2.6 Differential gene enrichment analysis and expression of immune cell infiltration in RA patient samples In order to clarify the biological functions and related signaling pathways between the core differential genes of RA samples and normal samples, we annotated the DEGs, aiming to understand the biological processes, molecular functions and cell composition of different levels of biological functions and signaling pathways. Statistical analysis of the data, under the condition of , was carried out using R packages including “clusterProfiler” and “enrichment”, with the results visualized in Bubble diagrams. Significantly, under the criteria of and corrected , Venn diagrams were employed to identify DEGs by intersecting core gene clusters. In this study, we utilized the “CIBERSORT” package in the R software to conduct 1000 simulation experiments on various types of immune cells. These simulation experiments not only yield precise data on the relative composition of immune cells but also establish a benchmark for quantifying immune cell quantities. To investigate differences in immune cells further, we employed the R packages “GSVA” and “GSA Base”. These packages offer a single-sample gene set enrichment analysis (ssGSEA) method, allowing us to compare immune cell content disparities between the healthy control and RA patient groups. Subsequently, to delve deeper into and validate the identified core genes, we conducted a correlation test between the core genes and ssGSEA score and visualized the correlation coefficient. 2.7 Clustering analysis of DEGs in RA patient samples The “ConsensusClusterPlus” package in R software was utilized to cluster RA samples based on core gene expression with k-means clustering, Euclidean distance, and other algorithms, generating up to 9 clusters. The resulting clusters were assessed by comparing their expression levels through heat maps and box plots. Additionally, principal component analysis was conducted to evaluate inter-cluster differences. 2.8 Construction of machine learning model and nomogram model for acupuncture treatment of RA In order to gain a deeper understanding of the pivotal genes involved in acupuncture treatment for RA, we utilized the “caret” R software package to construct machine learning models, encompassing the random forest model (RF), SVM model, generalized linear model (GLM), and extreme gradient boosting (XGB). RF is an ensemble machine learning technique that employs various independent decision trees to predict classification or regression [23]. The SVM algorithm is able to create a hyperplane with maximum margin in the feature space to differentiate between positive and negative instances [24]. GLM serves as an extension of multiple linear regression models, offering flexibility in evaluating the relationship between causal features of normal distribution and categorical or continuous independent features [25]. XGB, on the other hand, consists of boosting trees based on gradient boosting, enabling a careful comparison between classification error and model complexity [26]. The caret package automatically fine-tunes the parameters of these models through a grid search. All machine learning models were executed with default parameters and assessed using 5-fold cross-validation. Subsequently, the “DALEX” package was employed to explicate the aforementioned four machine learning models, visually representing the residual distribution and feature importance in these models. Moreover, the “pROC” R package was utilized to illustrate the area under the receiver operating characteristic (ROC) curve (AUC). Using the most effective machine learning model, we identified the top five significant variables as the key predictive genes associated with RA. Following the selection of the optimal model, we utilized the characteristic genes and their expression levels in both the normal and RA groups to create a nomogram model. Each predictor was assigned a corresponding score, with "Points” denoting the cumulative scores of the aforementioned predictors. 2.9 MR analysis and Colocalization Analysis of key genes and RA 2.9.1 MR analysis To delve deeper into the causal correlation between the key genes and the susceptibility to RA, we opted for a two-sample Mendelian randomization (MR) analysis, which is particularly adept at investigating causal effects [27]. Initially, we isolated the SNPs of the characteristic genes for the exposure factors as well as the SNPs linked to RA as the outcome variables from the Integrated Epidemiological Unit (IEU) database (https://gwas.mrcieu.ac.uk/). Leveraging the “Two Sample MR” software package, we conducted the MR analysis. Through the application of the inverse variance weighting (IVW) method, we obtained a more precise evaluation of the correlation between the expression levels of the characteristic genes and the RA risk [28]. Furthermore, Cochran’s Q statistic was employed to assess heterogeneity in the IVW outcomes, with a P-value under 0.05 signifying statistically significant heterogeneity. Finally, we employed MR-Egger regression and MR-PRESSO analysis to thoroughly evaluate potential pleiotropy [29] Any P-value below 0.05 in the IVW results indicated a significant level of pleiotropy. 2.9.2 Colocalisation analysis For genes that were signifcant in both cohorts, colocalisation analysis of RA risk was performed using the R package coloc [30]. Analyses were performed using SNPs harmonised by TwoSample MR package with default priori probabilities: p1=1E−4, p2=1E−4, p12=1E−5. P1, p2, and p12 are predefined probability that the SNP in the test area is substantially linked with gene expression, RA risk, or both. The posterior probabilities derived from the colocalization analysis correspond to one of five hypotheses: PPH0, SNPs are not associated with either trait; PPH1, SNPs are associated with gene expression but not with RA risk; PPH2, associated with RA risk but not with gene expression; PPH3, associated with RA risk and gene expression but driven by different SNPs; PPH4, associated with RA risk and gene expression, was driven by common SNPs. The threshold of significance for colocalisation was set at PPH4>0.80, and genes that colocalised with RA could be considered as potential acupuncture target genes. 3. Results 3.1 Collection of active ingredients and related genes after acupuncture treatment and collection of RA-related genes By systematically searching Web of Science, PubMed, and Chinese academic databases, this study identified 10 bioactive substances closely related to acupuncture treatment for RA. The analysis indicates that acupuncture intervention significantly modulates various neuroimmune modulatory substances, primarily including prostaglandins such as Dinoprostone and Corticosterone (CORT), monoamine neurotransmitters and their metabolites such as 5-Hydroxyindoleacetic acid (5-HIAA) and serotonin, as well as neuropeptides such as CCK-8 and MENK. These compounds may improve the characteristic joint inflammation and pain symptoms of RA by modulating the overactive immune response [31]. Thus, acupuncture may play an active role in RA-related inflammation. Dinoprostone, also known as prostaglandin E 2 (PGE2). As a pro-inflammatory mediator, PGE2 can enhance the antigen-presenting function of dendritic cells (DCs) and promote the production of IL-17 by CD4 + αβ T cells in RA patients with RA to aggravate joint inflammation [32]. The afferent nerves in acupoints activated by acupuncture and moxibustion transmit sensory signals to spinal cord, brainstem and hypothalamic neurons, further stimulating various neuroimmune pathways, and ultimately exerting anti-inflammatory effects by acting on immune cells to release key neurotransmitters and hormones, including reducing PGE2 levels [33]. The activation of peripheral nociceptors by PGE2 receptors can induce pain, and electroacupuncture can inhibit the activity of PGE2 receptors in arthritis models to exert its anti-inflammatory and analgesic effects [34]. CORT, as an endogenous glucocorticoid, exerts significant anti-inflammatory effects by downregulating pro-inflammatory mediators such as cyclooxygenase 2 (COX-2) and PGE2. Research indicates that electroacupuncture intervention can bidirectionally regulate the neuroendocrine system of RA rats. On one hand, it increases CORT levels to promote the secretion of anti-inflammatory factors such as IL-10 [35]; on the other hand, it activates the hypothalamic-pituitary-adrenal axis (HPA axis), enhancing the release of ACTH [36]. This regulatory mechanism effectively inhibits neutrophil infiltration, reduces the expression of pro-inflammatory factors such as TNF-α and IL-1β, thereby restoring homeostasis within the body. 5-HIAA is a metabolite of serotonin and plays a crucial role in pain regulation and inflammatory response. Research indicates that elevated 5-HIAA levels can improve the pathological process of RA by activating the aryl hydrocarbon receptor (AhR) pathway, thereby inhibiting the differentiation of germinal center B cells into plasma cells while maintaining the immunoregulatory function of regulatory B cells (Bregs) [37]. In the pathogenesis of RA, the regulatory role of the hypothalamic-pituitary-adrenal axis (HPAA) is particularly crucial. Acupuncture treatment may exert therapeutic effects through dual mechanisms: on one hand, it modulates HPAA function to influence serotonin metabolism; on the other hand, it increases the levels of serotonin and 5-HIAA in the circulatory system while reducing serotonin content within platelets, ultimately achieving pain relief and anti-inflammatory effects [38]. CCK-8 is a brain-gut peptide with dual neuroendocrine regulatory functions, exerting anti-inflammatory and immunomodulatory effects through specific binding to cholecystokinin receptors. Its molecular mechanisms involve regulating lymphocyte proliferation and differentiation; affecting immune cell migration and phagocytic function; modulating the secretion of inflammatory factors [39,40]. In the pathological process of RA, CCK-8 can significantly inhibit the abnormal activation of matrix metalloproteinases (MMPs) in synovial cells, thereby improving joint inflammation [41]. Clinical studies have shown that different acupuncture methods (cheek acupuncture/body acupuncture) can effectively enhance the expression levels of CCK-8 in the central nervous system, which may be an important material basis for acupuncture in alleviating RA pain [42-44]. Enkephalins, as endogenous opioid neuropeptides, exert immunomodulatory and anti-inflammatory effects by binding to receptors on the surface of immune cells. Research indicates that acupuncture promotes enkephalin secretion through regulation of the hypothalamic-pituitary axis, which is a crucial mechanism for its treatment of RA [45,46]. Experimental evidence demonstrates, The frequency of electroacupuncture stimulation exhibits a dose-dependent relationship with the release of enkephalins; in osteoarthritis models, acupuncture significantly increases spinal cord enkephalin levels and blocks pain transmission [47]; electroacupuncture intervention can upregulate the expression of enkephalins locally in the joint, effectively alleviating pain and inflammation in acute gouty arthritis [48]. These findings collectively confirm the central role of the enkephalin system in acupuncture analgesia. Catecholamine substances (epinephrine/norepinephrine) possess dual functions as neurotransmitters and hormones. Research indicates that acupuncture exerts anti-inflammatory effects by promoting the secretion of adrenaline from the adrenal medulla and regulating stress responses, regulating stress responses, and increasing norepinephrine levels [49], electroacupuncture intervention significantly improves the levels of norepinephrine in arthritis models, inhibits the levels of pro-inflammatory cytokines TNF-α, IL-1β, and IL-6 in synovial tissue, and alleviates synovitis [50]. Dopamine inhibits the production of cytokines [51] and the NLRP3 inflammasome [52] through D1 dopamine receptors, thereby suppressing systemic inflammation. Experimental evidence confirms that acupuncture effectively modulates systemic inflammatory responses by activating the vagus nerve-adrenal reflex arc and promoting dopamine release [51]. 5-hydroxytryptophan (5-HTP) is an endogenous amino acid and a precursor to serotonin (5-HT), which can be metabolized in the body to produce serotonin. In addition to its role as a neurotransmitter, 5-HT is involved in immune regulation and inflammatory responses. Histamine (HA) and 5-HT are both vasoactive amines that can induce vasodilation and enhance microvenous permeability, thereby promoting local tissue edema and inflammatory response. In severe cases, this can lead to tissue hypoxia or even necrosis [53]. However, 5-HT can also inhibit the release of pro-inflammatory factors such as TNF-α and IL-6 through its receptor-mediated signaling pathways, thereby alleviating inflammatory responses. Research indicates that acupuncture may activate the descending pain control system by increasing the levels of serotonin in the spinal cord or brain [54]. Following acupuncture intervention, mast cells migrate to local acupoints via small arterioles in the subcutaneous and subcutaneous tissues. The aggregated mast cells release more trypsin, histamine, and serotonin through degranulation, thereby modulating immune responses and inflammatory processes [55].Subsequently, using PharmMapper and Swiss Target Prediction databases, a total of 631 potential target proteins associated with acupuncture active components were identified. Furthermore, based on searches of the Genecards and CTD databases, 1333 and 15870 RA related genes were obtained, respectively. Further utilized the Venny online tool for intersection analysis, ultimately screening out 215 common targets. These overlapping genes may constitute key molecular targets for acupuncture treatment of RA, providing crucial clues to elucidate its mechanism of action. (Figure 3A). 3.2 “Acupuncture-ingredients-gene-disease” network analysis and PPI network analysis This study utilized STRING and Cytoscape software to construct a PPI network comprising 215 shared targets, consisting of 213 nodes and 3898 interaction edges, with an average connectivity of 36.601.Subsequently, the network was clustered into four groups using the MCODE (Molecular Complex Detection) plug-in (), resulting in four distinct cluster networks, as illustrated in Figure 3 B and 3 C . Core cluster genes were identified by selecting those with a degree value ≥ 1.5 times the median in each cluster network, totaling 65 targets such as TNF, IL6, JAK2,MAPK8, GAPDH, and AKT1, ALB. To delve deeper into the relationship between the central components of acupuncture, their associated genes, and RA, Cytoscape software (version 3.9.0) was employed to construct the “acupuncture-component-gene-disease” network (Figure 3D) . This network comprises 76 nodes, including 10 active component nodes, 65 gene nodes, and one disease node, interconnected by 456 connections. Network topology parameters were analyzed using the Network Analyzer plug-in, which indicated an average number of adjacent nodes of 11.844, network heterogeneity of 1.085, network density of 0.156, and network centrality of 0.718. Nodes with higher degrees were identified as core nodes within the network. The most active components based on degree were dinoprostone( ), CORT( ),5-hydroxyindole-3-acetic acid ( ), and CCK-8 ( ) (Table 1). The effective components of acupuncture for RA are postulated to be those that exhibit extensive action points and robust interactions that play pivotal roles in the network. Furthermore, a single component may simultaneously affect multiple genes, reflecting a multi-gene regulatory characteristics, and multiple components may concurrently correlate with a single gene concurrently. These findings highlight the multi-faceted, multi-gene regulatory nature of acupuncture in treating RA. Table 1 ． Ranking table of compounds Name Degree Betweenness Centrality Closeness Centrality Dinoprostone 51 0.128465062 0.660869565 Corticosterone 48 0.093978634 0.628099174 5-Hydroxyindole-3-acetic acid 42 0.054056222 0.571428571 CCK-8 40 0.060173298 0.554744526 MET-enkephalin 37 0.037893873 0.531468531 Epinephrine 36 0.034330837 0.524137931 Norepinephrine 33 0.026687875 0.503311258 Dopamine 22 0.01489401 0.439306358 Serotonin 22 0.013884685 0.439306358 Histamine 13 0.003288627 0.397905759 3.3 Acquisition of GEO dataset samples and correlation analysis of intersection genes Through integrated analysis of network topology, this study identified 65 key target sites for acupuncture intervention in RA (Supplementary Table 1) . Data set GSE89408 related to RA was retrieved from the GEO database, which includes 28 healthy controls and 152 RA samples. After standardization preprocessing and differential expression analysis, 49 DEGs were obtained by intersecting with the aforementioned targets, including core regulatory factors such as TNF, IL6, GAPDH, and STAT1. Clinical sample validation revealed that, apart from the significant downregulation of nine genes (AKT1, EGFR, SRC, MAPK3, ESR1, HRAS, ACE, PGR, and NOS3) in the model group, the remaining DEGs showed a significant upregulation trend in the RA group (Figures 4A-B) . The specific chromosomal locations of acupuncture-related DEGs are shown in Figure 4C . Correlation analysis among the DEGs in RA samples indicated a strong relationship between them, as displayed in Figures 4D and 4 E . It is indicating that these genes are related to biological functions and may be involved in the same cellular processes or pathways in the disease state. This expression pattern may be used as a biomarker for disease progression or prognosis, which would be helpful in the diagnosis and treatment of RA. 3.4 Differential gene enrichment analysis and immune cell infiltration analysis of normal samples and RA samples Through systematic biological analysis of 49 DEGs associated with RA treated by acupuncture, this study reveals the potential mechanisms of acupuncture intervention on RA through multi-level functional annotation. Gene Ontology(GO) enrichment analysis yielded 1729 significant entries ( Supplementary Table 2 ) , with biological processes (BP) accounting for 1582 items, primarily enriched in immune stress response pathways, including lipopolysaccharide stress response, chemical stress adaptability regulation, and maintenance of redox homeostasis; molecular functions (MF) identified 98 key entries, involving protein tyrosine kinase activity, phosphatase interaction networks, and cytokine receptor binding among other molecular interaction features; cellular components (CC) localized to 49 substructures, such as membrane rafts signaling platforms, lateral plasma membrane compartments, and fibrinogen-enriched granules (Figure 4F) . Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis further identified 159 significantly enriched signaling pathways , primarily involving three core regulatory modules: 1) inflammatory cascades (TNF signaling axis, NF-κB signal transduction); 2) immune cell function regulation (JAK-STAT signaling network); and 3) cellular stress adaptability pathways. (Figure 4G) (Supplementary Table Table 3) These findings systematically elucidate that acupuncture may exert therapeutic effects through the coordinated regulation of the inflammation-immune-stress triple network, providing molecular-level evidence to support its multi-target action mode. 3.5 Unsupervised clustering of DEGs in RA samples and enrichment analysis of DEGs after unsupervised clustering We performed unsupervised clustering of the samples based on the core genes. leading to the identification of two primary clusters with the highest accuracy. The RA samples were stratified into the Diffgene1 and Diffgene2 groups, as shown in Figure 5A . Using cluster comparison analysis, we sought to augment and validate our previous findings (Figure 5B) . Subsequently, we examined the expression patterns of core genes in the two distinct clusters, displayed in Figure 5C and 5 D . In total. 49 core genes were examined in clusters Diffgene1 and Diffgene2. Notably, there were no significant differences in the expression levels of genes such as IGF1, KDR, ANXA5, MAPK1, PIK3CA, MAPK14, ACE, NOS3, or LGALS3 between the two cluster subgroups, whereas significant differences were observed in the remaining 40 genes. Principal component analysis (PCA) demonstrated the discriminative ability of the core genes between Diffgene1 and Diffgene2, as indicated in Figure 5E . Through functional enrichment analysis, we thoroughly investigated the biological functions and signaling pathways involved in DEGs within the two cluster subgroups, thereby comprehensively elucidating the mechanisms of acupuncture treatment for RA. Enrichment analysis based on GO identified 564 significantly enriched functional terms (Supplementary Table 4) , of which 511 pertain to BP. These primarily involve immune-related processes such as lymphocyte differentiation, monocyte differentiation, and leukocyte chemotaxis. In terms of MF, 36 terms were enriched, including key molecular functions such as chemokine activity, cytokine activity, and chemokine receptor binding. CC are enriched with 17 terms, such as the external side of the plasma membrane, tertiary granules, and immunological synapse, which are closely related to immune response and subcellular structures (Figure 5F) . KEGG pathway analysis further identified 38 significantly enriched signaling pathways, including the TNF signaling pathway, T cell receptor signaling pathway, and JAK-STAT signaling pathway among key inflammatory regulatory pathways (Figure 5G) (Supplementary Table 5). It is noteworthy that both independent enrichment analyses revealed that the top 20 enriched entries were significantly associated with immune response and inflammatory regulation processes. In GO analysis, at the MF level, there is continuous enrichment of cytokine receptor binding and phospholipase activator activity; at the BP level, key processes such as regulation of inflammatory response, T cell activation, negative regulation of immune system process, response to molecule and leukocyte migration are consistently enriched.The results of KEGG pathway analysis also show high consistency, with significant enrichment in the TNF signaling pathway and the JAK-STAT signaling pathway (Figure 4G, Figure 5G) . These findings systematically elucidate the immunomodulatory mechanisms of acupuncture in treating RA, particularly through the regulation of key inflammatory signaling pathways and the function of immune cells. 3.6 Analysis of immune cell infiltration in normal samples and RA samples Inflammatory cell infiltration is an important pathological feature of RA. The interaction between synovial and infiltrating cells can produce a large number of pro-inflammatory mediators and cytokines, which, in turn, act on the synovium and cartilage, activate nociceptors, secrete cytokines, and cause joint tissue damage. The results of acupuncture against RA DEGs enrichment analysis showed that RA are closely related to the functions and signaling pathways of immune cells. To thoroughly investigate the differences in the immune microenvironment between RA patients and healthy individuals, this study employed the Siber sorting algorithm in R language for systematic analysis of immune cell infiltration, accurately quantifying the distribution proportions of various immune cell subpopulations in the samples (Figure 6A) . Furthermore, using the ssGSEA method, significant differences in immune cell types between the two groups were identified. The results showed that 22 types of immune cells were detected in synovial tissue, with eight subpopulations exhibiting significant differences between the RA group and the control group (Figure 6B) . Specifically, the healthy control group exhibited higher abundances of resting CD4 + T cells, germinal center T cells, regulatory T cells (Tregs), resting and activated natural killer (NK) cells, monocytes, and resting mast cells. In contrast, the RA group showed a significant increase in M1 macrophages, activated CD4 + T cells, memory B cells, plasma cells, resting dendritic cells, and neutrophils. Correlation analysis (Figure 6C) reveals significant associations ( P < 0.05) between core genes and specific immune cell subpopulations: M1 macrophages, activated CD4 + T cells, memory B cells, and mast cells show positive correlations with most core genes, while NK cells exhibit a negative correlation trend. These findings further confirm the crucial regulatory role of immune cells in the pathogenesis of RA. 3.7 Machine learning model analysis and RA nomogram model construction This study identified 49 DEGs related to acupuncture treatment for RA as candidate genes, further identifying core DEGs influenced by acupuncture. The aim was to evaluate the diagnostic potential of DEGs between RA patients and healthy individuals and to explore the impact of acupuncture on these genes.To achieve this goal, we developed four sophisticated machine learning models (RF, SVM, GLM, and XGB) to identify key genes in the RA dataset for accurate classification of patients. The DALEX software package was employed to interpret these models and visualize the residual distribution of each model in the test dataset. We evaluated the discriminatory performance of the four machine learning algorithms using 5-fold cross-validation to calculate the ROC curve. Among the models, XGB demonstrated the largest AUC (GLM, AUC = 0.939; SVM, AUC = 0.978; RF, AUC = 0.986; XGB, AUC = 0.994, ( Figure 7A) . Furthermore, the XGB and RF models exhibited relatively low residuals (see Figure 7B, C) . Subsequently, we ranked the top ten important feature variables of each model based on the root mean square error (RMSE) (Figure 7D) . In summary, the XGB model was the most adept at distinguishing distinct patient clusters.Using feature importance evaluation based on the XGBoost model, this study identified five key predictive factors (STAT1, GAPDH, JAK2, PTGS2, and MDM2) as core regulatory genes.Immune microenvironment correlation analysis reveals that the expression levels of STAT1 and JAK2 genes are significantly positively correlated with pro-inflammatory immune cell subpopulations (M1 macrophages, activated CD4 + T cells, and memory B cells) ( P <0.05), but negatively regulate immunosuppressive cells (resting mast cells, activated NK cells, and Treg cells ). GAPDH exhibits a unique biphasic regulatory pattern, with its expression positively correlated with M1 macrophages, resting mast cells, and resting NK cells, while it shows an inhibitory effect on activated mast cells. It is noteworthy that PTGS2 specifically promotes eosinophil infiltration, while MDM2 synergistically enhances the pro-inflammatory phenotype of M1 macrophages and activated CD4 + T cells, simultaneously inhibiting mast cell quiescence (Figure 3H) . These findings suggest that screening genes may participate in the pathogenesis of RA by regulating the dynamic balance of immune cells. To validate the clinical applicability of the XGBoost model, this study constructed a risk prediction nomogram based on this algorithm, used to quantitatively evaluate disease risk stratification in a cohort of 152 RA patients (Figure 7E) . Calibration analysis indicates good consistency between model-predicted risk and actual risk (Figure 7F) , with a Hosmer-Lemeshow test P > 0.05 indicating no significant calibration bias. Decision curve analysis (DCA) further confirms that the model has significant clinical net benefit across a wide range of thresholds (Figure 7G) , with the standardized net benefit exceeding traditional diagnostic methods by more than 30%, indicating that this tool can serve as an effective auxiliary system for evidence-based medical decision-making, providing reliable quantitative basis for personalized treatment of RA. 3.8 MR analysis and Colocalization Analysis results of key genes and RA This study employs a two-sample MR framework to elucidate the causal associations between DEGs and RA. Integrated resources from the IEU OpenGWAS database, which includes 14,361 RA patients and 43,923 European ancestry controls, we identified single nucleotide polymorphisms (SNPs) that meet stringent instrumental variable criteria ( P < 5e-8) for five key genes, including STAT1 JAK2 MDM2 PTGS2 and GAPDH (Figure 8A ). MR analysis showed that the expression levels of STAT1 (OR = 1.53,95 % CI: 1.31-1.77, P = 4.71e-08) and PTGS2 (OR = 1.16,95 % CI: 1.03-1.32, P = 0.015) had a significant positive causal relationship with the risk of RA (Figure 8A) . This study employed various MR methods, including IVW and MR-Egger regression, to investigate the causal effects of STAT1 and PTGS2 genes on RA. The robustness of these findings was confirmed through funnel plot symmetry analysis and Leave-one-out analysis. Furthermore, MR-Egger regression and MR-PRESSO analysis did not detect directional pleiotropy ( P > 0.05), further supporting the reliability of the research results, the above results are shown in Supplementary Figure 1 . Co-localization analysis reveals a significant causal relationship between the STAT1 locus and RA phenotype (PPH4 > 0.8), suggesting that variations in the STAT1 gene may play a crucial role in the pathogenesis of RA (Figure 8B) . Given the crucial role of STAT1 in the immune system, interventions targeting this gene may have potential therapeutic effects for RA. Furthermore, the fact that the other four genetic loci do not share causal variants with the RA phenotype underscores the uniqueness and importance of STAT1 as a therapeutic target for RA (Figure 8C-F ) . This study employs causal inference methods to systematically analyze the impact of acupuncture on gene expression in patients with rheumatoid arthritis, revealing that the expression of the STAT1 gene may be regulated. This finding not only elucidates how acupuncture might exert therapeutic effects by modulating these key genes but also offers new perspectives for developing precision treatments based on gene regulation in the future. Additionally, the study underscores the importance of integrating traditional medicine with modern scientific technology, providing valuable references for research into the treatment of other diseases. 4. Discussion In recent years, research on the molecular mechanisms of acupuncture intervention in RA has gradually become a hotspot in the field of translational medicine [56,57]. Current evidence suggests that acupuncture active components can influence the pathological process of RA by modulating the dynamic balance of immune cells and the network of inflammatory factors [58]. With the advancement of high-throughput sequencing technology, integrating multi-omics data with AI algorithms offers new technological pathways for elucidating the key regulatory networks of complex diseases. However, there are still limitations in the study of acupuncture mechanisms [59,60]: first, traditional bioinformatics methods struggle to systematically reveal the patterns of multi-target synergistic effects; second, there is a lack of genetic evidence based on causal inference to support target selection. This study systematically elucidates the molecular mechanisms of acupuncture treatment for RA by constructing a multi-layer interaction network involving 'acupuncture-active components-targets-disease'. Initially, 261 potential target genes were screened based on network topology, with 65 target genes identified through modular analysis. Differential expression validation using the GEO dataset (GSE89408) yielded 49 DEGs, which were significantly enriched in inflammation-related pathways (such as NF-κB, JAK/STAT) and immune cell regulatory networks (M1 macrophages, activated CD4 + T cells, Treg cells, B cells, NK cell). A clinical prediction model based on the XGBoost algorithm (AUC=0.994) and risk nomogram was further optimized using four machine learning algorithms to identify five core disease-related genes (STAT1, PTGS2, MDM2, GAPDH and JAK2), providing a quantitative assessment tool for personalized treatment of RA. A significant genetic association between STAT1 (OR = 1.53, P = 4.71 × 10 8 ) and PTGS2 (OR = 1.16, P = 0.015) and RA risk was confirmed by MR. Co-localization analysis indicates that the STAT1 locus (PPH4 > 0.8) may serve as a key regulatory target for acupuncture intervention. This study found that acupuncture active components may inhibit synovial inflammation by targeting the STAT1/JAK2 signaling axis, regulating T-cell differentiation, macrophage polarization, and NK cell function. This study is the first to integrate systems biology, machine learning, and causal inference methods to elucidate the network of acupuncture treatment for RA at multiple levels, including molecular, cellular, and clinical. It not only provides an innovative paradigm for the modern research of traditional therapies but also lays a theoretical foundation for the development of novel targeted treatment strategies. 4.1 Acupuncture (compounds) affects immune cells to exert anti-inflammatory effects Based on the multi-dimensional network analysis of 'acupuncture-ingredients-targets-diseases', this study systematically elucidated the immunoregulatory mechanisms of acupuncture in treating RA. Network topology analysis identified key active ingredients such as CCK-8, PGE2, and CORT, which exert their effects through synergistic action on multiple targets to regulate the immune-inflammatory-pain axis. Research has confirmed that acupuncture intervention can significantly regulate key neurotransmitter levels (such as upregulating CCK-8 and enkephalins, downregulating PGE2. It also involves multiple functions and signaling pathways related to the nervous system, immune response, and inflammation, which form the basis of acupuncture treatment for RA. As an immunomodulatory peptide highly expressed in the central nervous system, CCK-8 exerts bidirectional regulatory effects by binding to cholecystokinin receptors (CCKR). In inflammatory responses, CCK-8 can influence B cells [61] and T cells [62]. It inhibits the phenotypic and functional maturation of DCs and B cell IgG1 secretion, reduces the release of pro-inflammatory cytokines such as TNF-α and IL-1β from macrophages, while promoting the production of anti-inflammatory cytokines like IL-4, thereby alleviating inflammatory responses [63,64]. CCK-8 promotes Th1 cell polarization both in vitro and in vivo, and reshapes the Th1/Th17 cell balance by modulating the cytokine profile of dendritic cells (upregulating IL-12 and downregulating IL-6 and IL-23), significantly improving joint inflammation in RA model animals [65]. Studies show that acupuncture can increase the serum CCK-8 levels in RA model animals, with the concentration changes showing a significant positive correlation with the elevation of pain threshold [44]. Endogenous opioid peptides (such as enkephalins) mediate complex immune regulation through μ/δ receptors. MENK inhibits the differentiation of Foxp3 + Tregs induced by TGF-β, thereby blocking the formation of an immunosuppressive [66] microenvironment. Additionally, enkephalins can promote the maturation and function of DCs by increasing the expression of MHC II, CD86, and CD40 on the surface of mouse DCs, which in turn enhances the proliferation and polarization of CD4 + T cells and exacerbates inflammatory responses [67]. Acupuncture stimulation dynamically regulates the release of opioid peptides, with its analgesic effects closely related to the inhibition of pain transmission mediated by substances such as enkephalin and β-endorphin, as well as the downregulation of inflammatory cytokines like IL-6 and TNF-α [68]. PGE2 regulates Th cell differentiation through EP receptors, characterized by enhanced secretion of Th2-type cytokines (IL-4/IL-10/IL-13) and inhibition of IL-12 signaling [69]. In RA, PGE2 can enhance antigen presentation by DCs and promote IL-17 production by CD4 + αβ T cells [32,70]. Studies have shown that acupuncture intervention significantly reduces PGE2 levels in RA synovium, promoting polarization of M1 macrophages to M2 macrophages, thereby inhibiting the expression of key pro-inflammatory factors (IL-6, MCP-1, IL-1β, G-CSF, TNF-α), alleviating inflammation and pain [71]. Additionally, electroacupuncture stimulation can reduce PGE2 levels in RA rats by affecting the hypothalamic-pituitary axis, exerting anti-inflammatory effects [72]. The aforementioned findings elucidate that acupuncture dynamically regulates the neuroendocrineimmune system through a multidimensional network of “components-targets-pathways,” such as by remodeling the functional state of immune cells via mediators like CCK-8/PGE2, ultimately achieving multiple therapeutic effects including anti-inflammatory, immunomodulatory, and analgesic outcomes. This provides a theoretical framework for developing targeted treatment strategies for RA based on the mechanisms of acupuncture. 4.2The potential mechanism of acupuncture in the treatment of RA 4.2.1 Immune cells in RA A notable feature of the RA synovial microenvironment is the abnormal aggregation of various immune cells, including T lymphocytes, B lymphocytes, macrophages, natural killer cells, and mast cells [73,74]. Among these, synovial macrophages exhibit an M1 polarization phenotype during RA progression. They possess the potential to differentiate into osteoclasts and mediate the chemotactic migration of monocytes and neutrophils through the secretion of inflammatory cytokines. Additionally, they activate T cells and drive the aberrant proliferation of synoviocytes, thereby establishing a sustained pro-inflammatory effect [75]. Among the immune cells infiltrating the synovium, subsets of CD4+ T cells play a central role [76]. Activated CD4 + , CD8 + T cells, and Th17 cells significantly promote osteoclast differentiation by releasing key effector molecules such as nuclear factor kappa B ligand (RANKL), TNF-α, IL-1, IL-6, and IL-17. In contrast, IFN-γ and IL-4 secreted by Th1 and Th2 cells exert negative regulatory effects on osteoclastogenesis [77]. Notably, activated CD4 + T cells can further amplify inflammatory damage in joint tissues through cascaded activation of macrophages and B cells, whereas resting CD4 + T cells do not participate in this pathological process [78].Follicular helper T cells (Tfh) within the CD4+ T cell subsets play a unique role in RA, with their abnormal expression of surface markers CXCR5, ICOS, and PD1 being closely related to disease occurrence. These cells participate in autoimmune responses by regulating B cell antibody production. Concurrently, functional defects in Tregs may disrupt immune homeostasis, leading to excessive activation of autoreactive T cells [79]. The mechanism of B cell involvement in RA encompasses multi-pathway regulation: B cells promote osteoclast maturation by producing RANKL [80], activate memory B cells to enhance immune response, and induce synovial tissue to produce pro-inflammatory cytokines such as IL-1α, IL-23, IL-12, IL-6, and TNF-α, thereby exacerbating bone destruction [81]. NK cells participate in disease progression through cytotoxic activity and cytokine networks [82]. Research indicates that acupuncture may improve RA immune imbalance by modulating NK cell activity [83]. Mast cells, as resident components of the synovium in innate immunity, regulate T/B cell and APC function by secreting mediators such as TNF-α, IL-1β, IL-4, and IL-5 [84]. Their abnormal activation is significantly associated with worsened joint inflammation [85]. This study employed the Siebel algorithm in R language to analyze the immune cell infiltration characteristics of RA patients, revealing significant upregulation of M1 macrophages, activated CD4 + T cells, and memory B cells. This finding corroborates the theory of immune-mediated synovial damage, suggesting that abnormal immune responses may directly drive inflammation through metabolic remodeling and polarization changes in macrophages, while persistent activation of T cells leads to a vicious cycle of autoimmunity. In summary, different subpopulations of immune cells collectively constitute the core mechanism of RA pathogenesis through complex interaction networks. 4.2.2 The interaction between core genes (JAK2, STAT1, GAPDH, PTGS2, MDM2) and immune cells affected the pathogenesis of RA The JAK/STAT signaling pathway is widely expressed in various cells and is stimulated by multiple inflammatory stimuli, affecting the differentiation of macrophages and inflammatory responses and participating in many important biological processes, such as cell proliferation, differentiation, apoptosis, and immune regulation [86]. JAK2 kinase, as a core regulatory molecule in cytokine signaling, drives disease progression through multiple mechanisms. Clinical studies have demonstrated that T lymphocytes, macrophages, and fibroblast-like synoviocytes (FLS) in the peripheral blood and synovial microenvironment of RA patients exhibit abnormal overexpression of JAK2 [87], suggesting that this molecule has the potential to serve as a biomarker for disease monitoring. STAT1 serves as the core effector molecule of this pathway, exhibiting specific activation characteristics. Its functional regulation depends on JAK-mediated dual phosphorylation modifications at the Y701/S727 sites [88], which are particularly prominent in FLS [89]. Clinical research data indicate that the expression levels of STAT1 in RA synovial tissues are significantly higher than those in osteoarthritis (OA) control groups, and show characteristic downregulation following effective treatment [90]. De Hooge et al. [91]showed that granuloma formation and progressive arthritis were observed in the arthritis of STAT1 deficient mice induced by zymosan, indicating that the anti-inflammatory effect of STAT1 deficiency may be weakened. At the molecular level of RA, pro-inflammatory cytokines such as IL-6 and IL-12 activate the JAK1/JAK2 kinase complex by binding to T cell surface receptors like IL-6R and IL-12R. JAK2 initiates the phosphorylation cascade of downstream STAT3/STAT1 through its interaction with the intracellular domains of transmembrane receptors such as IL-6R and IL-23R. This process promotes the specific secretion of IL-17A by Th17 cells, disrupting Treg-mediated immune homeostasis and exacerbating inflammatory infiltration and cartilage destruction in the synovium [92,93]. Furthermore, STAT1 can promote the abnormal survival of autoreactive memory T cells by regulating key transcription factors such as BCL-6, thereby becoming an important mechanism for disease chronicity [94]. The IFN-γ-mediated JAK2-STAT1 signaling pathway can induce macrophages to convert to an M1 pro-inflammatory phenotype, significantly increasing the release levels of pro-inflammatory mediators such as TNF-α and IL-12, thereby creating a sustained inflammatory microenvironment [95]. In RA, cytokines such as IFN-γ and IL-6 mediate the activation of JAK1/JAK2 kinase cascades by binding to surface receptors on macrophages, such as IFN-γR and IL-6R. This activation further leads to the phosphorylation and activation of STAT1, which upregulates HIF-1α, enhancing glycolytic metabolism and thereby sustaining the energy requirements of the pro-inflammatory M1 macrophage phenotype [96], resulting in the self-maintenance and amplification of inflammatory responses [97]. Inhibition of the JAK/STAT signaling pathway can regulate macrophage polarization, promoting the transition of pro-inflammatory M1 macrophages to anti-inflammatory M2 macrophages [98]. The inhibition of JAK2 expression also reduces the production of proinflammatory cytokines and local inflammatory responses [99]Blocking STAT1 signaling effectively inhibits pro-inflammatory gene expression and reduces the degree of M1 polarization, confirming its central regulatory role in the inflammatory cascade [100]. Inhibitors of the JAK-STAT signaling pathway, such as tofacitinib, have demonstrated anti-inflammatory effects in rheumatoid arthritis [101]. In RA (B-cell-dependent pathogenic mechanisms), the sustained activation of the JAK-STAT pathway plays a crucial regulatory role. Experimental evidence indicates that pro-inflammatory mediators such as IFN-γ and IL-6, by binding to their specific receptors (IFN-γR and IL-6R), activate members of the JAK kinase family (primarily JAK1/JAK2), leading to the phosphorylation of STAT1. Phosphorylated STAT1 forms homodimers and translocates to the nucleus, where it initiates the transcription program of inflammatory cytokines such as CXCL10 and IRF1, as well as anti-apoptotic molecules like BCL-2, resulting in clonal expansion of B cells and enhanced autoimmune reactivity [102]. It is noteworthy that IFN-γ secreted by such activated B cells can act on adjacent macrophages via paracrine signaling, activating their JAK-STAT signaling cascade. This not only promotes the pro-inflammatory M1 phenotype polarization but also induces excessive release of matrix metalloproteinases (MMPs), thereby exacerbating synovial tissue degradation [103]. Mechanistic studies have revealed that STAT1 signaling significantly enhances T-cell antigen presentation efficiency by upregulating the expression levels of co-stimulatory molecules CD80/CD86 on B-cell surfaces. This aberrant intercellular interaction drives the persistent deterioration of local joint inflammatory responses [104]. PTGS2 is a key enzyme in the PGE2 biosynthesis pathway, and its activation leads to an increase in PGE2 production, which is involved in the regulation of inflammation and immune responses. In the acute inflammatory process, the expression of PTGS2 in macrophages increases, promoting the production of prostaglandins, thereby promoting the release of inflammatory factors (TNF-α, IL-1β) by macrophages, and the inhibition of PTGS2 can reduce the production of these inflammatory factors [105]. The expression of PTGS2 in the synovial tissue of patients with RA is significantly increased, accompanied by the synthesis of prostaglandin E2 and inflammatory factors. This leads to the infiltration of inflammatory cells, abnormal proliferation of synovial tissue, and formation of new blood vessels, thereby exacerbating inflammation and tissue damage. Inhibition of PTGS2 expression in fibroblast-like synovial cells reduces the synthesis of prostaglandin E2 and the inflammatory response in RA [106]. The reduction in PTGS2 expression promotes the activation of Tregs, maintains immune tolerance and an anti-inflammatory environment [107], and significantly inhibits mast cell degranulation, reducing vascular permeability and the expression of inflammatory cytokines [108]. Acupuncture may reduce the production of the proinflammatory prostaglandin PGE2 by inhibiting the activity of PTGS2 and affecting the HPAA, thereby reducing inflammatory responses [33]. This provides molecular evidence for its anti-inflammatory mechanism. The pathological effects of GAPDH are closely related to its metabolic regulatory functions. In an oxidative stress microenvironment, this enzyme enhances aerobic glycolysis in macrophages, driving M1 polarization [109]. In an immune-activated state, macrophages convert the intracellular metabolic enzyme GAPDH into an extracellular signaling molecule. Extracellular GAPDH acts as a ligand that can bind to the CD147 receptor, promoting Th17 differentiation and enhancing glycolysis in CD4 + T cells [110].Targeted inhibition of GAPDH not only regulates T-cell immune function but also alleviates tissue damage through metabolic reprogramming [111,112]. Acupuncture intervention may improve oxidative stress status by balancing GAPDH activity, thereby restoring immune homeostasis [113]. MDM2 is an E3 ubiquitin ligase involved in various cellular processes (cell cycle, apoptosis regulation) and glycolysis. Studies have shown that inhibition of MDM2 activity may affect the activity of inflammatory cells (macrophages and T cells) by inhibiting their survival and death, thereby reducing inflammation [114].MDM2 can exacerbate inflammatory responses by integrating the iNOS-NO and HIF-1α signaling networks [115], and its overactivation promotes glycolysis in M1 macrophages. Specific inhibition of MDM2 can block pro-inflammatory cell survival signals. MDM2 extends Th17 cell survival by inhibiting p53-dependent apoptosis, while impairing Treg immunosuppressive function through the ubiquitination and degradation of Foxp3. The MDM2-mTORC1 axis promotes glycolytic metabolism, providing energy support for sustained T-cell activation. MDM2 induces the expression of M1 polarization markers (iNOS, IL-12) by activating the NF-κB pathway [116]. MDM2 inhibits B cell apoptosis by degrading p53, thereby promoting the expansion of autoreactive B cell clones [117]. The activity of MDM2 may influence the activity of inflammatory cells by inhibiting the survival and death of various immune cells, such as macrophages and T cells, thereby exacerbating inflammation. Therefore, MDM2-mediated metabolic reprogramming leading to Th17/Treg imbalance could provide new intervention targets for the treatment of rheumatoid arthritis. These key genes play an anti-inflammatory role in RA by participating in cell signal transduction, regulating the expression of cytokines, affecting cell survival and apoptosis. In addition, the MR analysis of these five genes revealed a potential causal relationship between Increased STAT1 levels and increased risk of RA. Acupuncture can provide valuable insights into the pathogenesis of RA by targeting these genes. 4.2.3 Acupuncture may exert anti-inflammatory effects by modulating the JAK2/STAT1 signaling pathway and influencing the function of immune cells Acupuncture exhibits multi-target regulatory characteristics in the treatment of RA, having been proven to reduce the polarization of M1 macrophages in RA synovial tissue, promote the polarization of T cells towards anti-inflammatory cells [118,119], and inhibit pro-inflammatory cytokines (TNF-α, IL-6, and IL-17) levels to achieve anti-inflammatory and analgesic effects [120]. Acupuncture may improve joint inflammation by correcting the imbalance of Th1/Treg cell activity, characterized by decreased levels of IFN-γ, IL-17, and Increase levels of IL-10, IL-4, and TGF-β [10]. Acupuncture intervention can significantly inhibit the polarization process of M1 macrophages in synovial tissue, while promoting the differentiation of T cells into anti-inflammatory subpopulations, and achieving anti-inflammatory and analgesic effects by reducing the expression levels of pro-inflammatory cytokines such as TNF-α, IL-6, and IL-17 [121]. Acupuncture modulates the local cytokine environment at acupoints, including IL-1β and IL-6, activating the anti-inflammatory programs of macrophages and CD4 + T cells [10]. Given the central role of aberrant activation of the JAK/STAT signaling pathway in the pathogenesis of rheumatoid arthritis, this study focuses on the regulatory effects of acupuncture on this pathway and its immunomodulatory mechanisms. Previous studies have demonstrated that pro-inflammatory cytokines such as IL-6 and IFN-γ can activate JAK2 kinase by binding to their transmembrane receptors (IL-6R/IFN-γR), leading to the phosphorylation modification of STAT1 protein. Phosphorylated STAT1 is transported into the nucleus via nuclear localization signals, where it binds to the γ-activated sequence (GAS) in the promoter regions of target genes, initiating the expression of inflammation-related genes. This process aids the body in combating pathogens or repairing damage [122]. In the JAK2-STAT1 signaling pathway, METTL3-mediated m6A methylation modification enhances the accumulation and activity of STAT1 protein by increasing the stability of STAT1 mRNA, thereby elucidating the molecular basis for sustained activation of this signaling pathway at the epigenetic level [123]. In RA, aberrant activation of the JAK/STAT signaling pathway may lead to excessive production of pro-inflammatory cytokines, activating JAK1 and JAK2, which subsequently promotes the polarization of macrophages towards the M1 phenotype [124]. Acupuncture can reduce inflammatory responses in the synovial tissue of RA rats by inhibiting the release of inflammatory factor IL-1 in the serum of RA rats, increasing the levels of immune regulatory factor IL-2, and enhancing the expression of STAT1 and negative regulators of cytokine signaling in the synovial tissue of RA rats [125]. This indicates that acupuncture's inhibition of STAT1 phosphorylation involves the participation of the JAK2/STAT1 signaling pathway in immune cells, enriching our understanding of the JAK2-STAT1 signaling pathway and providing potential targets for developing new treatment strategies for related diseases. Based on the aforementioned evidence, acupuncture may exert its anti-inflammatory therapeutic effects by modulating the JAK2-STAT1 pathway, thereby influencing the ability of CD4 + T cells and macrophages to secrete pro-inflammatory cytokines such as IL-6 and TNF-α. 5. Conclusion This study integrates TCM's holistic philosophy with AI-driven precision medicine to systematically identify key active components, action targets, and biomarkers of acupuncture intervention in RA. The research found that endogenous active substances (CCK-8, PGE-2, CORT, enkephalin) and DEGs (STAT1, GAPDH, JAK2, PTGS2, MDM2) constitute the core molecular network of acupuncture treatment for RA. Enkephalin and PGE-2 may exert anti-inflammatory effects similar to glucocorticoids, playing a crucial regulatory role in acupuncture therapy, which provides theoretical support for the development of new anti-inflammatory drugs. Notably, differential gene screening results suggest that STAT1, GAPDH, JAK2, PTGS2, and MDM2 may be potential targets for acupuncture intervention in RA. Further analysis indicates that STAT1, as a key regulatory factor, can control the immune homeostasis mediated by M1 macrophages and CD4 + T cells. The study proposes that acupuncture may regulate the functions of immune cells such as M1 macrophages and CD4 + T cells by inhibiting the JAK2/STAT1 signaling axis, reshaping the RA immune microenvironment and exerting anti-inflammatory effects. The innovation of this study lies in constructing a multidimensional target network for acupuncture treatment of RA. It involves screening out 10 characteristic active components using literature mining methods and further identifying five core regulatory genes, including STAT1, through machine learning and Mendelian randomization. Subsequent studies require validation and screening of these substances and molecules through basic experiments. These findings not only deepen the scientific understanding of the anti-inflammatory mechanisms of acupuncture but also provide theoretical basis for developing acupuncture-assisted treatment plans in the era of precision medicine. Future research will focus on: 1) establishing clinical diagnostic indicators based on key targets; 2) developing biologic therapeutic strategies targeting the JAK2/STAT1 pathway; 3) exploring the potential of combining acupuncture with other applications. Abbreviations English abbreviations English full name RA Rheumatoid Arthritis GEO Gene Expression Omnibus English abbreviations English full name PPI Protein-Protein Interaction MCODE Molecular Complex Detection KEGG Kyoto Encyclopedia of Genes and Genomes GO Gene Ontology DEGs Differentially Expressed Genes ROC Receiver Operating Characteristic AUC Area Under the ROC Curve RF Random Forest SVM Support Vector Machine GLM Generalized Linear Model XGB Extreme Gradient Boosting MR Mendelian Randomization IVW Inverse Variance Weighting English abbreviations English full name GSEA Gene Set Enrichment Analysis HPAA Hypothalamic-Pituitary-Adrenal Axis JAK Janus Kinase STAT Signal Transducer and Activator of Transcription PTGS2 Prostaglandin-Endoperoxide Synthase 2 MDM2 Mouse Double Minute 2 CCK-8 Cholecystokinin Octapeptide MENK Methionine-enkephalin CORT Corticosterone PGE2 Prostaglandin E 2 AhR Aryl hydrocarbon receptor MMPs Matrix metalloproteinases 5-HTP 5-hydroxytryptophan English abbreviations English full name HPAA Hypothalamic-pituitary-adrenal axis HA Histamine MHC Major Histocompatibility Complex Tregs Regulatory T cells NK Natural Killer DCs Dendritic Cells IL Interleukin TNF Tumor Necrosis Factor IFN-γ Interferon Gamma Th1/Th2 T-helper 1/T-helper 2 Th17 T-helper 17 RANKL Receptor Activator of Nuclear Factor-κB Ligand NLRP3 NOD-like Receptor Family Pyrin Domain-containing Protein 3 Declarations Availability of data and materials Data will be made available on request. Ethics declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments We thank colleagues and institutions that supported the authors of this study for their contributions. Funding This study was supported in part by the National Natural Science Foundation of China (82274649,82004473 and 82205279). Tianjin Natural Science Foundation of China (22JCZXJC00070) Rights and permissions Open Access: All the samples used in this study were obtained from public databases, and the sample data were obtained from database (https://www.ncbi.nlm.nih.gov/geo/). Disease Gene Database were obtained from GeneCards (https://www.genecards.org/) and Comparative Toxicogenomics Database (CTD) (https://ctdbase.org/) Contributions FYL and ZL are joint first authors. YongMing G and YiNan G obtained funding. YX, YG and ZFX designed the study. GongMing Y, JYZ, PYL, RuiW, JH and XL collected the data. FYL and ZL analyzed the data. FYL drafted the manuscript. YiNan G and YX contributed to the interpretation of the results and critical revision of the manuscript for important intellectual content and approved the final version of the manuscript. All authors have read and approved the final manuscript. YongMing G and YG are the study guarantors. References Global, regional, and national burden of other musculoskeletal disorders, 1990-2020, and projections to 2050: a systematic analysis of the Global Burden of Disease Study 2021. The Lancet. Rheumatology 5 , e670-e682 (2023). Farhat, H. , et al. Increased Risk of Cardiovascular Diseases in Rheumatoid Arthritis: A Systematic Review. Cureus 14 , e32308 (2022). Sokka, T. , et al. Work disability remains a major problem in rheumatoid arthritis in the 2000s: data from 32 countries in the QUEST-RA study. Arthritis research & therapy 12 , R42 (2010). Chung, I.M., Ketharnathan, S., Thiruvengadam, M. & Rajakumar, G. Rheumatoid Arthritis: The Stride from Research to Clinical Practice. International journal of molecular sciences 17 (2016). Mazaud, C. & Fardet, L. Relative risk of and determinants for adverse events of methotrexate prescribed at a low dose: a systematic review and meta-analysis of randomized placebo-controlled trials. The British journal of dermatology 177 , 978-986 (2017). Kirwan, J.R. Glucocorticoid resistance in patients with rheumatoid arthritis. Scandinavian journal of rheumatology 36 , 165-166 (2007). Ling, S. & Jamali, F. The effect of infliximab on hepatic cytochrome P450 and pharmacokinetics of verapamil in rats with pre-adjuvant arthritis: a drug-disease and drug-drug interaction. Basic & clinical pharmacology & toxicology 105 , 24-29 (2009). González, C.M. , et al. Perceptions of patients with rheumatic diseases on the impact on daily life and satisfaction with their medications: RHEU-LIFE, a survey to patients treated with subcutaneous biological products. Patient preference and adherence 11 , 1243-1252 (2017). Sunil, D. & Kamath, P.R. Multi-Target Directed Indole Based Hybrid Molecules in Cancer Therapy : An Up-To-Date Evidence-Based Review. Current topics in medicinal chemistry 17 , 959-985 (2017). Li, N. , et al. The Anti-Inflammatory Actions and Mechanisms of Acupuncture from Acupoint to Target Organs via Neuro-Immune Regulation. Journal of inflammation research 14 , 7191-7224 (2021). Wang, Y. , et al. Effect of Moxibustion on β-EP and Dyn Levels of Pain-Related Indicators in Patients with Rheumatoid Arthritis. Evidence-based complementary and alternative medicine : eCAM 2021 , 6637554 (2021). Yang, F. , et al. ST36 Acupuncture Alleviates the Inflammation of Adjuvant-Induced Arthritic Rats by Targeting Monocyte/Macrophage Modulation. Evidence-based complementary and alternative medicine : eCAM 2021 , 9430501 (2021). Wooller, S.K., Benstead-Hume, G., Chen, X., Ali, Y. & Pearl, F.M.G. Bioinformatics in translational drug discovery. Bioscience reports 37 (2017). Ye, H., Wei, J., Tang, K., Feuers, R. & Hong, H. Drug Repositioning Through Network Pharmacology. Current topics in medicinal chemistry 16 , 3646-3656 (2016). Boezio, B., Audouze, K., Ducrot, P. & Taboureau, O.J.M.i. Network‐based approaches in pharmacology. 36 , 1700048 (2017). Fan, A.Y. Anti-inflammatory mechanism of electroacupuncture involves the modulation of multiple systems, levels and targets and is not limited to "driving the vagus-adrenal axis". Journal of integrative medicine 21 , 320-323 (2023). Han, Z., Zhang, Y., Wang, P., Tang, Q. & Zhang, K. Is acupuncture effective in the treatment of COVID-19 related symptoms? Based on bioinformatics/network topology strategy. Briefings in bioinformatics 22 (2021). Wang, X. , et al. PharmMapper 2017 update: a web server for potential drug target identification with a comprehensive target pharmacophore database. Nucleic acids research 45 , W356-w360 (2017). Daina, A., Michielin, O. & Zoete, V. SwissTargetPrediction: updated data and new features for efficient prediction of protein targets of small molecules. Nucleic acids research 47 , W357-w364 (2019). Safran, M. , et al. The genecards suite. 27-56 (2021). Wyatt, B. , et al. Transforming environmental health datasets from the comparative toxicogenomics database into chord diagrams to visualize molecular mechanisms. Frontiers in toxicology 6 , 1437884 (2024). Szklarczyk, D. , et al. The STRING database in 2021: customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets. Nucleic acids research 49 , D605-d612 (2021). Rigatti, S.J. Random Forest. Journal of insurance medicine (New York, N.Y.) 47 , 31-39 (2017). Gold, C. & Sollich, P. Model selection for support vector machine classification. Neurocomputing 55 , 221-249 (2003). Nelder, J.A. & Wedderburn, R.W.M. Generalized Linear Models. Royal Statistical Society. Journal. Series A: General 135 , 370-384 (1972). Chen, T.J.R.p.v.-. Xgboost: extreme gradient boosting. 1 (2015). Higgins, J.P., Thompson, S.G., Deeks, J.J. & Altman, D.G. Measuring inconsistency in meta-analyses. BMJ (Clinical research ed.) 327 , 557-560 (2003). Burgess, S. & Thompson, S.G. Interpreting findings from Mendelian randomization using the MR-Egger method. European journal of epidemiology 32 , 377-389 (2017). Zhang, B. , et al. m(6)A regulator-mediated methylation modification patterns and tumor microenvironment infiltration characterization in gastric cancer. Molecular cancer 19 , 53 (2020). Giambartolomei, C. , et al. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. 10 , e1004383 (2014). McInnes, I.B. & Schett, G. The pathogenesis of rheumatoid arthritis. The New England journal of medicine 365 , 2205-2219 (2011). Du, B., Zhu, M., Li, Y., Li, G. & Xi, X. The prostaglandin E2 increases the production of IL-17 and the expression of costimulatory molecules on γδ T cells in rheumatoid arthritis. Scandinavian journal of immunology 91 , e12872 (2020). Xu, Z.-F. , et al. Neuroendocrine-immune regulating mechanisms for the anti-inflammatory and analgesic actions of acupuncture. 6 , 384-392 (2020). Zhang, R., Lao, L., Ren, K. & Berman, B.M.J.A. Mechanisms of acupuncture–electroacupuncture on persistent pain. 120 , 482-503 (2014). Zhang, R.-X. , et al. Electroacupuncture attenuates inflammation in a rat model. 11 , 135-142 (2005). Wei, Y. , et al. Regulation of hypothalamic-pituitary-adrenal axis activity and immunologic function contributed to the anti-inflammatory effect of acupuncture in the OVA-induced murine asthma model. 636 , 177-183 (2017). Ying, Z.-H., Mao, C.-L., Xie, W. & Yu, C.-H.J.F.i.M. Postbiotics in rheumatoid arthritis: Emerging mechanisms and intervention perspectives. 14 , 1290015 (2023). Lee, E.J. & Warden, S.J.E.J.o.I.M. The effects of acupuncture on serotonin metabolism. 8 , 355-367 (2016). De la Fuente, M., Medina, S., Del Rio, M., Ferrández, M.D. & Hernanz, A. Effect of aging on the modulation of macrophage functions by neuropeptides. Life Sciences 67 , 2125-2135 (2000). Trejter, M. , et al. Studies on the involvement Histology and Histopathology Cellular and Molecular Biology of endogenous neuropeptides in the control of thymocyte proliferation in the rat. (2001). Lee, E.-G. , et al. Adrenomedullin inhibits IL-1β-induced rheumatoid synovial fibroblast proliferation and MMPs, COX-2 and PGE2 production. 34 , 335-343 (2011). Zhou, Y., Sun, Y.-H., Shen, J.-M. & Han, J.-S.J.N. Increased release of immunoreactive CCK-8 by electroacupuncture and enhancement of electroacupuncture analgesia by CCK-B antagonist in rat spinal cord. 24 , 139-144 (1993). Han, J., Ding, X. & Fan, S.J.P. Cholecystokinin octapeptide (CCK-8): antagonism to electroacupuncture analgesia and a possible role in electroacupuncture tolerance. 27 , 101-115 (1986). Jie, W. , et al. Analgesic effect of buccal acupuncture on acute arthritis in rabbits and underlying mechanisms. 42 , 517-521 (2017). Zhao, Z.-Q.J.P.i.n. Neural mechanism underlying acupuncture analgesia. 85 , 355-375 (2008). Cheng, L.-L. , et al. Effects of electroacupuncture of different frequencies on the release profile of endogenous opioid peptides in the central nerve system of goats. 2012 , 476457 (2012). Zheng, X., Lin, J., Wang, Z., Zeng, Z. & Chen, H.J.H. Research of the analgesic effects and central nervous system impact of electroacupuncture therapy in rats with knee osteoarthritis. 10 (2024). Fang, J. , et al. Involvement of peripheral beta-endorphin and MU, delta, kappa opioid receptors in electro acupuncture analgesia for prolonged inflammatory pain of rats. 11 , 375-383 (2013). Liu, J., Dong, S., Liu, S.J.A. & Medicine, H. Aberrant parasympathetic responses in acupuncture therapy for restoring immune homeostasis. 3 , 69-75 (2023). Chen, W. , et al. Electroacupuncture activated local sympathetic noradrenergic signaling to relieve synovitis and referred pain behaviors in knee osteoarthritis rats, Front. Mol. Neurosci. 16 (2023) 1069965. (Epub 2023/03/25. doi: 10.3389/fnmol, 2023). Torres-Rosas, R. , et al. Dopamine mediates vagal modulation of the immune system by electroacupuncture. 20 , 291-295 (2014). Yan, Y. , et al. Dopamine controls systemic inflammation through inhibition of NLRP3 inflammasome. 160 , 62-73 (2015). Shen, Y., Liu, F., Zhang, M.J.B. & Pharmacotherapy. Therapeutic potential of plant-derived natural compounds in Alzheimer’s disease: Targeting microglia-mediated neuroinflammation. 178 , 117235 (2024). Ma, X. , et al. Potential mechanisms of acupuncture for neuropathic pain based on somatosensory system. 16 , 940343 (2022). Wang, M., Liu, W., Ge, J. & Liu, S.J.F.i.i. The immunomodulatory mechanisms for acupuncture practice. 14 , 1147718 (2023). Zhang, Y. , et al. Pathological pathway analysis in an experimental rheumatoid arthritis model and the tissue repair effect of acupuncture at ST36. 14 , 1164157 (2023). Wang, J. , et al. Therapeutic effect and mechanism of acupuncture in autoimmune diseases. 50 , 639-652 (2022). Jang, S., Kwon, E.-J. & Lee, J.J.J.I.j.o.m.s. Rheumatoid arthritis: pathogenic roles of diverse immune cells. 23 , 905 (2022). Huang, Y. , et al. Identification of diagnostic genes and drug prediction in metabolic syndrome-associated rheumatoid arthritis by integrated bioinformatics analysis, machine learning, and molecular docking. 15 , 1431452 (2024). Zhou, J. , et al. Identification of aging-related biomarkers and immune infiltration characteristics in osteoarthritis based on bioinformatics analysis and machine learning. 14 , 1168780 (2023). Zhang, J.-G. , et al. Cholecystokinin octapeptide regulates lipopolysaccharide-activated B cells co-stimulatory molecule expression and cytokines production in vitro. 33 , 157-163 (2011). Zhang, J.-G. , et al. Cholecystokinin octapeptide regulates the differentiation and effector cytokine production of CD4+ T cells in vitro. 20 , 307-315 (2014). Crawley, J.N. & Corwin, R.L. Biological actions of cholecystokinin. Peptides 15 , 731-755 (1994). Zhang, J.G. , et al. Cholecystokinin octapeptide inhibits immunoglobulin G1 production of lipopolysaccharide-activated B cells. International immunopharmacology 11 , 1685-1690 (2011). Li, Q. , et al. Cholecystokinin octapeptide significantly suppresses collagen-induced arthritis in mice by inhibiting Th17 polarization primed by dendritic cells. Cellular immunology 272 , 53-60 (2011). Li, X. , et al. Methionine enkephalin (MENK) inhibits tumor growth through regulating CD4+ Foxp3+ regulatory T cells (Tregs) in mice. 16 , 450-459 (2015). Shan, F. , et al. Functional modulation of the pathway between dendritic cells (DCs) and CD4+ T cells by the neuropeptide: methionine enkephalin (MENK). 32 , 929-937 (2011). Xing, J., Xia, M., Wang, T. & Mu, J.J.Z.c.y.j.A.R. Study on the analgesic effect of acupuncture with opioid receptors agonist in induced arthritic rats. 14 , 375-378 (1989). Tsuge, K., Inazumi, T., Shimamoto, A. & Sugimoto, Y. Molecular mechanisms underlying prostaglandin E2-exacerbated inflammation and immune diseases. International immunology 31 , 597-606 (2019). Sreeramkumar, V., Fresno, M., Cuesta, N.J.I. & biology, c. Prostaglandin E2 and T cells: friends or foes? 90 , 579-586 (2012). Qiao, L.-n. , et al. Effect of Electroacupuncture Intervention on Expression of CGRP, SP, COX‐1, and PGE2 of Dorsal Portion of the Cervical Spinal Cord in Rats with Neck‐Incision Pain. 2013 , 294091 (2013). JIANG, J. , et al. Efficacy of electroacupuncture stimulating Zusanli (ST36) and Xuanzhong (GB39) on synovial angiogenesis in rats with adjuvant arthritis. 43 , 955 (2023). Ao, Y., Wang, Z., Hu, J., Yao, M. & Zhang, W.J.S.R. Identification of essential genes and immune cell infiltration in rheumatoid arthritis by bioinformatics analysis. 13 , 2032 (2023). Zhou, S., Lu, H. & Xiong, M.J.F.i.i. Identifying immune cell infiltration and effective diagnostic biomarkers in rheumatoid arthritis by bioinformatics analysis. 12 , 726747 (2021). Boutet, M.-A. , et al. Novel insights into macrophage diversity in rheumatoid arthritis synovium. 20 , 102758 (2021). Gao, Y. , et al. Immunosenescence of T cells: a key player in rheumatoid arthritis. Inflammation research : official journal of the European Histamine Research Society ... [et al.] 71 , 1449-1462 (2022). Tang, M., Tian, L., Luo, G. & Yu, X.J.F.i.i. Interferon-gamma-mediated osteoimmunology. 9 , 1508 (2018). Roberts, C.A., Dickinson, A.K. & Taams, L.S.J.F.i.i. The interplay between monocytes/macrophages and CD4+ T cell subsets in rheumatoid arthritis. 6 , 571 (2015). Kondo, Y. , et al. Transcriptional regulation of CD 4+ T cell differentiation in experimentally induced arthritis and rheumatoid arthritis. 70 , 653-661 (2018). Sun, W. , et al. B cells inhibit bone formation in rheumatoid arthritis by suppressing osteoblast differentiation. 9 , 5127 (2018). Wu, F. , et al. B cells in rheumatoid arthritis： pathogenic mechanisms and treatment prospects. 12 , 750753 (2021). Kucuksezer, U.C. , et al. The role of natural killer cells in autoimmune diseases. 12 , 622306 (2021). Liu, F. , et al. Acupuncture and its ability to restore and maintain immune homeostasis. 117 , 167-176 (2024). Lei, Y. , et al. Synovial microenvironment-influenced mast cells promote the progression of rheumatoid arthritis. 15 , 113 (2024). Loucks, A. , et al. The multifaceted role of mast cells in joint inflammation and arthritis. 31 , 567-575 (2023). Xue, C. , et al. Evolving cognition of the JAK-STAT signaling pathway: autoimmune disorders and cancer. 8 , 204 (2023). Xia, X. , et al. Single cell immunoprofile of synovial fluid in rheumatoid arthritis with TNF/JAK inhibitor treatment. Nature communications 16 , 2152 (2025). Qin, Y. , et al. Age-associated B cells contribute to the pathogenesis of rheumatoid arthritis by inducing activation of fibroblast-like synoviocytes via TNF-α-mediated ERK1/2 and JAK-STAT1 pathways. 81 , 1504-1514 (2022). Kasperkovitz, P. , et al. Activation of the STAT1 pathway in rheumatoid arthritis. 63 , 233-239 (2004). Monari, C. , et al. A microbial polysaccharide reduces the severity of rheumatoid arthritis by influencing Th17 differentiation and proinflammatory cytokines production. 183 , 191-200 (2009). de Hooge, A.S. , et al. Local activation of STAT‐1 and STAT‐3 in the inflamed synovium during zymosan‐induced arthritis: exacerbation of joint inflammation in STAT‐1 gene–knockout mice. 50 , 2014-2023 (2004). Zhang, M., Xu, M., Wang, K., Li, L. & Zhao, J. Effect of Inhibition of the JAK2/STAT3 Signaling Pathway on the Th17/IL-17 Axis in Acute Cellular Rejection After Heart Transplantation in Mice. Journal of cardiovascular pharmacology 77 , 614-620 (2021). Lv, Y. , et al. The JAK-STAT pathway: from structural biology to cytokine engineering. Signal transduction and targeted therapy 9 , 221 (2024). Choi, Y.S., Eto, D., Yang, J.A., Lao, C. & Crotty, S. Cutting edge: STAT1 is required for IL-6-mediated Bcl6 induction for early follicular helper cell differentiation. Journal of immunology (Baltimore, Md. : 1950) 190 , 3049-3053 (2013). Chen, R. , et al. Augmented PFKFB3-mediated glycolysis by interferon-γ promotes inflammatory M1 polarization through the JAK2/STAT1 pathway in local vascular inflammation in Takayasu arteritis. Arthritis research & therapy 24 , 266 (2022). Ciobanu, D.A. , et al. JAK/STAT pathway in pathology of rheumatoid arthritis (Review). Experimental and therapeutic medicine 20 , 3498-3503 (2020). Owen, K.L., Brockwell, N.K. & Parker, B.S. JAK-STAT Signaling: A Double-Edged Sword of Immune Regulation and Cancer Progression. Cancers 11 (2019). Yang, X. , et al. Cell volume regulation modulates macrophage-related inflammatory responses via JAK/STAT signaling pathways. Acta biomaterialia 186 , 286-299 (2024). Sarapultsev, A. , et al. JAK-STAT signaling in inflammation and stress-related diseases: implications for therapeutic interventions. 4 , 40 (2023). Liang, Y.B. , et al. Downregulated SOCS1 expression activates the JAK1/STAT1 pathway and promotes polarization of macrophages into M1 type. Molecular medicine reports 16 , 6405-6411 (2017). Suda, Y. , et al. Comparison of anti-inflammatory and anti-angiogenic effects of JAK inhibitors in IL-6 and TNFα-stimulated fibroblast-like synoviocytes derived from patients with RA. Scientific reports 15 , 9736 (2025). Santos, C.I. & Costa-Pereira, A.P. Signal transducers and activators of transcription-from cytokine signalling to cancer biology. Biochimica et biophysica acta 1816 , 38-49 (2011). Yin, X. , et al. Research progress on macrophage polarization during osteoarthritis disease progression: a review. Journal of orthopaedic surgery and research 19 , 584 (2024). Elbrashy, M.M., Metwally, H., Sakakibara, S. & Kishimoto, T. Threonine Phosphorylation and the Yin and Yang of STAT1: Phosphorylation-Dependent Spectrum of STAT1 Functionality in Inflammatory Contexts. Cells 13 (2024). Muñoz, A., Costa, M.J.O.m. & longevity, c. Nutritionally mediated oxidative stress and inflammation. 2013 , 610950 (2013). Tang, M. , et al. Pharmacological aspects of natural quercetin in rheumatoid arthritis. 2043-2053 (2023). Zhao, M., Burisch, J.J.D.d. & sciences. Impact of genes and the environment on the pathogenesis and disease course of inflammatory bowel disease. 64 , 1759-1769 (2019). Chen, Y. , et al. PTGS2: A potential immune regulator and therapeutic target for chronic spontaneous urticaria. 344 , 122582 (2024). Chen, P.C. , et al. Moonlighting glyceraldehyde-3-phosphate dehydrogenase (GAPDH) protein of Lactobacillus gasseri attenuates allergic asthma via immunometabolic change in macrophages. Journal of biomedical science 29 , 75 (2022). Yang, H. , et al. CD147 modulates the differentiation of T-helper 17 cells in patients with rheumatoid arthritis. APMIS : acta pathologica, microbiologica, et immunologica Scandinavica 125 , 24-31 (2017). Cui, Z. , et al. MYO1F regulates T-cell activation and glycolytic metabolism by promoting the acetylation of GAPDH. Cellular & molecular immunology 22 , 176-190 (2025). Kornberg, M.D. , et al. Dimethyl fumarate targets GAPDH and aerobic glycolysis to modulate immunity. Science (New York, N.Y.) 360 , 449-453 (2018). Guo, B.J., Sun, J.H. & Pei, L.X. Research progress on mechanisms of acupuncture and moxibustion underlying improvement of oxidative stress. Zhen ci yan jiu = Acupuncture research 49 , 307-314 (2024). Thomasova, D., Mulay, S.R., Bruns, H. & Anders, H.-J.J.N. p53-independent roles of MDM2 in NF-κB signaling: implications for cancer therapy, wound healing, and autoimmune diseases. 14 , 1097-1101 (2012). Wu, K.K.-l. , et al. MDM2 induces pro-inflammatory and glycolytic responses in M1 macrophages by integrating iNOS-nitric oxide and HIF-1α pathways in mice. 15 , 8624 (2024). Wu, K.K. , et al. MDM2 induces pro-inflammatory and glycolytic responses in M1 macrophages by integrating iNOS-nitric oxide and HIF-1α pathways in mice. Nature communications 15 , 8624 (2024). Li, Z. , et al. Functions and mechanisms of non-histone post-translational modifications in cancer progression. Cell death discovery 11 , 125 (2025). Yu, N. , et al. Manual acupuncture at ST36 attenuates rheumatoid arthritis by inhibiting M1 macrophage polarization and enhancing Treg cell populations in adjuvant-induced arthritic rats. 41 , 96-109 (2023). Zhu, J. , et al. Electroacupuncture attenuates collagen-induced arthritis in rats through vasoactive intestinal peptide signalling-dependent re-establishment of the regulatory T cell/T-helper 17 cell balance. 33 , 305-311 (2015). Yang, F. , et al. ST36 Acupuncture Alleviates the Inflammation of Adjuvant‐Induced Arthritic Rats by Targeting Monocyte/Macrophage Modulation. 2021 , 9430501 (2021). Oh, J.E. & Kim, S.N. Anti-Inflammatory Effects of Acupuncture at ST36 Point: A Literature Review in Animal Studies. Frontiers in immunology 12 , 813748 (2021). Jerke, U. , et al. Stat1 nuclear translocation by nucleolin upon monocyte differentiation. PloS one 4 , e8302 (2009). Fang, W. , et al. m6A methylation modification and immune infiltration analysis in osteonecrosis of the femoral head. Journal of orthopaedic surgery and research 19 , 183 (2024). Ivashkiv, L.B.J.N.R.I. IFNγ: signalling, epigenetics and roles in immunity, metabolism, disease and cancer immunotherapy. 18 , 545-558 (2018). Hao, F. , et al. Effect of moxibustion on autophagy and the inflammatory response of synovial cells in rheumatoid arthritis model rat. 42 , 73-82 (2022). Additional Declarations No competing interests reported. Supplementary Files supplementarytable1.docx supplementarytable2GO.html supplementarytable3KEGG.html supplementarytable4unsupervisedclusteringGO.html supplementarytable5unsupervisedclusteringKEGG.html SupplementaryFigure1.tif Cite Share Download PDF Status: Published Journal Publication published 07 Oct, 2025 Read the published version in Chinese Medicine → Version 1 posted Editorial decision: Revision requested 06 Jul, 2025 Reviews received at journal 04 Jul, 2025 Reviews received at journal 04 Jul, 2025 Reviewers agreed at journal 29 Jun, 2025 Reviewers agreed at journal 28 Jun, 2025 Reviewers invited by journal 25 Jun, 2025 Editor assigned by journal 30 Apr, 2025 Submission checks completed at journal 30 Apr, 2025 First submitted to journal 26 Apr, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6535408","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":477924239,"identity":"a5bb1258-e93a-4cfa-9939-905af19d6bba","order_by":0,"name":"Feiyang Li","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Feiyang","middleName":"","lastName":"Li","suffix":""},{"id":477924240,"identity":"ccf702fb-02a6-4769-9b14-8dbf2fba96d0","order_by":1,"name":"Zhen Liu","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Zhen","middleName":"","lastName":"Liu","suffix":""},{"id":477924241,"identity":"b4e897fd-c38b-4fa4-b484-fec1f75363f2","order_by":2,"name":"Yuan Xu","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yuan","middleName":"","lastName":"Xu","suffix":""},{"id":477924242,"identity":"af78452b-fceb-4271-a8f8-617c996cdbb9","order_by":3,"name":"Yi Guo","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yi","middleName":"","lastName":"Guo","suffix":""},{"id":477924243,"identity":"73fff71f-2757-4b4d-b7d6-480f1209b60f","order_by":4,"name":"Zhifang Xu","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Zhifang","middleName":"","lastName":"Xu","suffix":""},{"id":477924244,"identity":"31fa8613-cbb9-42c6-af98-0a9eff4dba17","order_by":5,"name":"Gongming Yuan","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Gongming","middleName":"","lastName":"Yuan","suffix":""},{"id":477924245,"identity":"84d628d7-996d-4327-a934-383dd21a15ad","order_by":6,"name":"Jiyu Zhao","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Jiyu","middleName":"","lastName":"Zhao","suffix":""},{"id":477924246,"identity":"25d0b540-d05a-4f74-911b-7f4552b471c0","order_by":7,"name":"Peiyun Li","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Peiyun","middleName":"","lastName":"Li","suffix":""},{"id":477924247,"identity":"b21a00fb-69ee-420e-a838-52e2906d2d20","order_by":8,"name":"Rui Wang","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Rui","middleName":"","lastName":"Wang","suffix":""},{"id":477924249,"identity":"3172580d-a934-4c6b-9359-8681624f95a7","order_by":9,"name":"Julie Howatson","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Julie","middleName":"","lastName":"Howatson","suffix":""},{"id":477924250,"identity":"cc560831-08ac-465c-ad2e-e3c6df98adcc","order_by":10,"name":"Xue Li","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Xue","middleName":"","lastName":"Li","suffix":""},{"id":477924251,"identity":"bb8871fc-1882-4383-9a69-6c23fb984ee7","order_by":11,"name":"Yongming Guo","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAArklEQVRIiWNgGAWjYBADOTb29gOkaTHm4zmTQJqWxHkSDgbEKZWPSD4mdbPtXnqbBEMCw4+KbYS1GN5IS5POOVOc2ybdeICx58xtIrTMyDGTzqlIyG2TOZDAzNhGtBaDhHQ2iQQD4rTIS0BsSSBeiwHPs2TrnDMJhm3AQD5IlF/k25MP3s5tS5CXb28/+OBHBTG2HGBgkYBxDhBWD7KlgYH5A1EqR8EoGAWjYOQCALYlOOEEwFD1AAAAAElFTkSuQmCC","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":true,"prefix":"","firstName":"Yongming","middleName":"","lastName":"Guo","suffix":""},{"id":477924254,"identity":"58d6b8f9-7f13-4a5a-a90a-262ed18960f0","order_by":12,"name":"Yinan Gong","email":"","orcid":"","institution":"Tianjin University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yinan","middleName":"","lastName":"Gong","suffix":""}],"badges":[],"createdAt":"2025-04-26 13:53:05","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6535408/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6535408/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s13020-025-01209-8","type":"published","date":"2025-10-07T15:57:43+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":85647193,"identity":"51e7f7e2-9e55-4272-a8c0-4a4cc98882b1","added_by":"auto","created_at":"2025-06-30 08:46:17","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":490803,"visible":true,"origin":"","legend":"\u003cp\u003eSearch results and study selection.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/06b74cbb9fb620d4fb9013b7.png"},{"id":85647195,"identity":"602a1893-82d9-4935-85bf-77f8e84abd54","added_by":"auto","created_at":"2025-06-30 08:46:18","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2901036,"visible":true,"origin":"","legend":"\u003cp\u003eFlow chart of this study design\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/073ca7439a149064d7e9e625.png"},{"id":85648939,"identity":"c36d3bb5-eaee-443a-8a89-1e09330d4ff8","added_by":"auto","created_at":"2025-06-30 08:54:18","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":5274445,"visible":true,"origin":"","legend":"\u003cp\u003e(A) Intersection diagram of Acupuncture targets and RA disease targets; (B) PPI network diagram of 215 intersection targets; (C) Further screening of MCODE gene; (D) Acupuncture-Ingredients-Genes-Diseases’ diagram.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/d1301e574cb8ab9d4ba4553c.png"},{"id":85647196,"identity":"c03af3ba-3bd9-44f4-ba69-36b13ed2202e","added_by":"auto","created_at":"2025-06-30 08:46:18","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":416916,"visible":true,"origin":"","legend":"\u003cp\u003e(A) Box plot of core gene expression difference analysis between normal samples andRA samples; (B) Core gene chromosome position circle diagram; (C) Heatmap of core geneexpression in normal and RA samples; (D) Core gene related network; (E) Correlation analysis between two core genes; (F) GO enrichment analysis of DEGs between normal and RA samples; (G) KEGG enrichment analysis of DEGs between normal and RA samples.*p \u0026lt; 0.05 ; ** p \u0026lt; 0.01;***p \u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/cc894b32472b95a1da7b4818.png"},{"id":85647201,"identity":"6d0a3956-e6b1-40d9-8460-e3801ffe2af3","added_by":"auto","created_at":"2025-06-30 08:46:18","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":231265,"visible":true,"origin":"","legend":"\u003cp\u003e(A) Consensus cumulative distribution map of core gene sample unsupervisedclustering; (B) Consensus matrix heat map of core gene sampleunsupervised clustering; (C) heat map of core gene expression between core gene unsupervisedclusters; (D) Core gene cluster expression difference analysis box diagram; (E) PCA scatter diagram between core gene unsupervised clusters; (F) The GO enrichment analysis diagram of unsupervised clustering DEGs; (G) KEGG enrichment analysis of unsupervisedclustering DEGs.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/b3ac86b06b047e08598382b9.png"},{"id":85647199,"identity":"c7cc51a9-f8aa-42f4-9ce3-556a8a86fce2","added_by":"auto","created_at":"2025-06-30 08:46:18","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":337226,"visible":true,"origin":"","legend":"\u003cp\u003e(A)The relative percentage histogram of each immune cell in the sample; (B) Box plot of immune cell fraction of normal samples and RA samples; (C)Heatmap of correlation analysis between core genes and immune cells.*p \u0026lt; 0.05 \u0026nbsp;** p \u0026lt; 0.01 \u0026nbsp;\u0026nbsp;***p \u0026lt; 0.001\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/3ba88eee1bc6822a4e09ece7.png"},{"id":85647203,"identity":"644ed617-152d-4b04-a2ca-8b12b6257034","added_by":"auto","created_at":"2025-06-30 08:46:18","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":294258,"visible":true,"origin":"","legend":"\u003cp\u003e(A) ROC analysis of four machine learning models based on 5-fold cross-validation inthe test cohort; (B) Cumulative residual distribution of each machine learning model; the boxplot; (C) shows the residuals of each machine learning model (the red dot represents the root mean square of the residuals); (D) Important features in RF, SVM, GLM and XGB machine models; (E) Nomogram of key genes; (F) Calibration curve of key gene nomogram for acupuncture treatment of RA; (G) Acupuncture treatment of key gene nomogram decision curve of RA.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/41442067e48ca00507e22229.png"},{"id":85647202,"identity":"2af12fb4-0b48-4c81-b83c-4b7c04a119c8","added_by":"auto","created_at":"2025-06-30 08:46:18","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1709741,"visible":true,"origin":"","legend":"\u003cp\u003e(A) Forest plot of the causal relationship between five key genes and RA under IVW method; (B-F) The association of two specific single nucleotide polymorphisms (SNP, rs10774624 and rs1077498) in STAT1, PTGS2, JAK2, MDM2, GAPDH gene with RA. The left diagram shows the correlation between STAT1 gene expression and RA risk. The color bar represents the-log10 (\u003cem\u003eP\u003c/em\u003e) value, and the darker the color, the more significant the correlation. The right figure shows the relationship between the RA risk of these two SNP loci and their position on chromosome.\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/02c484ff31f4be5c4c5cfb7d.png"},{"id":85647205,"identity":"4a6efdfd-3910-4c88-a512-f83680e7482a","added_by":"auto","created_at":"2025-06-30 08:46:18","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":694567,"visible":true,"origin":"","legend":"\u003cp\u003eThe possible mechanism of STAT1 involved in acupuncture anti-inflammatory\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/c1ca77c27eeb17cdd1dae15f.png"},{"id":93419718,"identity":"facc1bbb-5a62-4b83-a5e4-3981f90040c7","added_by":"auto","created_at":"2025-10-13 16:06:27","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":13468907,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/eb7dbff8-afad-447c-9582-825d1b1f6e2b.pdf"},{"id":85648938,"identity":"428404fd-7181-4f39-adb6-7b1a83272359","added_by":"auto","created_at":"2025-06-30 08:54:18","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":25782,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarytable1.docx","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/5aae8a17d0a2cb43a7877f99.docx"},{"id":85647204,"identity":"4331f3d2-3b7c-498a-923e-0e88de544f48","added_by":"auto","created_at":"2025-06-30 08:46:18","extension":"html","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":964675,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarytable2GO.html","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/e92c5cda3feba3926ed4d795.html"},{"id":85647194,"identity":"076ad42a-fd32-421b-b6db-5ae85557fe5a","added_by":"auto","created_at":"2025-06-30 08:46:18","extension":"html","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":112349,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarytable3KEGG.html","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/0b959797e437353f80a1f7d4.html"},{"id":85648937,"identity":"dd5aa25f-3c17-42e6-a1bd-010c66228c46","added_by":"auto","created_at":"2025-06-30 08:54:18","extension":"html","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":402837,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarytable4unsupervisedclusteringGO.html","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/8dad22cd378796611c40573d.html"},{"id":85647207,"identity":"c55e85ca-ae02-48a4-8dfe-f2c449f85122","added_by":"auto","created_at":"2025-06-30 08:46:18","extension":"html","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":44099,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarytable5unsupervisedclusteringKEGG.html","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/59ec0e3d4882780be7f054a9.html"},{"id":85647236,"identity":"2dfe5e0e-a68a-4744-877a-1a2f4e6b3025","added_by":"auto","created_at":"2025-06-30 08:46:22","extension":"tif","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":87311036,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigure1.tif","url":"https://assets-eu.researchsquare.com/files/rs-6535408/v1/6aaf84ee272050b96f001c33.tif"}],"financialInterests":"No competing interests reported.","formattedTitle":"Potential Mechanisms of Acupuncture Treatment for Rheumatoid Arthritis: A Study Based on Network Topology and Machine Learning","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eRheumatoid arthritis (RA), a prevalent chronic inflammatory disorder, profoundly impacts patients\u0026apos; quality of life through persistent joint damage and systemic complications [1]. Characterized by progressive musculoskeletal deterioration, the disease leads to functional impairment, reduced productivity, and substantial healthcare burdens [2,3]. Epidemiological studies indicate a global prevalence of 0.5-1.0%, with peak onset between 30-50 years and a 3:1 female predominance [4].\u003c/p\u003e\n\u003cp\u003eCurrent pharmacological management employs disease-modifying antirheumatic drugs (DMARDs), biologics, and anti-inflammatory agents. While these therapies demonstrate clinical efficacy, they present limitations including adverse effects (gastrointestinal, hepatotoxicity) [5], drug resistance [6], and increased infection risk with prolonged biologic use [7]. Notably, monoclonal antibodies targeting specific cytokines (TNF-\u0026alpha;, IL-6) show variable response rates, with some patients developing treatment resistance [8]. This therapeutic challenge stems from RA\u0026apos;s complex pathogenesis involving dysregulated immune networks rather than isolated cytokine dysfunction [9]. These clinical observations have prompted a paradigm shift in therapeutic development - from single-target approaches to multi-target strategies addressing interconnected inflammatory pathways. This transition reflects growing recognition of RA as a systemic immune network disorder requiring comprehensive intervention strategies.\u003c/p\u003e\n\u003cp\u003eAcupuncture, rooted in TCM\u0026apos;s holistic philosophy of \u0026ldquo;regulating the body as an organic whole\u0026rdquo;, has been empirically proven to alleviate RA symptoms [10-12]. However, its multi-target mechanisms\u0026mdash;particularly in immune-microenvironment modulation\u0026mdash;are poorly understood. Traditional reductionist approaches struggle to capture the complexity of acupuncture\u0026apos;s systemic effects, necessitating advanced computational strategies.\u003c/p\u003e\n\u003cp\u003eBioinformatics methods play a vital role in the study of acupuncture transformation in academia. The amount of data generated by each stage of acupuncture discovery is increasing, and the use of these data for calculation can solve the key challenges in this process [13]. Network topology analysis has proven particularly valuable for mechanistic studies and target prediction in traditional medicine [14,15]. Using machine learning algorithms such as clustering and support vector machine (SVM) to mine useful information from a large number of TCM data, optimize TCM research design, reduce clinical research costs, and improve research quality and efficiency. Mendelian randomization (MR) research is a method that uses genetic variation as a tool variable to explore the causal relationship between risk factors and diseases. In the field of TCM, MR can effectively overcome the bias caused by confounding factors and reverse causality, and provide a new causal inference method for TCM research. Recent studies have emphasized the overall and multi-target effects of acupuncture on various body systems. Recent advances in Network topology and artificial intelligence (AI) offer unprecedented opportunities to decode acupuncture\u0026apos;s \u0026ldquo;multi-component, multi-target\u0026rdquo; nature. By integrating topological analysis of compound-target networks with ensemble machine learning models, we can systematically identify hub genes that serve as convergence points of acupuncture\u0026apos;s systemic regulation. Therefore, it is of great clinical value and research significance to explore the potential mechanism of acupuncture in the treatment of RA with a new research perspective [16].\u003c/p\u003e\n\u003cp\u003eIn this study, database retrieval is used to identify the active components of acupuncture in the treatment of RA. Based on the previous study by Hanet al. [17], we employed network topology and machine learning to investigate the potential mechanism of acupuncture in treating RA and analyze its effectiveness. By exploring the interaction between acupuncture and disease, the therapeutic targets of acupuncture were clarified, and gene expression pattern were confirmed by bioinformatics analysis of the Gene Expression Omnibus (GEO) data set. Through GO/KEGG functional enrichment analysis, we systematically analyzed the key biological processes and signaling pathways of acupuncture intervention in RA, revealing its potential molecular mechanisms for restoring immune homeostasis through pathways such as JAK-STAT/TNF. We also identified the characteristics of immune cell infiltration in RA, providing valuable insights for future research. Additionally, machine learning algorithms were used to screen core therapeutic targets for RA. MR analysis and colocalization analysis established causal relationships between key targets and RA, validating the impact of the key gene STAT1 on RA risk. This study developed an AI-guided comprehensive framework that not only elucidated molecular targets for acupuncture treatment of RA (such as STAT1) but also clarified the overall logic of acupuncture therapy, offering promising prospects for current research and clinical interventions.\u003c/p\u003e"},{"header":"2. Material and methods","content":"\u003ch2\u003e2.1 Collection of potentially effective active compounds produced in the human body after acupuncture treatment for RA\u003c/h2\u003e\n\u003cp\u003eIn this analysis, we systematically searched four databases: Web of Science, PubMed, CNKI, and Wan Fang Database (last updated on June 30, 2024), using the following terms: \u0026ldquo;acupuncture\u0026rdquo;, \u0026ldquo;body acupuncture\u0026rdquo;, \u0026ldquo;electroacupuncture\u0026rdquo;, \u0026ldquo;warm needle\u0026rdquo;, \u0026ldquo;fire needle\u0026rdquo;,\u0026rdquo; blood-letting puncture\u0026rdquo;, \u0026ldquo;RA\u0026rdquo;, and \u0026ldquo;arthritis\u0026rdquo;. Acupuncture is defined here as the stimulation of acupuncture points on the skin with or without electrical stimulation. Studies involving other forms of stimulation, such as acupressure, transcutaneous electrical nerve stimulation, and laser acupuncture, were excluded. We included randomized controlled trials in humans or animals, as well as non-randomized comparative trials (prospective and retrospective), and excluded single-group observational studies that evaluated outcomes before and after interventions. Control interventions could involve placebo acupuncture, sham acupuncture, no treatment, another form of active treatment, or medication. Studies comparing only different forms of acupuncture were excluded. The reported outcomes of interest were also analyzed. Two researchers independently searched the databases using predefined inclusion and exclusion criteria and selected appropriate full-text articles. Disagreements among the researchers were resolved through discussion, with a third researcher resolving disputes in cases of non-consensus.\u003c/p\u003e\n\u003ch2\u003e2.2 Identification of protein targets for active components produced in the body after acupuncture and acquisition of genes related to RA\u003c/h2\u003e\n\u003cp\u003ePharmMapper is an integrated platform for pharmacophore matching that utilizes statistical methods to identify potential targets [18]. Swiss Target Prediction, a web server based on 2D and 3D similarity metrics and known ligand binding, accurately predicts the targets of bioactive molecules [19]. Target proteins with high binding affinity () to Endogenous compounds affected by acupuncture. 5-hydroxyindole-3-acetic acid, cell cholecystokinin octapeptide (CCK-8), corticosterone, dinoprostone, dopamine, epinephrine, histamine, methionine-enkephalin (MENK), norepinephrine, and serotonin were retrieved from the PharmMapper and Swiss Target Prediction databases. The targets of the active compounds were then identified by merging the search results from the two databases and removing duplicates. Subsequently, using \u0026ldquo;RA\u0026rdquo; as a keyword, relevant genes associated with RA were retrieved from two disease gene databases: GeneCards (https://www.genecards.org/) and Comparative Toxicogenomics Database (CTD) (https://ctdbase.org/) [20,21]. The data obtained from both databases were sorted and organized, finally entered as disease genes.\u003c/p\u003e\n\u003ch2\u003e2.3 Construction of the acupuncture-component-gene-disease network and protein-protein interaction network\u003c/h2\u003e\n\u003cp\u003eThe acupuncture-related genes and RA-related genes were obtained using the online software Venn Diagrams (https://bioinfogp.cnb.csic.es/tools/venny/). The intersection of these two components was used to identify the effective treatment genes for acupuncture intervention in RA. Subsequently, a network was constructed using Cytoscape 3.9.0, with disease, acupuncture, components, and related genes as nodes and their relationships as edges, for topological analysis to determine the core components within the network.\u003c/p\u003e\n\u003cp\u003eSTRING (Search Tool for the Retrieval of Interacting Genes/Proteins, https://cn.string-db.org/) is an online bioinformatics database aimed at providing information on gene and protein interactions [22]. The intersection genes were imported into the STRING database with a filtering criterion of \u0026ldquo;minimum required interaction score \u0026ge;0.4,\u0026rdquo; and the Protein-Protein Interaction (PPI) network graph and Tab-Separated Values (TSV) format file were downloaded and saved. Subsequently, Cytoscape software (version 3.9.0) was used for the visualization and multidimensional network construction of the PPI network for acupuncture-RA.\u003c/p\u003e\n\u003ch2\u003e2.4 Collection and preprocessing of GEO samples\u003c/h2\u003e\n\u003cp\u003eThe keyword \u0026ldquo;RA\u0026rdquo; was utilized to filter RA-related samples in the GEO database (https://www.ncbi.nlm.nih.gov/geo/). The data type was set as gene expression profiles, limited to human samples. Gene expression matrices and clinical grouping information were collected from these samples, followed by gene symbol annotation and data correction using Perl code. This process aimed to determine the intersecting gene expression levels between acupuncture treatment for RA in the normal and RA groups in the field of network pharmacology.\u003c/p\u003e\n\u003ch2\u003e2.5 The expression difference of intersection genes, chromosome location, and the expression correlation of core genes\u003c/h2\u003e\n\u003cp\u003eThe R software packages \u0026ldquo;limma\u0026rdquo;, \u0026ldquo;heatmap\u0026rdquo;, and \u0026ldquo;ggpubr\u0026rdquo; were utilized to analyze the expression of overlapping genes in individuals with RA and in healthy individuals. Genes with a significance level of\u003cimg width=\"58\" height=\"22\" src=\"data:image/wmf;base64,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\" alt=\"image\"\u003ewere considered as core genes, and their differential expression was visualized using box plots and heat maps. Perl scripting was used to identify the core genes, which were then represented on a circular map. The correlation coefficient of each core gene was visually assessed using the \u0026ldquo;Rcircos\u0026rdquo; package in the R programming language.\u003c/p\u003e\n\u003ch2\u003e2.6 Differential gene enrichment analysis and expression of immune cell infiltration in RA patient samples\u003c/h2\u003e\n\u003cp\u003eIn order to clarify the biological functions and related signaling pathways between the core differential genes of RA samples and normal samples, we annotated the\u0026nbsp;DEGs, aiming to understand the biological processes, molecular functions and cell composition of different levels of biological functions and signaling pathways. Statistical analysis of the data, under the condition of\u003cimg width=\"58\" height=\"22\" src=\"data:image/wmf;base64,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\" alt=\"image\"\u003e, was carried out using R packages including \u0026ldquo;clusterProfiler\u0026rdquo; and \u0026ldquo;enrichment\u0026rdquo;, with the results visualized in Bubble diagrams. Significantly, under the criteria of \u003cimg width=\"78\" height=\"26\" src=\"data:image/wmf;base64,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\" alt=\"image\"\u003e\u0026nbsp;and corrected\u003cimg width=\"58\" height=\"22\" src=\"data:image/wmf;base64,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\" alt=\"image\"\u003e, Venn diagrams were employed to identify DEGs by intersecting core gene clusters.\u003c/p\u003e\n\u003cp\u003eIn this study, we utilized the \u0026ldquo;CIBERSORT\u0026rdquo; package in the R software to conduct 1000 simulation experiments on various types of immune cells. These simulation experiments not only yield precise data on the relative composition of immune cells but also establish a benchmark for quantifying immune cell quantities. To investigate differences in immune cells further, we employed the R packages \u0026ldquo;GSVA\u0026rdquo; and \u0026ldquo;GSA Base\u0026rdquo;. These packages offer a single-sample gene set enrichment analysis\u0026nbsp;(ssGSEA) method, allowing us to compare immune cell content disparities between the healthy control and RA patient groups. Subsequently, to delve deeper into and validate the identified core genes, we conducted a correlation test between the core genes and ssGSEA score and visualized the correlation coefficient.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e2.7 Clustering analysis of DEGs in RA patient samples\u0026nbsp;\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe \u0026ldquo;ConsensusClusterPlus\u0026rdquo; package in R software was utilized to cluster RA samples based on core gene expression with k-means clustering, Euclidean distance, and other algorithms, generating up to 9 clusters. The resulting clusters were assessed by comparing their expression levels through heat maps and box plots. Additionally, principal component analysis was conducted to evaluate inter-cluster differences.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e2.8 Construction of machine learning model and nomogram model for acupuncture treatment of RA\u003c/h2\u003e\n\u003cp\u003eIn order to gain a deeper understanding of the pivotal genes involved in acupuncture treatment for RA, we utilized the \u0026ldquo;caret\u0026rdquo; R software package to construct machine learning models, encompassing the random forest model (RF), SVM model, generalized linear model (GLM), and extreme gradient boosting (XGB). RF is an ensemble machine learning technique that employs various independent decision trees to predict classification or regression [23]. The SVM algorithm is able to create a hyperplane with maximum margin in the feature space to differentiate between positive and negative instances [24]. GLM serves as an extension of multiple linear regression models, offering flexibility in evaluating the relationship between causal features of normal distribution and categorical or continuous independent features [25]. XGB, on the other hand, consists of boosting trees based on gradient boosting, enabling a careful comparison between classification error and model complexity [26]. The caret package automatically fine-tunes the parameters of these models through a grid search. All machine learning models were executed with default parameters and assessed using 5-fold cross-validation. Subsequently, the \u0026ldquo;DALEX\u0026rdquo; package was employed to explicate the aforementioned four machine learning models, visually representing the residual distribution and feature importance in these models. Moreover, the \u0026ldquo;pROC\u0026rdquo; R package was utilized to illustrate the area under the receiver operating characteristic (ROC) curve (AUC). Using the most effective machine learning model, we identified the top five significant variables as the key predictive genes associated with RA. Following the selection of the optimal model, we utilized the characteristic genes and their expression levels in both the normal and RA groups to create a nomogram model. Each predictor was assigned a corresponding score, with \u0026quot;Points\u0026rdquo; denoting the cumulative scores of the aforementioned predictors.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e2.9\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;MR analysis\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;and\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eColocalization Analysis\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eof key genes and RA\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003e2.9.1 MR analysis\u003c/p\u003e\n\u003cp\u003eTo delve deeper into the causal correlation between the key genes and the susceptibility to RA, we opted for a two-sample Mendelian randomization (MR) analysis, which is particularly adept at investigating causal effects [27]. Initially, we isolated the SNPs of the characteristic genes for the exposure factors as well as the SNPs linked to RA as the outcome variables from the Integrated Epidemiological Unit (IEU) database (https://gwas.mrcieu.ac.uk/). Leveraging the \u0026ldquo;Two Sample MR\u0026rdquo; software package, we conducted the MR analysis. Through the application of the inverse variance weighting (IVW) method, we obtained a more precise evaluation of the correlation between the expression levels of the characteristic genes and the RA risk [28]. Furthermore, Cochran\u0026rsquo;s Q statistic was employed to assess heterogeneity in the IVW outcomes, with a P-value under 0.05 signifying statistically significant heterogeneity. Finally, we employed MR-Egger regression and MR-PRESSO analysis to thoroughly evaluate potential pleiotropy [29] Any P-value below 0.05 in the IVW results indicated a significant level of pleiotropy.\u003c/p\u003e\n\u003cp\u003e2.9.2 Colocalisation analysis\u003c/p\u003e\n\u003cp\u003eFor genes that were signifcant in both cohorts, colocalisation analysis of RA risk was performed using the R package coloc [30]. Analyses were performed using SNPs harmonised by TwoSample MR package with default priori probabilities: p1=1E\u0026minus;4, p2=1E\u0026minus;4, p12=1E\u0026minus;5. P1, p2, and p12 are predefined probability that the SNP in the test area is substantially linked with gene expression, RA risk, or both. The posterior probabilities derived from the colocalization analysis correspond to one of five hypotheses: PPH0, SNPs are not associated with either trait; PPH1, SNPs are associated with gene expression but not with RA risk; PPH2, associated with RA risk but not with gene expression; PPH3, associated with RA risk and gene expression but driven by different SNPs; PPH4, associated with RA risk and gene expression, was driven by common SNPs. The threshold of significance for colocalisation was set at PPH4\u0026gt;0.80, and genes that colocalised with RA could be considered as potential acupuncture target genes.\u003c/p\u003e"},{"header":"3. Results","content":"\u003ch2\u003e3.1 Collection of active ingredients and related genes after acupuncture treatment and collection of RA-related genes\u003c/h2\u003e\n\u003cp\u003eBy systematically searching Web of Science, PubMed, and Chinese academic databases, this study identified 10 bioactive substances closely related to acupuncture treatment for RA. The analysis indicates that acupuncture intervention\u0026nbsp;significantly modulates various neuroimmune modulatory substances, primarily including prostaglandins such as Dinoprostone\u0026nbsp;and Corticosterone (CORT), monoamine neurotransmitters and their metabolites such as 5-Hydroxyindoleacetic acid (5-HIAA) and serotonin, as well as neuropeptides such as CCK-8 and MENK. These compounds may improve the characteristic joint inflammation and pain symptoms of RA by modulating the overactive immune response [31]. Thus, acupuncture may play an active role in RA-related inflammation.\u003c/p\u003e\n\u003cp\u003eDinoprostone, also known as prostaglandin E 2 (PGE2). As a pro-inflammatory mediator, PGE2 can enhance the antigen-presenting function of dendritic cells (DCs) and promote the production of IL-17 by CD4\u003csup\u003e+\u003c/sup\u003e \u0026alpha;\u0026beta; T cells in RA patients with RA to aggravate joint inflammation [32]. The afferent nerves in acupoints activated by acupuncture and moxibustion transmit sensory signals to spinal cord, brainstem and hypothalamic neurons, further stimulating various neuroimmune pathways, and ultimately exerting anti-inflammatory effects by acting on immune cells to release key neurotransmitters and hormones, including reducing PGE2 levels [33]. The activation of peripheral nociceptors by PGE2 receptors can induce pain, and electroacupuncture can inhibit the activity of PGE2 receptors in arthritis models to exert its anti-inflammatory and analgesic effects [34].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCORT, as an endogenous glucocorticoid, exerts significant anti-inflammatory effects by downregulating pro-inflammatory mediators such as cyclooxygenase 2\u0026nbsp;(COX-2) and PGE2. Research indicates that electroacupuncture intervention can bidirectionally regulate the neuroendocrine system of RA rats. On one hand, it increases CORT levels to promote the secretion of anti-inflammatory factors such as IL-10 [35]; on the other hand, it activates the hypothalamic-pituitary-adrenal axis (HPA axis), enhancing the release of ACTH [36]. This regulatory mechanism effectively inhibits neutrophil infiltration, reduces the expression of pro-inflammatory factors such as TNF-\u0026alpha; and IL-1\u0026beta;, thereby restoring homeostasis within the body.\u003c/p\u003e\n\u003cp\u003e5-HIAA is a metabolite of serotonin and plays a crucial role in pain regulation and inflammatory response. Research indicates that elevated 5-HIAA levels can improve the pathological process of RA by activating the aryl hydrocarbon receptor (AhR) pathway, thereby inhibiting the differentiation of germinal center B cells into plasma cells while maintaining the immunoregulatory function of regulatory B cells (Bregs) [37]. In the pathogenesis of RA, the regulatory role of the hypothalamic-pituitary-adrenal axis (HPAA) is particularly crucial. Acupuncture treatment may exert therapeutic effects through dual mechanisms: on one hand, it modulates HPAA function to influence serotonin metabolism; on the other hand, it increases the levels of serotonin and 5-HIAA in the circulatory system while reducing serotonin content within platelets, ultimately achieving pain relief and anti-inflammatory effects [38].\u003c/p\u003e\n\u003cp\u003eCCK-8 is a brain-gut peptide with dual neuroendocrine regulatory functions, exerting anti-inflammatory and immunomodulatory effects through specific binding to cholecystokinin receptors. Its molecular mechanisms involve regulating lymphocyte proliferation and differentiation; affecting immune cell migration and phagocytic function; modulating the secretion of inflammatory factors [39,40]. In the pathological process of RA, CCK-8 can significantly inhibit the abnormal activation of matrix metalloproteinases (MMPs) in synovial cells, thereby improving joint inflammation [41]. Clinical studies have shown that different acupuncture methods (cheek acupuncture/body acupuncture) can effectively enhance the expression levels of CCK-8 in the central nervous system, which may be an important material basis for acupuncture in alleviating RA pain [42-44].\u003c/p\u003e\n\u003cp\u003eEnkephalins, as endogenous opioid neuropeptides, exert immunomodulatory and anti-inflammatory effects by binding to receptors on the surface of immune cells. Research indicates that acupuncture promotes enkephalin secretion through regulation of the hypothalamic-pituitary axis, which is a crucial mechanism for its treatment of RA [45,46]. Experimental evidence demonstrates, The frequency of electroacupuncture stimulation exhibits a dose-dependent relationship with the release of enkephalins; in osteoarthritis models, acupuncture significantly increases spinal cord enkephalin levels and blocks pain transmission [47]; electroacupuncture intervention can upregulate the expression of enkephalins locally in the joint, effectively alleviating pain and inflammation in acute gouty arthritis [48]. These findings collectively confirm the central role of the enkephalin system in acupuncture analgesia.\u003c/p\u003e\n\u003cp\u003eCatecholamine substances (epinephrine/norepinephrine) possess dual functions as neurotransmitters and hormones. Research indicates that acupuncture exerts anti-inflammatory effects by promoting the secretion of adrenaline from the adrenal medulla and regulating stress responses, regulating stress responses, and increasing norepinephrine levels [49], electroacupuncture intervention significantly improves the levels of norepinephrine in arthritis models, inhibits the levels of pro-inflammatory cytokines TNF-\u0026alpha;, IL-1\u0026beta;, and IL-6 in synovial tissue, and alleviates synovitis [50].\u003c/p\u003e\n\u003cp\u003eDopamine inhibits the production of cytokines [51] and the NLRP3 inflammasome [52] through D1 dopamine receptors, thereby suppressing systemic inflammation. Experimental evidence confirms that acupuncture effectively modulates systemic inflammatory responses by activating the vagus nerve-adrenal reflex arc and promoting dopamine release [51].\u003c/p\u003e\n\u003cp\u003e5-hydroxytryptophan (5-HTP) is an endogenous amino acid and a precursor to serotonin (5-HT), which can be metabolized in the body to produce serotonin. In addition to its role as a neurotransmitter, 5-HT is involved in immune regulation and inflammatory responses. Histamine (HA) and 5-HT are both vasoactive amines that can induce vasodilation and enhance microvenous permeability, thereby promoting local tissue edema and inflammatory response. In severe cases, this can lead to tissue hypoxia or even necrosis [53]. However, 5-HT can also inhibit the release of pro-inflammatory factors such as TNF-\u0026alpha; and IL-6 through its receptor-mediated signaling pathways, thereby alleviating inflammatory responses. Research indicates that acupuncture may activate the descending pain control system by increasing the levels of serotonin in the spinal cord or brain [54]. Following acupuncture intervention, mast cells migrate to local acupoints via small arterioles in the subcutaneous and subcutaneous tissues. The aggregated mast cells release more trypsin, histamine, and serotonin through degranulation, thereby modulating immune responses and inflammatory processes [55].Subsequently, using PharmMapper and Swiss Target Prediction databases, a total of 631 potential target proteins associated with acupuncture active components were identified. Furthermore, based on searches of the Genecards and CTD databases, 1333 and 15870 RA related genes were obtained, respectively. Further utilized the Venny online tool for intersection analysis, ultimately screening out 215 common targets. These overlapping genes may constitute key molecular targets for acupuncture treatment of RA, providing crucial clues to elucidate its mechanism of action. \u003cstrong\u003e(Figure 3A).\u003c/strong\u003e\u003c/p\u003e\n\u003ch2\u003e3.2 \u0026ldquo;Acupuncture-ingredients-gene-disease\u0026rdquo;\u0026nbsp;network analysis and\u0026nbsp;PPI\u0026nbsp;network analysis\u003c/h2\u003e\n\u003cp\u003eThis study utilized STRING and Cytoscape software to construct a PPI network comprising 215 shared targets, consisting of 213 nodes and 3898 interaction edges, with an average connectivity of 36.601.Subsequently, the network was clustered into four groups using the MCODE (Molecular Complex Detection) plug-in (), resulting in four distinct cluster networks, as illustrated in \u003cstrong\u003eFigure\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003cstrong\u003eB\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;and 3\u003c/strong\u003e\u003cstrong\u003eC\u003c/strong\u003e. Core cluster genes were identified by selecting those with a degree value \u0026ge; 1.5 times the median in each cluster network, totaling 65 targets such as TNF, IL6, JAK2,MAPK8, GAPDH, and AKT1, ALB. To delve deeper into the relationship between the central components of acupuncture, their associated genes, and RA, Cytoscape software (version 3.9.0) was employed to construct the \u0026ldquo;acupuncture-component-gene-disease\u0026rdquo; network \u003cstrong\u003e(Figure 3D)\u003c/strong\u003e. This network comprises 76 nodes, including 10 active component nodes, 65 gene nodes, and one disease node, interconnected by 456 connections. Network topology parameters were analyzed using the Network Analyzer plug-in, which indicated an average number of adjacent nodes of 11.844, network heterogeneity of 1.085, network density of 0.156, and network centrality of 0.718. Nodes with higher degrees were identified as core nodes within the network. The most active components based on degree were dinoprostone(\u003cimg width=\"83\" height=\"21\" src=\"data:image/wmf;base64,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\" alt=\"image\"\u003e), CORT(\u003cimg width=\"84\" height=\"21\" src=\"data:image/wmf;base64,R0lGODlhVAAVAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAIAAwBQAA4AhQAAAAAAAAAdHR0AAAAAHR0AHQAAMx0dNAAdMgAcSB0dSB00NB0zWgAzWh4zRx1IWx1GbDMAADIAHTIdADIAMjMeRzMzWzMzSDVIWzVbbjNbgEgcAEceM1ozAEczHlszM0lJHVtIHVtINVtISEhbbkhZf11uXVl/WV1/f0huf1lublluf0RubmxGHW5bNW5GM25dXW5uRH9/XX9uSH9uWW5uWW5ugGaIiIBbM4iIZgECAwECAwECAwECAwECAwECAwb/QICwEygGBMKkcslsOp/QJS6hJAachmQkyu16nRHqMNlhdhrK63fNdoLEgA0kCRdGNMoIus1n4+EbSS1MWUobewAeRUoDRgELA4UDS0VzQopqbBYALXUDiFVMenYPnEkEp0kKACZ5pYN2dGSORqAAH7JphYaWSZNxex1zgUmFLQ5Lh0nDZH2cdRwAEbtCBUsdhag1AAxJiIilStrcADgEKQATfLQBeGVjSnUAai0BBAdLE/bhABRM9pTumWEXwJYgOFuS+LPTywOBDEkQJMEn5IKUJhKF4MPhjAmgJBw50UK1BFM3IfUoRilyMtGijp0MFZG3RkSSERMMdtzZkaSQDRc8gwYN0eiI0KNJggAAOw==\" alt=\"image\"\u003e),5-hydroxyindole-3-acetic acid (\u003cimg width=\"84\" height=\"21\" src=\"data:image/wmf;base64,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\" alt=\"image\"\u003e), and CCK-8 (\u003cimg width=\"84\" height=\"21\" src=\"data:image/wmf;base64,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\" alt=\"image\"\u003e) \u003cstrong\u003e(Table 1).\u003c/strong\u003e The effective components of acupuncture for RA are postulated to be those that exhibit extensive action points and robust interactions that play pivotal roles in the network. Furthermore, a single component may simultaneously affect multiple genes, reflecting a multi-gene regulatory characteristics, and multiple components may concurrently correlate with a single gene concurrently. These findings highlight the multi-faceted, multi-gene regulatory nature of acupuncture in treating RA.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1\u003c/strong\u003e\u003cstrong\u003e．\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;Ranking table of compounds\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"624\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eName\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDegree\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBetweenness Centrality\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCloseness Centrality\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003eDinoprostone\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e0.128465062\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e0.660869565\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003eCorticosterone\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e0.093978634\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e0.628099174\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003e5-Hydroxyindole-3-acetic acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e0.054056222\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e0.571428571\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003eCCK-8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e0.060173298\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e0.554744526\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003eMET-enkephalin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e0.037893873\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e0.531468531\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003eEpinephrine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e0.034330837\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e0.524137931\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003eNorepinephrine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e0.026687875\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e0.503311258\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003eDopamine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e0.01489401\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e0.439306358\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003eSerotonin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e0.013884685\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e0.439306358\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 33.9744%;\"\u003e\n \u003cp\u003eHistamine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9.9359%;\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 30.2885%;\"\u003e\n \u003cp\u003e0.003288627\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.8013%;\"\u003e\n \u003cp\u003e0.397905759\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003ch2\u003e3.3 Acquisition of\u0026nbsp;GEO\u0026nbsp;dataset samples and correlation analysis of intersection genes\u003c/h2\u003e\n\u003cp\u003eThrough integrated analysis of network topology, this study identified 65 key target sites for acupuncture intervention in RA \u003cstrong\u003e(Supplementary Table 1)\u003c/strong\u003e. Data set GSE89408 related to RA was retrieved from the GEO database, which includes 28 healthy controls and 152 RA samples. After standardization preprocessing and differential expression analysis, 49 DEGs were obtained by intersecting with the aforementioned targets, including core regulatory factors such as TNF, IL6, GAPDH, and STAT1. Clinical sample validation revealed that, apart from the significant downregulation of nine genes (AKT1, EGFR, SRC, MAPK3, ESR1, HRAS, ACE, PGR, and NOS3) in the model group, the remaining DEGs showed a significant upregulation trend in the RA group \u003cstrong\u003e(Figures 4A-B)\u003c/strong\u003e. The specific chromosomal locations of acupuncture-related DEGs are shown in \u003cstrong\u003eFigure 4C\u003c/strong\u003e. Correlation analysis among the DEGs in RA samples indicated a strong relationship between them, as displayed in\u003cstrong\u003e\u0026nbsp;Figures 4D and\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e4\u003c/strong\u003e\u003cstrong\u003eE\u003c/strong\u003e. It is indicating that these genes are related to biological functions and may be involved in the same cellular processes or pathways in the disease state. This expression pattern may be used as a biomarker for disease progression or prognosis, which would be helpful in the diagnosis and treatment of RA.\u003c/p\u003e\n\u003ch2\u003e3.4 Differential gene enrichment analysis and immune cell infiltration analysis of normal samples and RA samples\u003c/h2\u003e\n\u003cp\u003eThrough systematic biological analysis of 49 DEGs associated with RA treated by acupuncture, this study reveals the potential mechanisms of acupuncture intervention on RA through multi-level functional annotation. Gene Ontology(GO) enrichment analysis yielded 1729 significant entries (\u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e)\u003cstrong\u003e,\u003c/strong\u003e with biological processes (BP) accounting for 1582 items, primarily enriched in immune stress response pathways, including lipopolysaccharide stress response, chemical stress adaptability regulation, and maintenance of redox homeostasis; molecular functions (MF) identified 98 key entries, involving protein tyrosine kinase activity, phosphatase interaction networks, and cytokine receptor binding among other molecular interaction features; cellular components (CC) localized to 49 substructures, such as membrane rafts signaling platforms, lateral plasma membrane compartments, and fibrinogen-enriched granules \u003cstrong\u003e(Figure 4F)\u003c/strong\u003e. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis further identified 159 significantly enriched signaling pathways\u003cstrong\u003e,\u003c/strong\u003e primarily involving three core regulatory modules: 1) inflammatory cascades (TNF signaling axis, NF-\u0026kappa;B signal transduction); 2) immune cell function regulation (JAK-STAT signaling network); and 3) cellular stress adaptability pathways. \u003cstrong\u003e(Figure 4G) (Supplementary Table Table 3)\u0026nbsp;\u003c/strong\u003eThese findings systematically elucidate that acupuncture may exert therapeutic effects through the coordinated regulation of the inflammation-immune-stress triple network, providing molecular-level evidence to support its multi-target action mode.\u003c/p\u003e\n\u003ch2\u003e3.5 Unsupervised clustering of DEGs in RA samples and enrichment analysis of DEGs after unsupervised clustering\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eWe performed unsupervised clustering of the samples based on the core genes. leading to the identification of two primary clusters with the highest accuracy. The RA samples were stratified into the Diffgene1 and Diffgene2 groups, as shown in\u003cstrong\u003e\u0026nbsp;Figure 5A\u003c/strong\u003e. Using cluster comparison analysis, we sought to augment and validate our previous findings\u003cstrong\u003e\u0026nbsp;(Figure 5B)\u003c/strong\u003e. Subsequently, we examined the expression patterns of core genes in the two distinct clusters, displayed in\u003cstrong\u003e\u0026nbsp;Figure 5C and\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e5\u003c/strong\u003e\u003cstrong\u003eD\u003c/strong\u003e. In total. 49 core genes were examined in clusters Diffgene1 and Diffgene2. Notably, there were no significant differences in the expression levels of genes such as IGF1, KDR, ANXA5, MAPK1, PIK3CA, MAPK14, ACE, NOS3, or LGALS3 between the two cluster subgroups, whereas significant differences were observed in the remaining 40 genes. Principal component analysis (PCA) demonstrated the discriminative ability of the core genes between Diffgene1 and Diffgene2, as indicated in \u003cstrong\u003eFigure 5E\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003eThrough functional enrichment analysis, we thoroughly investigated the biological functions and signaling pathways involved in DEGs within the two cluster subgroups, thereby comprehensively elucidating the mechanisms of acupuncture treatment for RA. Enrichment analysis based on GO identified 564 significantly enriched functional terms \u003cstrong\u003e(Supplementary Table 4)\u003c/strong\u003e, of which 511 pertain to BP. These primarily involve immune-related processes such as lymphocyte differentiation, monocyte differentiation, and leukocyte chemotaxis. In terms of MF, 36 terms were enriched, including key molecular functions such as chemokine activity, cytokine activity, and chemokine receptor binding. CC are enriched with 17 terms, such as the external side of the plasma membrane, tertiary granules, and immunological synapse, which are closely related to immune response and subcellular structures \u003cstrong\u003e(Figure 5F)\u003c/strong\u003e. KEGG pathway analysis further identified 38 significantly enriched signaling pathways, including the TNF signaling pathway, T cell receptor signaling pathway, and JAK-STAT signaling pathway among key inflammatory regulatory pathways \u003cstrong\u003e(Figure 5G) (Supplementary Table 5).\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIt is noteworthy that both independent enrichment analyses revealed that the top 20 enriched entries were significantly associated with immune response and inflammatory regulation processes. In GO analysis, at the MF level, there is continuous enrichment of cytokine receptor binding and phospholipase activator activity; at the BP level, key processes such as regulation of inflammatory response, T cell activation, negative regulation of immune system process, response to molecule and leukocyte migration are consistently enriched.The results of KEGG pathway analysis also show high consistency, with significant enrichment in the TNF signaling pathway and the JAK-STAT signaling pathway \u003cstrong\u003e(Figure 4G, Figure 5G)\u003c/strong\u003e. These findings systematically elucidate the immunomodulatory mechanisms of acupuncture in treating RA, particularly through the regulation of key inflammatory signaling pathways and the function of immune cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.6 Analysis of immune cell infiltration in normal samples and RA samples\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eInflammatory cell infiltration is an important pathological feature of RA. The interaction between synovial and infiltrating cells can produce a large number of pro-inflammatory mediators and cytokines, which, in turn, act on the synovium and cartilage, activate nociceptors, secrete cytokines, and cause joint tissue damage. The results of acupuncture against RA DEGs enrichment analysis showed that RA are closely related to the functions and signaling pathways of immune cells. To thoroughly investigate the differences in the immune microenvironment between RA patients and healthy individuals, this study employed the Siber sorting algorithm in R language for systematic analysis of immune cell infiltration, accurately quantifying the distribution proportions of various immune cell subpopulations in the samples\u003cstrong\u003e\u0026nbsp;(Figure 6A)\u003c/strong\u003e. Furthermore, using the ssGSEA method, significant differences in immune cell types between the two groups were identified. The results showed that 22 types of immune cells were detected in synovial tissue, with eight subpopulations exhibiting significant differences between the RA group and the control group \u003cstrong\u003e(Figure 6B)\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003eSpecifically, the healthy control group exhibited higher abundances of resting CD4\u003csup\u003e+\u003c/sup\u003e T cells, germinal center T cells, regulatory T cells (Tregs), resting and activated natural killer (NK) cells, monocytes, and resting mast cells. In contrast, the RA group showed a significant increase in M1 macrophages, activated CD4\u003csup\u003e+\u003c/sup\u003e T cells, memory B cells, plasma cells, resting dendritic cells, and neutrophils. Correlation analysis \u003cstrong\u003e(Figure 6C)\u003c/strong\u003e reveals significant associations (\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u0026lt; 0.05) between core genes and specific immune cell subpopulations: M1 macrophages, activated CD4\u003csup\u003e+\u003c/sup\u003e T cells, memory B cells, and mast cells show positive correlations with most core genes, while NK cells exhibit a negative correlation trend. These findings further confirm the crucial regulatory role of immune cells in the pathogenesis of RA.\u003c/p\u003e\n\u003ch2\u003e3.7 Machine learning model analysis and RA nomogram model construction\u003c/h2\u003e\n\u003cp\u003eThis study identified 49 DEGs related to acupuncture treatment for RA as candidate genes, further identifying core DEGs influenced by acupuncture. The aim was to evaluate the diagnostic potential of DEGs between RA patients and healthy individuals and to explore the impact of acupuncture on these genes.To achieve this goal, we developed four sophisticated machine learning models (RF, SVM, GLM, and XGB) to identify key genes in the RA dataset for accurate classification of patients. The DALEX software package was employed to interpret these models and visualize the residual distribution of each model in the test dataset. We evaluated the discriminatory performance of the four machine learning algorithms using 5-fold cross-validation to calculate the ROC curve. Among the models, XGB demonstrated the largest AUC (GLM, AUC = 0.939; SVM, AUC = 0.978; RF, AUC = 0.986; XGB, AUC = 0.994, (\u003cstrong\u003eFigure 7A)\u003c/strong\u003e.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eFurthermore, the XGB and RF models exhibited relatively low residuals \u003cstrong\u003e(see Figure 7B, C)\u003c/strong\u003e.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eSubsequently, we ranked the top ten important feature variables of each model based on the root mean square error (RMSE) \u003cstrong\u003e(Figure 7D)\u003c/strong\u003e. In summary, the XGB model was the most adept at distinguishing distinct patient clusters.Using feature importance evaluation based on the XGBoost model, this study identified five key predictive factors (STAT1, GAPDH, JAK2, PTGS2, and MDM2) as core regulatory genes.Immune microenvironment correlation analysis reveals that the expression levels of STAT1 and JAK2 genes are significantly positively correlated with pro-inflammatory immune cell subpopulations (M1 macrophages, activated CD4\u003csup\u003e+\u003c/sup\u003e T cells, and memory B cells) (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05), but negatively regulate immunosuppressive cells (resting mast cells, activated NK cells, and Treg cells ).\u0026nbsp;GAPDH exhibits a unique biphasic regulatory pattern, with its expression positively correlated with M1 macrophages, resting mast cells, and resting NK cells, while it shows an inhibitory effect on activated mast cells.\u0026nbsp;It is noteworthy that PTGS2 specifically promotes eosinophil infiltration, while MDM2 synergistically enhances the pro-inflammatory phenotype of M1 macrophages and activated CD4\u003csup\u003e+\u003c/sup\u003e T cells, simultaneously inhibiting mast cell quiescence \u003cstrong\u003e(Figure 3H)\u003c/strong\u003e.\u0026nbsp;These findings suggest that screening genes may participate in the pathogenesis of RA by regulating the dynamic balance of immune cells.\u003c/p\u003e\n\u003cp\u003eTo validate the clinical applicability of the XGBoost model, this study constructed a risk prediction nomogram based on this algorithm, used to quantitatively evaluate disease risk stratification in a cohort of 152 RA patients \u003cstrong\u003e(Figure 7E)\u003c/strong\u003e. Calibration analysis indicates good consistency between model-predicted risk and actual risk \u003cstrong\u003e(Figure 7F)\u003c/strong\u003e, with a Hosmer-Lemeshow test \u003cem\u003eP\u003c/em\u003e \u0026gt; 0.05 indicating no significant calibration bias. Decision curve analysis (DCA) further confirms that the model has significant clinical net benefit across a wide range of thresholds \u003cstrong\u003e(Figure 7G)\u003c/strong\u003e, with the standardized net benefit exceeding traditional diagnostic methods by more than 30%, indicating that this tool can serve as an effective auxiliary system for evidence-based medical decision-making, providing reliable quantitative basis for personalized treatment of RA.\u003c/p\u003e\n\u003ch2\u003e3.8 MR analysis and Colocalization Analysis\u0026nbsp;results of key genes and RA\u003c/h2\u003e\n\u003cp\u003eThis study employs a two-sample MR framework to elucidate the causal associations between DEGs and RA.\u0026nbsp;Integrated resources from the IEU OpenGWAS database, which includes 14,361 RA patients and 43,923 European ancestry controls, we identified single nucleotide polymorphisms (SNPs) that meet stringent instrumental variable criteria (\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u0026lt; 5e-8) for five key genes, including STAT1 JAK2 MDM2 PTGS2 and GAPDH \u003cstrong\u003e(Figure 8A\u003c/strong\u003e).\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eMR analysis showed that the expression levels of STAT1 (OR = 1.53,95 % CI: 1.31-1.77, \u003cem\u003eP\u003c/em\u003e = 4.71e-08) and PTGS2 (OR = 1.16,95 % CI: 1.03-1.32, \u003cem\u003eP\u003c/em\u003e = 0.015) had a significant positive causal relationship with the risk of RA \u003cstrong\u003e(Figure 8A)\u003c/strong\u003e. This study employed various MR methods, including IVW and MR-Egger regression, to investigate the causal effects of STAT1 and PTGS2 genes on RA. The robustness of these findings was confirmed through funnel plot symmetry analysis and Leave-one-out analysis. Furthermore, MR-Egger regression and MR-PRESSO analysis did not detect directional pleiotropy (\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u0026gt; 0.05), further supporting the reliability of the research results, the above results are shown in\u003cstrong\u003e\u0026nbsp;Supplementary Figure 1\u003c/strong\u003e. Co-localization analysis reveals a significant causal relationship between the STAT1 locus and RA phenotype (PPH4 \u0026gt; 0.8), suggesting that variations in the STAT1 gene may play a crucial role in the pathogenesis of RA \u003cstrong\u003e(Figure 8B)\u003c/strong\u003e. Given the crucial role of STAT1 in the immune system, interventions targeting this gene may have potential therapeutic effects for RA. Furthermore, the fact that the other four genetic loci do not share causal variants with the RA phenotype underscores the uniqueness and importance of STAT1 as a therapeutic target for RA \u003cstrong\u003e(Figure\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;8C-F\u003c/strong\u003e\u003cstrong\u003e)\u003c/strong\u003e. This study employs causal inference methods to systematically analyze the impact of acupuncture on gene expression in patients with rheumatoid arthritis, revealing that the expression of the STAT1 gene may be regulated. This finding not only elucidates how acupuncture might exert therapeutic effects by modulating these key genes but also offers new perspectives for developing precision treatments based on gene regulation in the future. Additionally, the study underscores the importance of integrating traditional medicine with modern scientific technology, providing valuable references for research into the treatment of other diseases.\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eIn recent years, research on the molecular mechanisms of acupuncture intervention in RA has gradually become a hotspot in the field of translational medicine [56,57]. Current evidence suggests that acupuncture active components can influence the pathological process of RA by modulating the dynamic balance of immune cells and the network of inflammatory factors [58]. With the advancement of high-throughput sequencing technology, integrating multi-omics data with AI algorithms offers new technological pathways for elucidating the key regulatory networks of complex diseases. However, there are still limitations in the study of acupuncture mechanisms [59,60]: first, traditional bioinformatics methods struggle to systematically reveal the patterns of multi-target synergistic effects; second, there is a lack of genetic evidence based on causal inference to support target selection.\u003c/p\u003e\n\u003cp\u003eThis study systematically elucidates the molecular mechanisms of acupuncture treatment for RA by constructing a multi-layer interaction network involving \u0026apos;acupuncture-active components-targets-disease\u0026apos;. Initially, 261 potential target genes were screened based on network topology, with 65 target genes identified through modular analysis. Differential expression validation using the GEO dataset (GSE89408) yielded 49 DEGs, which were significantly enriched in inflammation-related pathways (such as NF-\u0026kappa;B, JAK/STAT) and immune cell regulatory networks (M1 macrophages, activated CD4\u003csup\u003e+\u003c/sup\u003e T cells, Treg cells, B cells, NK cell). A clinical prediction model based on the XGBoost algorithm (AUC=0.994) and risk nomogram was further optimized using four machine learning algorithms to identify five core disease-related genes (STAT1, PTGS2, MDM2, GAPDH and JAK2), providing a quantitative assessment tool for personalized treatment of RA. A significant genetic association between STAT1 (OR = 1.53, \u003cem\u003eP\u0026nbsp;\u003c/em\u003e= 4.71 \u0026times; 10\u003csup\u003e8\u003c/sup\u003e) and PTGS2 (OR = 1.16, \u003cem\u003eP\u003c/em\u003e = 0.015) and RA risk was confirmed by MR. Co-localization analysis indicates that the STAT1 locus (PPH4 \u0026gt; 0.8) may serve as a key regulatory target for acupuncture intervention.\u003c/p\u003e\n\u003cp\u003eThis study found that acupuncture active components may inhibit synovial inflammation by targeting the STAT1/JAK2 signaling axis, regulating T-cell differentiation, macrophage polarization, and NK cell function. This study is the first to integrate systems biology, machine learning, and causal inference methods to elucidate the network of acupuncture treatment for RA at multiple levels, including molecular, cellular, and clinical. It not only provides an innovative paradigm for the modern research of traditional therapies but also lays a theoretical foundation for the development of novel targeted treatment strategies.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.1 Acupuncture (compounds) affects immune cells to exert anti-inflammatory effects\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on the multi-dimensional network analysis of \u0026apos;acupuncture-ingredients-targets-diseases\u0026apos;, this study systematically elucidated the immunoregulatory mechanisms of acupuncture in treating RA. Network topology analysis identified key active ingredients such as CCK-8, PGE2, and CORT, which exert their effects through synergistic action on multiple targets to regulate the immune-inflammatory-pain axis. Research has confirmed that acupuncture intervention can significantly regulate key neurotransmitter levels (such as upregulating CCK-8 and enkephalins, downregulating PGE2. It also involves multiple functions and signaling pathways related to the nervous system, immune response, and inflammation, which form the basis of acupuncture treatment for RA.\u003c/p\u003e\n\u003cp\u003eAs an immunomodulatory peptide highly expressed in the central nervous system, CCK-8 exerts bidirectional regulatory effects by binding to cholecystokinin receptors (CCKR). In inflammatory responses, CCK-8 can influence B cells [61] and T cells [62]. It inhibits the phenotypic and functional maturation of DCs and B cell IgG1 secretion, reduces the release of pro-inflammatory cytokines such as TNF-\u0026alpha; and IL-1\u0026beta; from macrophages, while promoting the production of anti-inflammatory cytokines like IL-4, thereby alleviating inflammatory responses [63,64]. CCK-8 promotes Th1 cell polarization both in vitro and in vivo, and reshapes the Th1/Th17 cell balance by modulating the cytokine profile of dendritic cells (upregulating IL-12 and downregulating IL-6 and IL-23), significantly improving joint inflammation in RA model animals [65]. Studies show that acupuncture can increase the serum CCK-8 levels in RA model animals, with the concentration changes showing a significant positive correlation with the elevation of pain threshold [44].\u003c/p\u003e\n\u003cp\u003eEndogenous opioid peptides (such as enkephalins) mediate complex immune regulation through \u0026mu;/\u0026delta; receptors. MENK inhibits the differentiation of Foxp3\u003csup\u003e+\u003c/sup\u003e Tregs induced by TGF-\u0026beta;, thereby blocking the formation of an immunosuppressive [66] microenvironment. Additionally, enkephalins can promote the maturation and function of DCs by increasing the expression of MHC II, CD86, and CD40 on the surface of mouse DCs, which in turn enhances the proliferation and polarization of CD4\u003csup\u003e+\u003c/sup\u003e T cells and exacerbates inflammatory responses [67]. Acupuncture stimulation dynamically regulates the release of opioid peptides, with its analgesic effects closely related to the inhibition of pain transmission mediated by substances such as enkephalin and \u0026beta;-endorphin, as well as the downregulation of inflammatory cytokines like IL-6 and TNF-\u0026alpha; [68].\u003c/p\u003e\n\u003cp\u003ePGE2 regulates Th cell differentiation through EP receptors, characterized by enhanced secretion of Th2-type cytokines (IL-4/IL-10/IL-13) and inhibition of IL-12 signaling [69]. In RA, PGE2 can enhance antigen presentation by DCs and promote IL-17 production by CD4\u003csup\u003e+\u003c/sup\u003e \u0026alpha;\u0026beta; T cells [32,70]. Studies have shown that\u0026nbsp;acupuncture intervention significantly reduces PGE2 levels in RA synovium, promoting polarization of M1 macrophages to M2 macrophages, thereby inhibiting the expression of key pro-inflammatory factors (IL-6, MCP-1, IL-1\u0026beta;, G-CSF, TNF-\u0026alpha;), alleviating inflammation and pain [71]. Additionally, electroacupuncture stimulation can reduce PGE2 levels in RA rats by affecting the hypothalamic-pituitary axis, exerting anti-inflammatory effects [72].\u003c/p\u003e\n\u003cp\u003eThe aforementioned findings elucidate that acupuncture dynamically regulates the neuroendocrineimmune system through a multidimensional network of \u0026ldquo;components-targets-pathways,\u0026rdquo; such as by remodeling the functional state of immune cells via mediators like CCK-8/PGE2, ultimately achieving multiple therapeutic effects including anti-inflammatory, immunomodulatory, and analgesic outcomes. This provides a theoretical framework for developing targeted treatment strategies for RA based on the mechanisms of acupuncture.\u003c/p\u003e\n\u003ch2\u003e4.2The potential mechanism of acupuncture in the treatment of RA\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003e4.2.1 Immune cells in RA\u003c/p\u003e\n\u003cp\u003eA notable feature of the RA synovial microenvironment is the abnormal aggregation of various immune cells, including T lymphocytes, B lymphocytes, macrophages, natural killer cells, and mast cells [73,74]. Among these, synovial macrophages exhibit an M1 polarization phenotype during RA progression. They possess the potential to differentiate into osteoclasts and mediate the chemotactic migration of monocytes and neutrophils through the secretion of inflammatory cytokines. Additionally, they activate T cells and drive the aberrant proliferation of synoviocytes, thereby establishing a sustained pro-inflammatory effect [75].\u003c/p\u003e\n\u003cp\u003eAmong the immune cells infiltrating the synovium, subsets of CD4+ T cells play a central role [76]. Activated CD4\u003csup\u003e+\u003c/sup\u003e, CD8\u003csup\u003e+\u003c/sup\u003e T cells, and Th17 cells significantly promote osteoclast differentiation by releasing key effector molecules such as nuclear factor kappa B ligand (RANKL), TNF-\u0026alpha;, IL-1, IL-6, and IL-17. In contrast, IFN-\u0026gamma; and IL-4 secreted by Th1 and Th2 cells exert negative regulatory effects on osteoclastogenesis [77]. Notably, activated CD4\u003csup\u003e+\u003c/sup\u003e T cells can further amplify inflammatory damage in joint tissues through cascaded activation of macrophages and B cells, whereas resting CD4\u003csup\u003e+\u003c/sup\u003e T cells do not participate in this pathological process [78].Follicular helper T cells (Tfh) within the CD4+ T cell subsets play a unique role in RA, with their abnormal expression of surface markers CXCR5, ICOS, and PD1 being closely related to disease occurrence. These cells participate in autoimmune responses by regulating B cell antibody production. Concurrently, functional defects in Tregs may disrupt immune homeostasis, leading to excessive activation of autoreactive T cells [79]. The mechanism of B cell involvement in RA encompasses multi-pathway regulation: B cells promote osteoclast maturation by producing RANKL [80], activate memory B cells to enhance immune response, and induce synovial tissue to produce pro-inflammatory cytokines such as IL-1\u0026alpha;, IL-23, IL-12, IL-6, and TNF-\u0026alpha;, thereby exacerbating bone destruction [81]. NK cells participate in disease progression through cytotoxic activity and cytokine networks [82]. Research indicates that acupuncture may improve RA immune imbalance by modulating NK cell activity [83]. Mast cells, as resident components of the synovium in innate immunity, regulate T/B cell and APC function by secreting mediators such as TNF-\u0026alpha;, IL-1\u0026beta;, IL-4, and IL-5 [84]. Their abnormal activation is significantly associated with worsened joint inflammation [85].\u003c/p\u003e\n\u003cp\u003eThis study employed the Siebel algorithm in R language to analyze the immune cell infiltration characteristics of RA patients, revealing significant upregulation of M1 macrophages, activated CD4\u003csup\u003e+\u003c/sup\u003e T cells, and memory B cells. This finding corroborates the theory of immune-mediated synovial damage, suggesting that abnormal immune responses may directly drive inflammation through metabolic remodeling and polarization changes in macrophages, while persistent activation of T cells leads to a vicious cycle of autoimmunity. In summary, different subpopulations of immune cells collectively constitute the core mechanism of RA pathogenesis through complex interaction networks.\u003c/p\u003e\n\u003cp\u003e4.2.2 The interaction between core genes (JAK2, STAT1, GAPDH, PTGS2, MDM2) and immune cells affected the pathogenesis of RA\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe JAK/STAT signaling pathway is widely expressed in various cells and is stimulated by multiple inflammatory stimuli, affecting the differentiation of macrophages and inflammatory responses and participating in many important biological processes, such as cell proliferation, differentiation, apoptosis, and immune regulation [86].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eJAK2 kinase, as a core regulatory molecule in cytokine signaling, drives disease progression through multiple mechanisms. Clinical studies have demonstrated that T lymphocytes, macrophages, and fibroblast-like synoviocytes (FLS) in the peripheral blood and synovial microenvironment of RA patients exhibit abnormal overexpression of JAK2 [87], suggesting that this molecule has the potential to serve as a biomarker for disease monitoring.\u0026nbsp;STAT1 serves as the core effector molecule of this pathway, exhibiting specific activation characteristics. Its functional regulation depends on JAK-mediated dual phosphorylation modifications at the Y701/S727 sites [88], which are particularly prominent in FLS [89]. Clinical research data indicate that the expression levels of STAT1 in RA synovial tissues are significantly higher than those in osteoarthritis (OA) control groups, and show characteristic downregulation following effective treatment [90]. De Hooge et al. [91]showed that granuloma formation and progressive arthritis were observed in the arthritis of STAT1 deficient mice induced by zymosan, indicating that the anti-inflammatory effect of STAT1 deficiency may be weakened.\u0026nbsp;At the molecular level of RA, pro-inflammatory cytokines such as IL-6 and IL-12 activate the JAK1/JAK2 kinase complex by binding to T cell surface receptors like IL-6R and IL-12R. JAK2 initiates the phosphorylation cascade of downstream STAT3/STAT1 through its interaction with the intracellular domains of transmembrane receptors such as IL-6R and IL-23R. This process promotes the specific secretion of IL-17A by Th17 cells, disrupting Treg-mediated immune homeostasis and exacerbating inflammatory infiltration and cartilage destruction in the synovium [92,93]. Furthermore, STAT1 can promote the abnormal survival of autoreactive memory T cells by regulating key transcription factors such as BCL-6, thereby becoming an important mechanism for disease chronicity [94]. The IFN-\u0026gamma;-mediated JAK2-STAT1 signaling pathway can induce macrophages to convert to an M1 pro-inflammatory phenotype, significantly increasing the release levels of pro-inflammatory mediators such as TNF-\u0026alpha; and IL-12, thereby creating a sustained inflammatory microenvironment [95].\u003c/p\u003e\n\u003cp\u003eIn RA, cytokines such as IFN-\u0026gamma; and IL-6 mediate the activation of JAK1/JAK2 kinase cascades by binding to surface receptors on macrophages, such as IFN-\u0026gamma;R and IL-6R. This activation further leads to the phosphorylation and activation of STAT1, which upregulates HIF-1\u0026alpha;, enhancing glycolytic metabolism and thereby sustaining the energy requirements of the pro-inflammatory M1 macrophage phenotype [96], resulting in the self-maintenance and amplification of inflammatory responses [97]. Inhibition of the JAK/STAT signaling pathway can regulate macrophage polarization, promoting the transition of pro-inflammatory M1 macrophages to anti-inflammatory M2 macrophages [98]. The inhibition of JAK2 expression also reduces the production of proinflammatory cytokines and local inflammatory responses\u0026nbsp;[99]Blocking STAT1 signaling effectively inhibits pro-inflammatory gene expression and reduces the degree of M1 polarization, confirming its central regulatory role in the inflammatory cascade [100]. Inhibitors of the JAK-STAT signaling pathway, such as tofacitinib, have demonstrated anti-inflammatory effects in rheumatoid arthritis [101].\u003c/p\u003e\n\u003cp\u003eIn RA (B-cell-dependent pathogenic mechanisms), the sustained activation of the JAK-STAT pathway plays a crucial regulatory role. Experimental evidence indicates that pro-inflammatory mediators such as IFN-\u0026gamma; and IL-6, by binding to their specific receptors (IFN-\u0026gamma;R and IL-6R), activate members of the JAK kinase family (primarily JAK1/JAK2), leading to the phosphorylation of STAT1. Phosphorylated STAT1 forms homodimers and translocates to the nucleus, where it initiates the transcription program of inflammatory cytokines such as CXCL10 and IRF1, as well as anti-apoptotic molecules like BCL-2, resulting in clonal expansion of B cells and enhanced autoimmune reactivity [102]. It is noteworthy that IFN-\u0026gamma; secreted by such activated B cells can act on adjacent macrophages via paracrine signaling, activating their JAK-STAT signaling cascade. This not only promotes the pro-inflammatory M1 phenotype polarization but also induces excessive release of matrix metalloproteinases (MMPs), thereby exacerbating synovial tissue degradation [103]. Mechanistic studies have revealed that STAT1 signaling significantly enhances T-cell antigen presentation efficiency by upregulating the expression levels of co-stimulatory molecules CD80/CD86 on B-cell surfaces. This aberrant intercellular interaction drives the persistent deterioration of local joint inflammatory responses [104].\u003c/p\u003e\n\u003cp\u003ePTGS2 is a key enzyme in the PGE2 biosynthesis pathway, and its activation leads to an increase in PGE2 production, which is involved in the regulation of inflammation and immune responses. In the acute inflammatory process, the expression of PTGS2 in macrophages increases, promoting the production of prostaglandins, thereby promoting the release of inflammatory factors (TNF-\u0026alpha;, IL-1\u0026beta;) by macrophages, and the inhibition of PTGS2 can reduce the production of these inflammatory factors [105]. The expression of PTGS2 in the synovial tissue of patients with RA is significantly increased, accompanied by the synthesis of prostaglandin E2 and inflammatory factors. This leads to the infiltration of inflammatory cells, abnormal proliferation of synovial tissue, and formation of new blood vessels, thereby exacerbating inflammation and tissue damage. Inhibition of PTGS2 expression in fibroblast-like synovial cells reduces the synthesis of prostaglandin E2 and the inflammatory response in RA [106]. The reduction in PTGS2 expression promotes the activation of Tregs, maintains immune tolerance and an anti-inflammatory environment [107], and significantly inhibits mast cell degranulation, reducing vascular permeability and the expression of inflammatory cytokines [108]. Acupuncture may reduce the production of the proinflammatory prostaglandin PGE2 by inhibiting the activity of PTGS2 and affecting the HPAA, thereby reducing inflammatory responses [33].\u0026nbsp;This provides molecular evidence for its anti-inflammatory mechanism.\u003c/p\u003e\n\u003cp\u003eThe pathological effects of GAPDH are closely related to its metabolic regulatory functions. In an oxidative stress microenvironment, this enzyme enhances aerobic glycolysis in macrophages, driving M1 polarization [109]. In an immune-activated state, macrophages convert the intracellular metabolic enzyme GAPDH into an extracellular signaling molecule. Extracellular GAPDH acts as a ligand that can bind to the CD147 receptor, promoting Th17 differentiation and enhancing glycolysis in CD4\u003csup\u003e+\u003c/sup\u003e T cells [110].Targeted inhibition of GAPDH not only regulates T-cell immune function but also alleviates tissue damage through metabolic reprogramming [111,112]. Acupuncture intervention may improve oxidative stress status by balancing GAPDH activity, thereby restoring immune homeostasis [113].\u003c/p\u003e\n\u003cp\u003eMDM2 is an E3 ubiquitin ligase involved in various cellular processes (cell cycle, apoptosis regulation) and glycolysis. Studies have shown that inhibition of MDM2 activity may affect the activity of inflammatory cells (macrophages and T cells) by inhibiting their survival and death, thereby reducing inflammation [114].MDM2 can exacerbate inflammatory responses by integrating the iNOS-NO and HIF-1\u0026alpha; signaling networks [115], and its overactivation promotes glycolysis in M1 macrophages. Specific inhibition of MDM2 can block pro-inflammatory cell survival signals. MDM2 extends Th17 cell survival by inhibiting p53-dependent apoptosis, while impairing Treg immunosuppressive function through the ubiquitination and degradation of Foxp3. The MDM2-mTORC1 axis promotes glycolytic metabolism, providing energy support for sustained T-cell activation. MDM2 induces the expression of M1 polarization markers (iNOS, IL-12) by activating the NF-\u0026kappa;B pathway [116].\u003c/p\u003e\n\u003cp\u003eMDM2 inhibits B cell apoptosis by degrading p53, thereby promoting the expansion of autoreactive B cell clones [117]. The activity of MDM2 may influence the activity of inflammatory cells by inhibiting the survival and death of various immune cells, such as macrophages and T cells, thereby exacerbating inflammation. Therefore, MDM2-mediated metabolic reprogramming leading to Th17/Treg imbalance could provide new intervention targets for the treatment of rheumatoid arthritis.\u003c/p\u003e\n\u003cp\u003eThese key genes play an anti-inflammatory role in RA by participating in cell signal transduction, regulating the expression of cytokines, affecting cell survival and apoptosis. In addition, the MR analysis of these five genes revealed a potential causal relationship between Increased STAT1 levels and increased risk of RA. Acupuncture can provide valuable insights into the pathogenesis of RA by targeting these genes.\u003c/p\u003e\n\u003cp\u003e4.2.3 Acupuncture may exert anti-inflammatory effects by modulating the JAK2/STAT1 signaling pathway and influencing the function of immune cells\u003c/p\u003e\n\u003cp\u003eAcupuncture exhibits multi-target regulatory characteristics in the treatment of RA, having been proven to reduce the polarization of M1 macrophages in RA synovial tissue, promote the polarization of T cells towards anti-inflammatory cells [118,119], and inhibit pro-inflammatory cytokines (TNF-\u0026alpha;, IL-6, and IL-17) levels to achieve anti-inflammatory and analgesic effects [120]. Acupuncture may improve joint inflammation by correcting the imbalance of Th1/Treg cell activity, characterized by decreased\u0026nbsp;levels of\u0026nbsp;IFN-\u0026gamma;, IL-17, and\u0026nbsp;Increase levels of IL-10, IL-4, and TGF-\u0026beta; [10].\u0026nbsp;Acupuncture intervention can significantly inhibit the polarization process of M1 macrophages in synovial tissue, while promoting the differentiation of T cells into anti-inflammatory subpopulations, and achieving anti-inflammatory and analgesic effects by reducing the expression levels of pro-inflammatory cytokines such as TNF-\u0026alpha;, IL-6, and IL-17 [121].\u0026nbsp;Acupuncture modulates the local cytokine environment at acupoints, including IL-1\u0026beta; and IL-6, activating the anti-inflammatory programs of macrophages and CD4\u003csup\u003e+\u003c/sup\u003e T cells [10].\u0026nbsp;Given the central role of aberrant activation of the JAK/STAT signaling pathway in the pathogenesis of rheumatoid arthritis, this study focuses on the regulatory effects of acupuncture on this pathway and its immunomodulatory mechanisms.\u003c/p\u003e\n\u003cp\u003ePrevious studies have demonstrated that pro-inflammatory cytokines such as IL-6 and IFN-\u0026gamma; can activate JAK2 kinase by binding to their transmembrane receptors (IL-6R/IFN-\u0026gamma;R), leading to the phosphorylation modification of STAT1 protein. Phosphorylated STAT1 is transported into the nucleus via nuclear localization signals, where it binds to the \u0026gamma;-activated sequence (GAS) in the promoter regions of target genes, initiating the expression of inflammation-related genes. This process aids the body in combating pathogens or repairing damage [122].\u0026nbsp;In the JAK2-STAT1 signaling pathway, METTL3-mediated m6A methylation modification enhances the accumulation and activity of STAT1 protein by increasing the stability of STAT1 mRNA, thereby elucidating the molecular basis for sustained activation of this signaling pathway at the epigenetic level [123].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn RA, aberrant activation of the JAK/STAT signaling pathway may lead to excessive production of pro-inflammatory cytokines, activating JAK1 and JAK2, which subsequently promotes the polarization of macrophages towards the M1 phenotype [124].\u0026nbsp;Acupuncture can reduce inflammatory responses in the synovial tissue of RA rats by inhibiting the release of inflammatory factor IL-1 in the serum of RA rats, increasing the levels of immune regulatory factor IL-2, and enhancing the expression of STAT1 and negative regulators of cytokine signaling in the synovial tissue of RA rats [125].\u0026nbsp;This indicates that acupuncture\u0026apos;s inhibition of STAT1 phosphorylation involves the participation of the JAK2/STAT1 signaling pathway in immune cells, enriching our understanding of the JAK2-STAT1 signaling pathway and providing potential targets for developing new treatment strategies for related diseases. Based on the aforementioned evidence, acupuncture may exert its anti-inflammatory therapeutic effects by modulating the JAK2-STAT1 pathway, thereby influencing the ability of CD4\u003csup\u003e+\u003c/sup\u003e T cells and macrophages to secrete pro-inflammatory cytokines such as IL-6 and TNF-\u0026alpha;.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eThis study integrates TCM\u0026apos;s holistic philosophy with AI-driven precision medicine to systematically identify key active components, action targets, and biomarkers of acupuncture intervention in RA. The research found that endogenous active substances (CCK-8, PGE-2, CORT, enkephalin) and DEGs (STAT1, GAPDH, JAK2, PTGS2, MDM2) constitute the core molecular network of acupuncture treatment for RA. Enkephalin and PGE-2 may exert anti-inflammatory effects similar to glucocorticoids, playing a crucial regulatory role in acupuncture therapy, which provides theoretical support for the development of new anti-inflammatory drugs. Notably, differential gene screening results suggest that STAT1, GAPDH, JAK2, PTGS2, and MDM2 may be potential targets for acupuncture intervention in RA. Further analysis indicates that STAT1, as a key regulatory factor, can control the immune homeostasis mediated by M1 macrophages and CD4\u003csup\u003e+\u003c/sup\u003e T cells. The study proposes that acupuncture may regulate the functions of immune cells such as M1 macrophages and CD4\u003csup\u003e+\u003c/sup\u003e T cells by inhibiting the JAK2/STAT1 signaling axis, reshaping the RA immune microenvironment and exerting anti-inflammatory effects.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe innovation of this study lies in constructing a multidimensional target network for acupuncture treatment of RA. It involves screening out 10 characteristic active components using literature mining methods and further identifying five core regulatory genes, including STAT1, through machine learning and Mendelian randomization. Subsequent studies require validation and screening of these substances and molecules through basic experiments. These findings not only deepen the scientific understanding of the anti-inflammatory mechanisms of acupuncture but also provide theoretical basis for developing acupuncture-assisted treatment plans in the era of precision medicine. Future research will focus on: 1) establishing clinical diagnostic indicators based on key targets; 2) developing biologic therapeutic strategies targeting the JAK2/STAT1 pathway; 3) exploring the potential of combining acupuncture with other applications.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"556\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eEnglish abbreviations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eEnglish full name\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eRA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eRheumatoid Arthritis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eGEO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eGene Expression Omnibus\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eEnglish abbreviations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eEnglish full name\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003ePPI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eProtein-Protein Interaction\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eMCODE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eMolecular Complex Detection\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eKEGG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eKyoto Encyclopedia of Genes and Genomes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eGO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eGene Ontology\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eDEGs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eDifferentially Expressed Genes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eROC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eReceiver Operating Characteristic\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eAUC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eArea Under the ROC Curve\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eRF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eRandom Forest\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eSVM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eSupport Vector Machine\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eGLM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eGeneralized Linear Model\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eXGB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eExtreme Gradient Boosting\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eMR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eMendelian Randomization\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eIVW\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eInverse Variance Weighting\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eEnglish abbreviations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eEnglish full name\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eGSEA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eGene Set Enrichment Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eHPAA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eHypothalamic-Pituitary-Adrenal Axis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eJAK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eJanus Kinase\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eSTAT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eSignal Transducer and Activator of Transcription\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eProstaglandin-Endoperoxide Synthase 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eMDM2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eMouse Double Minute 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eCCK-8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eCholecystokinin Octapeptide\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eMENK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eMethionine-enkephalin\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eCORT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eCorticosterone\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003ePGE2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eProstaglandin E 2\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003e\u0026nbsp;AhR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eAryl hydrocarbon receptor\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eMMPs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eMatrix metalloproteinases\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003e5-HTP\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003e5-hydroxytryptophan\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eEnglish abbreviations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eEnglish full name\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eHPAA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eHypothalamic-pituitary-adrenal axis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eHA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eHistamine\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eMHC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eMajor Histocompatibility Complex\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eTregs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eRegulatory T cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eNK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eNatural Killer\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eDCs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eDendritic Cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eIL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eInterleukin\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eTNF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eTumor Necrosis Factor\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eIFN-\u0026gamma;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eInterferon Gamma\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eTh1/Th2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eT-helper 1/T-helper 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eTh17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eT-helper 17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eRANKL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eReceptor Activator of Nuclear Factor-\u0026kappa;B Ligand\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 30.3957%;\"\u003e\n \u003cp\u003eNLRP3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 69.6043%;\"\u003e\n \u003cp\u003eNOD-like Receptor Family Pyrin Domain-containing Protein 3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData will be made available on request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank colleagues and institutions that supported the authors of this study for their contributions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported in part by the National Natural Science Foundation of China (82274649,82004473 and 82205279). Tianjin Natural Science Foundation of China (22JCZXJC00070)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRights and permissions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOpen Access: All the samples used in this study were obtained from public databases, and the sample data were obtained from database (https://www.ncbi.nlm.nih.gov/geo/). Disease Gene Database were obtained from GeneCards (https://www.genecards.org/) and Comparative Toxicogenomics Database (CTD) (https://ctdbase.org/)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eContributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFYL and ZL are joint first authors. YongMing G and YiNan G obtained funding. YX, YG and ZFX designed the study. GongMing Y, JYZ, PYL, RuiW, JH and XL collected the data. FYL and ZL analyzed the data. FYL drafted the manuscript. YiNan G and YX contributed to the interpretation of the results and critical revision of the manuscript for important intellectual content and approved the final version of the manuscript. All authors have read and approved the final manuscript. YongMing G and YG are the study guarantors.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eGlobal, regional, and national burden of other musculoskeletal disorders, 1990-2020, and projections to 2050: a systematic analysis of the Global Burden of Disease Study 2021. \u003cem\u003eThe Lancet. Rheumatology\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, e670-e682 (2023).\u003c/li\u003e\n\u003cli\u003eFarhat, H.\u003cem\u003e, et al.\u003c/em\u003e Increased Risk of Cardiovascular Diseases in Rheumatoid Arthritis: A Systematic Review. \u003cem\u003eCureus\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, e32308 (2022).\u003c/li\u003e\n\u003cli\u003eSokka, T.\u003cem\u003e, et al.\u003c/em\u003e Work disability remains a major problem in rheumatoid arthritis in the 2000s: data from 32 countries in the QUEST-RA study. \u003cem\u003eArthritis research \u0026amp; therapy\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, R42 (2010).\u003c/li\u003e\n\u003cli\u003eChung, I.M., Ketharnathan, S., Thiruvengadam, M. \u0026amp; Rajakumar, G. Rheumatoid Arthritis: The Stride from Research to Clinical Practice. \u003cem\u003eInternational journal of molecular sciences\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e(2016).\u003c/li\u003e\n\u003cli\u003eMazaud, C. \u0026amp; Fardet, L. Relative risk of and determinants for adverse events of methotrexate prescribed at a low dose: a systematic review and meta-analysis of randomized placebo-controlled trials. \u003cem\u003eThe British journal of dermatology\u003c/em\u003e \u003cstrong\u003e177\u003c/strong\u003e, 978-986 (2017).\u003c/li\u003e\n\u003cli\u003eKirwan, J.R. Glucocorticoid resistance in patients with rheumatoid arthritis. \u003cem\u003eScandinavian journal of rheumatology\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 165-166 (2007).\u003c/li\u003e\n\u003cli\u003eLing, S. \u0026amp; Jamali, F. The effect of infliximab on hepatic cytochrome P450 and pharmacokinetics of verapamil in rats with pre-adjuvant arthritis: a drug-disease and drug-drug interaction. \u003cem\u003eBasic \u0026amp; clinical pharmacology \u0026amp; toxicology\u003c/em\u003e \u003cstrong\u003e105\u003c/strong\u003e, 24-29 (2009).\u003c/li\u003e\n\u003cli\u003eGonz\u0026aacute;lez, C.M.\u003cem\u003e, et al.\u003c/em\u003e Perceptions of patients with rheumatic diseases on the impact on daily life and satisfaction with their medications: RHEU-LIFE, a survey to patients treated with subcutaneous biological products. \u003cem\u003ePatient preference and adherence\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1243-1252 (2017).\u003c/li\u003e\n\u003cli\u003eSunil, D. \u0026amp; Kamath, P.R. Multi-Target Directed Indole Based Hybrid Molecules in Cancer Therapy : An Up-To-Date Evidence-Based Review. \u003cem\u003eCurrent topics in medicinal chemistry\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 959-985 (2017).\u003c/li\u003e\n\u003cli\u003eLi, N.\u003cem\u003e, et al.\u003c/em\u003e The Anti-Inflammatory Actions and Mechanisms of Acupuncture from Acupoint to Target Organs via Neuro-Immune Regulation. \u003cem\u003eJournal of inflammation research\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 7191-7224 (2021).\u003c/li\u003e\n\u003cli\u003eWang, Y.\u003cem\u003e, et al.\u003c/em\u003e Effect of Moxibustion on \u0026beta;-EP and Dyn Levels of Pain-Related Indicators in Patients with Rheumatoid Arthritis. \u003cem\u003eEvidence-based complementary and alternative medicine : eCAM\u003c/em\u003e \u003cstrong\u003e2021\u003c/strong\u003e, 6637554 (2021).\u003c/li\u003e\n\u003cli\u003eYang, F.\u003cem\u003e, et al.\u003c/em\u003e ST36 Acupuncture Alleviates the Inflammation of Adjuvant-Induced Arthritic Rats by Targeting Monocyte/Macrophage Modulation. \u003cem\u003eEvidence-based complementary and alternative medicine : eCAM\u003c/em\u003e \u003cstrong\u003e2021\u003c/strong\u003e, 9430501 (2021).\u003c/li\u003e\n\u003cli\u003eWooller, S.K., Benstead-Hume, G., Chen, X., Ali, Y. \u0026amp; Pearl, F.M.G. Bioinformatics in translational drug discovery. \u003cem\u003eBioscience reports\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e(2017).\u003c/li\u003e\n\u003cli\u003eYe, H., Wei, J., Tang, K., Feuers, R. \u0026amp; Hong, H. Drug Repositioning Through Network Pharmacology. \u003cem\u003eCurrent topics in medicinal chemistry\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 3646-3656 (2016).\u003c/li\u003e\n\u003cli\u003eBoezio, B., Audouze, K., Ducrot, P. \u0026amp; Taboureau, O.J.M.i. Network‐based approaches in pharmacology. \u003cstrong\u003e36\u003c/strong\u003e, 1700048 (2017).\u003c/li\u003e\n\u003cli\u003eFan, A.Y. Anti-inflammatory mechanism of electroacupuncture involves the modulation of multiple systems, levels and targets and is not limited to \u0026quot;driving the vagus-adrenal axis\u0026quot;. \u003cem\u003eJournal of integrative medicine\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 320-323 (2023).\u003c/li\u003e\n\u003cli\u003eHan, Z., Zhang, Y., Wang, P., Tang, Q. \u0026amp; Zhang, K. Is acupuncture effective in the treatment of COVID-19 related symptoms? Based on bioinformatics/network topology strategy. \u003cem\u003eBriefings in bioinformatics\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e(2021).\u003c/li\u003e\n\u003cli\u003eWang, X.\u003cem\u003e, et al.\u003c/em\u003e PharmMapper 2017 update: a web server for potential drug target identification with a comprehensive target pharmacophore database. \u003cem\u003eNucleic acids research\u003c/em\u003e \u003cstrong\u003e45\u003c/strong\u003e, W356-w360 (2017).\u003c/li\u003e\n\u003cli\u003eDaina, A., Michielin, O. \u0026amp; Zoete, V. SwissTargetPrediction: updated data and new features for efficient prediction of protein targets of small molecules. \u003cem\u003eNucleic acids research\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, W357-w364 (2019).\u003c/li\u003e\n\u003cli\u003eSafran, M.\u003cem\u003e, et al.\u003c/em\u003e The genecards suite. 27-56 (2021).\u003c/li\u003e\n\u003cli\u003eWyatt, B.\u003cem\u003e, et al.\u003c/em\u003e Transforming environmental health datasets from the comparative toxicogenomics database into chord diagrams to visualize molecular mechanisms. \u003cem\u003eFrontiers in toxicology\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 1437884 (2024).\u003c/li\u003e\n\u003cli\u003eSzklarczyk, D.\u003cem\u003e, et al.\u003c/em\u003e The STRING database in 2021: customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets. \u003cem\u003eNucleic acids research\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, D605-d612 (2021).\u003c/li\u003e\n\u003cli\u003eRigatti, S.J. Random Forest. \u003cem\u003eJournal of insurance medicine (New York, N.Y.)\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, 31-39 (2017).\u003c/li\u003e\n\u003cli\u003eGold, C. \u0026amp; Sollich, P. Model selection for support vector machine classification. \u003cem\u003eNeurocomputing\u003c/em\u003e \u003cstrong\u003e55\u003c/strong\u003e, 221-249 (2003).\u003c/li\u003e\n\u003cli\u003eNelder, J.A. \u0026amp; Wedderburn, R.W.M. Generalized Linear Models. \u003cem\u003eRoyal Statistical Society. Journal. Series A: General\u003c/em\u003e \u003cstrong\u003e135\u003c/strong\u003e, 370-384 (1972).\u003c/li\u003e\n\u003cli\u003eChen, T.J.R.p.v.-. Xgboost: extreme gradient boosting. \u003cstrong\u003e1\u003c/strong\u003e(2015).\u003c/li\u003e\n\u003cli\u003eHiggins, J.P., Thompson, S.G., Deeks, J.J. \u0026amp; Altman, D.G. Measuring inconsistency in meta-analyses. \u003cem\u003eBMJ (Clinical research ed.)\u003c/em\u003e \u003cstrong\u003e327\u003c/strong\u003e, 557-560 (2003).\u003c/li\u003e\n\u003cli\u003eBurgess, S. \u0026amp; Thompson, S.G. Interpreting findings from Mendelian randomization using the MR-Egger method. \u003cem\u003eEuropean journal of epidemiology\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 377-389 (2017).\u003c/li\u003e\n\u003cli\u003eZhang, B.\u003cem\u003e, et al.\u003c/em\u003e m(6)A regulator-mediated methylation modification patterns and tumor microenvironment infiltration characterization in gastric cancer. \u003cem\u003eMolecular cancer\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 53 (2020).\u003c/li\u003e\n\u003cli\u003eGiambartolomei, C.\u003cem\u003e, et al.\u003c/em\u003e Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. \u003cstrong\u003e10\u003c/strong\u003e, e1004383 (2014).\u003c/li\u003e\n\u003cli\u003eMcInnes, I.B. \u0026amp; Schett, G. The pathogenesis of rheumatoid arthritis. \u003cem\u003eThe New England journal of medicine\u003c/em\u003e \u003cstrong\u003e365\u003c/strong\u003e, 2205-2219 (2011).\u003c/li\u003e\n\u003cli\u003eDu, B., Zhu, M., Li, Y., Li, G. \u0026amp; Xi, X. The prostaglandin E2 increases the production of IL-17 and the expression of costimulatory molecules on \u0026gamma;\u0026delta; T cells in rheumatoid arthritis. \u003cem\u003eScandinavian journal of immunology\u003c/em\u003e \u003cstrong\u003e91\u003c/strong\u003e, e12872 (2020).\u003c/li\u003e\n\u003cli\u003eXu, Z.-F.\u003cem\u003e, et al.\u003c/em\u003e Neuroendocrine-immune regulating mechanisms for the anti-inflammatory and analgesic actions of acupuncture. \u003cstrong\u003e6\u003c/strong\u003e, 384-392 (2020).\u003c/li\u003e\n\u003cli\u003eZhang, R., Lao, L., Ren, K. \u0026amp; Berman, B.M.J.A. Mechanisms of acupuncture\u0026ndash;electroacupuncture on persistent pain. \u003cstrong\u003e120\u003c/strong\u003e, 482-503 (2014).\u003c/li\u003e\n\u003cli\u003eZhang, R.-X.\u003cem\u003e, et al.\u003c/em\u003e Electroacupuncture attenuates inflammation in a rat model. \u003cstrong\u003e11\u003c/strong\u003e, 135-142 (2005).\u003c/li\u003e\n\u003cli\u003eWei, Y.\u003cem\u003e, et al.\u003c/em\u003e Regulation of hypothalamic-pituitary-adrenal axis activity and immunologic function contributed to the anti-inflammatory effect of acupuncture in the OVA-induced murine asthma model. \u003cstrong\u003e636\u003c/strong\u003e, 177-183 (2017).\u003c/li\u003e\n\u003cli\u003eYing, Z.-H., Mao, C.-L., Xie, W. \u0026amp; Yu, C.-H.J.F.i.M. Postbiotics in rheumatoid arthritis: Emerging mechanisms and intervention perspectives. \u003cstrong\u003e14\u003c/strong\u003e, 1290015 (2023).\u003c/li\u003e\n\u003cli\u003eLee, E.J. \u0026amp; Warden, S.J.E.J.o.I.M. The effects of acupuncture on serotonin metabolism. \u003cstrong\u003e8\u003c/strong\u003e, 355-367 (2016).\u003c/li\u003e\n\u003cli\u003eDe la Fuente, M., Medina, S., Del Rio, M., Ferr\u0026aacute;ndez, M.D. \u0026amp; Hernanz, A. Effect of aging on the modulation of macrophage functions by neuropeptides. \u003cem\u003eLife Sciences\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e, 2125-2135 (2000).\u003c/li\u003e\n\u003cli\u003eTrejter, M.\u003cem\u003e, et al.\u003c/em\u003e Studies on the involvement Histology and Histopathology Cellular and Molecular Biology of endogenous neuropeptides in the control of thymocyte proliferation in the rat. (2001).\u003c/li\u003e\n\u003cli\u003eLee, E.-G.\u003cem\u003e, et al.\u003c/em\u003e Adrenomedullin inhibits IL-1\u0026beta;-induced rheumatoid synovial fibroblast proliferation and MMPs, COX-2 and PGE2 production. \u003cstrong\u003e34\u003c/strong\u003e, 335-343 (2011).\u003c/li\u003e\n\u003cli\u003eZhou, Y., Sun, Y.-H., Shen, J.-M. \u0026amp; Han, J.-S.J.N. Increased release of immunoreactive CCK-8 by electroacupuncture and enhancement of electroacupuncture analgesia by CCK-B antagonist in rat spinal cord. \u003cstrong\u003e24\u003c/strong\u003e, 139-144 (1993).\u003c/li\u003e\n\u003cli\u003eHan, J., Ding, X. \u0026amp; Fan, S.J.P. Cholecystokinin octapeptide (CCK-8): antagonism to electroacupuncture analgesia and a possible role in electroacupuncture tolerance. \u003cstrong\u003e27\u003c/strong\u003e, 101-115 (1986).\u003c/li\u003e\n\u003cli\u003eJie, W.\u003cem\u003e, et al.\u003c/em\u003e Analgesic effect of buccal acupuncture on acute arthritis in rabbits and underlying mechanisms. \u003cstrong\u003e42\u003c/strong\u003e, 517-521 (2017).\u003c/li\u003e\n\u003cli\u003eZhao, Z.-Q.J.P.i.n. Neural mechanism underlying acupuncture analgesia. \u003cstrong\u003e85\u003c/strong\u003e, 355-375 (2008).\u003c/li\u003e\n\u003cli\u003eCheng, L.-L.\u003cem\u003e, et al.\u003c/em\u003e Effects of electroacupuncture of different frequencies on the release profile of endogenous opioid peptides in the central nerve system of goats. \u003cstrong\u003e2012\u003c/strong\u003e, 476457 (2012).\u003c/li\u003e\n\u003cli\u003eZheng, X., Lin, J., Wang, Z., Zeng, Z. \u0026amp; Chen, H.J.H. Research of the analgesic effects and central nervous system impact of electroacupuncture therapy in rats with knee osteoarthritis. \u003cstrong\u003e10\u003c/strong\u003e(2024).\u003c/li\u003e\n\u003cli\u003eFang, J.\u003cem\u003e, et al.\u003c/em\u003e Involvement of peripheral beta-endorphin and MU, delta, kappa opioid receptors in electro acupuncture analgesia for prolonged inflammatory pain of rats. \u003cstrong\u003e11\u003c/strong\u003e, 375-383 (2013).\u003c/li\u003e\n\u003cli\u003eLiu, J., Dong, S., Liu, S.J.A. \u0026amp; Medicine, H. Aberrant parasympathetic responses in acupuncture therapy for restoring immune homeostasis. \u003cstrong\u003e3\u003c/strong\u003e, 69-75 (2023).\u003c/li\u003e\n\u003cli\u003eChen, W.\u003cem\u003e, et al.\u003c/em\u003e Electroacupuncture activated local sympathetic noradrenergic signaling to relieve synovitis and referred pain behaviors in knee osteoarthritis rats, Front. Mol. Neurosci. 16 (2023) 1069965. (Epub 2023/03/25. doi: 10.3389/fnmol, 2023).\u003c/li\u003e\n\u003cli\u003eTorres-Rosas, R.\u003cem\u003e, et al.\u003c/em\u003e Dopamine mediates vagal modulation of the immune system by electroacupuncture. \u003cstrong\u003e20\u003c/strong\u003e, 291-295 (2014).\u003c/li\u003e\n\u003cli\u003eYan, Y.\u003cem\u003e, et al.\u003c/em\u003e Dopamine controls systemic inflammation through inhibition of NLRP3 inflammasome. \u003cstrong\u003e160\u003c/strong\u003e, 62-73 (2015).\u003c/li\u003e\n\u003cli\u003eShen, Y., Liu, F., Zhang, M.J.B. \u0026amp; Pharmacotherapy. Therapeutic potential of plant-derived natural compounds in Alzheimer\u0026rsquo;s disease: Targeting microglia-mediated neuroinflammation. \u003cstrong\u003e178\u003c/strong\u003e, 117235 (2024).\u003c/li\u003e\n\u003cli\u003eMa, X.\u003cem\u003e, et al.\u003c/em\u003e Potential mechanisms of acupuncture for neuropathic pain based on somatosensory system. \u003cstrong\u003e16\u003c/strong\u003e, 940343 (2022).\u003c/li\u003e\n\u003cli\u003eWang, M., Liu, W., Ge, J. \u0026amp; Liu, S.J.F.i.i. The immunomodulatory mechanisms for acupuncture practice. \u003cstrong\u003e14\u003c/strong\u003e, 1147718 (2023).\u003c/li\u003e\n\u003cli\u003eZhang, Y.\u003cem\u003e, et al.\u003c/em\u003e Pathological pathway analysis in an experimental rheumatoid arthritis model and the tissue repair effect of acupuncture at ST36. \u003cstrong\u003e14\u003c/strong\u003e, 1164157 (2023).\u003c/li\u003e\n\u003cli\u003eWang, J.\u003cem\u003e, et al.\u003c/em\u003e Therapeutic effect and mechanism of acupuncture in autoimmune diseases. \u003cstrong\u003e50\u003c/strong\u003e, 639-652 (2022).\u003c/li\u003e\n\u003cli\u003eJang, S., Kwon, E.-J. \u0026amp; Lee, J.J.J.I.j.o.m.s. Rheumatoid arthritis: pathogenic roles of diverse immune cells. \u003cstrong\u003e23\u003c/strong\u003e, 905 (2022).\u003c/li\u003e\n\u003cli\u003eHuang, Y.\u003cem\u003e, et al.\u003c/em\u003e Identification of diagnostic genes and drug prediction in metabolic syndrome-associated rheumatoid arthritis by integrated bioinformatics analysis, machine learning, and molecular docking. \u003cstrong\u003e15\u003c/strong\u003e, 1431452 (2024).\u003c/li\u003e\n\u003cli\u003eZhou, J.\u003cem\u003e, et al.\u003c/em\u003e Identification of aging-related biomarkers and immune infiltration characteristics in osteoarthritis based on bioinformatics analysis and machine learning. \u003cstrong\u003e14\u003c/strong\u003e, 1168780 (2023).\u003c/li\u003e\n\u003cli\u003eZhang, J.-G.\u003cem\u003e, et al.\u003c/em\u003e Cholecystokinin octapeptide regulates lipopolysaccharide-activated B cells co-stimulatory molecule expression and cytokines production in vitro. \u003cstrong\u003e33\u003c/strong\u003e, 157-163 (2011).\u003c/li\u003e\n\u003cli\u003eZhang, J.-G.\u003cem\u003e, et al.\u003c/em\u003e Cholecystokinin octapeptide regulates the differentiation and effector cytokine production of CD4+ T cells in vitro. \u003cstrong\u003e20\u003c/strong\u003e, 307-315 (2014).\u003c/li\u003e\n\u003cli\u003eCrawley, J.N. \u0026amp; Corwin, R.L. Biological actions of cholecystokinin. \u003cem\u003ePeptides\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 731-755 (1994).\u003c/li\u003e\n\u003cli\u003eZhang, J.G.\u003cem\u003e, et al.\u003c/em\u003e Cholecystokinin octapeptide inhibits immunoglobulin G1 production of lipopolysaccharide-activated B cells. \u003cem\u003eInternational immunopharmacology\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1685-1690 (2011).\u003c/li\u003e\n\u003cli\u003eLi, Q.\u003cem\u003e, et al.\u003c/em\u003e Cholecystokinin octapeptide significantly suppresses collagen-induced arthritis in mice by inhibiting Th17 polarization primed by dendritic cells. \u003cem\u003eCellular immunology\u003c/em\u003e \u003cstrong\u003e272\u003c/strong\u003e, 53-60 (2011).\u003c/li\u003e\n\u003cli\u003eLi, X.\u003cem\u003e, et al.\u003c/em\u003e Methionine enkephalin (MENK) inhibits tumor growth through regulating CD4+ Foxp3+ regulatory T cells (Tregs) in mice. \u003cstrong\u003e16\u003c/strong\u003e, 450-459 (2015).\u003c/li\u003e\n\u003cli\u003eShan, F.\u003cem\u003e, et al.\u003c/em\u003e Functional modulation of the pathway between dendritic cells (DCs) and CD4+ T cells by the neuropeptide: methionine enkephalin (MENK). \u003cstrong\u003e32\u003c/strong\u003e, 929-937 (2011).\u003c/li\u003e\n\u003cli\u003eXing, J., Xia, M., Wang, T. \u0026amp; Mu, J.J.Z.c.y.j.A.R. Study on the analgesic effect of acupuncture with opioid receptors agonist in induced arthritic rats. \u003cstrong\u003e14\u003c/strong\u003e, 375-378 (1989).\u003c/li\u003e\n\u003cli\u003eTsuge, K., Inazumi, T., Shimamoto, A. \u0026amp; Sugimoto, Y. Molecular mechanisms underlying prostaglandin E2-exacerbated inflammation and immune diseases. \u003cem\u003eInternational immunology\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 597-606 (2019).\u003c/li\u003e\n\u003cli\u003eSreeramkumar, V., Fresno, M., Cuesta, N.J.I. \u0026amp; biology, c. Prostaglandin E2 and T cells: friends or foes? \u003cstrong\u003e90\u003c/strong\u003e, 579-586 (2012).\u003c/li\u003e\n\u003cli\u003eQiao, L.-n.\u003cem\u003e, et al.\u003c/em\u003e Effect of Electroacupuncture Intervention on Expression of CGRP, SP, COX‐1, and PGE2 of Dorsal Portion of the Cervical Spinal Cord in Rats with Neck‐Incision Pain. \u003cstrong\u003e2013\u003c/strong\u003e, 294091 (2013).\u003c/li\u003e\n\u003cli\u003eJIANG, J.\u003cem\u003e, et al.\u003c/em\u003e Efficacy of electroacupuncture stimulating Zusanli (ST36) and Xuanzhong (GB39) on synovial angiogenesis in rats with adjuvant arthritis. \u003cstrong\u003e43\u003c/strong\u003e, 955 (2023).\u003c/li\u003e\n\u003cli\u003eAo, Y., Wang, Z., Hu, J., Yao, M. \u0026amp; Zhang, W.J.S.R. Identification of essential genes and immune cell infiltration in rheumatoid arthritis by bioinformatics analysis. \u003cstrong\u003e13\u003c/strong\u003e, 2032 (2023).\u003c/li\u003e\n\u003cli\u003eZhou, S., Lu, H. \u0026amp; Xiong, M.J.F.i.i. Identifying immune cell infiltration and effective diagnostic biomarkers in rheumatoid arthritis by bioinformatics analysis. \u003cstrong\u003e12\u003c/strong\u003e, 726747 (2021).\u003c/li\u003e\n\u003cli\u003eBoutet, M.-A.\u003cem\u003e, et al.\u003c/em\u003e Novel insights into macrophage diversity in rheumatoid arthritis synovium. \u003cstrong\u003e20\u003c/strong\u003e, 102758 (2021).\u003c/li\u003e\n\u003cli\u003eGao, Y.\u003cem\u003e, et al.\u003c/em\u003e Immunosenescence of T cells: a key player in rheumatoid arthritis. \u003cem\u003eInflammation research : official journal of the European Histamine Research Society ... [et al.]\u003c/em\u003e \u003cstrong\u003e71\u003c/strong\u003e, 1449-1462 (2022).\u003c/li\u003e\n\u003cli\u003eTang, M., Tian, L., Luo, G. \u0026amp; Yu, X.J.F.i.i. Interferon-gamma-mediated osteoimmunology. \u003cstrong\u003e9\u003c/strong\u003e, 1508 (2018).\u003c/li\u003e\n\u003cli\u003eRoberts, C.A., Dickinson, A.K. \u0026amp; Taams, L.S.J.F.i.i. The interplay between monocytes/macrophages and CD4+ T cell subsets in rheumatoid arthritis. \u003cstrong\u003e6\u003c/strong\u003e, 571 (2015).\u003c/li\u003e\n\u003cli\u003eKondo, Y.\u003cem\u003e, et al.\u003c/em\u003e Transcriptional regulation of CD 4+ T cell differentiation in experimentally induced arthritis and rheumatoid arthritis. \u003cstrong\u003e70\u003c/strong\u003e, 653-661 (2018).\u003c/li\u003e\n\u003cli\u003eSun, W.\u003cem\u003e, et al.\u003c/em\u003e B cells inhibit bone formation in rheumatoid arthritis by suppressing osteoblast differentiation. \u003cstrong\u003e9\u003c/strong\u003e, 5127 (2018).\u003c/li\u003e\n\u003cli\u003eWu, F.\u003cem\u003e, et al.\u003c/em\u003e B cells in rheumatoid arthritis： pathogenic mechanisms and treatment prospects. \u003cstrong\u003e12\u003c/strong\u003e, 750753 (2021).\u003c/li\u003e\n\u003cli\u003eKucuksezer, U.C.\u003cem\u003e, et al.\u003c/em\u003e The role of natural killer cells in autoimmune diseases. \u003cstrong\u003e12\u003c/strong\u003e, 622306 (2021).\u003c/li\u003e\n\u003cli\u003eLiu, F.\u003cem\u003e, et al.\u003c/em\u003e Acupuncture and its ability to restore and maintain immune homeostasis. \u003cstrong\u003e117\u003c/strong\u003e, 167-176 (2024).\u003c/li\u003e\n\u003cli\u003eLei, Y.\u003cem\u003e, et al.\u003c/em\u003e Synovial microenvironment-influenced mast cells promote the progression of rheumatoid arthritis. \u003cstrong\u003e15\u003c/strong\u003e, 113 (2024).\u003c/li\u003e\n\u003cli\u003eLoucks, A.\u003cem\u003e, et al.\u003c/em\u003e The multifaceted role of mast cells in joint inflammation and arthritis. \u003cstrong\u003e31\u003c/strong\u003e, 567-575 (2023).\u003c/li\u003e\n\u003cli\u003eXue, C.\u003cem\u003e, et al.\u003c/em\u003e Evolving cognition of the JAK-STAT signaling pathway: autoimmune disorders and cancer. \u003cstrong\u003e8\u003c/strong\u003e, 204 (2023).\u003c/li\u003e\n\u003cli\u003eXia, X.\u003cem\u003e, et al.\u003c/em\u003e Single cell immunoprofile of synovial fluid in rheumatoid arthritis with TNF/JAK inhibitor treatment. \u003cem\u003eNature communications\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 2152 (2025).\u003c/li\u003e\n\u003cli\u003eQin, Y.\u003cem\u003e, et al.\u003c/em\u003e Age-associated B cells contribute to the pathogenesis of rheumatoid arthritis by inducing activation of fibroblast-like synoviocytes via TNF-\u0026alpha;-mediated ERK1/2 and JAK-STAT1 pathways. \u003cstrong\u003e81\u003c/strong\u003e, 1504-1514 (2022).\u003c/li\u003e\n\u003cli\u003eKasperkovitz, P.\u003cem\u003e, et al.\u003c/em\u003e Activation of the STAT1 pathway in rheumatoid arthritis. \u003cstrong\u003e63\u003c/strong\u003e, 233-239 (2004).\u003c/li\u003e\n\u003cli\u003eMonari, C.\u003cem\u003e, et al.\u003c/em\u003e A microbial polysaccharide reduces the severity of rheumatoid arthritis by influencing Th17 differentiation and proinflammatory cytokines production. \u003cstrong\u003e183\u003c/strong\u003e, 191-200 (2009).\u003c/li\u003e\n\u003cli\u003ede Hooge, A.S.\u003cem\u003e, et al.\u003c/em\u003e Local activation of STAT‐1 and STAT‐3 in the inflamed synovium during zymosan‐induced arthritis: exacerbation of joint inflammation in STAT‐1 gene\u0026ndash;knockout mice. \u003cstrong\u003e50\u003c/strong\u003e, 2014-2023 (2004).\u003c/li\u003e\n\u003cli\u003eZhang, M., Xu, M., Wang, K., Li, L. \u0026amp; Zhao, J. Effect of Inhibition of the JAK2/STAT3 Signaling Pathway on the Th17/IL-17 Axis in Acute Cellular Rejection After Heart Transplantation in Mice. \u003cem\u003eJournal of cardiovascular pharmacology\u003c/em\u003e \u003cstrong\u003e77\u003c/strong\u003e, 614-620 (2021).\u003c/li\u003e\n\u003cli\u003eLv, Y.\u003cem\u003e, et al.\u003c/em\u003e The JAK-STAT pathway: from structural biology to cytokine engineering. \u003cem\u003eSignal transduction and targeted therapy\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 221 (2024).\u003c/li\u003e\n\u003cli\u003eChoi, Y.S., Eto, D., Yang, J.A., Lao, C. \u0026amp; Crotty, S. Cutting edge: STAT1 is required for IL-6-mediated Bcl6 induction for early follicular helper cell differentiation. \u003cem\u003eJournal of immunology (Baltimore, Md. : 1950)\u003c/em\u003e \u003cstrong\u003e190\u003c/strong\u003e, 3049-3053 (2013).\u003c/li\u003e\n\u003cli\u003eChen, R.\u003cem\u003e, et al.\u003c/em\u003e Augmented PFKFB3-mediated glycolysis by interferon-\u0026gamma; promotes inflammatory M1 polarization through the JAK2/STAT1 pathway in local vascular inflammation in Takayasu arteritis. \u003cem\u003eArthritis research \u0026amp; therapy\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 266 (2022).\u003c/li\u003e\n\u003cli\u003eCiobanu, D.A.\u003cem\u003e, et al.\u003c/em\u003e JAK/STAT pathway in pathology of rheumatoid arthritis (Review). \u003cem\u003eExperimental and therapeutic medicine\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 3498-3503 (2020).\u003c/li\u003e\n\u003cli\u003eOwen, K.L., Brockwell, N.K. \u0026amp; Parker, B.S. JAK-STAT Signaling: A Double-Edged Sword of Immune Regulation and Cancer Progression. \u003cem\u003eCancers\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e(2019).\u003c/li\u003e\n\u003cli\u003eYang, X.\u003cem\u003e, et al.\u003c/em\u003e Cell volume regulation modulates macrophage-related inflammatory responses via JAK/STAT signaling pathways. \u003cem\u003eActa biomaterialia\u003c/em\u003e \u003cstrong\u003e186\u003c/strong\u003e, 286-299 (2024).\u003c/li\u003e\n\u003cli\u003eSarapultsev, A.\u003cem\u003e, et al.\u003c/em\u003e JAK-STAT signaling in inflammation and stress-related diseases: implications for therapeutic interventions. \u003cstrong\u003e4\u003c/strong\u003e, 40 (2023).\u003c/li\u003e\n\u003cli\u003eLiang, Y.B.\u003cem\u003e, et al.\u003c/em\u003e Downregulated SOCS1 expression activates the JAK1/STAT1 pathway and promotes polarization of macrophages into M1 type. \u003cem\u003eMolecular medicine reports\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 6405-6411 (2017).\u003c/li\u003e\n\u003cli\u003eSuda, Y.\u003cem\u003e, et al.\u003c/em\u003e Comparison of anti-inflammatory and anti-angiogenic effects of JAK inhibitors in IL-6 and TNF\u0026alpha;-stimulated fibroblast-like synoviocytes derived from patients with RA. \u003cem\u003eScientific reports\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 9736 (2025).\u003c/li\u003e\n\u003cli\u003eSantos, C.I. \u0026amp; Costa-Pereira, A.P. Signal transducers and activators of transcription-from cytokine signalling to cancer biology. \u003cem\u003eBiochimica et biophysica acta\u003c/em\u003e \u003cstrong\u003e1816\u003c/strong\u003e, 38-49 (2011).\u003c/li\u003e\n\u003cli\u003eYin, X.\u003cem\u003e, et al.\u003c/em\u003e Research progress on macrophage polarization during osteoarthritis disease progression: a review. \u003cem\u003eJournal of orthopaedic surgery and research\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 584 (2024).\u003c/li\u003e\n\u003cli\u003eElbrashy, M.M., Metwally, H., Sakakibara, S. \u0026amp; Kishimoto, T. Threonine Phosphorylation and the Yin and Yang of STAT1: Phosphorylation-Dependent Spectrum of STAT1 Functionality in Inflammatory Contexts. \u003cem\u003eCells\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e(2024).\u003c/li\u003e\n\u003cli\u003eMu\u0026ntilde;oz, A., Costa, M.J.O.m. \u0026amp; longevity, c. Nutritionally mediated oxidative stress and inflammation. \u003cstrong\u003e2013\u003c/strong\u003e, 610950 (2013).\u003c/li\u003e\n\u003cli\u003eTang, M.\u003cem\u003e, et al.\u003c/em\u003e Pharmacological aspects of natural quercetin in rheumatoid arthritis. 2043-2053 (2023).\u003c/li\u003e\n\u003cli\u003eZhao, M., Burisch, J.J.D.d. \u0026amp; sciences. Impact of genes and the environment on the pathogenesis and disease course of inflammatory bowel disease. \u003cstrong\u003e64\u003c/strong\u003e, 1759-1769 (2019).\u003c/li\u003e\n\u003cli\u003eChen, Y.\u003cem\u003e, et al.\u003c/em\u003e PTGS2: A potential immune regulator and therapeutic target for chronic spontaneous urticaria. \u003cstrong\u003e344\u003c/strong\u003e, 122582 (2024).\u003c/li\u003e\n\u003cli\u003eChen, P.C.\u003cem\u003e, et al.\u003c/em\u003e Moonlighting glyceraldehyde-3-phosphate dehydrogenase (GAPDH) protein of Lactobacillus gasseri attenuates allergic asthma via immunometabolic change in macrophages. \u003cem\u003eJournal of biomedical science\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 75 (2022).\u003c/li\u003e\n\u003cli\u003eYang, H.\u003cem\u003e, et al.\u003c/em\u003e CD147 modulates the differentiation of T-helper 17 cells in patients with rheumatoid arthritis. \u003cem\u003eAPMIS : acta pathologica, microbiologica, et immunologica Scandinavica\u003c/em\u003e \u003cstrong\u003e125\u003c/strong\u003e, 24-31 (2017).\u003c/li\u003e\n\u003cli\u003eCui, Z.\u003cem\u003e, et al.\u003c/em\u003e MYO1F regulates T-cell activation and glycolytic metabolism by promoting the acetylation of GAPDH. \u003cem\u003eCellular \u0026amp; molecular immunology\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 176-190 (2025).\u003c/li\u003e\n\u003cli\u003eKornberg, M.D.\u003cem\u003e, et al.\u003c/em\u003e Dimethyl fumarate targets GAPDH and aerobic glycolysis to modulate immunity. \u003cem\u003eScience (New York, N.Y.)\u003c/em\u003e \u003cstrong\u003e360\u003c/strong\u003e, 449-453 (2018).\u003c/li\u003e\n\u003cli\u003eGuo, B.J., Sun, J.H. \u0026amp; Pei, L.X. Research progress on mechanisms of acupuncture and moxibustion underlying improvement of oxidative stress. \u003cem\u003eZhen ci yan jiu = Acupuncture research\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, 307-314 (2024).\u003c/li\u003e\n\u003cli\u003eThomasova, D., Mulay, S.R., Bruns, H. \u0026amp; Anders, H.-J.J.N. p53-independent roles of MDM2 in NF-\u0026kappa;B signaling: implications for cancer therapy, wound healing, and autoimmune diseases. \u003cstrong\u003e14\u003c/strong\u003e, 1097-1101 (2012).\u003c/li\u003e\n\u003cli\u003eWu, K.K.-l.\u003cem\u003e, et al.\u003c/em\u003e MDM2 induces pro-inflammatory and glycolytic responses in M1 macrophages by integrating iNOS-nitric oxide and HIF-1\u0026alpha; pathways in mice. \u003cstrong\u003e15\u003c/strong\u003e, 8624 (2024).\u003c/li\u003e\n\u003cli\u003eWu, K.K.\u003cem\u003e, et al.\u003c/em\u003e MDM2 induces pro-inflammatory and glycolytic responses in M1 macrophages by integrating iNOS-nitric oxide and HIF-1\u0026alpha; pathways in mice. \u003cem\u003eNature communications\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 8624 (2024).\u003c/li\u003e\n\u003cli\u003eLi, Z.\u003cem\u003e, et al.\u003c/em\u003e Functions and mechanisms of non-histone post-translational modifications in cancer progression. \u003cem\u003eCell death discovery\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 125 (2025).\u003c/li\u003e\n\u003cli\u003eYu, N.\u003cem\u003e, et al.\u003c/em\u003e Manual acupuncture at ST36 attenuates rheumatoid arthritis by inhibiting M1 macrophage polarization and enhancing Treg cell populations in adjuvant-induced arthritic rats. \u003cstrong\u003e41\u003c/strong\u003e, 96-109 (2023).\u003c/li\u003e\n\u003cli\u003eZhu, J.\u003cem\u003e, et al.\u003c/em\u003e Electroacupuncture attenuates collagen-induced arthritis in rats through vasoactive intestinal peptide signalling-dependent re-establishment of the regulatory T cell/T-helper 17 cell balance. \u003cstrong\u003e33\u003c/strong\u003e, 305-311 (2015).\u003c/li\u003e\n\u003cli\u003eYang, F.\u003cem\u003e, et al.\u003c/em\u003e ST36 Acupuncture Alleviates the Inflammation of Adjuvant‐Induced Arthritic Rats by Targeting Monocyte/Macrophage Modulation. \u003cstrong\u003e2021\u003c/strong\u003e, 9430501 (2021).\u003c/li\u003e\n\u003cli\u003eOh, J.E. \u0026amp; Kim, S.N. Anti-Inflammatory Effects of Acupuncture at ST36 Point: A Literature Review in Animal Studies. \u003cem\u003eFrontiers in immunology\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 813748 (2021).\u003c/li\u003e\n\u003cli\u003eJerke, U.\u003cem\u003e, et al.\u003c/em\u003e Stat1 nuclear translocation by nucleolin upon monocyte differentiation. \u003cem\u003ePloS one\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, e8302 (2009).\u003c/li\u003e\n\u003cli\u003eFang, W.\u003cem\u003e, et al.\u003c/em\u003e m6A methylation modification and immune infiltration analysis in osteonecrosis of the femoral head. \u003cem\u003eJournal of orthopaedic surgery and research\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 183 (2024).\u003c/li\u003e\n\u003cli\u003eIvashkiv, L.B.J.N.R.I. IFN\u0026gamma;: signalling, epigenetics and roles in immunity, metabolism, disease and cancer immunotherapy. \u003cstrong\u003e18\u003c/strong\u003e, 545-558 (2018).\u003c/li\u003e\n\u003cli\u003eHao, F.\u003cem\u003e, et al.\u003c/em\u003e Effect of moxibustion on autophagy and the inflammatory response of synovial cells in rheumatoid arthritis model rat. \u003cstrong\u003e42\u003c/strong\u003e, 73-82 (2022).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"chinese-medicine","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"cmed","sideBox":"Learn more about [Chinese Medicine](http://cmjournal.biomedcentral.com)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/cmed/default.aspx","title":"Chinese Medicine","twitterHandle":"@BioMedCentral","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Acupuncture, Rheumatoid Arthritis, Machine Learning, Network Topology, Mendelian Randomization","lastPublishedDoi":"10.21203/rs.3.rs-6535408/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6535408/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eObjective: \u003c/strong\u003eRheumatoid arthritis (RA) is a systemic autoimmune disease requiring multi-target therapeutic strategies. Acupuncture, a holistic therapy in traditional Chinese medicine (TCM), has shown clinical efficacy in RA, yet its molecular mechanisms remain elusive. By integrating network topology and machine learning methods, decode the systemic regulatory effects of acupuncture on RA of acupuncture on RA.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods: \u003c/strong\u003eData on the interactions between acupuncture-affected endogenous compounds and RA-related targets were extracted from databases, built a multi-dimensional interaction network to map the interactions between acupuncture and RA. screened RA differentially expressed genes (DEGs) from GEO, intersecting with acupuncture-responsive genes. Utilize clusterProfiler for KEGG/GO enrichment analysis of these DEGs, Analyze the immune microenvironment using CIBERSORTx and xCell algorithms. Utilize ConsensusClusterPlus (R package) for unsupervised clustering to obtain DEGs. Subsequently, identify key genes using an ensemble machine learning model (GLM/SVM/XGB/RF) and create nomograms. Apply TwoSample MR and colocalization analysis to validate the causal relationship between core acupuncture-affected DEGs and RA risk.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults: \u003c/strong\u003eThis study identified 10 acupuncture-regulated metabolites and 49 RA-related DEGs. KEGG analysis showed DEGs enriched in immune pathways including the JAK/STAT pathway mediating inflammatory responses, the T-cell receptor signaling pathway involved in T cell differentiation and the TNF signaling pathway. Immunome profiling based on the CIBERSORT algorithm indicated that DEGs were primarily enriched in key immune cell subpopulations such as M1 macrophages, activated CD4⁺ T cells, Tregs, and B lymphocytes. Machine learning identified five key genes associated with immune infiltration (STAT1, GAPDH, JAK2, PTGS2, MDM2). MR/colocalization confirmed acupuncture-regulated STAT1 expression positively correlated with RA genetic susceptibility, highlighting STAT1-mediated JAK/STAT pathway in immune remodeling.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion: \u003c/strong\u003eSTAT1, GAPDH, JAK2, PTGS2, and MDM2 may be potential targets for acupuncture treatment of RA. Acupuncture may achieve systemic immune regulation by synergistically targeting multiple pathways (JAK/STAT, TNF) and immune cells (M1 macrophages, CD4\u003csup\u003e+\u003c/sup\u003e T cells), This initiative integrates the holistic philosophy of TCM with the precision of AI-driven medical science.\u003c/p\u003e","manuscriptTitle":"Potential Mechanisms of Acupuncture Treatment for Rheumatoid Arthritis: A Study Based on Network Topology and Machine Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-30 08:46:12","doi":"10.21203/rs.3.rs-6535408/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-07T01:53:33+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-05T03:52:10+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-04T04:23:56+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"724973371606104551359773495942507469","date":"2025-06-29T10:07:25+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"191564172325844236148143596277554239558","date":"2025-06-29T03:02:31+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-06-25T15:16:23+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-04-30T12:40:09+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-04-30T12:35:22+00:00","index":"","fulltext":""},{"type":"submitted","content":"Chinese Medicine","date":"2025-04-26T13:40:15+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"chinese-medicine","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"cmed","sideBox":"Learn more about [Chinese Medicine](http://cmjournal.biomedcentral.com)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/cmed/default.aspx","title":"Chinese Medicine","twitterHandle":"@BioMedCentral","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"521e8d64-fdb7-43cd-9ea5-b302148bb6f8","owner":[],"postedDate":"June 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-10-13T16:01:15+00:00","versionOfRecord":{"articleIdentity":"rs-6535408","link":"https://doi.org/10.1186/s13020-025-01209-8","journal":{"identity":"chinese-medicine","isVorOnly":false,"title":"Chinese Medicine"},"publishedOn":"2025-10-07 15:57:43","publishedOnDateReadable":"October 7th, 2025"},"versionCreatedAt":"2025-06-30 08:46:12","video":"","vorDoi":"10.1186/s13020-025-01209-8","vorDoiUrl":"https://doi.org/10.1186/s13020-025-01209-8","workflowStages":[]},"version":"v1","identity":"rs-6535408","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6535408","identity":"rs-6535408","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}