Network Toxicology and Bioinformatics Reveal Synovial Immune-Inflammatory Mechanisms of DEHP/DBP in Osteoarthritis

Network Toxicology and Bioinformatics Reveal Synovial Immune-Inflammatory Mechanisms of DEHP/DBP in Osteoarthritis | 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 Network Toxicology and Bioinformatics Reveal Synovial Immune-Inflammatory Mechanisms of DEHP/DBP in Osteoarthritis Xiaopeng Ling, Yang Liu, Chong Li, Xinmin Yang, Fuheng Ma, Fan Wang, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7815618/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: Osteoarthritis (OA) is increasingly recognized as an immune-inflammatory disease of the joint, with synovial dysfunction contributing to cartilage degeneration. Phthalate plasticizers such as di(2-ethylhexyl) phthalate (DEHP) and dibutyl phthalate (DBP) are ubiquitous environmental toxicants; however, their potential association with synovial immune perturbation in OA remains unclear. Methods: We integrated four synovial-tissue transcriptome datasets from GEO (GSE1919, GSE82107, GSE55235, GSE55457) and performed batch correction and differential expression analysis, followed by weighted gene co-expression network analysis (WGCNA) to identify phenotype-associated modules. Three machine-learning algorithms (LASSO, SVM-RFE, Random Forest) were used to select robust markers, which were validated in an independent dataset (GSE12021). Immune infiltration was profiled by ssGSEA. DEHP/DBP targets were predicted using cheminformatics resources and intersected with phenotype-related genes. Molecular docking and all-atom molecular dynamics (MD) simulations evaluated the binding modes and stability of DEHP with core protein targets. Results: We identified 1,394 differentially expressed genes and OA-associated WGCNA modules enriched for cytokine signaling and leukocyte activation. Intersecting phthalate-predicted targets with phenotype-related and differentially expressed genes yielded plasticizer-linked OA candidate genes. Convergent machine-learning selection nominated six core genes—ATF3, SLC2A3, PIM1, VEGFA, RGS4 and TSPO—showing consistent dysregulation and diagnostic utility across training and validation cohorts. ssGSEA indicated altered synovial immune-cell landscapes, and several core genes correlated with pro-inflammatory cell subsets. Docking and MD suggested comparatively stable binding of DEHP to TSPO, SLC2A3 and PIM1, supporting a plausible molecular interface between phthalate exposure and synovial immune dysregulation in OA. Conclusions: This integrative network-toxicology study links DEHP/DBP exposure to synovial immune–inflammatory signatures in OA and prioritizes six core genes as potential biomarkers or mechanistic candidates. The computational findings generate testable hypotheses for experimental validation and may inform exposure-mitigation and therapeutic strategies. Phthalates DEHP DBP Osteoarthritis Synovium Immune inflammation WGCNA Machine learning Molecular docking Molecular dynamics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 1. Introduction Osteoarthritis (OA) is a widespread and disabling chronic joint disease characterized by a pathological characteristic that is far from simple mechanical wear and tear, but involves a complex process involving a persistent immuno-inflammatory response and synovial lesions1 2 . OA, traditionally considered a "degenerative disease", has now been redefined by scientific consensus as a disease involving both systemic and local immune responses 3 . In addition to classic manifestations such as cartilage degeneration and osteophyte formation, synovitis has been confirmed to be one of the core pathological features of OA, and studies have shown that synovitis is prevalent in OA patients, and its severity is closely related to the patient's pain perception and disease progression rate 4 5 . In the occurrence and development of OA, the synovial membrane, as the main source of intra-articular inflammation, not only intensifies cartilage catabolism by secreting a variety of inflammatory mediators and cytokines, but also plays a key role in the formation of joint dysfunction and chronic pain 6 7 . Despite the growing clinical importance of synovial inflammation, the precise molecular mechanisms and key triggers that drive its sustained development and lead to immune microenvironment disturbances remain unfully elucidated, which largely limits the development of effective targeted therapeutic strategies for the nature of OA inflammation 8 . The understanding of the pathogenesis of OA has shifted from a "mechanical wear" paradigm to an "immune-inflammatory driven" paradigm 2 . During this process, fragments of the cartilage matrix caused by joint injury or stress, damage-associated molecular patterns (DAMPs) released by cells, and possibly autoantigens are released into the joint cavity that are recognized by intrinsic immune cells in the synovium (e.g., macrophages) through pattern recognition receptors (e.g., Toll-like receptors), initiating a non-infectious, low-grade chronic inflammatory response 9 2 . The persistence of this inflammatory state involves the infiltration and activation of a variety of immune cells, including macrophages, T cells, B cells, etc. Through a complex network of interactions, these cells release core inflammatory factors including interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α), and activate downstream key signaling pathways such as NF-κB and MAPK. This cascade of reactions forms a vicious cycle of self-amplification, constantly exacerbating inflammation and ultimately leading to structural disruption of joint tissue 10 11 . However, exactly what initial factors most strongly trigger and maintain the immuno-inflammatory state of the synovium and how this process interacts with lesions in other joint tissues, such as cartilage and subchondral bone, remains a core scientific question in the field 12 . In the immuno-inflammatory network of OA, the synovial membrane plays a particularly prominent role, serving not only as the main place for the aggregation and activation of inflammatory cells, but also as a central hub for regulating the entire joint microenvironment 13 14 . In-depth analysis of the immune cell atlas within the synovium reveals an imbalance in the polarization state of synovial macrophages (i.e., the predominance of pro-inflammatory M1 macrophages), the activation of T lymphocyte subsets (particularly interferon-γ-producing Th1 cells and IL-17-producing Th17 cells), and the potential antibody production and antigen presentation functions of B cells collectively constitute a complex immune regulatory ecosystem that drives disease progression16. Notably, a growing body of evidence in recent years suggests that synovial inflammation may occur before significant morphological changes in articular cartilage, suggesting that synovitis may not only be a result of OA, but more likely an early driver of its occurrence15. Therefore, an in-depth investigation of the origin and regulatory mechanism of synovial immune inflammation is of decisive significance for fundamentally understanding the pathogenesis of OA and developing new intervention strategies that can block or delay the disease process at the source 16 . Among the many factors that can trigger and exacerbate synovial immune inflammation, the role of environmental factors, particularly the ubiquitous endocrine-disrupting chemicals (EDCs), is gaining scientific attention 17 . Plasticizers, such as di(2-ethylhexyl) phthalate (DEHP) and dibutyl phthalate (DBP), as one of the most common EDCs, are widely found in everyday items such as plastic products, personal care products, and food packaging, and are capable of entering the body through multiple routes and interfering with normal physiological functions 18 19 . Available toxicological studies have shown that exposure to such chemicals is associated with the development of several immune dysfunctions and inflammatory diseases 20 21 . However, whether and how these common environmental pollutants specifically target joint synovial tissues to promote or exacerbate the inflammatory pathological process of OA by disrupting their sophisticated immune homeostasis remains a blind spot that has not been deeply explored. Elucidating whether these environmental pollutants can serve as independent triggers for OA immunoinflammation and revealing the molecular and cellular bridges in which they act will provide a fresh perspective on understanding the environmental etiology of OA and provide a key scientific basis for the development of prevention-based public health strategies and new therapeutic interventions. Based on the above background, this study aims to systematically reveal the central role of synovial immune inflammation in OA, and to explore for the first time whether and how environmental plasticizers DEHP and DBP promote OA by disturbing the synovial immune microenvironment. In this study, a series of differentially expressed genes and co-expression modules closely related to the immune inflammatory response were accurately located in OA synovial tissue by integrating multi-omics data. The results of bioinformatics analysis further confirm the key position of classical inflammatory signaling pathways such as IL-17, TNF and NF-κB in the pathological process of OA synovial membrane, which is consistent with the existing literature reports. Through innovative intersection analysis, the intersection of potential plasticizer targets and OA immunoinflammation-related genes was screened, and six core genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO) were accurately located using machine learning methods. Not only have these genes been validated to have good diagnostic value in independent clinical sample cohorts, but their expression levels are closely related to the degree of infiltration of key immune cells such as macrophages and T cells. Subsequent immunoinfiltration analysis further revealed the unique immune cell profile in the OA synovium and its complex regulatory relationship with these six core genes. In order to explore the direct interaction at the molecular level, molecular docking and molecular dynamics simulation were used to analyze the binding ability of plasticizers to core target proteins, and the results confirmed that DEHP has high affinity and stable binding patterns with some core target proteins. Finally, this study integrates all the findings and constructs a theoretical framework of Adverse Outcome Pathway (AOP) linking plasticizer exposure, synovial immune inflammation perturbation and OA pathological outcomes, emphasizing the central role of synovial immunoinflammation in environmental factors driving the occurrence and development of OA, and providing a new theoretical model and potential intervention target for research in this field. 2. Materials and methods 2.1. Data acquisition and preprocessing Four synovial tissue transcriptome datasets were obtained from the gene expression synthesis (GEO) database GSE1919 (5 normal synovial membranes, 5 OA synovial membranes), GSE82107 (7 normal synovial membranes, 10 OA synovial membranes), and GSE55457 (10 normal synovial and 10 OA synoviums).and GSE55235 (10 normal synovial and 10 OA synovium), including 35 osteoarthritis (OA) patients and 32 healthy control samples 22 . The original CEL file was downloaded where available and background correction and probe aggregation were performed using the oligo package (version 1.62.0) in R. According to the platform annotation file, the probe ID is converted into the official gene symbol, and if multiple probes match the same gene, the average expression value of the probe is taken. To reduce abiotic variation between different datasets, the ComBat algorithm in the R language sva package (version 3.46.0) was adopted, and empirical Bayesian correction was applied to adjust for batch effects (Fig. 1 A). Subsequently, the standardized expression matrix was log2 transformed and quantile normalized using the limma package (version 3.54.0) to ensure data uniformity and comparability (Fig. 1 B). 2.2. Differential expression analysis Differentially expressed genes (DEGs) were identified between the osteoarthritis (OA) group and the control group using the limma package. A moderated t-test was conducted, followed by the Benjamini–Hochberg multiple testing correction. The threshold for identifying significant DEGs was set at an adjusted p-value (ADJ. P.Val) < of less than 0.05 and an absolute log fold change (|log₂FC|) greater than 0.5. The results are depicted in a volcano plot generated using the ggplot2 package (Fig. 1 C). Furthermore, a heatmap of the top 50 DEGs, ranked by adjusted p-value, was created with the pheatmap package, where red indicates upregulation and blue signifies downregulation (Fig. 1 D). 2.3. Weighted Gene Co-expression Network Analysis (WGCNA) Before constructing the network, low-expression genes were filtered out, and 8,055 genes exhibiting the greatest expression variance were retained. Sample clustering was conducted to detect and eliminate outliers. The soft-thresholding power was chosen based on the scale-free topology criterion, (with a fitting index R² >greater than 0.9, resulting in a value of 7 (Fig. 2 A). A signed adjacency matrix was then constructed and transformed into a topological overlap matrix (TOM). Hierarchical clustering was performed using a dynamic pruning algorithm (with deepSplit set to 2 and minModuleSize set to 50) to identify co-expression modules. Modules with a characteristic gene similarity > greater than 0.75 were merged (Fig. 2 B). The module-phenotype association was assessed by calculating the Pearson correlation coefficient between the module signature genes and the disease phenotype (osteoarthritis vs. control). Modules with an absolute correlation value (|r|) greater than 0.7 and a p-value less than 0.05 were considered significantly correlated (blue modules indicating a negative correlation; cyan module indicating a positive correlation) (Figs. 2 C–D). 2.4. Functional enrichment analysis Three sets of genes were obtained: WGCNA⁻: 244 genes, which are the intersecting genes in the blue module (negatively correlated with OA). WGCNA⁺: The intersection genes of the cyan module (positively correlated) with DEGs (505). Two WGCNA modules intersect with DEGs (748 genes) (Figs. 3 A–C). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed using the clusterProfiler package (version 4.6.0). Entries with a corrected p-value of < 0.05 were considered significantly enriched, and the results were displayed by bar and dot plots (Fig. 3 D–F). 2.5. Plasticizer target prediction and validation The 2D structure and SMILES of di(2-ethylhexyl) phthalate (DEHP) and dibutyl phthalate (DBP) were obtained by Pubchem. SwissTargetPrediction, PharmMapper, STITCH, and SEA databases were used to predict potential targets for plasticizers di(2-ethylhexyl) phthalate (DEHP) and dibutyl phthalate (DBP), and a total of 600 candidate targets were obtained 23 24 25 26 . These targets were intersected with phenotype-related DEGs, and 42 key targets associated with the pathogenesis of OA were obtained (Fig. 4 A). GO and KEGG enrichment analyses were performed on these intersecting targets to elucidate their biological functions (Figs. 4 B–C). 2.6. Machine Learning Identifies Core Genes Three machine learning algorithms were employed to identify robust diagnostic biomarkers: LASSO regression, which screened out 9 genes based on the minimum binomial bias (λ = -3.2) using 5-fold cross-validation; SVM-RFE, which selected the feature set corresponding to the highest accuracy (14 features, accuracy = 0.918, error rate = 0.0817) using the radial basis kernel function; and Random Forest, which set up 500 trees and retained genes with an importance score of > 0.5 (24 genes). The genes identified by these three methods were combined, resulting in six core genes: ATF3, SLC2A3, PIM1, VEGFA, RGS4, and TSPO. 2.7. Validation with an Independent Dataset The core genes were validated using an independent dataset, GSE12021, which included 9 normal and 10 osteoarthritis (OA) synovial samples. Receiver operating characteristic (ROC) curves were plotted, and the area under the curve (AUC) was calculated to assess diagnostic performance (Fig. 6 B). The expression differences were verified using the Student's t-test (p < 0.05), and the expression patterns were visualized (Figs. 6 C–D). 2.8. Immunoinfiltration Analysis Single-sample gene set enrichment analysis (ssGSEA) was utilized to quantify the relative enrichment of 22 immune cell types in each sample using the GSVA package. A Spearman correlation coefficient was calculated for core gene expression and immune cell enrichment fraction, with a value > 0.3 and p < 0.05 considered significantly correlated. The results were visualized using a heatmap (Fig. 7 ). 2.9. Molecular Docking and Kinetic Simulation AutoDock Vina was used to perform molecular docking of DEHP and DBP with the protein products of the 6 core genes. Binding energy <-5.0 kcal/mol was considered indicative of stable binding (Fig. 8 ) 27 . Molecular dynamics (MD) simulations were conducted using GROMACS software and AMBER force fields 28 . The system was solvated and neutralized with ions in the TIP3P water model, and after energy minimization and equilibration, a 100 ns production run was executed. Trajectory analyses included the root mean square deviation (RMSD), radius of gyration (Rg), solvent accessible surface area (SASA), root mean square fluctuation (RMSF), and the number of hydrogen bonds (Figs. 9 A–E). The combined free energy was calculated using the MM/PBSA method (Figs. 9 F–I). 2.10. Adverse Outcome Pathway (AOP) Construction Following the AOP development principles articulated by Villeneuve et al. (2014), we assembled a conceptual AOP linking phthalate exposure, synovial immune dysregulation and OA-related outcomes 29 . The network aims to systematically elucidate the sequence of associations from molecular initiation events (plasticizer exposure), cellular responses, organ-level effects, to adverse outcomes. 3. Results 3.1. Osteoarthritis dataset collection and difference analysis To explore differential gene expression in synovial tissues of osteoarthritis, we processed datasets GSE1919, GSE82107, GSE55457, and GSE55235 from the GEO database by integrating and de-batching via R language (Fig. 1A), followed by standardization (Fig. 1B). We performed differential expression analysis using thresholds of adjusted p-value 0.5, identifying 1394 differentially expressed genes (DEGs); specifically, 707 genes were up-regulated (depicted in red on the volcano plot, Fig. 1C), and 687 were down-regulated (depicted in blue). Additionally, we generated a heatmap for the top 50 DEGs, with red signifying positive correlation and blue signifying negative correlation (Fig. 1D). These findings indicate that these genes may contribute to osteoarthritis mechanisms. 3.2. The most relevant targets of osteoarthritis phenotype were obtained by WGCNA analysis To elucidate genes strongly linked to osteoarthritis, we conducted WGCNA analysis on the integrated dataset. We initially removed outlier samples and applied a soft-thresholding power of 7, chosen as the scale-free topology fit index exceeded 0.9 (Fig. 2A), yielding co-expression gene modules (Fig. 2B). This identified five modules with over 50 genes each. Notably, the blue module displayed the strongest negative disease correlation (p < 0.05), whereas the turquoise module manifested the highest positive correlation (p < 0.05) (Fig. 2C). Module-trait association analysis confirmed that both modules correlated with phenotypic traits beyond 0.7 (p < 0.05) (Fig. 2D). Subsequently, we isolated 610 genes from the blue module and 928 genes from the turquoise module, indicating that these genes potentially contribute critically to osteoarthritis mechanisms. 3.3. intersection targets and GO and KEGG analysis To establish phenotypic-genetic linkages in osteoarthritis, we intersected disease-negative correlated (WGCNA-), positive correlated (WGCNA+), and dual-module integrated genes with differentially expressed genes (DEGs) (Fig. 3A-C), obtaining 244, 505, and 748 core genes respectively. Functional enrichment analysis demonstrated: Negative-correlated DEGs exhibited significant enrichment in biological processes including positive regulation of T cell-mediated immunity, leukocyte proliferation, and positive regulation of inflammatory response (P<0.05, Fig. 3D). Positive-correlated and dual-module DEGs consistently enriched in lymphocyte activation, positive regulation of T cell activation, and leukocyte activation involved in immune response (P<0.05, Fig. 3E-F). KEGG pathway analysis revealed negative-correlated DEGs primarily enriched in IL-17 signaling, TNF signaling, NF-kappa B signaling, and cytokine-cytokine receptor interaction (Fig. 3D); positive-correlated DEGs in chemokine signaling, Th17 cell differentiation, and NF-kappa B signaling (Fig. 3E); while dual-module DEGs showed prominent enrichment in IL-17 signaling, TNF signaling, NF-kappa B signaling, and Toll-like receptor signaling (P<0.05, Fig. 3F). These results collectively implicate osteoarthritis as an immune-inflammatory pathology fundamentally driven by dysregulated immune responses and sustained inflammatory activation. 3.4. targets of plasticizer-induced osteoarthritis and GO and KEGG enrichment analysis To investigate the toxic mechanisms underlying DEHP- and DBP-induced osteoarthritis, we identified 600 potential targets via SwissTargetPrediction, PharmMapper, STITCH, and Similarity Ensemble Approach databases, and intersected them with phenotype-linked differential genes to yield 42 key targets (Fig. 4A). Subsequent GO and KEGG enrichment analyses demonstrated: Biological processes (BP) significantly enriched in regulation of apoptotic signaling pathway, positive regulation of T cell activation, positive regulation of leukocyte cell-cell adhesion, tumor necrosis factor production, and regulation of T cell activation; cellular components (CC) enriched in ficolin-1-rich granule, secretory granule lumen, cytoplasmic vesicle lumen, and vesicle lumen; and molecular functions (MF) enriched in oxidoreductase activity, protein kinase regulator activity, Toll-like receptor binding, endopeptidase activity, and growth factor activity (Fig. 4B). KEGG pathway analysis revealed enrichment in IL-17 signaling pathway, TNF signaling pathway, NF-kappa B signaling pathway, Toll-like receptor signaling pathway, and Th17 cell differentiation (Fig. 4C). These findings imply that DEHP and DBP potentially exacerbate osteoarthritis by dysregulating immune-inflammatory networks, particularly through sustained activation of pathways like IL-17 and TNF signaling. 3.5. genes are screened through machine learning methods To identify core genes tightly linked to the immune-inflammatory phenotype in plasticizer (DEHP and DBP)-induced osteoarthritis, we applied three machine learning algorithms to 42 DEHP/DBP-disease crossover targets. LASSO regression yielded the lowest model deviation (BD = 0.7) at a penalty coefficient λ = -3.2, selecting 9 genes(Fig. 5A). SVM-RFE achieved peak cross-validation accuracy (0.918) and a minimum error rate (0.0817) with 14 features, yielding 14 genes(Fig. 5B). Random forest analysis displayed a stable error rate (≈0.3) at 300 features, identifying 24 genes with importance scores >0.5(Fig. 4C). Integration of these results revealed six core genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, and TSPO), indicating their role as a core molecular network in osteoarthritis(Fig. D). 3.6. Independent datasets were used to verify the differential expression of core genes and core genes To assess the diagnostic utility of machine learning-derived core genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO) in osteoarthritis, we conducted ROC curve validation using an independent cohort, GSE12021 (9 normal vs. 10 osteoarthritis synovial samples). All genes achieved AUC values exceeding 0.7 (ATF3=0.878, SLC2A3=0.767, PIM1=0.867, VEGFA=0.889, RGS4=0.822, TSPO=0.789), demonstrating robust discriminative capacity (Fig. 6B). Differential expression analysis consistently revealed significant downregulation of ATF3, SLC2A3, PIM1, and VEGFA (P<0.05), alongside upregulation of RGS4 and TSPO (P<0.05) in osteoarthritic synovium across both primary and validation datasets (Fig. 6C-D). These results underscore the viability of these genes as diagnostic or phenotypic biomarkers for osteoarthritis. 3.7. Immunoinfiltration analysis of core genes To delineate differences in synovial immune cell composition between healthy individuals and osteoarthritis (OA) patients and their association with core genes, we conducted single-sample gene set enrichment analysis (ssGSEA) across 22 immune cell types. Our analysis identified significant differences (P<0.05) between healthy and OA synovium for plasma cells, CD8+ T cells, CD4+ memory resting T cells, T follicular helper cells, gamma delta T cells, M0 macrophages, activated dendritic cells, resting mast cells, activated mast cells, and eosinophils (Figure 7A, B). Specific correlations among immune cell types were also evident (Figure 7C); notably, CD4+ memory resting T cells correlated positively with activated mast cells, whereas plasma cells correlated positively with both T follicular helper cells and gamma delta T cells (P<0.05). In contrast, CD4+ memory resting T cells correlated negatively with T follicular helper cells, M0 macrophages, and resting mast cells (P<0.05). Immune infiltration analysis of core genes implicated in phthalate-induced OA pathogenesis (Figure 7D) demonstrated that ATF3, PIM1, SLC2A3, and VEGFA each correlated positively with CD4+ memory resting T cells, activated mast cells, and activated dendritic cells, but negatively with plasma cells and resting mast cells (P<0.05). Specifically, ATF3 also showed negative correlations with T follicular helper cells and M0 macrophages; SLC2A3 correlated negatively with T follicular helper cells; and VEGFA correlated negatively with gamma delta T cells, T follicular helper cells, and M0 macrophages, yet correlated positively with eosinophils (P<0.05). Both RGS4 and TSPO correlated negatively with activated dendritic cells (P<0.05); further, RGS4 correlated positively with gamma delta T cells and T follicular helper cells, while TSPO correlated negatively with eosinophils (P<0.05). These findings collectively delineate a distinct immune cell landscape characteristic of OA synovium and establish direct functional connections between core pathogenic genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO), identified via phthalate toxicity screening, and dysregulated key immune cell populations in OA. 3.8. molecular docking To evaluate potential interactions between the plasticizer di(2-ethylhexyl) phthalate (DEHP) and six key osteoarthritis (OA)-related target proteins (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO), we conducted molecular docking analysis. Our results revealed strong binding affinities of DEHP for TSPO, PIM1, and SLC2A3, with binding energies of -7.7 kcal/mol, -7.7 kcal/mol, and -7.6 kcal/mol, respectively (Figure 8A-C). DEHP also demonstrated potential binding to ATF3, RGS4, and VEGFA, yielding binding energies of -5.2 kcal/mol, -5.2 kcal/mol, and -5.0 kcal/mol, respectively (Figure 8D-E). As all calculated binding energies fell below the threshold of -5.0 kcal/mol, the docking poses confirmed spontaneous interactions between DEHP and each of these target proteins (Figure 8). Applying standard interpretation (binding energy < -5.0 kcal/mol suggests potential binding; < -7.0 kcal/mol indicates strong binding), DEHP exhibited particularly robust binding affinities for TSPO, PIM1, and SLC2A3 (all energies < -7.0 kcal/mol). This molecular evidence confirms the core toxicological mechanism: DEHP directly binds to and disrupts key target proteins (notably TSPO, PIM1, and SLC2A3), thereby perturbing immune microenvironment homeostasis and driving OA pathogenesis. 3.9. Molecular dynamics simulation To evaluate the conformational stability of DEHP complexes with the key osteoarthritis-related target proteins TSPO, SLC2A3, and PIM1, we performed molecular dynamics (MD) simulations coupled with MM/PBSA binding free energy calculations. MD trajectories demonstrated stable complex formation: RMSD values equilibrated at 0.6 ± 0.05 nm after ~65 ns (DEHP-TSPO), 0.25 ± 0.05 nm after ~30 ns (DEHP-SLC2A3), and 0.23 ± 0.05 nm after ~10 ns (DEHP-PIM1) (Figure 9A). Radius of gyration (Rg) analysis revealed convergence: Rg decreased from 1.8 nm to stabilize at 1.70 ± 0.04 nm after ~60 ns (DEHP-TSPO), while remaining constant at 2.26 ± 0.02 nm (DEHP-SLC2A3) and 1.90 ± 0.02 nm (DEHP-PIM1) throughout the simulations (Figure 9B). Solvent-accessible surface area (SASA) profiles indicated stability: SASA stabilized at 100 ± 5 nm² after 5 ns (DEHP-TSPO), 200 ± 10 nm² after 55 ns (DEHP-SLC2A3), and remained at 140 ± 8 nm² (DEHP-PIM1) (Figure 9C). Root-mean-square fluctuation (RMSF) was low (< 0.4 nm) for key residues across all complexes, excluding termini (Figure 9D). Hydrogen bond occupancy analysis showed the DEHP-TSPO complex predominantly maintained two bonds, DEHP-SLC2A3 maintained one bond, and DEHP-PIM1 maintained one bond only ~33% of the time (Figure 9E). Free energy landscapes identified minima corresponding to stable conformations at Rg/RMSD values of 1.67-1.70 nm / 0.55-0.60 nm (DEHP-TSPO), 2.25-2.27 nm / 0.23-0.27 nm (DEHP-SLC2A3), and 1.89-1.92 nm / 0.19-0.24 nm (TRPC3) (Figure 9G-I). MM/PBSA calculations confirmed strong binding, yielding ΔG values of -185.302 kJ/mol (TSPO), -175.321 kJ/mol (SLC2A3), and -186.664 kJ/mol (PIM1) (Figure 9F). Energy decomposition attributed binding primarily to favorable van der Waals interactions (VDW: -303.206, -257.841, -250.976 kJ/mol), significantly opposed by polar solvation penalties (PB: +68.448, +63.846, +64.451 kJ/mol). Nonpolar solvation energy (SA: -26.343, -27.691, -25.899 kJ/mol) provided stabilization, while entropic contributions (-TΔS: +83.543, +51.293, +28.939 kJ/mol) partially counteracted the enthalpic gain. The molecular mechanics energy (MM: -309.83, -261.706, -255.365 kJ/mol) was dominated by VDW, with minor Coulombic components (COU: -6.625, -3.865, -4.39 kJ/mol). Collectively, these MD and MM/PBSA results provide robust evidence for the direct, strong, and stable binding of DEHP to the core OA-associated targets TSPO, SLC2A3, and PIM1. 3.10. Outcome Pathways (AOP) framework Subsequently, we establish a new adverse outcome pathway (AOP) framework based on the method proposed by Daniel et al. In this AOP network, the expression and activity of ATF3, SLC2A3, PIM1, TSPO, VEGFA, and RGS4 may be affected by plasticizers (DEHP and DBP). In addition, the IL-17/TNF/NF-κB signaling pathway may also be involved, leading to synovial immune dysregulation, which in turn leads to synovial inflammation and oxidative stress, and ultimately to the occurrence and development of osteoarthritis. This study successfully constructed a comprehensive AOP framework by systematically integrating the identified key central genes, which provides a theoretical basis for elucidating the mechanism of arthritis exacerbated by plasticizers (DEHP and DBP) (Fig. 10). 4. Discussion In this study, we systematically applied multi-omics analysis, machine learning approaches, and computational simulations to elucidate the contribution of the environmental plasticizers DEHP and DBP to the pathogenesis of osteoarthritis (OA). For the first time, we demonstrate that these widespread endocrine-disrupting chemicals may exacerbate OA development through perturbations of synovial immune homeostasis. Six core genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO) were identified as central to synovial immune inflammation and demonstrated strong diagnostic utility. Furthermore, molecular docking and molecular dynamics simulations confirmed stable binding of DEHP to proteins such as TSPO, PIM1, and SLC2A3, suggesting a direct molecular mechanism of interference with synovial immune microenvironment balance. Finally, by integrating genetic, pathway, and toxicological evidence, we constructed an adverse outcome pathway (AOP) that conceptually links phthalate exposure to immune dysregulation, synovial inflammation, and the progression of OA. These results highlight the central role of immune dysfunction, particularly involving macrophage and T-cell subsets, in OA synovitis 30 31 . Both DEGs and predicted phthalate targets were strongly enriched in canonical inflammatory cascades, including IL-17, TNF, and NF-κB signaling, consistent with established models of chronic OA inflammation 32 . In contrast to most prior studies that have focused on biomechanical stress or endogenous damage-associated signals (DAMPs) as the initiating factors of OA, our findings suggest that environmental plasticizers may act as important exogenous drivers of synovial inflammation. For example, TSPO, PIM1, and SLC2A3 displayed significant expression changes in OA synovium and bound to DEHP with high affinity, indicating that chemical exposures may directly trigger or modulate OA-associated immune responses—a mechanistic link not previously reported. Some of the results further expand on the knowledge of previous studies. The downregulation of ATF3 and VEGFA in the synovial membrane of OA is consistent with their known inhibitory effects on inflammation and regulation of angiogenesis 33 34 . However, they also serve as potential targets of DEHP/DBP, which provides a new basis for understanding the mechanistic connection between environmental exposure and OA occurrence. Similarly, although infiltration of immune cells (such as M0 macrophages and activated dendritic cells) in the OA synovial membrane has been reported 35 36 , its significant association with plasticizer-sensitive genes provides a new perspective on how chemical exposure alters the synovial immune microenvironment. Future research should therefore focus on experimental verification of the effects of DEHP/DBP on core immune-related genes and signaling pathways within OA synovial tissue models. Prospective studies are also needed to evaluate the epidemiological association between plasticizer exposure and OA development. In conclusion, this study proposes a novel mechanism by which environmental plasticizers promote OA through disruption of synovial immune-inflammatory pathways. By identifying core molecular targets and constructing an AOP framework, our findings extend current understanding of environmental contributions to OA etiology and provide potential biomarkers and therapeutic targets for interventions aimed at reducing plasticizer-related health risks. Limitations First, although four public synovial cohorts were integrated, overall sample size remained modest and cross-platform heterogeneity may influence generalizability. Second, immune-cell inference from bulk transcriptomes (ssGSEA with predefined gene sets) provides relative enrichment rather than absolute cell fractions and is sensitive to gene-set choice. Third, target prediction, docking and MM/PBSA are in-silico and contain known approximations; they support hypotheses but do not establish binding or causality. Finally, the proposed AOP is conceptual and requires experimental and epidemiological corroboration. Abbreviations OA osteoarthritis DEHP di(2—ethylhexyl) phthalate DBP dibutyl phthalate EDCs endocrine—disrupting chemicals DAMPs damage—associated molecular patterns GEO Gene Expression Omnibus GSE GEO Series DEGs differentially expressed genes WGCNA weighted gene co—expression network analysis TOM topological overlap matrix GO Gene Ontology BP biological process CC cellular component MF molecular function KEGG Kyoto Encyclopedia of Genes and Genomes PCA principal component analysis LASSO least absolute shrinkage and selection operator SVM RFE—support vector machine—recursive feature elimination ROC receiver operating characteristic AUC area under the curve FDR false discovery rate CI confidence interval SD standard deviation SEM standard error of the mean ssGSEA single—sample gene set enrichment analysis MD molecular dynamics RMSD root—mean—square deviation RMSF root—mean—square fluctuation SASA solvent accessible surface area Rg radius of gyration FEL free—energy landscape MM/PBSA molecular mechanics/Poisson—Boltzmann surface area AOP adverse outcome pathway FNIH Foundation for the National Institutes of Health NIH National Institutes of Health TNF tumor necrosis factor IL 17—interleukin—17 Declarations Conflict of interest: There are no conflicts of interest for this manuscript. Consent for publication: All authors got the consent for publication. Ethics declaration: not applicable . Ethics approval and consent to participate Not applicable. This study analyzed publicly available, de-identified gene-expression datasets and did not involve human participants, human tissue, or animals. Consent for publication Informed consent for publication of the manuscript was obtained from all authors. Availability of data and materials All transcriptome datasets analyzed are publicly available from the Gene Expression Omnibus (GEO): GSE1919, GSE82107, GSE55235, GSE55457 and validation set GSE12021. Processed data and analysis scripts are available from the corresponding author on reasonable request; dataset accession numbers are cited in the Methods. Competing interests The authors declare that they have no competing interests. Funding This work was supported by the Nanchang University Graduate Innovation Special Fund, Jiangxi Province (Grant No. YC2025-S305). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. YC2025-S305， 2025 Nanchang University Graduate Innovation Special Fund, Jiangxi Province Authors’ contributions Xiaopeng Lin conceived and supervised the study.Yang Liu and Chong Li curated datasets and performed bioinformatics analyses. Yang Liu conducted molecular docking and molecular dynamics. Yang Liu contributed to data interpretation and revisions. Yang Liu drafted the manuscript. All authors read and approved the final manuscript. Acknowledgements We sincerely thank GEO for generously sharing their data. We also gratefully acknowledge the financial support from Nanchang University Graduate Student Innovation Fund. Authors’ information Affiliated Rehabilitation Hospital of Nanchang University, Address: No. 133 South Square Road, Nanchang City, Nanchang (330003), Jiangxi,China，Xiaopeng Lin ，Yang Liu 1 , Chong Li 2 ,Xinmin Yang 3 ,Fuheng Ma 4 ,Fan Wang 5 ,Qixin Liu 6 NanchangUniversity,Address:999 Xuefu Road, Honggutan District, Nanchang(330019), Jiangxi, China，Xiaopeng Lin ，Yang Liu 1 , Chong Li 2 ,Xinmin Yang 3 ,Fuheng Ma 4 ,Fan Wang 5 ,Qixin Liu 6 References Zhang, Q., Sun, C., Liu, X., Zhu, C., Ma, C., & Feng, R. (2023). Mechanism of immune infiltration in synovial tissue of osteoarthritis: a gene expression-based study. Journal of Orthopaedic Surgery and Research, 18(1). https://doi.org/10.1186/s13018-023-03541-x Woodell‐May, J. E., & Sommerfeld, S. D. (2019). Role of Inflammation and the Immune System in the Progression of Osteoarthritis. Journal of Orthopaedic Research, 38(2), 253–257. https://doi.org/10.1002/jor.24457 Knights, A. J., Redding, S. J., & Maerz, T. (2022). Inflammation in osteoarthritis: the latest progress and ongoing challenges. Current Opinion in Rheumatology, 35(2), 128–134. https://doi.org/10.1097/bor.0000000000000923 Xuan, Q., Ruan, Y., Cao, C., Yin, Z., Du, J., Lv, C., Fang, F., & Gu, W. (2022). Effect of Ultrasonic Penetration with Volatile Oil of Olibanum and Chuanxiong Rhizoma on Acute Knee Synovitis Induced by Sports Training: An Open-Label Randomized Controlled Study. Pain Research and Management, 2022, 1–10. https://doi.org/10.1155/2022/6806565 Ramezanpour, S., Kanthawang, T., Lynch, J., McCulloch, C. E., Nevitt, M. C., Link, T. M., & Joseph, G. B. (2022). Impact of Sustained Synovitis on Knee Joint Structural Degeneration: 4‐Year MRI Data from the Osteoarthritis Initiative. Journal of Magnetic Resonance Imaging, 57(1), 153–164. https://doi.org/10.1002/jmri.28223 Panichi, V., Costantini, S., Grasso, M., Arciola, C. R., & Dolzani, P. (2024). Innate Immunity and Synovitis: Key Players in Osteoarthritis Progression. International Journal of Molecular Sciences, 25(22), 12082. https://doi.org/10.3390/ijms252212082 Cheng, J.-H., Hsu, C.-C., Hsu, S.-L., Chou, W.-Y., Wu, Y.-N., Kuo, C.-E. A., Hsu, T.-C., Shiu, L.-Y., & Jhan, S.-W. (2021). Adipose-Derived Mesenchymal Stem Cells-Conditioned Medium Modulates the Expression of Inflammation Induced Bone Morphogenetic Protein-2, -5 and -6 as Well as Compared with Shockwave Therapy on Rat Knee Osteoarthritis. Biomedicines, 9(10), 1399. https://doi.org/10.3390/biomedicines9101399 Li, Y.-N., Fan, M.-L., Liu, H.-Q., Ma, B., Dai, W.-L., Yu, B.-Y., & Liu, J.-H. (2019). Dihydroartemisinin derivative DC32 inhibits inflammatory response in osteoarthritic synovium through regulating Nrf2/NF-κB pathway. International Immunopharmacology, 74, 105701. https://doi.org/10.1016/j.intimp.2019.105701 Jiang, H., Zhang, Y., Hu, G., Shang, X., Ming, J., Deng, M., Li, Y., Ma, Y., Liu, S., & Zhou, Y. (2022). Innate/Inflammatory Bioregulation of Surfactant Protein D Alleviates Rat Osteoarthritis by Inhibiting Toll-Like Receptor 4 Signaling. Frontiers in Immunology, 13. https://doi.org/10.3389/fimmu.2022.913901 Cheleschi, S., Gallo, I., Barbarino, M., Giannotti, S., Mondanelli, N., Giordano, A., Tenti, S., & Fioravanti, A. (2019). MicroRNA Mediate Visfatin and Resistin Induction of Oxidative Stress in Human Osteoarthritic Synovial Fibroblasts Via NF-κB Pathway. International Journal of Molecular Sciences, 20(20), 5200. https://doi.org/10.3390/ijms20205200 Wang, X., Chen, T., Liang, W., Fan, T., Zhu, Z., Cao, P., Ruan, G., Zhang, Y., Chen, S., Wang, Q., Li, S., Huang, Y., Zeng, M., Hunter, D. J., Li, J., & Ding, C. (2022). Synovitis mediates the association between bone marrow lesions and knee pain in osteoarthritis: data from the Foundation for the National Institute of Health (FNIH) Osteoarthritis Biomarkers Consortium. Osteoarthritis and Cartilage, 30(9), 1270–1277. https://doi.org/10.1016/j.joca.2022.06.004 Knights, A. J., Redding, S. J., & Maerz, T. (2022). Inflammation in osteoarthritis: the latest progress and ongoing challenges. Current Opinion in Rheumatology, 35(2), 128–134. https://doi.org/10.1097/bor.0000000000000923 Thoenen, J., MacKay, J. W., Sandford, H. J. C., Gold, G. E., & Kogan, F. (2022). Imaging of Synovial Inflammation in Osteoarthritis, From the AJR Special Series on Inflammation. American Journal of Roentgenology, 218(3), 405–417. https://doi.org/10.2214/ajr.21.26170 Theeuwes, W. F., van den Bosch, M. H. J., Thurlings, R. M., Blom, A. B., & van Lent, P. L. E. M. (2021). The role of inflammation in mesenchymal stromal cell therapy in osteoarthritis, perspectives for post-traumatic osteoarthritis: a review. Rheumatology, 60(3), 1042–1053. https://doi.org/10.1093/rheumatology/keaa910 Jiang, T., Weng, Q., Liu, K., He, H., Zhang, Y., Zhang, W., Doherty, M., Xie, J., Yang, T., Li, J., Yang, Z., Chen, Q., Long, H., Wang, Y., Wei, J., Lei, G., & Zeng, C. (2024). Bidirectional association identified between synovitis and knee and hand osteoarthritis: a general population-based study. The Lancet Regional Health - Western Pacific, 50, 101169. https://doi.org/10.1016/j.lanwpc.2024.101169 Werner, D. M., Golightly, Y. M., Tao, M., Post, A., & Wellsandt, E. (2022). Environmental Risk Factors for Osteoarthritis: The Impact on Individuals with Knee Joint Injury. Rheumatic Disease Clinics of North America, 48(4), 907–930. https://doi.org/10.1016/j.rdc.2022.06.010 Werner, D. M., Golightly, Y. M., Tao, M., Post, A., & Wellsandt, E. (2022). Environmental Risk Factors for Osteoarthritis: The Impact on Individuals with Knee Joint Injury. Rheumatic Disease Clinics of North America, 48(4), 907–930. https://doi.org/10.1016/j.rdc.2022.06.010 Li, L., Huang, L., Lei, R., Zhang, P., Yang, Y., Liu, H., & Zhang, Y. (2024). DEHP and DBP, common phthalates, induce glucose metabolism disorders in rats via oxidative damage of PI3K/Akt/GLUT4 signaling. Environmental Pollution, 341, 122948. https://doi.org/10.1016/j.envpol.2023.122948 Sun, C., Li, K., Zhang, J., & Huang, C. (2021). Prediction of phthalates concentration in household dust based on back propagation neural network. Indoor and Built Environment, 31(1), 230–244. https://doi.org/10.1177/1420326x211010212 Cui, Y., Yin, K., Zheng, Y., Wang, B., Qu, Y., Li, S., & Lin, H. (2021). Mixed plasticizers aggravated apoptosis by NOD2-RIP2-NF-κB pathway in grass carp hepatocytes. Journal of Hazardous Materials, 402, 123527. https://doi.org/10.1016/j.jhazmat.2020.123527 Wang, S., Xu, K., Du, W., Gao, X., Ma, P., Yang, X., & Chen, M. (2024). Exposure to environmental doses of DEHP causes phenotypes of polycystic ovary syndrome. Toxicology, 509, 153952. https://doi.org/10.1016/j.tox.2024.153952 Sherry, S. T., Ward, M. H., Kholodov, M., Baker, J., Phan, L., Smigielski, E. M., & Sirotkin, K. (2001). dbSNP: the NCBI database of genetic variation. Nucleic acids research, 29(1), 308–311. https://doi.org/10.1093/nar/29.1.308 Daina, A., Michielin, O., & Zoete, V. (2019). SwissTargetPrediction: updated data and new features for efficient prediction of protein targets of small molecules. Nucleic acids research, 47(W1), W357–W364. https://doi.org/10.1093/nar/gkz382 Liu, X., Ouyang, S., Yu, B., Liu, Y., Huang, K., Gong, J., Zheng, S., Li, Z., Li, H., & Jiang, H. (2010). PharmMapper server: a web server for potential drug target identification using pharmacophore mapping approach. Nucleic acids research, 38(Web Server issue), W609–W614. https://doi.org/10.1093/nar/gkq300 Heather, J. M., Spindler, M. J., Alonso, M. H., Shui, Y. I., Millar, D. G., Johnson, D. S., Cobbold, M., & Hata, A. N. (2022). Stitchr: stitching coding TCR nucleotide sequences from V/J/CDR3 information. Nucleic acids research, 50(12), e68. https://doi.org/10.1093/nar/gkac190 Wei, Y., Zhang, S., Shang, S., Zhang, B., Li, S., Wang, X., Wang, F., Su, J., Wu, Q., Liu, H., & Zhang, Y. (2016). SEA: a super-enhancer archive. Nucleic acids research, 44(D1), D172–D179. https://doi.org/10.1093/nar/gkv1243 Bugnon, M., Röhrig, U. F., Goullieux, M., Perez, M. A. S., Daina, A., Michielin, O., & Zoete, V. (2024). SwissDock 2024: major enhancements for small-molecule docking with Attracting Cavities and AutoDock Vina. Nucleic acids research, 52(W1), W324–W332. https://doi.org/10.1093/nar/gkae300 Serçinoglu, O., & Ozbek, P. (2018). gRINN: a tool for calculation of residue interaction energies and protein energy network analysis of molecular dynamics simulations. Nucleic acids research, 46(W1), W554–W562. https://doi.org/10.1093/nar/gky381 Villeneuve, D. L., Crump, D., Garcia-Reyero, N., Hecker, M., Hutchinson, T. H., LaLone, C. A., Landesmann, B., Lettieri, T., Munn, S., Nepelska, M., Ottinger, M. A., Vergauwen, L., & Whelan, M. (2014). Adverse outcome pathway (AOP) development I: strategies and principles. Toxicological sciences : an official journal of the Society of Toxicology, 142(2), 312–320. https://doi.org/10.1093/toxsci/kfu199 Li, W., Liu, Y., Wei, M., Yang, Z., Li, Z., Guo, Z., Yan, L., Lu, Y., Tang, H., Li, B., & Huang, W. (2025). Functionalized Biomimetic Nanoparticles Targeting the IL‐10/IL‐10Rα/Glycolytic Axis in Synovial Macrophages Alleviate Cartilage Degeneration in Osteoarthritis. Advanced Science, 12(27). https://doi.org/10.1002/advs.202504768 Zhu, J., Wei, J., Lin, Y., Tang, Y., Su, Z., Li, L., Liu, B., & Cai, X. (2024). Inhibition of IL-17 signaling in macrophages underlies the anti-arthritic effects of halofuginone hydrobromide: Network pharmacology, molecular docking, and experimental validation. BMC Complementary Medicine and Therapies, 24(1). https://doi.org/10.1186/s12906-024-04397-2 Sharma, P., Malhotra, H., Garg, V., Singh, G., Ashri, A., & Kamal, M. A. (2022). MECHANISTIC TARGETING OF TNF-α AND NF-κB IN INFLAMMATORY DISEASES. In International Journal of Cell Biology and Cellular Processes 8 (2), 14-23, 2022. Liang, Y., Jiang, Y., Jin, X., Chen, P., Heng, Y., Cai, L., Zhang, W., Li, L., & Jia, L. (2020). Neddylation inhibition activates the protective autophagy through NF-κB-catalase-ATF3 Axis in human esophageal cancer cells. Cell Communication and Signaling, 18(1). https://doi.org/10.1186/s12964-020-00576-z Abdelnaby, R., Sonbol, Y. T., Dardeer, K. T., ELgenidy, A., Ebrahim, M. A., Elmenawi, K. A., Abdallah, A. A., Ballut, O. O., Metwally, S. A., Arafat, M. S., Elsayed, M., Abouelhamd, A. M., Farahat, R. A., Karkour, A. M., Mohamed, K. A., & Rauen, T. (2023). Could Osteopontin be a useful biomarker in the diagnosis and severity assessment of osteoarthritis? A systematic review and meta-analysis of recent evidence. Clinical Immunology, 246, 109187. https://doi.org/10.1016/j.clim.2022.109187 Li, W., Liu, Y., Wei, M., Yang, Z., Li, Z., Guo, Z., Yan, L., Lu, Y., Tang, H., Li, B., & Huang, W. (2025). Functionalized Biomimetic Nanoparticles Targeting the IL‐10/IL‐10Rα/Glycolytic Axis in Synovial Macrophages Alleviate Cartilage Degeneration in Osteoarthritis. Advanced Science, 12(27). https://doi.org/10.1002/advs.202504768 Macečková, Z., Pergner, J., Krbec, M., Urban, M., & Zahradníček, M. (2025). Application of amniotic membrane in osteoarthritis management. Journal of Cartilage & Joint Preservation, 5(1), 100174. https://doi.org/10.1016/j.jcjp.2024.100174 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted 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-7815618","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":545585757,"identity":"45cdd92c-dc58-4bef-a567-8bfcc3a52f1c","order_by":0,"name":"Xiaopeng Ling","email":"data:image/png;base64,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","orcid":"","institution":"Affiliated Rehabilitation Hospital Nanchang University","correspondingAuthor":true,"prefix":"","firstName":"Xiaopeng","middleName":"","lastName":"Ling","suffix":""},{"id":545585758,"identity":"9ebd5926-982f-4ab5-bdb3-01320dd30170","order_by":1,"name":"Yang Liu","email":"","orcid":"","institution":"Affiliated Rehabilitation Hospital Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Yang","middleName":"","lastName":"Liu","suffix":""},{"id":545585759,"identity":"7affadbf-12b5-4361-8e04-0b46f597b1c0","order_by":2,"name":"Chong Li","email":"","orcid":"","institution":"Affiliated Rehabilitation Hospital Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Chong","middleName":"","lastName":"Li","suffix":""},{"id":545585760,"identity":"267ef481-951e-41cb-a9dc-dc596f9b8a3d","order_by":3,"name":"Xinmin Yang","email":"","orcid":"","institution":"Affiliated Rehabilitation Hospital Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Xinmin","middleName":"","lastName":"Yang","suffix":""},{"id":545585761,"identity":"a2420b49-3f85-4950-b34b-8f252b67e340","order_by":4,"name":"Fuheng Ma","email":"","orcid":"","institution":"Affiliated Rehabilitation Hospital Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Fuheng","middleName":"","lastName":"Ma","suffix":""},{"id":545585762,"identity":"ac8ae385-2645-4df3-a294-cbcfe28d794f","order_by":5,"name":"Fan Wang","email":"","orcid":"","institution":"Affiliated Rehabilitation Hospital Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Fan","middleName":"","lastName":"Wang","suffix":""},{"id":545585763,"identity":"4af3f70c-5b22-461e-a6bf-4ffe333587c4","order_by":6,"name":"Qixin Liu","email":"","orcid":"","institution":"Affiliated Rehabilitation Hospital Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Qixin","middleName":"","lastName":"Liu","suffix":""}],"badges":[],"createdAt":"2025-10-09 09:23:34","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7815618/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7815618/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96136396,"identity":"40360d7a-d24b-44e6-a5ea-bced6d4ee2cb","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35391607,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/45e1cbe1fe99e8103f0971b2.docx"},{"id":96249933,"identity":"e43b4919-82b5-42c9-8ec4-b6d558ceb32b","added_by":"auto","created_at":"2025-11-19 07:36:47","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8600,"visible":true,"origin":"","legend":"","description":"","filename":"1a1c3754fdc547c488ac59ede338ade2.json","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/d91cee46d2c111d9cae20d9d.json"},{"id":96250987,"identity":"8354f0d7-3b68-474d-a6ad-dfcee784a87f","added_by":"auto","created_at":"2025-11-19 07:39:11","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":134331,"visible":true,"origin":"","legend":"","description":"","filename":"1a1c3754fdc547c488ac59ede338ade21enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/1d95b460e56db2642a19c4fc.xml"},{"id":96136373,"identity":"6381664b-5313-4533-a3be-2b228666cea1","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"jpeg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5559652,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/617280b1713e2321ee5f263e.jpeg"},{"id":96248821,"identity":"868c517f-53f4-4887-9c0c-30f5d82dcf91","added_by":"auto","created_at":"2025-11-19 07:29:23","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":68646,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage10.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/d0b3740ec34a0d62065a9ccd.jpeg"},{"id":96251083,"identity":"ae82c295-6e61-4a3a-b2d0-c40f54282aaf","added_by":"auto","created_at":"2025-11-19 07:39:17","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2876630,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/b762eca196020bf7d2a2cf78.jpeg"},{"id":96251159,"identity":"8bd64f12-d1df-456f-a8ff-8e6525c668b9","added_by":"auto","created_at":"2025-11-19 07:39:26","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3120656,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/c2d83c7efa92c5c57148536e.jpeg"},{"id":96136390,"identity":"73ed39d9-7560-4283-9294-402846c8805f","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1775554,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/ce95db9666ef0d1b686c080b.jpeg"},{"id":96248844,"identity":"ee9236e3-1242-4b96-a7ff-6b627a2a3ce4","added_by":"auto","created_at":"2025-11-19 07:29:29","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2721732,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/2e127d18fa576e07b2b4993e.jpeg"},{"id":96136374,"identity":"50a581b7-38f5-4196-a191-50485535e8ad","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"jpeg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1897606,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/b6fa9112f476305e52593151.jpeg"},{"id":96136384,"identity":"4eafcb2e-6a0a-4853-9af2-45a282394059","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"jpeg","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4937550,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/6a4f36b5437ee7a748e1cf83.jpeg"},{"id":96136395,"identity":"e5dfc608-8343-46b9-89cc-5a97395999af","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"jpeg","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7054528,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/2feb1a592cfae57636445200.jpeg"},{"id":96248458,"identity":"b4aec023-fe4f-4d4f-991a-e0c6c8d3fc22","added_by":"auto","created_at":"2025-11-19 07:28:28","extension":"jpeg","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5328014,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/c9d025ef7534534665fb43d6.jpeg"},{"id":96136381,"identity":"3371d671-7b31-4c65-ab4c-ed35d1ab85a6","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":359553,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/2c3ee5c73d84cfe9aa45c51b.png"},{"id":96136378,"identity":"0f1f1e8e-7d32-48a1-9151-e887db1ce462","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":14130,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/72196767df3e08b4bf20e5ea.png"},{"id":96248586,"identity":"8958dd17-7f9d-4baa-b58c-f50c18d63641","added_by":"auto","created_at":"2025-11-19 07:28:44","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":293355,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/4e1679b9748a979df48bc6a8.png"},{"id":96136391,"identity":"6d7aed0f-2943-47a1-8cb8-108e595f87a9","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":399651,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/51360fc0a5c781d525883af7.png"},{"id":96136392,"identity":"ef32e22f-21d8-443a-bea5-9716f5bb6ed5","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":217213,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/b25d3bb732e68b0e1e9d4911.png"},{"id":96136387,"identity":"617983e9-1769-4acf-a98c-361be4d740c3","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":278733,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/a23605eb38157f52515e3b45.png"},{"id":96136393,"identity":"3be86bf8-2b95-41d9-8e50-61ed2be18ec8","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":173068,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/45c7c97c670b8bf8ef162a82.png"},{"id":96251217,"identity":"e2970c73-bab9-47d6-86e0-8f9fcb568664","added_by":"auto","created_at":"2025-11-19 07:39:31","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":499231,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/afc75b30cae08173751b5298.png"},{"id":96136385,"identity":"9c71c4fe-afe3-4d6d-a5a2-f3cc3b911990","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":875898,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/0b6086a17059fd8a37374dbd.png"},{"id":96136380,"identity":"c100d645-a374-4f83-b047-a51442c997b0","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":528785,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/ef821c459ec7d92971dd5a49.png"},{"id":96251273,"identity":"5019a550-9be5-4241-a3aa-7a673be581cf","added_by":"auto","created_at":"2025-11-19 07:39:35","extension":"xml","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":133171,"visible":true,"origin":"","legend":"","description":"","filename":"1a1c3754fdc547c488ac59ede338ade21structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/2ad924cc71887983753681ca.xml"},{"id":96136397,"identity":"ddc6ce14-e138-4f40-91ab-84b28a1ad685","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"html","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":141476,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/8adcec6cc2528f8c68422601.html"},{"id":96136365,"identity":"0ae0baea-a029-40b2-93fd-8175cb5b9399","added_by":"auto","created_at":"2025-11-18 04:02:49","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":333151,"visible":true,"origin":"","legend":"\u003cp\u003eDataset integration and differential expression in synovial tissue. (A) PCA of the merged cohort after batch correction. (B) Distribution of expression values before/after normalization. (C) Volcano plot of differentially expressed genes (DEGs). (D) Heatmap of the top DEGs across samples. Unless otherwise stated, DEGs were defined by adjusted P \u0026lt; 0.05 (Benjamini–Hochberg) with an absolute log2 fold-change threshold as specified in Methods.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/d5b011155b8fb5b09d922704.png"},{"id":96136363,"identity":"cb1216d1-f733-4bc1-86fb-357cea224965","added_by":"auto","created_at":"2025-11-18 04:02:49","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":132255,"visible":true,"origin":"","legend":"\u003cp\u003eWGCNA and module–trait relationships in osteoarthritis (OA). (A) Scale-free topology fit and mean connectivity across soft-threshold powers; the chosen β is indicated. (B) Gene dendrogram with dynamic tree cutting. (C) Module–trait correlation heatmap (Pearson r; cells show r and adjusted P). (D) Gene significance vs module membership in key OA-associated modules.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/0a22ec3f4c13a56e79d166d5.png"},{"id":96136368,"identity":"5b3683a3-e834-4916-8550-88b7f3a3a811","added_by":"auto","created_at":"2025-11-18 04:02:49","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":197291,"visible":true,"origin":"","legend":"\u003cp\u003eIntersection analysis and functional enrichment of OA-related genes. (A) Venn of WGCNA⁻/WGCNA⁺ with DEGs. (B–C) GO and KEGG enrichment (top FDR terms). (D–F) Bar/dot plots of representative immune-inflammatory terms; adjusted P shown.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/ef71b0d6911db15bc5459c9c.png"},{"id":96136364,"identity":"96eb1060-4fc5-4b9b-b162-33246b585bf0","added_by":"auto","created_at":"2025-11-18 04:02:49","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":107699,"visible":true,"origin":"","legend":"\u003cp\u003ePlasticizer-linked targets and enrichment analysis. (A) Venn intersecting predicted DEHP/DBP targets with phenotype-related genes and DEGs. (B–C) GO/KEGG enrichment highlighting cytokine/chemokine and TNF/IL-17 signaling. Target prediction resources and parameters are detailed in Methods.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/ba516121c1dc43056469aa0f.png"},{"id":96251142,"identity":"d65ebe5e-e010-40f0-98d4-08fdfe4b25fd","added_by":"auto","created_at":"2025-11-19 07:39:24","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":114312,"visible":true,"origin":"","legend":"\u003cp\u003eMachine-learning selection of core genes. (A) LASSO coefficient profiles and optimal λ. (B) SVM-RFE performance across feature counts. (C) Random-forest importance. (D) Intersection yielding six core genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO).\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/6256b84bf14db6b24b3cd5c3.png"},{"id":96136371,"identity":"d14e17de-d0e4-4535-b59d-55379c428a0a","added_by":"auto","created_at":"2025-11-18 04:02:49","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":91074,"visible":true,"origin":"","legend":"\u003cp\u003eDiagnostic performance and expression validation. (A) ROC curves in training/pooled discovery dataset. (B) ROC curves in GSE12021. (C–D) Expression distributions by group. AUCs with 95% CIs are reported where applicable; P values are two-sided unless stated.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/98773da6374c4910fb412635.png"},{"id":96136376,"identity":"42e3fc17-5698-47b2-84b3-07042febab9a","added_by":"auto","created_at":"2025-11-18 04:02:50","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":282715,"visible":true,"origin":"","legend":"\u003cp\u003eImmune infiltration and gene–immune correlations. (A–B) ssGSEA abundance of 22 immune cell types in normal vs OA synovium; differentially infiltrated cell types highlighted. (C) Correlation heatmap between core genes and immune-cell signatures (Spearman r; FDR-adjusted P).\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/6de694d1f4aeb37b3d48c8c3.png"},{"id":96136369,"identity":"57afe1ad-13ee-44f5-b1df-a50ea8b455a7","added_by":"auto","created_at":"2025-11-18 04:02:49","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":429628,"visible":true,"origin":"","legend":"\u003cp\u003eMolecular docking of DEHP with core protein targets. (A–F) Predicted binding poses for DEHP in the pockets of PIM1, TSPO, SLC2A3, RGS4, VEGFA and others as indicated. Key interacting residues and hydrogen-bond/π contacts annotated; docking protocol/scoring in Methods.\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/5e46f005e817add1254d2a8f.png"},{"id":96136372,"identity":"e088362b-4b06-47fa-8382-fbf8c6e8b695","added_by":"auto","created_at":"2025-11-18 04:02:49","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":242029,"visible":true,"origin":"","legend":"\u003cp\u003eMolecular dynamics (MD) simulations of DEHP–protein complexes. (A) RMSD of protein and ligand heavy atoms. (B) Protein RMSF. (C) Radius of gyration (Rg). (D) Hydrogen-bond occupancy / contact frequency. (E) Free-energy landscapes (FEL) by principal components. (F) If performed, MM/PBSA components (mean ± SD). Simulation settings (force field, ensemble, time step, thermostat/barostat) are given in Methods.\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/a9b6c9767bd9ef39af0c417d.png"},{"id":96248763,"identity":"b72346cb-b78b-45b1-826a-71477e5dd20e","added_by":"auto","created_at":"2025-11-19 07:29:09","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":38033,"visible":true,"origin":"","legend":"\u003cp\u003eAdverse Outcome Pathway (AOP) framework linking phthalate exposure to OA exacerbation. The schematic summarizes molecular initiating events (e.g., DEHP/DBP interacting with core targets), key cellular events (synovial immune dysregulation and inflammatory cascades), tissue effects, and the adverse outcome (OA progression).\u003c/p\u003e","description":"","filename":"10.png","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/4f5cc21f82fdbffb90d3b935.png"},{"id":101713079,"identity":"dd3388b5-3628-4ae2-84d9-dbc29823b0ff","added_by":"auto","created_at":"2026-02-02 23:09:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2800400,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7815618/v1/7be40ed8-5d4b-4a99-bbbb-7aede5699c7c.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Network Toxicology and Bioinformatics Reveal Synovial Immune-Inflammatory Mechanisms of DEHP/DBP in Osteoarthritis","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eOsteoarthritis (OA) is a widespread and disabling chronic joint disease characterized by a pathological characteristic that is far from simple mechanical wear and tear, but involves a complex process involving a persistent immuno-inflammatory response and synovial lesions1 \u003csup\u003e2\u003c/sup\u003e. OA, traditionally considered a \"degenerative disease\", has now been redefined by scientific consensus as a disease involving both systemic and local immune responses\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. In addition to classic manifestations such as cartilage degeneration and osteophyte formation, synovitis has been confirmed to be one of the core pathological features of OA, and studies have shown that synovitis is prevalent in OA patients, and its severity is closely related to the patient's pain perception and disease progression rate \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. In the occurrence and development of OA, the synovial membrane, as the main source of intra-articular inflammation, not only intensifies cartilage catabolism by secreting a variety of inflammatory mediators and cytokines, but also plays a key role in the formation of joint dysfunction and chronic pain \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Despite the growing clinical importance of synovial inflammation, the precise molecular mechanisms and key triggers that drive its sustained development and lead to immune microenvironment disturbances remain unfully elucidated, which largely limits the development of effective targeted therapeutic strategies for the nature of OA inflammation\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eThe understanding of the pathogenesis of OA has shifted from a \"mechanical wear\" paradigm to an \"immune-inflammatory driven\" paradigm \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. During this process, fragments of the cartilage matrix caused by joint injury or stress, damage-associated molecular patterns (DAMPs) released by cells, and possibly autoantigens are released into the joint cavity that are recognized by intrinsic immune cells in the synovium (e.g., macrophages) through pattern recognition receptors (e.g., Toll-like receptors), initiating a non-infectious, low-grade chronic inflammatory response \u003csup\u003e9 2\u003c/sup\u003e. The persistence of this inflammatory state involves the infiltration and activation of a variety of immune cells, including macrophages, T cells, B cells, etc. Through a complex network of interactions, these cells release core inflammatory factors including interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α), and activate downstream key signaling pathways such as NF-κB and MAPK. This cascade of reactions forms a vicious cycle of self-amplification, constantly exacerbating inflammation and ultimately leading to structural disruption of joint tissue\u003csup\u003e10 11\u003c/sup\u003e. However, exactly what initial factors most strongly trigger and maintain the immuno-inflammatory state of the synovium and how this process interacts with lesions in other joint tissues, such as cartilage and subchondral bone, remains a core scientific question in the field\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. In the immuno-inflammatory network of OA, the synovial membrane plays a particularly prominent role, serving not only as the main place for the aggregation and activation of inflammatory cells, but also as a central hub for regulating the entire joint microenvironment\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. In-depth analysis of the immune cell atlas within the synovium reveals an imbalance in the polarization state of synovial macrophages (i.e., the predominance of pro-inflammatory M1 macrophages), the activation of T lymphocyte subsets (particularly interferon-γ-producing Th1 cells and IL-17-producing Th17 cells), and the potential antibody production and antigen presentation functions of B cells collectively constitute a complex immune regulatory ecosystem that drives disease progression16. Notably, a growing body of evidence in recent years suggests that synovial inflammation may occur before significant morphological changes in articular cartilage, suggesting that synovitis may not only be a result of OA, but more likely an early driver of its occurrence15. Therefore, an in-depth investigation of the origin and regulatory mechanism of synovial immune inflammation is of decisive significance for fundamentally understanding the pathogenesis of OA and developing new intervention strategies that can block or delay the disease process at the source\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eAmong the many factors that can trigger and exacerbate synovial immune inflammation, the role of environmental factors, particularly the ubiquitous endocrine-disrupting chemicals (EDCs), is gaining scientific attention\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Plasticizers, such as di(2-ethylhexyl) phthalate (DEHP) and dibutyl phthalate (DBP), as one of the most common EDCs, are widely found in everyday items such as plastic products, personal care products, and food packaging, and are capable of entering the body through multiple routes and interfering with normal physiological functions\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Available toxicological studies have shown that exposure to such chemicals is associated with the development of several immune dysfunctions and inflammatory diseases\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. However, whether and how these common environmental pollutants specifically target joint synovial tissues to promote or exacerbate the inflammatory pathological process of OA by disrupting their sophisticated immune homeostasis remains a blind spot that has not been deeply explored. Elucidating whether these environmental pollutants can serve as independent triggers for OA immunoinflammation and revealing the molecular and cellular bridges in which they act will provide a fresh perspective on understanding the environmental etiology of OA and provide a key scientific basis for the development of prevention-based public health strategies and new therapeutic interventions.\u003c/p\u003e\u003cp\u003eBased on the above background, this study aims to systematically reveal the central role of synovial immune inflammation in OA, and to explore for the first time whether and how environmental plasticizers DEHP and DBP promote OA by disturbing the synovial immune microenvironment. In this study, a series of differentially expressed genes and co-expression modules closely related to the immune inflammatory response were accurately located in OA synovial tissue by integrating multi-omics data. The results of bioinformatics analysis further confirm the key position of classical inflammatory signaling pathways such as IL-17, TNF and NF-κB in the pathological process of OA synovial membrane, which is consistent with the existing literature reports. Through innovative intersection analysis, the intersection of potential plasticizer targets and OA immunoinflammation-related genes was screened, and six core genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO) were accurately located using machine learning methods. Not only have these genes been validated to have good diagnostic value in independent clinical sample cohorts, but their expression levels are closely related to the degree of infiltration of key immune cells such as macrophages and T cells. Subsequent immunoinfiltration analysis further revealed the unique immune cell profile in the OA synovium and its complex regulatory relationship with these six core genes. In order to explore the direct interaction at the molecular level, molecular docking and molecular dynamics simulation were used to analyze the binding ability of plasticizers to core target proteins, and the results confirmed that DEHP has high affinity and stable binding patterns with some core target proteins. Finally, this study integrates all the findings and constructs a theoretical framework of Adverse Outcome Pathway (AOP) linking plasticizer exposure, synovial immune inflammation perturbation and OA pathological outcomes, emphasizing the central role of synovial immunoinflammation in environmental factors driving the occurrence and development of OA, and providing a new theoretical model and potential intervention target for research in this field.\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1. Data acquisition and preprocessing\u003c/h2\u003e\u003cp\u003eFour synovial tissue transcriptome datasets were obtained from the gene expression synthesis (GEO) database GSE1919 (5 normal synovial membranes, 5 OA synovial membranes), GSE82107 (7 normal synovial membranes, 10 OA synovial membranes), and GSE55457 (10 normal synovial and 10 OA synoviums).and GSE55235 (10 normal synovial and 10 OA synovium), including 35 osteoarthritis (OA) patients and 32 healthy control samples\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. The original CEL file was downloaded where available and background correction and probe aggregation were performed using the oligo package (version 1.62.0) in R. According to the platform annotation file, the probe ID is converted into the official gene symbol, and if multiple probes match the same gene, the average expression value of the probe is taken. To reduce abiotic variation between different datasets, the ComBat algorithm in the R language sva package (version 3.46.0) was adopted, and empirical Bayesian correction was applied to adjust for batch effects (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). Subsequently, the standardized expression matrix was log2 transformed and quantile normalized using the limma package (version 3.54.0) to ensure data uniformity and comparability (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2. Differential expression analysis\u003c/h2\u003e\u003cp\u003eDifferentially expressed genes (DEGs) were identified between the osteoarthritis (OA) group and the control group using the limma package. A moderated t-test was conducted, followed by the Benjamini\u0026ndash;Hochberg multiple testing correction. The threshold for identifying significant DEGs was set at an adjusted p-value (ADJ. P.Val)\u0026thinsp;\u0026lt;\u0026thinsp;of less than 0.05 and an absolute log fold change (|log₂FC|) greater than 0.5. The results are depicted in a volcano plot generated using the ggplot2 package (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). Furthermore, a heatmap of the top 50 DEGs, ranked by adjusted p-value, was created with the pheatmap package, where red indicates upregulation and blue signifies downregulation (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3. Weighted Gene Co-expression Network Analysis (WGCNA)\u003c/h2\u003e\u003cp\u003eBefore constructing the network, low-expression genes were filtered out, and 8,055 genes exhibiting the greatest expression variance were retained. Sample clustering was conducted to detect and eliminate outliers. The soft-thresholding power was chosen based on the scale-free topology criterion, (with a fitting index R\u0026sup2; \u0026gt;greater than 0.9, resulting in a value of 7 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). A signed adjacency matrix was then constructed and transformed into a topological overlap matrix (TOM). Hierarchical clustering was performed using a dynamic pruning algorithm (with deepSplit set to 2 and minModuleSize set to 50) to identify co-expression modules. Modules with a characteristic gene similarity\u0026thinsp;\u0026gt;\u0026thinsp;greater than 0.75 were merged (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). The module-phenotype association was assessed by calculating the Pearson correlation coefficient between the module signature genes and the disease phenotype (osteoarthritis vs. control). Modules with an absolute correlation value (|r|) greater than 0.7 and a p-value less than 0.05 were considered significantly correlated (blue modules indicating a negative correlation; cyan module indicating a positive correlation) (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC\u0026ndash;D).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.4. Functional enrichment analysis\u003c/h2\u003e\u003cp\u003eThree sets of genes were obtained: WGCNA⁻: 244 genes, which are the intersecting genes in the blue module (negatively correlated with OA). WGCNA⁺: The intersection genes of the cyan module (positively correlated) with DEGs (505). Two WGCNA modules intersect with DEGs (748 genes) (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA\u0026ndash;C). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed using the clusterProfiler package (version 4.6.0). Entries with a corrected p-value of \u0026lt;\u0026thinsp;0.05 were considered significantly enriched, and the results were displayed by bar and dot plots (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD\u0026ndash;F).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e2.5. Plasticizer target prediction and validation\u003c/h2\u003e\u003cp\u003eThe 2D structure and SMILES of di(2-ethylhexyl) phthalate (DEHP) and dibutyl phthalate (DBP) were obtained by Pubchem. SwissTargetPrediction, PharmMapper, STITCH, and SEA databases were used to predict potential targets for plasticizers di(2-ethylhexyl) phthalate (DEHP) and dibutyl phthalate (DBP), and a total of 600 candidate targets were obtained \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. These targets were intersected with phenotype-related DEGs, and 42 key targets associated with the pathogenesis of OA were obtained (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). GO and KEGG enrichment analyses were performed on these intersecting targets to elucidate their biological functions (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB\u0026ndash;C).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e2.6. Machine Learning Identifies Core Genes\u003c/h2\u003e\u003cp\u003eThree machine learning algorithms were employed to identify robust diagnostic biomarkers: LASSO regression, which screened out 9 genes based on the minimum binomial bias (λ = -3.2) using 5-fold cross-validation; SVM-RFE, which selected the feature set corresponding to the highest accuracy (14 features, accuracy\u0026thinsp;=\u0026thinsp;0.918, error rate\u0026thinsp;=\u0026thinsp;0.0817) using the radial basis kernel function; and Random Forest, which set up 500 trees and retained genes with an importance score of \u0026gt;\u0026thinsp;0.5 (24 genes). The genes identified by these three methods were combined, resulting in six core genes: ATF3, SLC2A3, PIM1, VEGFA, RGS4, and TSPO.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e2.7. Validation with an Independent Dataset\u003c/h2\u003e\u003cp\u003e\u003cb\u003eThe\u003c/b\u003e core genes were validated using an independent dataset, GSE12021, which included 9 normal and 10 osteoarthritis (OA) synovial samples. Receiver operating characteristic (ROC) curves were plotted, and the area under the curve (AUC) was calculated to assess diagnostic performance (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB). The expression differences were verified using the Student's t-test (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), and the expression patterns were visualized (Figs.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC\u0026ndash;D).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e2.8. Immunoinfiltration Analysis\u003c/h2\u003e\u003cp\u003eSingle-sample gene set enrichment analysis (ssGSEA) was utilized to quantify the relative enrichment of 22 immune cell types in each sample using the GSVA package. A Spearman correlation coefficient was calculated for core gene expression and immune cell enrichment fraction, with a value\u0026thinsp;\u0026gt;\u0026thinsp;0.3 and p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 considered significantly correlated. The results were visualized using a heatmap (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e2.9. Molecular Docking and Kinetic Simulation\u003c/h2\u003e\u003cp\u003eAutoDock Vina was used to perform molecular docking of DEHP and DBP with the protein products of the 6 core genes. Binding energy \u0026lt;-5.0 kcal/mol was considered indicative of stable binding (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e) \u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Molecular dynamics (MD) simulations were conducted using GROMACS software and AMBER force fields \u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. The system was solvated and neutralized with ions in the TIP3P water model, and after energy minimization and equilibration, a 100 ns production run was executed. Trajectory analyses included the root mean square deviation (RMSD), radius of gyration (Rg), solvent accessible surface area (SASA), root mean square fluctuation (RMSF), and the number of hydrogen bonds (Figs.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eA\u0026ndash;E). The combined free energy was calculated using the MM/PBSA method (Figs.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eF\u0026ndash;I).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e2.10. Adverse Outcome Pathway (AOP) Construction\u003c/h2\u003e\u003cp\u003eFollowing the AOP development principles articulated by Villeneuve et al. (2014), we assembled a conceptual AOP linking phthalate exposure, synovial immune dysregulation and OA-related outcomes\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. The network aims to systematically elucidate the sequence of associations from molecular initiation events (plasticizer exposure), cellular responses, organ-level effects, to adverse outcomes.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Results","content":"\u003cp\u003e\u003cstrong\u003e3.1. Osteoarthritis dataset collection and difference analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore differential gene expression in synovial tissues of osteoarthritis, we processed datasets GSE1919, GSE82107, GSE55457, and GSE55235 from the GEO database by integrating and de-batching via R language (Fig. 1A), followed by standardization (Fig. 1B). We performed differential expression analysis using thresholds of adjusted p-value \u0026lt;0.05 and |log2FC| \u0026gt;0.5, identifying 1394 differentially expressed genes (DEGs); specifically, 707 genes were up-regulated (depicted in red on the volcano plot, Fig. 1C), and 687 were down-regulated (depicted in blue). Additionally, we generated a heatmap for the top 50 DEGs, with red signifying positive correlation and blue signifying negative correlation (Fig. 1D). These findings indicate that these genes may contribute to osteoarthritis mechanisms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.2.\u003c/strong\u003e\u003cstrong\u003eThe most relevant targets of osteoarthritis phenotype were obtained by WGCNA analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo elucidate genes strongly linked to osteoarthritis, we conducted WGCNA analysis on the integrated dataset. We initially removed outlier samples and applied a soft-thresholding power of 7, chosen as the scale-free topology fit index exceeded 0.9 (Fig. 2A), yielding co-expression gene modules (Fig. 2B). This identified five modules with over 50 genes each. Notably, the blue module displayed the strongest negative disease correlation (p \u0026lt; 0.05), whereas the turquoise module manifested the highest positive correlation (p \u0026lt; 0.05) (Fig. 2C). Module-trait association analysis confirmed that both modules correlated with phenotypic traits beyond 0.7 (p \u0026lt; 0.05) (Fig. 2D). Subsequently, we isolated 610 genes from the blue module and 928 genes from the turquoise module, indicating that these genes potentially contribute critically to osteoarthritis mechanisms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.3.\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eintersection targets and GO and KEGG analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo establish phenotypic-genetic linkages in osteoarthritis, we intersected disease-negative correlated (WGCNA-), positive correlated (WGCNA+), and dual-module integrated genes with differentially expressed genes (DEGs) (Fig. 3A-C), obtaining 244, 505, and 748 core genes respectively. Functional enrichment analysis demonstrated: Negative-correlated DEGs exhibited significant enrichment in biological processes including positive regulation of T cell-mediated immunity, leukocyte proliferation, and positive regulation of inflammatory response (P\u0026lt;0.05, Fig. 3D). Positive-correlated and dual-module DEGs consistently enriched in lymphocyte activation, positive regulation of T cell activation, and leukocyte activation involved in immune response (P\u0026lt;0.05, Fig. 3E-F). KEGG pathway analysis revealed negative-correlated DEGs primarily enriched in IL-17 signaling, TNF signaling, NF-kappa B signaling, and cytokine-cytokine receptor interaction (Fig. 3D); positive-correlated DEGs in chemokine signaling, Th17 cell differentiation, and NF-kappa B signaling (Fig. 3E); while dual-module DEGs showed prominent enrichment in IL-17 signaling, TNF signaling, NF-kappa B signaling, and Toll-like receptor signaling (P\u0026lt;0.05, Fig. 3F). These results collectively implicate osteoarthritis as an immune-inflammatory pathology fundamentally driven by dysregulated immune responses and sustained inflammatory activation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.4.\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003etargets of plasticizer-induced osteoarthritis and GO and KEGG enrichment analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo investigate the toxic mechanisms underlying DEHP- and DBP-induced osteoarthritis, we identified 600 potential targets via SwissTargetPrediction, PharmMapper, STITCH, and Similarity Ensemble Approach databases, and intersected them with phenotype-linked differential genes to yield 42 key targets (Fig. 4A). Subsequent GO and KEGG enrichment analyses demonstrated: Biological processes (BP) significantly enriched in regulation of apoptotic signaling pathway, positive regulation of T cell activation, positive regulation of leukocyte cell-cell adhesion, tumor necrosis factor production, and regulation of T cell activation; cellular components (CC) enriched in ficolin-1-rich granule, secretory granule lumen, cytoplasmic vesicle lumen, and vesicle lumen; and molecular functions (MF) enriched in oxidoreductase activity, protein kinase regulator activity, Toll-like receptor binding, endopeptidase activity, and growth factor activity (Fig. 4B). KEGG pathway analysis revealed enrichment in IL-17 signaling pathway, TNF signaling pathway, NF-kappa B signaling pathway, Toll-like receptor signaling pathway, and Th17 cell differentiation (Fig. 4C). These findings imply that DEHP and DBP potentially exacerbate osteoarthritis by dysregulating immune-inflammatory networks, particularly through sustained activation of pathways like IL-17 and TNF signaling.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.5.\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003egenes are screened through machine learning methods\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify core genes tightly linked to the immune-inflammatory phenotype in plasticizer (DEHP and DBP)-induced osteoarthritis, we applied three machine learning algorithms to 42 DEHP/DBP-disease crossover targets. LASSO regression yielded the lowest model deviation (BD = 0.7) at a penalty coefficient \u0026lambda; = -3.2, selecting 9 genes(Fig. 5A). SVM-RFE achieved peak cross-validation accuracy (0.918) and a minimum error rate (0.0817) with 14 features, yielding 14 genes(Fig. 5B). Random forest analysis displayed a stable error rate (\u0026asymp;0.3) at 300 features, identifying 24 genes with importance scores \u0026gt;0.5(Fig. 4C). Integration of these results revealed six core genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, and TSPO), indicating their role as a core molecular network in osteoarthritis(Fig. D).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.6.\u003c/strong\u003e\u003cstrong\u003eIndependent datasets were used to verify the differential expression of core genes and core genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo assess the diagnostic utility of machine learning-derived core genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO) in osteoarthritis, we conducted ROC curve validation using an independent cohort, GSE12021 (9 normal vs. 10 osteoarthritis synovial samples). All genes achieved AUC values exceeding 0.7 (ATF3=0.878, SLC2A3=0.767, PIM1=0.867, VEGFA=0.889, RGS4=0.822, TSPO=0.789), demonstrating robust discriminative capacity (Fig. 6B). Differential expression analysis consistently revealed significant downregulation of ATF3, SLC2A3, PIM1, and VEGFA (P\u0026lt;0.05), alongside upregulation of RGS4 and TSPO (P\u0026lt;0.05) in osteoarthritic synovium across both primary and validation datasets (Fig. 6C-D). These results underscore the viability of these genes as diagnostic or phenotypic biomarkers for osteoarthritis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.7.\u003c/strong\u003e\u003cstrong\u003eImmunoinfiltration analysis of core genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo delineate differences in synovial immune cell composition between healthy individuals and osteoarthritis (OA) patients and their association with core genes, we conducted single-sample gene set enrichment analysis (ssGSEA) across 22 immune cell types. Our analysis identified significant differences (P\u0026lt;0.05) between healthy and OA synovium for plasma cells, CD8+ T cells, CD4+ memory resting T cells, T follicular helper cells, gamma delta T cells, M0 macrophages, activated dendritic cells, resting mast cells, activated mast cells, and eosinophils (Figure 7A, B). Specific correlations among immune cell types were also evident (Figure 7C); notably, CD4+ memory resting T cells correlated positively with activated mast cells, whereas plasma cells correlated positively with both T follicular helper cells and gamma delta T cells (P\u0026lt;0.05). In contrast, CD4+ memory resting T cells correlated negatively with T follicular helper cells, M0 macrophages, and resting mast cells (P\u0026lt;0.05). Immune infiltration analysis of core genes implicated in phthalate-induced OA pathogenesis (Figure 7D) demonstrated that ATF3, PIM1, SLC2A3, and VEGFA each correlated positively with CD4+ memory resting T cells, activated mast cells, and activated dendritic cells, but negatively with plasma cells and resting mast cells (P\u0026lt;0.05). Specifically, ATF3 also showed negative correlations with T follicular helper cells and M0 macrophages; SLC2A3 correlated negatively with T follicular helper cells; and VEGFA correlated negatively with gamma delta T cells, T follicular helper cells, and M0 macrophages, yet correlated positively with eosinophils (P\u0026lt;0.05). Both RGS4 and TSPO correlated negatively with activated dendritic cells (P\u0026lt;0.05); further, RGS4 correlated positively with gamma delta T cells and T follicular helper cells, while TSPO correlated negatively with eosinophils (P\u0026lt;0.05). These findings collectively delineate a distinct immune cell landscape characteristic of OA synovium and establish direct functional connections between core pathogenic genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO), identified via phthalate toxicity screening, and dysregulated key immune cell populations in OA.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.8.\u003c/strong\u003e\u003cstrong\u003emolecular docking\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate potential interactions between the plasticizer di(2-ethylhexyl) phthalate (DEHP) and six key osteoarthritis (OA)-related target proteins (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO), we conducted molecular docking analysis. Our results revealed strong binding affinities of DEHP for TSPO, PIM1, and SLC2A3, with binding energies of -7.7 kcal/mol, -7.7 kcal/mol, and -7.6 kcal/mol, respectively (Figure 8A-C). DEHP also demonstrated potential binding to ATF3, RGS4, and VEGFA, yielding binding energies of -5.2 kcal/mol, -5.2 kcal/mol, and -5.0 kcal/mol, respectively (Figure 8D-E). As all calculated binding energies fell below the threshold of -5.0 kcal/mol, the docking poses confirmed spontaneous interactions between DEHP and each of these target proteins (Figure 8). Applying standard interpretation (binding energy \u0026lt; -5.0 kcal/mol suggests potential binding; \u0026lt; -7.0 kcal/mol indicates strong binding), DEHP exhibited particularly robust binding affinities for TSPO, PIM1, and SLC2A3 (all energies \u0026lt; -7.0 kcal/mol). This molecular evidence confirms the core toxicological mechanism: DEHP directly binds to and disrupts key target proteins (notably TSPO, PIM1, and SLC2A3), thereby perturbing immune microenvironment homeostasis and driving OA pathogenesis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.9.\u003c/strong\u003e\u003cstrong\u003eMolecular dynamics simulation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the conformational stability of DEHP complexes with the key osteoarthritis-related target proteins TSPO, SLC2A3, and PIM1, we performed molecular dynamics (MD) simulations coupled with MM/PBSA binding free energy calculations. MD trajectories demonstrated stable complex formation: RMSD values equilibrated at 0.6 \u0026plusmn; 0.05 nm after ~65 ns (DEHP-TSPO), 0.25 \u0026plusmn; 0.05 nm after ~30 ns (DEHP-SLC2A3), and 0.23 \u0026plusmn; 0.05 nm after ~10 ns (DEHP-PIM1) (Figure 9A). Radius of gyration (Rg) analysis revealed convergence: Rg decreased from 1.8 nm to stabilize at 1.70 \u0026plusmn; 0.04 nm after ~60 ns (DEHP-TSPO), while remaining constant at 2.26 \u0026plusmn; 0.02 nm (DEHP-SLC2A3) and 1.90 \u0026plusmn; 0.02 nm (DEHP-PIM1) throughout the simulations (Figure 9B). Solvent-accessible surface area (SASA) profiles indicated stability: SASA stabilized at 100 \u0026plusmn; 5 nm\u0026sup2; after 5 ns (DEHP-TSPO), 200 \u0026plusmn; 10 nm\u0026sup2; after 55 ns (DEHP-SLC2A3), and remained at 140 \u0026plusmn; 8 nm\u0026sup2; (DEHP-PIM1) (Figure 9C). Root-mean-square fluctuation (RMSF) was low (\u0026lt; 0.4 nm) for key residues across all complexes, excluding termini (Figure 9D). Hydrogen bond occupancy analysis showed the DEHP-TSPO complex predominantly maintained two bonds, DEHP-SLC2A3 maintained one bond, and DEHP-PIM1 maintained one bond only ~33% of the time (Figure 9E). Free energy landscapes identified minima corresponding to stable conformations at Rg/RMSD values of 1.67-1.70 nm / 0.55-0.60 nm (DEHP-TSPO), 2.25-2.27 nm / 0.23-0.27 nm (DEHP-SLC2A3), and 1.89-1.92 nm / 0.19-0.24 nm (TRPC3) (Figure 9G-I). MM/PBSA calculations confirmed strong binding, yielding \u0026Delta;G values of -185.302 kJ/mol (TSPO), -175.321 kJ/mol (SLC2A3), and -186.664 kJ/mol (PIM1) (Figure 9F). Energy decomposition attributed binding primarily to favorable van der Waals interactions (VDW: -303.206, -257.841, -250.976 kJ/mol), significantly opposed by polar solvation penalties (PB: +68.448, +63.846, +64.451 kJ/mol). Nonpolar solvation energy (SA: -26.343, -27.691, -25.899 kJ/mol) provided stabilization, while entropic contributions (-T\u0026Delta;S: +83.543, +51.293, +28.939 kJ/mol) partially counteracted the enthalpic gain. The molecular mechanics energy (MM: -309.83, -261.706, -255.365 kJ/mol) was dominated by VDW, with minor Coulombic components (COU: -6.625, -3.865, -4.39 kJ/mol). Collectively, these MD and MM/PBSA results provide robust evidence for the direct, strong, and stable binding of DEHP to the core OA-associated targets TSPO, SLC2A3, and PIM1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.10.\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eOutcome Pathways (AOP) framework\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSubsequently, we establish a new adverse outcome pathway (AOP) framework based on the method proposed by Daniel et al. In this AOP network, the expression and activity of ATF3, SLC2A3, PIM1, TSPO, VEGFA, and RGS4 may be affected by plasticizers (DEHP and DBP). In addition, the IL-17/TNF/NF-\u0026kappa;B signaling pathway may also be involved, leading to synovial immune dysregulation, which in turn leads to synovial inflammation and oxidative stress, and ultimately to the occurrence and development of osteoarthritis. This study successfully constructed a comprehensive AOP framework by systematically integrating the identified key central genes, which provides a theoretical basis for elucidating the mechanism of arthritis exacerbated by plasticizers (DEHP and DBP) (Fig. 10).\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eIn this study, we systematically applied multi-omics analysis, machine learning approaches, and computational simulations to elucidate the contribution of the environmental plasticizers DEHP and DBP to the pathogenesis of osteoarthritis (OA). For the first time, we demonstrate that these widespread endocrine-disrupting chemicals may exacerbate OA development through perturbations of synovial immune homeostasis. Six core genes (ATF3, SLC2A3, PIM1, VEGFA, RGS4, TSPO) were identified as central to synovial immune inflammation and demonstrated strong diagnostic utility. Furthermore, molecular docking and molecular dynamics simulations confirmed stable binding of DEHP to proteins such as TSPO, PIM1, and SLC2A3, suggesting a direct molecular mechanism of interference with synovial immune microenvironment balance. Finally, by integrating genetic, pathway, and toxicological evidence, we constructed an adverse outcome pathway (AOP) that conceptually links phthalate exposure to immune dysregulation, synovial inflammation, and the progression of OA.\u003c/p\u003e\u003cp\u003eThese results highlight the central role of immune dysfunction, particularly involving macrophage and T-cell subsets, in OA synovitis \u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Both DEGs and predicted phthalate targets were strongly enriched in canonical inflammatory cascades, including IL-17, TNF, and NF-κB signaling, consistent with established models of chronic OA inflammation \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. In contrast to most prior studies that have focused on biomechanical stress or endogenous damage-associated signals (DAMPs) as the initiating factors of OA, our findings suggest that environmental plasticizers may act as important exogenous drivers of synovial inflammation. For example, TSPO, PIM1, and SLC2A3 displayed significant expression changes in OA synovium and bound to DEHP with high affinity, indicating that chemical exposures may directly trigger or modulate OA-associated immune responses\u0026mdash;a mechanistic link not previously reported.\u003c/p\u003e\u003cp\u003eSome of the results further expand on the knowledge of previous studies. The downregulation of ATF3 and VEGFA in the synovial membrane of OA is consistent with their known inhibitory effects on inflammation and regulation of angiogenesis\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. However, they also serve as potential targets of DEHP/DBP, which provides a new basis for understanding the mechanistic connection between environmental exposure and OA occurrence. Similarly, although infiltration of immune cells (such as M0 macrophages and activated dendritic cells) in the OA synovial membrane has been reported\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e, its significant association with plasticizer-sensitive genes provides a new perspective on how chemical exposure alters the synovial immune microenvironment.\u003c/p\u003e\u003cp\u003eFuture research should therefore focus on experimental verification of the effects of DEHP/DBP on core immune-related genes and signaling pathways within OA synovial tissue models. Prospective studies are also needed to evaluate the epidemiological association between plasticizer exposure and OA development.\u003c/p\u003e\u003cp\u003eIn conclusion, this study proposes a novel mechanism by which environmental plasticizers promote OA through disruption of synovial immune-inflammatory pathways. By identifying core molecular targets and constructing an AOP framework, our findings extend current understanding of environmental contributions to OA etiology and provide potential biomarkers and therapeutic targets for interventions aimed at reducing plasticizer-related health risks.\u003c/p\u003e\u003cp\u003e\u003cb\u003eLimitations\u003c/b\u003e\u003c/p\u003e\u003cp\u003eFirst, although four public synovial cohorts were integrated, overall sample size remained modest and cross-platform heterogeneity may influence generalizability. Second, immune-cell inference from bulk transcriptomes (ssGSEA with predefined gene sets) provides relative enrichment rather than absolute cell fractions and is sensitive to gene-set choice. Third, target prediction, docking and MM/PBSA are in-silico and contain known approximations; they support hypotheses but do not establish binding or causality. Finally, the proposed AOP is conceptual and requires experimental and epidemiological corroboration.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eOA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eosteoarthritis\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eDEHP\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003edi(2\u0026mdash;ethylhexyl) phthalate\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eDBP\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003edibutyl phthalate\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eEDCs\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eendocrine\u0026mdash;disrupting chemicals\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eDAMPs\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003edamage\u0026mdash;associated molecular patterns\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eGEO\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eGene Expression Omnibus\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eGSE\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eGEO Series\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eDEGs\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003edifferentially expressed genes\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eWGCNA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eweighted gene co\u0026mdash;expression network analysis\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eTOM\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003etopological overlap matrix\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eGO\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eGene Ontology\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eBP\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003ebiological process\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eCC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003ecellular component\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eMF\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003emolecular function\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eKEGG\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eKyoto Encyclopedia of Genes and Genomes\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePCA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eprincipal component analysis\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eLASSO\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eleast absolute shrinkage and selection operator\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eSVM\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eRFE\u0026mdash;support vector machine\u0026mdash;recursive feature elimination\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eROC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003ereceiver operating characteristic\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eAUC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003earea under the curve\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eFDR\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003efalse discovery rate\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eCI\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003econfidence interval\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eSD\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003estandard deviation\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eSEM\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003estandard error of the mean\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003essGSEA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003esingle\u0026mdash;sample gene set enrichment analysis\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eMD\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003emolecular dynamics\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eRMSD\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eroot\u0026mdash;mean\u0026mdash;square deviation\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eRMSF\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eroot\u0026mdash;mean\u0026mdash;square fluctuation\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eSASA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003esolvent accessible surface area\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eRg\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eradius of gyration\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eFEL\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003efree\u0026mdash;energy landscape\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eMM/PBSA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003emolecular mechanics/Poisson\u0026mdash;Boltzmann surface area\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eAOP\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eadverse outcome pathway\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eFNIH\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eFoundation for the National Institutes of Health\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eNIH\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eNational Institutes of Health\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eTNF\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003etumor necrosis factor\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eIL\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003e17\u0026mdash;interleukin\u0026mdash;17\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eConflict of interest:\u0026nbsp;\u003c/strong\u003eThere are no conflicts of interest for this manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication:\u0026nbsp;\u003c/strong\u003eAll authors got the consent for publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declaration:\u0026nbsp;\u003c/strong\u003enot applicable\u003cstrong\u003e.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;Not applicable. This study analyzed publicly available, de-identified gene-expression datasets and did not involve human participants, human tissue, or animals.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;Informed consent for publication of the manuscript was obtained from all authors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;All transcriptome datasets analyzed are publicly available from the Gene Expression Omnibus (GEO): GSE1919, GSE82107, GSE55235, GSE55457 and validation set GSE12021. Processed data and analysis scripts are available from the corresponding author on reasonable request; dataset accession numbers are cited in the Methods.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;The authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;This work was supported by the Nanchang University Graduate Innovation Special Fund, Jiangxi Province (Grant No. YC2025-S305). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\u003c/p\u003e\n\u003cp\u003eYC2025-S305， 2025 Nanchang University Graduate Innovation Special Fund, Jiangxi Province\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;Xiaopeng Lin conceived and supervised the study.Yang Liu and Chong Li curated datasets and performed bioinformatics analyses. Yang Liu conducted molecular docking and molecular dynamics. Yang Liu contributed to data interpretation and revisions. Yang Liu drafted the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;We sincerely thank GEO for generously sharing their data. We also gratefully acknowledge the financial support from Nanchang University Graduate Student Innovation Fund.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; information\u003c/strong\u003e\u003cbr\u003e Affiliated Rehabilitation Hospital of Nanchang University, Address: No. 133 South Square Road, Nanchang City, Nanchang (330003), Jiangxi,China，Xiaopeng Lin ，Yang Liu\u003csup\u003e1\u003c/sup\u003e, Chong Li\u003csup\u003e2\u003c/sup\u003e,Xinmin Yang \u003csup\u003e3\u003c/sup\u003e,Fuheng Ma \u003csup\u003e4\u003c/sup\u003e,Fan Wang \u003csup\u003e5\u003c/sup\u003e,Qixin Liu\u003csup\u003e6\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eNanchangUniversity,Address:999 Xuefu Road, Honggutan District, Nanchang(330019), Jiangxi, China，Xiaopeng Lin ，Yang Liu\u003csup\u003e1\u003c/sup\u003e, Chong Li\u003csup\u003e2\u003c/sup\u003e,Xinmin Yang \u003csup\u003e3\u003c/sup\u003e,Fuheng Ma \u003csup\u003e4\u003c/sup\u003e,Fan Wang \u003csup\u003e5\u003c/sup\u003e,Qixin Liu\u003csup\u003e6\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eZhang, Q., Sun, C., Liu, X., Zhu, C., Ma, C., \u0026amp; Feng, R. (2023). Mechanism of immune infiltration in synovial tissue of osteoarthritis: a gene expression-based study. Journal of Orthopaedic Surgery and Research, 18(1). https://doi.org/10.1186/s13018-023-03541-x\u003c/li\u003e\n\u003cli\u003eWoodell‐May, J. E., \u0026amp; Sommerfeld, S. D. (2019). Role of Inflammation and the Immune System in the Progression of Osteoarthritis. Journal of Orthopaedic Research, 38(2), 253\u0026ndash;257. https://doi.org/10.1002/jor.24457\u003c/li\u003e\n\u003cli\u003eKnights, A. J., Redding, S. J., \u0026amp; Maerz, T. (2022). Inflammation in osteoarthritis: the latest progress and ongoing challenges. Current Opinion in Rheumatology, 35(2), 128\u0026ndash;134. https://doi.org/10.1097/bor.0000000000000923\u003c/li\u003e\n\u003cli\u003eXuan, Q., Ruan, Y., Cao, C., Yin, Z., Du, J., Lv, C., Fang, F., \u0026amp; Gu, W. (2022). Effect of Ultrasonic Penetration with Volatile Oil of Olibanum and Chuanxiong Rhizoma on Acute Knee Synovitis Induced by Sports Training: An Open-Label Randomized Controlled Study. Pain Research and Management, 2022, 1\u0026ndash;10. https://doi.org/10.1155/2022/6806565\u003c/li\u003e\n\u003cli\u003eRamezanpour, S., Kanthawang, T., Lynch, J., McCulloch, C. E., Nevitt, M. C., Link, T. M., \u0026amp; Joseph, G. B. (2022). Impact of Sustained Synovitis on Knee Joint Structural Degeneration: \u0026amp;#60;scp\u0026amp;#62;4‐Year MRI\u0026amp;#60;/scp\u0026amp;#62; Data from the Osteoarthritis Initiative. Journal of Magnetic Resonance Imaging, 57(1), 153\u0026ndash;164. https://doi.org/10.1002/jmri.28223\u003c/li\u003e\n\u003cli\u003ePanichi, V., Costantini, S., Grasso, M., Arciola, C. R., \u0026amp; Dolzani, P. (2024). Innate Immunity and Synovitis: Key Players in Osteoarthritis Progression. International Journal of Molecular Sciences, 25(22), 12082. https://doi.org/10.3390/ijms252212082\u003c/li\u003e\n\u003cli\u003eCheng, J.-H., Hsu, C.-C., Hsu, S.-L., Chou, W.-Y., Wu, Y.-N., Kuo, C.-E. A., Hsu, T.-C., Shiu, L.-Y., \u0026amp; Jhan, S.-W. (2021). Adipose-Derived Mesenchymal Stem Cells-Conditioned Medium Modulates the Expression of Inflammation Induced Bone Morphogenetic Protein-2, -5 and -6 as Well as Compared with Shockwave Therapy on Rat Knee Osteoarthritis. Biomedicines, 9(10), 1399. https://doi.org/10.3390/biomedicines9101399\u003c/li\u003e\n\u003cli\u003eLi, Y.-N., Fan, M.-L., Liu, H.-Q., Ma, B., Dai, W.-L., Yu, B.-Y., \u0026amp; Liu, J.-H. (2019). Dihydroartemisinin derivative DC32 inhibits inflammatory response in osteoarthritic synovium through regulating Nrf2/NF-\u0026kappa;B pathway. International Immunopharmacology, 74, 105701. https://doi.org/10.1016/j.intimp.2019.105701\u003c/li\u003e\n\u003cli\u003eJiang, H., Zhang, Y., Hu, G., Shang, X., Ming, J., Deng, M., Li, Y., Ma, Y., Liu, S., \u0026amp; Zhou, Y. (2022). Innate/Inflammatory Bioregulation of Surfactant Protein D Alleviates Rat Osteoarthritis by Inhibiting Toll-Like Receptor 4 Signaling. Frontiers in Immunology, 13. https://doi.org/10.3389/fimmu.2022.913901\u003c/li\u003e\n\u003cli\u003eCheleschi, S., Gallo, I., Barbarino, M., Giannotti, S., Mondanelli, N., Giordano, A., Tenti, S., \u0026amp; Fioravanti, A. (2019). MicroRNA Mediate Visfatin and Resistin Induction of Oxidative Stress in Human Osteoarthritic Synovial Fibroblasts Via NF-\u0026kappa;B Pathway. International Journal of Molecular Sciences, 20(20), 5200. https://doi.org/10.3390/ijms20205200\u003c/li\u003e\n\u003cli\u003eWang, X., Chen, T., Liang, W., Fan, T., Zhu, Z., Cao, P., Ruan, G., Zhang, Y., Chen, S., Wang, Q., Li, S., Huang, Y., Zeng, M., Hunter, D. J., Li, J., \u0026amp; Ding, C. (2022). Synovitis mediates the association between bone marrow lesions and knee pain in osteoarthritis: data from the Foundation for the National Institute of Health (FNIH) Osteoarthritis Biomarkers Consortium. Osteoarthritis and Cartilage, 30(9), 1270\u0026ndash;1277. https://doi.org/10.1016/j.joca.2022.06.004\u003c/li\u003e\n\u003cli\u003eKnights, A. J., Redding, S. J., \u0026amp; Maerz, T. (2022). Inflammation in osteoarthritis: the latest progress and ongoing challenges. Current Opinion in Rheumatology, 35(2), 128\u0026ndash;134. https://doi.org/10.1097/bor.0000000000000923\u003c/li\u003e\n\u003cli\u003eThoenen, J., MacKay, J. W., Sandford, H. J. C., Gold, G. E., \u0026amp; Kogan, F. (2022). Imaging of Synovial Inflammation in Osteoarthritis, From the AJR Special Series on Inflammation. American Journal of Roentgenology, 218(3), 405\u0026ndash;417. https://doi.org/10.2214/ajr.21.26170\u003c/li\u003e\n\u003cli\u003eTheeuwes, W. F., van den Bosch, M. H. J., Thurlings, R. M., Blom, A. B., \u0026amp; van Lent, P. L. E. M. (2021). The role of inflammation in mesenchymal stromal cell therapy in osteoarthritis, perspectives for post-traumatic osteoarthritis: a review. Rheumatology, 60(3), 1042\u0026ndash;1053. https://doi.org/10.1093/rheumatology/keaa910\u003c/li\u003e\n\u003cli\u003eJiang, T., Weng, Q., Liu, K., He, H., Zhang, Y., Zhang, W., Doherty, M., Xie, J., Yang, T., Li, J., Yang, Z., Chen, Q., Long, H., Wang, Y., Wei, J., Lei, G., \u0026amp; Zeng, C. (2024). Bidirectional association identified between synovitis and knee and hand osteoarthritis: a general population-based study. The Lancet Regional Health - Western Pacific, 50, 101169. https://doi.org/10.1016/j.lanwpc.2024.101169\u003c/li\u003e\n\u003cli\u003eWerner, D. M., Golightly, Y. M., Tao, M., Post, A., \u0026amp; Wellsandt, E. (2022). Environmental Risk Factors for Osteoarthritis: The Impact on Individuals with Knee Joint Injury. Rheumatic Disease Clinics of North America, 48(4), 907\u0026ndash;930. https://doi.org/10.1016/j.rdc.2022.06.010\u003c/li\u003e\n\u003cli\u003eWerner, D. M., Golightly, Y. M., Tao, M., Post, A., \u0026amp; Wellsandt, E. (2022). Environmental Risk Factors for Osteoarthritis: The Impact on Individuals with Knee Joint Injury. Rheumatic Disease Clinics of North America, 48(4), 907\u0026ndash;930. https://doi.org/10.1016/j.rdc.2022.06.010\u003c/li\u003e\n\u003cli\u003eLi, L., Huang, L., Lei, R., Zhang, P., Yang, Y., Liu, H., \u0026amp; Zhang, Y. (2024). DEHP and DBP, common phthalates, induce glucose metabolism disorders in rats via oxidative damage of PI3K/Akt/GLUT4 signaling. Environmental Pollution, 341, 122948. https://doi.org/10.1016/j.envpol.2023.122948\u003c/li\u003e\n\u003cli\u003eSun, C., Li, K., Zhang, J., \u0026amp; Huang, C. (2021). Prediction of phthalates concentration in household dust based on back propagation neural network. Indoor and Built Environment, 31(1), 230\u0026ndash;244. https://doi.org/10.1177/1420326x211010212\u003c/li\u003e\n\u003cli\u003eCui, Y., Yin, K., Zheng, Y., Wang, B., Qu, Y., Li, S., \u0026amp; Lin, H. (2021). Mixed plasticizers aggravated apoptosis by NOD2-RIP2-NF-\u0026kappa;B pathway in grass carp hepatocytes. Journal of Hazardous Materials, 402, 123527. https://doi.org/10.1016/j.jhazmat.2020.123527\u003c/li\u003e\n\u003cli\u003eWang, S., Xu, K., Du, W., Gao, X., Ma, P., Yang, X., \u0026amp; Chen, M. (2024). Exposure to environmental doses of DEHP causes phenotypes of polycystic ovary syndrome. Toxicology, 509, 153952. https://doi.org/10.1016/j.tox.2024.153952\u003c/li\u003e\n\u003cli\u003eSherry, S. T., Ward, M. H., Kholodov, M., Baker, J., Phan, L., Smigielski, E. M., \u0026amp; Sirotkin, K. (2001). dbSNP: the NCBI database of genetic variation. Nucleic acids research, 29(1), 308\u0026ndash;311. https://doi.org/10.1093/nar/29.1.308\u003c/li\u003e\n\u003cli\u003eDaina, A., Michielin, O., \u0026amp; Zoete, V. (2019). SwissTargetPrediction: updated data and new features for efficient prediction of protein targets of small molecules. Nucleic acids research, 47(W1), W357\u0026ndash;W364. https://doi.org/10.1093/nar/gkz382\u003c/li\u003e\n\u003cli\u003eLiu, X., Ouyang, S., Yu, B., Liu, Y., Huang, K., Gong, J., Zheng, S., Li, Z., Li, H., \u0026amp; Jiang, H. (2010). PharmMapper server: a web server for potential drug target identification using pharmacophore mapping approach. Nucleic acids research, 38(Web Server issue), W609\u0026ndash;W614. https://doi.org/10.1093/nar/gkq300\u003c/li\u003e\n\u003cli\u003eHeather, J. M., Spindler, M. J., Alonso, M. H., Shui, Y. I., Millar, D. G., Johnson, D. S., Cobbold, M., \u0026amp; Hata, A. N. (2022). Stitchr: stitching coding TCR nucleotide sequences from V/J/CDR3 information. Nucleic acids research, 50(12), e68. https://doi.org/10.1093/nar/gkac190\u003c/li\u003e\n\u003cli\u003eWei, Y., Zhang, S., Shang, S., Zhang, B., Li, S., Wang, X., Wang, F., Su, J., Wu, Q., Liu, H., \u0026amp; Zhang, Y. (2016). SEA: a super-enhancer archive. Nucleic acids research, 44(D1), D172\u0026ndash;D179. https://doi.org/10.1093/nar/gkv1243\u003c/li\u003e\n\u003cli\u003eBugnon, M., R\u0026ouml;hrig, U. F., Goullieux, M., Perez, M. A. S., Daina, A., Michielin, O., \u0026amp; Zoete, V. (2024). SwissDock 2024: major enhancements for small-molecule docking with Attracting Cavities and AutoDock Vina. Nucleic acids research, 52(W1), W324\u0026ndash;W332. https://doi.org/10.1093/nar/gkae300\u003c/li\u003e\n\u003cli\u003eSer\u0026ccedil;inoglu, O., \u0026amp; Ozbek, P. (2018). gRINN: a tool for calculation of residue interaction energies and protein energy network analysis of molecular dynamics simulations. Nucleic acids research, 46(W1), W554\u0026ndash;W562. https://doi.org/10.1093/nar/gky381\u003c/li\u003e\n\u003cli\u003eVilleneuve, D. L., Crump, D., Garcia-Reyero, N., Hecker, M., Hutchinson, T. H., LaLone, C. A., Landesmann, B., Lettieri, T., Munn, S., Nepelska, M., Ottinger, M. A., Vergauwen, L., \u0026amp; Whelan, M. (2014). Adverse outcome pathway (AOP) development I: strategies and principles. Toxicological sciences : an official journal of the Society of Toxicology, 142(2), 312\u0026ndash;320. https://doi.org/10.1093/toxsci/kfu199\u003c/li\u003e\n\u003cli\u003eLi, W., Liu, Y., Wei, M., Yang, Z., Li, Z., Guo, Z., Yan, L., Lu, Y., Tang, H., Li, B., \u0026amp; Huang, W. (2025). Functionalized Biomimetic Nanoparticles Targeting the IL‐10/IL‐10R\u0026alpha;/Glycolytic Axis in Synovial Macrophages Alleviate Cartilage Degeneration in Osteoarthritis. Advanced Science, 12(27). https://doi.org/10.1002/advs.202504768\u003c/li\u003e\n\u003cli\u003eZhu, J., Wei, J., Lin, Y., Tang, Y., Su, Z., Li, L., Liu, B., \u0026amp; Cai, X. (2024). Inhibition of IL-17 signaling in macrophages underlies the anti-arthritic effects of halofuginone hydrobromide: Network pharmacology, molecular docking, and experimental validation. BMC Complementary Medicine and Therapies, 24(1). https://doi.org/10.1186/s12906-024-04397-2\u003c/li\u003e\n\u003cli\u003eSharma, P., Malhotra, H., Garg, V., Singh, G., Ashri, A., \u0026amp; Kamal, M. A. (2022). MECHANISTIC TARGETING OF TNF-\u0026alpha; AND NF-\u0026kappa;B IN INFLAMMATORY DISEASES. In International Journal of Cell Biology and Cellular Processes 8 (2), 14-23, 2022.\u003c/li\u003e\n\u003cli\u003eLiang, Y., Jiang, Y., Jin, X., Chen, P., Heng, Y., Cai, L., Zhang, W., Li, L., \u0026amp; Jia, L. (2020). Neddylation inhibition activates the protective autophagy through NF-\u0026kappa;B-catalase-ATF3 Axis in human esophageal cancer cells. Cell Communication and Signaling, 18(1). https://doi.org/10.1186/s12964-020-00576-z\u003c/li\u003e\n\u003cli\u003eAbdelnaby, R., Sonbol, Y. T., Dardeer, K. T., ELgenidy, A., Ebrahim, M. A., Elmenawi, K. A., Abdallah, A. A., Ballut, O. O., Metwally, S. A., Arafat, M. S., Elsayed, M., Abouelhamd, A. M., Farahat, R. A., Karkour, A. M., Mohamed, K. A., \u0026amp; Rauen, T. (2023). Could Osteopontin be a useful biomarker in the diagnosis and severity assessment of osteoarthritis? A systematic review and meta-analysis of recent evidence. Clinical Immunology, 246, 109187. https://doi.org/10.1016/j.clim.2022.109187\u003c/li\u003e\n\u003cli\u003eLi, W., Liu, Y., Wei, M., Yang, Z., Li, Z., Guo, Z., Yan, L., Lu, Y., Tang, H., Li, B., \u0026amp; Huang, W. (2025). Functionalized Biomimetic Nanoparticles Targeting the IL‐10/IL‐10R\u0026alpha;/Glycolytic Axis in Synovial Macrophages Alleviate Cartilage Degeneration in Osteoarthritis. Advanced Science, 12(27). https://doi.org/10.1002/advs.202504768\u003c/li\u003e\n\u003cli\u003eMacečkov\u0026aacute;, Z., Pergner, J., Krbec, M., Urban, M., \u0026amp; Zahradn\u0026iacute;ček, M. (2025). Application of amniotic membrane in osteoarthritis management. Journal of Cartilage \u0026amp;amp; Joint Preservation, 5(1), 100174. https://doi.org/10.1016/j.jcjp.2024.100174\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Phthalates, DEHP, DBP, Osteoarthritis, Synovium, Immune inflammation, WGCNA, Machine learning, Molecular docking, Molecular dynamics","lastPublishedDoi":"10.21203/rs.3.rs-7815618/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7815618/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground: \u003c/strong\u003eOsteoarthritis (OA) is increasingly recognized as an immune-inflammatory disease of the joint, with synovial dysfunction contributing to cartilage degeneration. Phthalate plasticizers such as di(2-ethylhexyl) phthalate (DEHP) and dibutyl phthalate (DBP) are ubiquitous environmental toxicants; however, their potential association with synovial immune perturbation in OA remains unclear.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods: \u003c/strong\u003eWe integrated four synovial-tissue transcriptome datasets from GEO (GSE1919, GSE82107, GSE55235, GSE55457) and performed batch correction and differential expression analysis, followed by weighted gene co-expression network analysis (WGCNA) to identify phenotype-associated modules. Three machine-learning algorithms (LASSO, SVM-RFE, Random Forest) were used to select robust markers, which were validated in an independent dataset (GSE12021). Immune infiltration was profiled by ssGSEA. DEHP/DBP targets were predicted using cheminformatics resources and intersected with phenotype-related genes. Molecular docking and all-atom molecular dynamics (MD) simulations evaluated the binding modes and stability of DEHP with core protein targets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults: \u003c/strong\u003eWe identified 1,394 differentially expressed genes and OA-associated WGCNA modules enriched for cytokine signaling and leukocyte activation. Intersecting phthalate-predicted targets with phenotype-related and differentially expressed genes yielded plasticizer-linked OA candidate genes. Convergent machine-learning selection nominated six core genes—ATF3, SLC2A3, PIM1, VEGFA, RGS4 and TSPO—showing consistent dysregulation and diagnostic utility across training and validation cohorts. ssGSEA indicated altered synovial immune-cell landscapes, and several core genes correlated with pro-inflammatory cell subsets. Docking and MD suggested comparatively stable binding of DEHP to TSPO, SLC2A3 and PIM1, supporting a plausible molecular interface between phthalate exposure and synovial immune dysregulation in OA.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions: \u003c/strong\u003eThis integrative network-toxicology study links DEHP/DBP exposure to synovial immune–inflammatory signatures in OA and prioritizes six core genes as potential biomarkers or mechanistic candidates. The computational findings generate testable hypotheses for experimental validation and may inform exposure-mitigation and therapeutic strategies.\u003c/p\u003e","manuscriptTitle":"Network Toxicology and Bioinformatics Reveal Synovial Immune-Inflammatory Mechanisms of DEHP/DBP in Osteoarthritis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-18 04:02:45","doi":"10.21203/rs.3.rs-7815618/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"cac67996-b178-4328-b8fb-606fa3c5eb6b","owner":[],"postedDate":"November 18th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-02T23:08:15+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-18 04:02:45","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7815618","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7815618","identity":"rs-7815618","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}