Integrated Metabolomic and Transcriptomic Analysis Identifies Candidate Regulatory Networks in Bone Metabolic Dysregulation of Osteoarthritis

Integrated Metabolomic and Transcriptomic Analysis Identifies Candidate Regulatory Networks in Bone Metabolic Dysregulation of 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 Article Integrated Metabolomic and Transcriptomic Analysis Identifies Candidate Regulatory Networks in Bone Metabolic Dysregulation of Osteoarthritis Yang Lu, Shasha Wang, Jinkun Li, Jianpin Gan, Fenglin Zhu, Qin Shao, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6359791/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 Osteoarthritis (OA) represents a prevalent articular condition characterized by significant disturbances in bone metabolic balance, yet the molecular pathways governing these skeletal alterations remain inadequately defined.This study conducted metabolomic and transcriptomic analyses of 30 OA patients and 30 controls, revealing significant alterations in purine, lipid, and amino acid metabolism in OA patients. Retinol metabolism and ubiquinone biosynthesis showed increased activity, while histidine metabolism decreased. Key metabolites correlated with bone turnover markers (β-CTX, PINP, N-MID). Transcriptomic analysis identified differentially expressed mRNAs (DEGs) involved in apoptosis, inflammation, and lipid metabolism. Integrated analysis revealed candidate regulators connecting IL-17, MAPK, histidine, and fatty acid signaling pathways, providing potential intervention targets for OA bone metabolic abnormalities. Health sciences/Rheumatology/Osteoimmunology Health sciences/Biomarkers Health sciences/Pathogenesis Health sciences/Rheumatology Osteoarthritis Bone homeostasis Metabolomics Transcriptomics Regulatory networks Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction OA constitutes a frequently encountered deteriorative disorder of the locomotor system that typically manifests with articular nociception, compromised joint function, and architectural derangements of affected articulations 1 . Recent Global Burden of Disease studies indicate that over the last 30 years (1990–2020), the number of cases of OA has increased by 339 million (132%), currently affecting approximately 7.6% of the world's population, constituting a major public health challenge. Females are more likely than males to have this disorder, and its prevalence increases notably with age 2 , 3 . Multiple risk factors including aging, joint injuries, metabolic imbalances, changes in bone density, obesity, and sex variations collectively contribute to OA development 4 . From a pathological standpoint, OA distinguishes itself through several hallmark tissue alterations: aberrant osseous projections at joint peripheries, pathological hardening of bone beneath cartilage surfaces, inflammatory transformation of synovial linings, and structural compromise of articular cartilage, all functioning within a sophisticated network of reciprocal pathological interactions 5 . While OA has long been seen as a disease of cartilage, growing evidence shows that abnormal bone metabolism serves as a critical determinant in OA progression 6 . Clinical studies have found that bone metabolic markers in OA patients differ significantly from healthy individuals, including changes in standardized uptake values (SUV) and bone metabolic kinetic parameter K 7 . Some OA patients show increased levels of bone sialoprotein, osteocalcin, and abnormal RANKL/OPG ratios 8 , 9 . Notably, these bone changes often occur before cartilage damage and are closely related to disease severity 10 . However, despite these significant clinical observations, the specific metabolic pathways and gene regulatory networks that drive these bone metabolic abnormalities remain incompletely elucidated, impeding the development of targeted therapeutic interventions. To explore the molecular basis of these bone metabolic abnormalities, metabolomics as a sophisticated analytical technique that systematically quantifies small molecule metabolites in biological systems and identifies relationships between metabolic profiles and disease states 11 , offers a new perspective for OA research. Previous metabolomic investigations have revealed distinct metabolic signatures in OA patients, particularly involving altered lipid metabolism, amino acid processing, and energy metabolism pathways 12 . However, these studies have primarily examined synovial fluid or cartilage tissue 13 , while systemic metabolic changes that may drive bone metabolic abnormalities remain poorly characterized. Concurrently, with advancements in transcriptomics, analysis of gene expression regulation in OA has gained significant attention. These studies have identified candidate biomarkers and implicated several signaling cascades in OA pathophysiology 14 . However, conventional transcriptomic approaches are often restricted to localized tissue examination, limiting their ability to detect how systemic metabolic alterations affect bone homeostasis. Thus, single-omics approaches struggle to comprehensively reveal the complex mechanisms underlying OA bone metabolic abnormalities. Given this research gap, the integration of metabolomic and transcriptomic datasets provides a powerful research paradigm. This approach allows for concurrent examination of both metabolite fluctuations and gene expression changes. It utilizes the complementary nature of these omics platforms to thoroughly assess the interrelationship between transcriptional regulation and metabolic dysfunction. This systems biology approach can uncover mechanistic insights not apparent through single-omics analysis. In particular, such multi-omics integration, through cross-validation, can establish causal relationships between metabolism, gene expression, and phenotype, providing more robust evidence regarding the biological mechanisms underlying bone metabolic disturbances in OA. In this study, we investigated the metabolic and transcriptional drivers of OA bone metabolic abnormalities by analyzing peripheral blood samples. Peripheral blood as a research subject offers multiple advantages, not only providing a minimally invasive sampling method but also comprehensively reflecting systemic metabolic changes 15 , while containing circulating factors directly involved in bone remodeling processes 16 , and has been validated by multiple studies to effectively capture OA-specific molecular signatures 17 . We conducted untargeted metabolomic analysis and whole-transcriptome sequencing on a screened study cohort. Using a comprehensive data analysis approach, we first categorized genes by their biological functions, then identified important biochemical pathways, measured relationships between different molecular components, and created visual maps of gene and metabolite interactions. This multi-step analysis helped us uncover the underlying molecular processes and disease patterns involved in bone metabolism disorders. In summary, this study provides a comprehensive multi-omics analysis of OA bone metabolic abnormalities, with potential to establish blood-based diagnostic and prognostic biomarker systems, identify molecular targets driving bone metabolic abnormalities, offer new perspectives on the pathogenesis of this disease, and contribute to the development of personalized medical approaches for OA patients. Materials & methods Participants This exploratory pilot study recruited 60 participants, 30 OA patients and 30 Controls, from the Department of Rheumatology and Immunology at Chongqing Hospital of Traditional Chinese Medicine. Participants were selected for the study based on several criteria: (1) with the primary demographic requirement limiting participation to adults aged 40-70 years; (2) BMI between 18.5 and 30 kg/m²; and (3) a diagnosis of OA on the basis of the revised criteria of the American College of Rheumatology (ACR) 18 . Study exclusion was implemented for individuals who met any of these disqualifying criteria: (1) comorbid bone or joint diseases, such as osteoporosis, rheumatoid arthritis, or traumatic arthritis; (2) metabolic disorders, including type 2 diabetes; (3) use of oral or intra-articular glucocorticoids within the past 3 months; and (4) joint replacement surgery within the past 6 months. Given the exploratory nature of this multiomics study, the sample size was determined based on similar pilot studies published in the field 19,20 and practical resource constraints. Existing research has demonstrated that 20-30 samples per group can effectively identify major metabolic alterations 21 . When evaluated retrospectively, our participant numbers demonstrated adequate sensitivity (70% power) for detecting pronounced effect magnitudes (Cohen's d ≥ 0.8) while maintaining standard type I error rates of 0.05, which is sufficient for identifying candidate pathways and generating hypotheses for future large-scale validation studies. Demographic characteristics (sex, age, and BMI) were comparable between the two groups, with no statistically significant differences identified (p > 0.05). All subjects provided written informed consent prior to study participation. This investigation was conducted in accordance with the ethical standards delineated in the Declaration of Helsinki and received approval from the Ethics Committee of Chongqing Hospital of Traditional Chinese Medicine (approval number: 2022-KY-YJS-LY). Sample Collection and Marker Measurement Ten milliliters of fasting venous blood was collected from each participant. Five milliliters were transferred to procoagulant tubes and centrifuged (3,000 rpm, 10 min, 4°C) for serum separation. The obtained serum was stored at -80°C for subsequent biochemical and metabolomic analyses. The remaining 5 mL was collected in EDTA tubes for peripheral blood mononuclear cell (PBMC) isolation using Lymphoprep™ density gradient centrifugation (1.077 g/mL; STEMCELL Technologies, Canada). Isolated PBMCs were washed twice with sterile phosphate-buffered saline (PBS) at 4°C and pelleted by centrifugation (1,000 rpm, 5 min). For long-term preservation, the cell pellets were resuspended in fetal bovine serum containing 10% dimethyl sulfoxide (DMSO), gradually cooled (1°C/min) to -80°C overnight, and then transferred to liquid nitrogen until RNA extraction and transcriptomic analysis. Serum levels of PINP, N-MID, and β-CTX were measured via enzyme-linked immunosorbent assay (ELISA) kits (Wuhan Elabscience Biotechnology Co., Ltd.) in accordance with the manufacturer’s instructions. An automatic microplate reader (RT-6100, Rayto, USA) was used to quantify the results, enabling an assessment of bone metabolic markers. Untargeted Metabolite Extraction and Analysis After the serum samples were thawed on ice, 400 μL of extraction solution (methanol:water, 8:2, v/v) was combined with 100 μL aliquots of serum or plasma. After two minutes of vortexing, the mixture was centrifuged for 10 min at 4°C at 12,000 rpm. The supernatants were subsequently collected and placed into liquid chromatography–mass spectrometry (LC‒MS) vials for further examination. Equal quantities from each of the ten samples were combined to create a quality control (QC) sample, which was used to guarantee the dependability of the analytical platform. Blank samples (containing pure solvent) were also included to avoid contamination during sample preparation. Metabolomic analysis was performed using ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) on a Vanquish UHPLC system coupled with a Q Exactive™ HF-X mass spectrometer (Thermo Fisher Scientific, Germany) at Novogene Co., Ltd. (Beijing, China). Chromatographic separation was achieved using an ACQUITY UPLC BEH amide column (2.1 mm × 100 mm, 1.7 μm; Waters). The acquired LC-MS data were processed using Compound Discoverer v3.1 software. Metabolite identification and quantification were performed following peak alignment across samples, applying stringent filtering criteria including a retention time deviation tolerance of 0.2 min and a mass accuracy threshold of 5 ppm 22 . Target peaks were identified on the basis of reference standards and databases. Compounds were accurately extracted by refining parameters such as mass deviation, signal intensity variation, and signal-to-noise ratio thresholds. The peak areas, which represent the relative abundances of the compounds, were quantified through established quantitative analysis methods. Molecular formulas were predicted according to molecular ion peaks as well as fragment ion data, and compounds were identified by comparison with databases such as mzCloud, mzVault, and MassList. Background ion interference was minimized via the use of blank samples, and the quantitative results were standardized to calculate relative peak areas. Compounds with a coefficient of variation (CV) exceeding 30% in the QC samples were excluded to ensure reliable metabolite quantification. Principal component analysis (PCA) and orthogonal projections to latent structures discriminant analysis (OPLS-DA) were employed to examine metabolite variations between the two groups 23,24 . The OPLS-DA model's robustness was confirmed by permutation testing, and the ideal number of principal components was ascertained using 10-fold cross-validation. P values 1 were used to identify differential metabolites (DMs). Kyoto Encyclopedia of Genes and Genomes (KEGG), the Human Metabolome Database (HMDB), and the LIPID Metabolites and Pathways Strategy (LIPID MAPS) were used to annotate these metabolites. For visualization of the differences in metabolites, a thorough analysis of metabolite differences and sample groups was performed via the pheatmap tool for clustering heatmaps and the ggplot2 R package for creating volcano plots. The significantly changed metabolites were subjected to Z score normalization, and the pathways most affected by these alterations were identified using pathway enrichment analysis using MetaboAnalyst 6.0 25 . R was also used to perform correlation analysis of the DMs. Correlation of DMs with OA Clinical Markers Spearman correlation coefficients were calculated to investigate the connection between DMs and OA clinical markers; values greater than 0.7 indicate strong relationships. The glmnet package in R was used to identify metabolites that could effectively differentiate OA patients from controls via least absolute shrinkage and selection operator (LASSO) regression, a regularization strategy intended to increase prediction accuracy 26 . For analysis of the diagnostic potential for important metabolites, receiver operating characteristic (ROC) analysis was conducted via the pROC package in R. The classification efficacy was measured using the area under the curve (AUC), and a P value < 0.05 was considered statistically significant. Whole-Transcriptome Sequencing and Analysis Following the manufacturer's instructions, total RNA was extracted using the TRIzol reagent (Invitrogen, Carlsbad, California, USA) and then purified. A NanoDrop spectrophotometer (Thermo Fisher Scientific, USA) was used to determine the concentration and purity of the RNA, while electrophoresis on 1% agarose gels was used to evaluate RNA integrity as well as contamination. An Agilent 2100 Bioanalyzer system (Agilent Technologies, USA) was used to further evaluate the quality of the RNA. For library preparation, only samples with a 28S/18S ratio > 1.8 and an RNA integrity number (RIN) > 7.5 were employed to ensure high-quality starting material. The NEBNext® Ultra™ II RNA Library Prep Kit for Illumina® (New England Biolabs [NEB], Ipswich, Massachusetts, USA) was used to construct long noncoding RNA (lncRNA) libraries, whereas the NEBNext® Multiplex Small RNA Library Prep Set for Illumina® (NEB) was used to create small RNA libraries. Each sample was given a unique index code for sequence identification after ribosomal RNA (rRNA) was eliminated. The AMPure XP system (Beckman Coulter, Brea, California, USA) was harnessed to purify the library.To minimize technical variation, all samples were processed in parallel using the same reagent lots. The Illumina NovaSeq 6000 platform was harnessed for sequencing, producing paired-end reads of 150 bp. The target fragment size and effective concentration were used to pool the libraries. Clean reads were utilized for further analysis after adapter sequences and low-quality reads were removed from the raw data. Library preparation and sequencing were conducted by Novogene Bioinformatics Technology Co., Ltd. (Beijing, China). High-throughput sequencing data (Illumina PE150/SE50) were processed with CASAVA base calling, converting the raw image files into FASTQ format. FastQC software (version 0.11.5) was used for QC as well as data filtering. Reads with a Phred quality score < 20 or lengths < 50 nucleotides were discarded. This included assessing sample-to-sample correlation matrices and PCA to identify potential outliers or batch effects. To address potential experimental batch effects, we employed the ComBat algorithm implemented in the sva R package 27 , which uses an empirical Bayes framework to adjust for known batch effects while preserving biological variation of interest. The reference genome index was constructed via HISAT2 (v2.0.5), which was also employed to align filtered paired-end reads to the reference genome. Transcript assembly was carried out via StringTie software (version 1.3.3b), and transcript abundance was quantified as fragments per kilobase of transcript per million mapped reads (FPKM) 28 . For small RNA analysis, length-filtered reads were aligned to the reference sequence via Bowtie (version 1.0.1) 29 , and matches were compared with miRBase to identify small RNAs. Novel miRNAs were predicted using miRNA detection tools such as miREvo 30 and miRDeep2 31 . The expression levels of known and novel miRNAs were normalized and quantified as transcripts per million (TPM) values. Differential expression analysis of RNAs was conducted via DESeq2, which applies a negative binomial distribution model 32 . Differentially expressed mRNAs, lncRNAs, and miRNAs (DEGs, DE-lncRNAs, and DE-miRNAs) were identified via a Benjamini‒Hochberg adjusted P value (padj) ≤ 0.05 and a |log2-fold change (FC)| ≥ 1 as thresholds. Heatmaps were generated with the heatmap R package, and volcano plots were created via the ggplot2 R package. Gene Ontology (GO) and KEGG enrichment analyses were conducted to study the functional relevance of the DEGs. KEGG included information on pathways, illnesses, and medications, whereas GO analysis covered three primary categories: biological process (BP), molecular function (MF), and cellular component (CC). The clusterProfiler R package 33 was used for both analyses, and a P value < 0.05 was deemed significant. Further research into the BPs associated with OA bone metabolism was conducted using gene set enrichment analysis (GSEA). The "symbols.gmt" gene set from the Molecular Signatures Database (MSigDB, v7.5.1) 34 was employed using GSEA software (v4.2.0). Gene sets were considered highly enriched if their P value was less than 0.05 and if their false discovery rate (FDR) was less than 0.25 after 1,000 permutations. Protein‒protein interaction (PPI) networks were constructed with the STRING database 35 and Cytoscape software 36 . Hub genes were identified via the CytoHubba and Molecular Complex Detection (MCODE) plugins. With the MCODE settings set to degree cutoff = 2, node score cutoff = 0.2, k-core = 2, and max depth = 100, DEGs were included in the hub gene networks. The maximum clique centrality (MCC) technique was utilized by the CytoHubba plugin to rank the hub genes, and the PPI network was visualized via Cytoscape (v3.9.0). For prediction of additional PPIs, the GeneMANIA 37 web tool was utilized. This tool incorporates bioinformatic methods, including physical interaction, coexpression, genetic interaction, gene enrichment, and colocalization, to construct comprehensive interaction networks. In this study, GeneMANIA was used to predict interaction networks for the hub genes and proteins. ceRNA Network Construction By binding microRNAs (miRNAs), competing endogenous RNAs (ceRNAs) control post-transcriptional gene expression. lncRNAs function as miRNA sponges in this situation, binding to miRNAs and modifying the expression of downstream target mRNAs. In light of current database limitations, we attempted an initial exploration using correlation-based ceRNA network analysis in this research 38 . First, differential expression analysis was employed to identify mRNAs, miRNAs, and lncRNAs that were differentially expressed. Considerably negative correlations between miRNAs and their target RNAs (mRNAs and lncRNAs) and considerably positive correlations between RNA pairs that share the same miRNA (e.g., lncRNA and mRNA) were required for the calculation of expression correlations between these three RNA types. These connections led to the construction of ceRNA ternary regulatory modules using lncRNA‒miRNA-mRNA pairs that shared the same miRNA. Finally, all significant ternary regulatory units were integrated into a ceRNA regulatory network, and key nodes and critical regulatory modules were identified through network topology analysis. The entire network was constructed and visualized via Cytoscape software, which may suggest possible regulatory relationships that could potentially inform future functional validation studies. Joint Metabolomic and Transcriptomic Analysis DMs and DEGs linked to markers of bone metabolism were chosen according to predetermined standards to conduct an integrated analysis of metabolomic and transcriptomic data. P 1 were used to indicate DEGs, whereas a VIP score > 1 and P < 0.05 were employed to identify DMs. Spearman's correlation analysis was used to study the connections among bone metabolic markers, DEGs, and DMs. Only those that were deemed substantially associated and included in the integrated analysis had an absolute correlation coefficient (|r|) > 0.7 and P < 0.05. Heatmaps were created to show the relationships between genes and metabolites. These selected DMs and DEGs were further used to construct networks via Cytoscape software. This visualization enabled a detailed exploration of metabolite‒gene relationships and provided an overview of the composite network. MetaboAnalyst 6.0 was used to identify enriched metabolic pathways and KEGG pathways represented in both the metabolomic and transcriptomic data. Additionally, the summarized data facilitated the classification of DEGs and metabolites, providing a comprehensive view of the integrated analysis. Real-Time Quantitative PCR (RT-qPCR) RNA extraction and gene expression validation were performed on PBMCs obtained from 20 subjects (10 OA patients and 10 controls). Total RNA was isolated using a TRIzol-based protocol. Briefly, samples in 1.5-mL Eppendorf tubes were treated with 1 mL TRIzol reagent and 200 µL chloroform, followed by ice incubation (10 min) and centrifugation (12,000 ×g, 15 min). The aqueous phase (400 µL) was precipitated with an equivalent volume of isopropanol at ambient temperature for 10 min. Following centrifugation (12,000 ×g, 10 min), the resulting RNA precipitate underwent dual washing procedures with 75% ice-cold ethanol (1 mL) and was subsequently solubilized in RNase-free water. RNA integrity was evaluated using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific), with A260/A280 ratios between 1.8-2.0 considered acceptable. Reverse transcription was conducted using 2× SPARKscript II All-in-One RT SuperMix (SparkJade; Dongsi Kejie Biotechnology Co., Ltd., China). Quantitative PCR was performed using SYBR Green qPCR Master Mix (without ROX; Applied Biosystems) under the following thermal parameters: initial denaturation (95°C, 30 s) followed by 60 amplification cycles (95°C for 10 s, 62°C for 30 s). Gene expression was normalized to GAPDH using the comparative threshold (2−ΔΔCt) method. Target gene primers (Table S1) were synthesized by Shanghai Bioengineering Technology Service Company (Shanghai, China). All analyses were performed in triplicate with GAPDH as the reference standard. Statistical Analysis Data analysis was conducted using SPSS (version 24.0; IBM, Armonk, NY, USA), GraphPad Prism (version 9.0; GraphPad Software, San Diego, CA, USA), and R (version 4.3.0; R Foundation for Statistical Computing, Vienna, Austria). Statistical summaries included tabular formats, frequency distributions, and graphical illustrations. The Shapiro-Wilk method was employed to evaluate distribution normality of continuous variables. For normally distributed parameters, results were presented as mean ± standard deviation and analyzed via independent-samples t tests. Non-parametric data were expressed as median with interquartile range (25th–75th percentile) and evaluated using the Mann-Whitney U test. Categorical variables were reported as frequency counts with percentages, with between-group comparisons performed using either Pearson's chi-square analysis or Fisher's exact test when appropriate. Statistical significance threshold was established at P < 0.05. Results Clinical Characteristics This study enrolled 30 patients with primary osteoarthritis (OA) and 30 age- and sex-matched controls. To mitigate potential confounding factors, we excluded subjects with inflammatory arthritis, secondary OA, disorders affecting bone metabolism, or those receiving bone-active medications. As demonstrated in Table 1, no statistically significant differences were observed between groups regarding demographic characteristics (p > 0.05). The distribution of comorbidities (hypertension, diabetes mellitus, dyslipidemia) and medication use (non-steroidal anti-inflammatory drugs, analgesics) was relatively balanced between OA patients and controls, with no significant differences detected. To rigorously control for potential confounding influences, we constructed multivariate linear regression models with bone turnover markers as dependent variables and OA status as the primary independent variable, while adjusting for demographic characteristics, comorbidities, and medication use as covariates. The adjusted analyses (Table S2) revealed that the significant association between OA and elevated bone turnover marker levels persisted after controlling for these confounding factors (p < 0.001), with negligible attenuation of effect sizes, indicating that this association reflects a genuine biological relationship between OA and altered bone metabolism rather than an artifact of differences in comorbidity patterns or medication effects. Table 1. Baseline Characteristics of Study Participants Variable Osteoarthritis (n = 30) Control (n = 30) P-value Demographics Sex, n (%) 0.792 a Male 19 (63.3) 17 (56.7) Female 11 (36.7) 13 (43.3) Age, years 56.7 ± 6.9 54.7 ± 4.9 0.211 b BMI, kg/m² 22.1 ± 2.8 21.9 ± 2.9 0.749 b Bone turnover markers β-CTX, pg/mL 88.2 [75.0-96.1] 64.1 [49.6-69.2] <0.001 c N-MID, ng/mL 1.88 ± 0.20 1.30 ± 0.24 <0.001 b PINP, ng/mL 1.89 ± 0.27 1.24 ± 0.31 <0.001 b Comorbidities, n (%) Hypertension 12 (40.0) 8 (26.7) 0.273 a Diabetes mellitus 11 (36.7) 9 (30.0) 0.584 a Dyslipidemia 14 (46.7) 11 (36.7) 0.432 a Current medications, n (%) NSAIDs 7 (23.3) 10 (33.3) 0.390 a Analgesics 9 (30.0) 6 (20.0) 0.371 a Data are presented as mean ± standard deviation, median [interquartile range], or number (percentage). BMI: body mass index; β-CTX: β-isomerized C-terminal telopeptide of type I collagen; N-MID: N-terminal mid-fragment osteocalcin; PINP: procollagen type I N-terminal propeptide; NSAIDs: non-steroidal anti-inflammatory drugs. a Chi-square test or Fisher's exact test; b Independent t-test; c Mann-Whitney U test. Untargeted Metabolomic Analysis We investigated metabolic alterations between OA patients and controls through untargeted metabolomic analysis. LC-MS/MS analysis was performed on serum samples from 60 participants (30 OA patients and 30 age-sex matched controls). Following data processing, we identified 488 metabolic features using HDBM and KEGG annotation databases, with 263 detected in positive ion mode and 225 in negative ion mode. The database-annotated metabolites were categorized into chemical groups by the integration of positive and negative ion modes. The largest fractions were composed of lipids as well as lipid-like molecules, organic acids and their derivatives, organoheterocyclic compounds, benzoenoids, and other groups (Fig. 1A). Differences in metabolites between the positive and negative ion modes were detected via univariate analysis, and a volcano plot representing typical DMs (VIP > 1.0, P < 0.05) is displayed (Fig. 1B). OPLS-DA modeling revealed clear separation between OA patients and controls (R2Y=0.94, Q2Y=0.88) (Fig. 1C), with permutation testing validating model reliability (Fig. 1D). The matchstick plot displayed the top 20 upregulated and downregulated metabolites in OA patients (Fig. 1E). Notably, benzoic acid derivatives, fatty acids, and glycerophosphocholines were significantly upregulated, while amino acids, peptides, bile acids, and arachidonic acid derivatives were markedly downregulated. KEGG pathway enrichment analysis revealed significant involvement of multiple metabolic pathways in OA (Fig. 1F). Upregulated metabolites were enriched in retinol metabolism, ubiquinone biosynthesis, and arachidonic acid metabolism pathways, while downregulated metabolites were primarily associated with histidine metabolism, steroid hormone biosynthesis, and bile acid production. Further analysis of specific metabolites revealed elevated inflammatory mediators (prostaglandin G2) and decreased chondroprotective compounds (estradiol) in OA patients. Finally, we examined metabolites that play biological roles in eukaryotes using the HDMB database. Prostaglandin G2 (PGG2), hydroquinone, vitamin A, 3,4-dihydroxycinnamic acid, hydroxybenzoic acid, gonadotropin, and phenylacetaldehyde were among the representative metabolites whose levels substantially increased in the OA group. Metabolite levels in the OA group, in contrast, were considerably lower for estradiol, histidine, arginine, hydroxydodecanoic acid, cholic acid, lysine, glycochenodeoxycholic acid, 6-keto-prostaglandin, glutamic acid, lithocholic acid, and phosphatidylcholine. Overall, our findings indicate that OA patients have metabolic traits distinct from those of Controls.Our findings indicate that OA is associated with significant metabolic reprogramming, characterized by enhanced inflammatory pathways and suppressed bone anabolic processes, providing a potential strategy for early diagnosis and targeted treatment of OA. Correlation Analysis of Metabolites and Clinical Markers To screen for high-confidence metabolic features related to bone homeostasis, we selected significant differentially expressed metabolites (P 1) and analyzed their associations with bone metabolic markers using Spearman correlation analysis. Results are shown in Fig. 2A. Among upregulated metabolites, phenylglyoxylic acid, tropine, and PGG2 correlated with β-CTX levels, tropine correlated with PINP levels, and phenylglyoxylic acid and tropine correlated with N-MID levels. Among downregulated metabolites, multiple compounds showed significant correlations with bone metabolic markers, with 4-nitrophenol, stearamide, 6-hydroxymelatonin, lithocholic acid, and anserine correlating with all three bone metabolic markers. Comparing bone homeostasis markers, taurolithocholic acid was associated with bone formation markers PINP and N-MID, while 3-hydroxy-3-methylbutanoic acid and propylparaben were associated with the bone resorption marker β-CTX. These metabolic profile variations demonstrated the characteristics of abnormal bone homeostasis in OA patients. Using bone metabolism-related differential metabolites as markers, we assessed their predictive performance for OA through ROC analysis. Serum biomarkers with AUC ≥ 80%, including seven metabolites such as 1-methylguanosine, were screened (Fig. 2B). After LASSO algorithm feature selection (Fig. 2C and D), a model comprising six metabolites demonstrated excellent diagnostic efficacy (Fig. 2E, AUC = 0.991). These metabolites—1-methylguanosine, 4-nitrophenol, anserine, tropine, 20-carboxy-leukotriene B4, and thymopentin—show potential as effective diagnostic markers for OA. These differential metabolites align with our previous pathway analysis results. For instance, PGG2 upregulation supports the role of inflammatory pathways in OA, bile acids reflect lipid metabolism disorders, and changes in amino acid-related molecules confirm the importance of protein metabolism in OA pathology. By integrating metabolomic and clinical phenotype data, we established a metabolite-clinical feature association network, providing multi-level evidence for the molecular pathological mechanisms of OA. Transcriptomic Analysis Based on our metabolomics cohort, we selected 20 matched pairs for transcriptomic analysis to investigate transcriptional variations in OA. Through RNA-seq and bioinformatic analysis, we identified 694 DEGs, 53 DE-miRNAs, and 36 DE-lncRNAs in PBMCs. Volcano plots displayed the distribution of differentially expressed genes (Fig. 3A), with 482 downregulated and 212 upregulated mRNAs, 1 upregulated and 35 downregulated lncRNAs, and 35 upregulated and 18 downregulated miRNAs in the OA group. GO enrichment analysis revealed DEGs primarily enriched in T cell activation and differentiation pathways, mRNA splicing, and RNA polymerase II-related processes (Fig. 3B, left). DE-miRNAs were enriched in cytoskeletal organization, chromatin assembly, and signaling receptor binding (Fig. 3B, middle), while DE-lncRNAs were enriched in cytokine-mediated signaling, receptor complexes, and apoptotic processes (Fig. 3B, right). KEGG pathway analysis revealed DEGs were enriched in MAPK signaling, Spliceosome, and IL-17 signaling pathways (Fig. 3C), indicating inflammatory and RNA processing mechanisms in OA. DE-miRNAs were enriched in viral infection, cytokine-receptor interaction, and Huntington disease pathways (Fig. S1A), suggesting roles in immune regulation. DE-lncRNAs showed significant enrichment in cancer pathways, neurodegeneration, and Alzheimer disease (Fig. S1B), pointing to involvement in cellular stress responses. GSEA demonstrated multiple downregulated pathways in OA (Fig. 3D), including apoptosis, inflammatory response, and TNF-α signaling via NF-κB, with the latter showing particularly pronounced downregulation. Notably, cytokine-cytokine receptor interaction was enriched across all RNA types, suggesting this represents a core mechanism in OA pathophysiology regulated at multiple transcriptional levels. PPI Network and Hub Gene Identification To deeply investigate functional interactions among DEGs, we constructed protein-protein interaction (PPI) networks. Using the STRING database, we created a PPI network with 103 nodes and 524 edges, demonstrating functional associations between DEGs (Fig. 4A). Using the MCODE method in Cytoscape software, we identified the top-ranked module Cluster 1 with a score of 14.286 (29 nodes and 200 edges) (Fig. 4B). Cluster 1 comprised 14 genes, including 2 upregulated genes (CCR2 and CXCR6) and 12 downregulated genes (FOSL1, CCL20, JUND, JUNB, JUN, CXCR2, CXCL8, IL6, IL1B, IFNG, FOSB, and FOS). Further analysis through the MCC algorithm of the cytoHubba plugin identified the top 20 core genes (Fig. 4C). The intersection of nodes ranked by MCODE and cytoHubba analyses determined the final set of hub genes (Fig. 4D). Among these genes, the expression levels of FOS, IL1B, IL6, and JUN were significantly reduced in OA samples (P < 0.05) (Fig. S1C), highlighting their potential as biomarkers for OA. Additionally, we employed the GeneMANIA database to predict interaction networks of these hub genes, revealing that most DEGs were closely related to the hub genes (Fig. 4E). These core genes form regulatory modules that play crucial roles in OA-related inflammation and cartilage degradation processes, influencing OA development by regulating downstream gene expression. Noncoding RNA Regulatory Network LncRNAs can function as upstream regulators of miRNAs to influence gene expression. To reveal the regulatory roles of non-coding RNAs in OA, we constructed a ceRNA regulatory network based on expression profile data. First, we processed all RNA-seq data uniformly to reduce the effect of sample-to-sample sequencing depth. Strict screening criteria (Spearman correlation coefficient threshold |r| > 0.7, FDR 0.7, miRNA-lncRNA interaction pairs were screened. Finally, we utilized Cytoscape to create and visualize the lncRNA-miRNA-mRNA ceRNA network according to expression patterns of miRNAs, lncRNAs, and mRNAs from OA patients and controls. According to network topology analysis, the network exhibited a notable core-periphery structure and typical scale-free properties. The core nodes were composed primarily of highly connected miRNAs that might play important regulatory roles in OA onset and progression (Fig. 5A). We used the cytoHubba plugin in Cytoscape software to perform topological analysis of the complete network, identifying major regulatory modules in the ceRNA network. We then used the MCC technique to extract the core subnetwork, consisting of the top 20 most important hub genes. For example, HLA-E, the most important node in this subnetwork, directly regulates miRNA molecules including hsa-miR-136-3p, hsa-miR-487b-3p, hsa-miR-299-5p, and hsa-miR-410-3p. The subnetwork contained 11 miRNA nodes (shown by diamonds) and 7 target gene and lncRNA nodes (shown by ellipses). Notably, lncRNA molecules, including LINC00880, NUTM2G, and RUSC1-AS1, may exert critical synergistic functions in the regulatory network by competitively interacting with specific miRNAs (Fig. 5B). This ceRNA network reveals the complex roles of lncRNA-miRNA-mRNA three-layer regulatory relationships in OA pathogenesis, providing new perspectives for understanding non-coding RNA functions in OA. Integration and Crosstalk between the Metabolome and Transcriptome To further explore the pathogenic mechanisms of bone homeostasis imbalance in OA, we performed a multi-dimensional integrated analysis of whole transcriptome and metabolome data from the same biological samples. First, we employed GSEA using the C2 and C7 collections from the MSigDB to identify key pathways related to inflammation and metabolism. Subsequently, using Spearman correlation analysis, we screened differentially expressed genes and DMs associated with bone metabolism markers, establishing the connection between OA DEGs and metabolites. We visualized the overall relationships between metabolites and phenotypes using a heatmap (Fig. 6A). Based on stringent selection criteria (|cor| > 0.7, P < 0.05), we constructed a completely linked network of DMs and DEGs (Fig. 6B), clearly identifying transcription-metabolism co-regulatory modules.Within this network, we identified key co-regulatory feature clusters, including genes linked to inflammation, such as TNFAIP3; cell signaling pathways, such as PRDM3B and SORL1; and receptors, such as ROR and PPX. Notably, genes including TNFAIP3, LINC00905, and CKB6 exhibited significant positive or negative correlations with metabolites such as isoleucylglycerol, propylparaben, and panaxanthine. Next, we used the MetaboAnalyst 6.0 platform combined with GSEA methodology to perform a joint pathway analysis of DEGs and DMs to identify metabolic pathways co-enriched with genes and metabolites (raw data provided in Integration_DMs_DEGs_Pathways.csv). The results showed significant enrichment (P < 0.05) in metabolic pathways of histidine, caffeine, β-alanine, taurine, hypotaurine, glycerolipid, and aminoacyl-tRNA biosynthesis. Among these, histidine metabolism emerged as the most significantly enriched metabolic pathway, showing notable regulatory abnormalities in the gene-metabolite network analysis (Fig. 6C). As shown in Fig. 6C, the histidine metabolic pathway forms a complex metabolic network involving multiple key metabolites and enzymes. L-histidine can be metabolized through three main pathways: (1) entering the histidine catabolic pathway (blue box in the lower right) via catalysis by HAL (histidine ammonia-lyase), ultimately producing L-glutamate; (2) conversion to histamine via HDC (histidine decarboxylase), which subsequently enters the methylation process (blue box in the middle) under the action of HNMT (histamine N-methyltransferase), generating methylimidazole acetic acid; (3) participation in the carnosine metabolic cycle (circle in the upper right), forming anserine through catalysis by CARNS1 and CARNMT1. Additionally, this network connects to aspartate metabolism (left side) through enzymes such as ASPA and ALDH2. In OA patients, all three main metabolic pathways exhibit significant abnormalities, particularly the methylation process of histamine (indicated by the red arrow) showing an upregulation trend, which may lead to inflammation and bone metabolic disorders. We also examined the regulatory pathways associated with DEGs and metabolites using combined analysis (Fig. 6D). Based on analysis results from MSigDB's C2 and C7 collections, the modulation of OA bone metabolism involves signaling pathways such as the TNF signaling pathway, the MAPK signaling pathway, the legionellosis signaling pathway, and the IL-17 signaling pathway (P < 0.05). In particular, we deeply explored the key genes in the IL-17 signaling pathway and found that the previously identified hub genes were key nodes in this pathway. For example, genes such as JUND, CCL20, FOSL1, IL1B, JUN, FOS, CXCR2, IL6, JUNB, CXCL8, and CCR2 were significantly differentially expressed (Fig. 6E). Through analysis of MSigDB's C3 collection, we further discovered highly overlapping binding sites for STAT3 and NF-κB transcription factors in these pathways, indicating their synergistic role in transcription-metabolism co-regulation. Our multi-omics integrated analysis revealed abnormalities in the transcription-metabolism co-regulatory network in OA, particularly the coordinated dysregulation of histidine metabolism and IL-17 signaling pathways may constitute a key link in the pathogenic mechanism of OA. The close connection between histidine metabolism pathways and inflammatory responses provides potential new targets for precision treatment of OA. Validation of Key Genes in Independent Datasets and Clinical Samples To enhance the reliability of our findings, we employed a comprehensive cross-validation strategy using both public datasets and independent clinical samples. The mRNA expression levels of DEGs linked to bone metabolism in OA were validated by analyzing two independent RNA-seq datasets from the Gene Expression Omnibus (GEO) database: GSE48556 (106 OA samples and 33 normal controls) and GSE114007 (20 OA samples and 18 normal controls). In the GSE48556 dataset, we observed significant differential expression of IL1B, JUND, FOS, IL6, and JUNB between OA and normal samples (Fig. 7A). The validation in the GSE114007 dataset showed consistent downregulation of JUND, FOS, IL6, and JUNB in OA samples, though IL1B did not reach statistical significance in this cohort (Fig. 7B). To further validate these findings in clinical specimens, we performed RT-qPCR analysis on peripheral blood samples from an independent cohort of OA patients and controls (n = 10 per group). This analysis confirmed significantly reduced expression of all five genes in the OA group compared to the normal control group. The most pronounced differences were observed for IL1B, JUND, and JUNB (P < 0.001), while FOS (P < 0.05) and IL6 (P < 0.01) also showed significant downregulation (Fig. 7C). These consistent results across multiple independent datasets and clinical samples provide robust validation of our multi-omics analysis. They confirm that the dysregulation of inflammatory genes plays a critical role in OA pathogenesis. The convergent evidence from computational analysis, public datasets, and clinical specimens strengthens our conclusion that these inflammation-related genes are central to regulating bone metabolism homeostasis in OA and influencing disease progression. Moreover, the differential expression patterns observed across these validation approaches align with our findings from the histidine metabolism and IL-17 signaling pathway analyses, supporting the proposed transcription-metabolism co-regulatory network as a key mechanism in OA development. Discussion OA is a common degenerative joint disorder characterized by cartilage degradation and osteophyte formation. These pathological changes lead to significant functional impairment and disability, reducing patients' quality of life and adversely affecting long-term prognosis 39 . Despite extensive research, the exact pathomechanism of OA remains incompletely understood. Current evidence indicates that typical histopathological features of OA include cartilage matrix degradation, abnormal subchondral bone remodeling, and synovial inflammation 40 , 41 . Several theories attempt to explain these pathological changes, including mechanical stress theory, inflammatory cytokine theory, and metabolic dysregulation hypothesis 42 , 43 . As the disease progresses, articular cartilage is gradually eroded, subchondral bone undergoes sclerosis and cystic changes, and osteophytes form, ultimately resulting in joint pain, stiffness, and limited mobility 44 . Existing research suggests that OA results from the combined effects of genetic factors, lifestyle, and environmental influences, with certain OA phenotypes potentially attributable to specific metabolic and transcriptional alterations 45 , 46 . Bone metabolic imbalance plays a central role in OA pathogenesis. Under physiological conditions, osteoblasts synthesize and secrete bone matrix, while osteoclasts mediate bone resorption and degradation, maintaining a dynamic equilibrium within the local microenvironment. However, factors such as aging and altered estrogen levels can disrupt this balance, leading to abnormal bone metabolism and accelerated disease progression. Therefore, this study employed an integrated multi-omics approach to elucidate the metabolic-transcriptional regulatory networks involved in OA bone homeostasis dysregulation and explore potential biomarkers. In this study, we collected 60 peripheral blood samples (30 from OA patients and 30 from controls) and subsequently performed untargeted metabolomic analysis and whole transcriptome sequencing, identifying 127 DMs and 694 DEGs. Concurrently, we observed significantly elevated bone turnover markers (N-MID, PINP, and β-CTX) in OA patients, confirming the importance of bone metabolic imbalance in OA. Through metabolomic analysis, we found that proinflammatory metabolites (such as PGG2) were significantly elevated in OA patients, while protective metabolites (such as estriol and 3-indolepropionic acid) were markedly reduced. These findings align with previous reports 47 . PGG2, as an intermediate in the prostaglandin synthesis pathway, not only participates in inflammatory responses but also directly affects cartilage degradation by regulating matrix metalloproteinase expression 48 . The decreased estriol levels may reflect weakened bone protective mechanisms, consistent with previous clinical observations regarding estrogen replacement therapy delaying OA progression in women 49 . These metabolite alterations are not isolated phenomena but represent coordinated dysregulation across multiple metabolic pathways, potentially forming the metabolic foundation of OA pathogenesis. In pathway enrichment analysis, we identified multiple metabolic and signaling pathways associated with OA. Metabolomic analysis revealed that upregulated metabolites were primarily enriched in retinol metabolism, ubiquinone biosynthesis, and arachidonic acid metabolism pathways, while downregulated metabolites were mainly associated with histidine metabolism, steroid hormone biosynthesis, and bile acid production. Transcriptomic analysis revealed that DEGs were primarily enriched in DNA repair pathways, cell junctions, and actin cytoskeleton-related processes. Notably, the IL-17 signaling pathway was significantly dysregulated in OA, consistent with previous studies 50 . Through PPI network analysis, we identified core genes including IL1B, JUND, FOS, IL6, and JUNB, which have been implicated in OA pathogenesis in multiple studies 51 . The activation of the IL-17 signaling pathway involves not only the release of inflammatory mediators but may also directly regulate bone metabolism by influencing the differentiation balance between osteoblasts and osteoclasts 52 . In particular, core genes JUND and FOS, as transcription factors, can simultaneously regulate multiple downstream targets, forming complex regulatory networks. These genes are not simply effector molecules but key nodes integrating multiple signal inputs, potentially playing coordinating roles in the pathological cascade of OA. Furthermore, by integrating metabolomic and transcriptomic data, we constructed a transcription-metabolism co-regulatory network and discovered that the interaction between histidine metabolism and IL-17 signaling pathways might be a key driver of bone metabolic imbalance in OA. Specifically, the histidine metabolism pathway participates in OA pathogenesis through three main routes: (1) entering the histidine degradation pathway via HAL catalysis, ultimately producing L-glutamate; (2) conversion to histamine via HDC, subsequently entering methylation processes under HNMT action; and (3) participation in the carnosine metabolism cycle. In OA patients, all three major metabolic pathways exhibited significant abnormalities, particularly an upward trend in histamine methylation, potentially leading to inflammation and bone metabolic disorders. Our study has several limitations. First, our use of peripheral blood rather than joint tissues may not accurately reflect local OA pathological processes. Muhammad et al. 53 demonstrated significant molecular differences between systemic circulation and joint tissues in OA, highlighting the need for integrated multi-omics approaches across biological compartments. Second, our small sample size may have limited detection of subtle metabolic and transcriptional changes. Third, our correlation-based analysis lacks functional validation, preventing causal inference, with potential confounding factors possibly influencing results. Future research should include tissue-specific comparisons between matched blood and joint samples, larger longitudinal studies, functional validation experiments, and exploration of therapeutic approaches targeting histidine metabolism and IL-17 signaling pathways. In conclusion, our exploratory multi-omics study provides preliminary evidence of metabolic and transcriptional signatures associated with bone metabolic imbalance in OA, identifying potential interactions between histidine metabolism and IL-17 signaling pathways. While these findings require further validation in larger studies and functional experiments, they offer new perspectives on the molecular mechanisms of bone metabolism abnormalities in OA, potentially contributing to the development of novel diagnostic markers and therapeutic strategies. Conclusions In this study, we explored the metabolic-transcriptional regulatory networks involved in bone metabolic imbalance in OA through an integrated multi-omics approach. We analyzed peripheral blood samples from OA patients and identified differentially expressed metabolites and genes, observing significantly elevated bone turnover markers (N-MID, PINP, and β-CTX) in OA patients, which validated the importance of bone metabolic imbalance in OA pathogenesis. Our findings revealed potential interactions between histidine metabolism and IL-17 signaling pathways in OA, with abnormal histidine metabolism potentially playing a crucial role in OA pathomechanisms. The study identified core genes including IL1B, JUND, FOS, IL6, and JUNB, which may serve coordinating functions in the pathological cascade of OA. Our results suggest that increased proinflammatory metabolites coupled with decreased protective metabolites constitute the metabolic signature of OA, potentially providing molecular targets for early diagnosis and treatment of this condition. Nevertheless, larger-scale studies and functional validation experiments are still needed to further elucidate the clinical significance of these findings and their exact roles in OA pathophysiology. Abbreviations AUC Area Under the Curve β-CTX Beta-CrossLaps of Type I Collagen ceRNA Competing Endogenous RNA DEGs Differentially Expressed Genes DMs Differential Metabolites ELISA Enzyme-Linked Immunosorbent Assay FPKM Fragments Per Kilobase per Million Mapped Reads GO Gene Ontology GSEA Gene Set Enrichment Analysis CON Controls IL-17 Interleukin-17 KEGG Kyoto Encyclopedia of Genes and Genomes LASSO Least Absolute Shrinkage and Selection Operator MAPK Mitogen-Activated Protein Kinase N-MID Mid-Regional Fragment of Osteocalcin OA Osteoarthritis PBMCs Peripheral Blood Mononuclear Cells PINP Procollagen I N-Terminal Propeptide PPI Protein-Protein Interaction QC Quality Control RIN RNA Integrity Number ROC Receiver Operating Characteristic TNF Tumor Necrosis Factor UHPLC-MS Ultrahigh-Performance Liquid Chromatography-Tandem Mass Spectrometry VIP Variable Importance in Projection Declarations Acknowledgements We extend our gratitude to all study participants for their invaluable contributions to this research. Author Contributions YL, BW, and WC designed the study. YL, JG, SW, QS, and FZ conducted patient recruitment and data collection. YL and JL performed experiments. YL, JL, and SW analyzed data. YL wrote the manuscript. BW and WC supervised the project and acquired funding. All authors critically reviewed and approved the final manuscript. Data Availability The datasets analyzed during this study are available in the NCBI GEO repository under accession numbers GSE48556 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE48556) and GSE114007 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE114007). The datasets generated during this study, including integrated gene-metabolite pathway analysis results, are provided in Additional file 3: Integration_DMs_DEGs_Pathways.csv. Consent for publication All participants provided written informed consent for the publication of anonymized data. Competing interests The authors declare that they have no competing interests. Funding This research received no specific grant from funding agencies in the public, commercial, or not-for-profit sectors.Declarations Ethics approval and consent to participate This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Chongqing Hospital of Traditional Chinese Medicine (Approval No. 2022-KY-YJS-LY). Written informed consent was obtained from all participants or their legal guardians. Author details 1 College of Traditional Chinese Medicine, Chongqing Medical University, Chongqing 400016, China. 2 Department of Rheumatology, Chongqing Hospital of Traditional Chinese Medicine, Chongqing 400021, China. 3 Department of Nephrology and Rheumatology, Chongqing Jiangjin District Hospital of Traditional Chinese Medicine,Chongqing 402260, China. 4 Chongqing Key Laboratory of Traditional Chinese Medicine to Prevent and Treat Autoimmune Diseases, Chongqing, 400021, China. References Bijlsma, J. W. J., Berenbaum, F. & Lafeber, F. P. J. G. Osteoarthritis: an update with relevance for clinical practice. Lancet 377 , 2115-2126, doi:10.1016/s0140-6736(11)60243-2 (2011). Steinmetz, J. D. et al. Global, regional, and national burden of osteoarthritis, 1990-2020 and projections to 2050: a systematic analysis for the Global Burden of Disease Study 2021. Lancet Rheumatology 5 , e508-e522, doi:10.1016/s2665-9913(23)00163-7 (2023). Long, H. et al. Prevalence Trends of Site-Specific Osteoarthritis From 1990 to 2019: Findings From the Global Burden of Disease Study 2019. Arthritis & Rheumatology 74 , 1172-1183, doi:10.1002/art.42089 (2022). He, Y. et al. Pathogenesis of Osteoarthritis: Risk Factors, Regulatory Pathways in Chondrocytes, and Experimental Models. Biology-Basel 9 , doi:10.3390/biology9080194 (2020). Macfarlane, E., Seibel, M. J. & Zhou, H. Arthritis and the role of endogenous glucocorticoids. Bone research 8 , 33-33, doi:10.1038/s41413-020-00112-2 (2020). Loeser, R. F., Goldring, S. R., Scanzello, C. R. & Goldring, M. B. Osteoarthritis: A disease of the joint as an organ. Arthritis and Rheumatism 64 , 1697-1707, doi:10.1002/art.34453 (2012). Watkins, L. et al. Assessment of quantitative 18F Sodium fluoride PET measures of knee subchondral bone perfusion and mineralization in osteoarthritic and healthy subjects. Osteoarthritis and Cartilage 29 , 849-858, doi:10.1016/j.joca.2021.02.563 (2021). Hwang, J. S., Lee, H. S. & Gong, H. S. Three-Dimensional Analysis of the Trapezium Subchondral Bone and its Association with Trapeziometacarpal Joint Osteoarthritis. Calcified Tissue International 112 , 320-327, doi:10.1007/s00223-022-01040-5 (2023). Muratovic, D., Atkins, G. J. & Findlay, D. M. Is RANKL a potential molecular target in osteoarthritis? Osteoarthritis and Cartilage 32 , 493-500, doi:10.1016/j.joca.2023.10.010 (2024). Sahu, S., Naik, S., Bandyopadhyay, D. & Tripathy, S. Serum levels of osteoprotegerin, RANK-L & vitamin D in different stages of osteoarthritis of the knee. Indian Journal of Medical Research 154 , 491-496, doi:10.4103/ijmr.IJMR_873_19 (2021). Zhou, J. & Zhong, L. Applications of liquid chromatography-mass spectrometry based metabolomics in predictive and personalized medicine. Frontiers in Molecular Biosciences 9 , doi:10.3389/fmolb.2022.1049016 (2022). Van Pevenage, P. M., Birchmier, J. T. & June, R. K. Utilizing metabolomics to identify potential biomarkers and perturbed metabolic pathways in osteoarthritis: A systematic review. Seminars in Arthritis and Rheumatism 59 , doi:10.1016/j.semarthrit.2023.152163 (2023). Liao, Z. et al. Differential Metabolites in Osteoarthritis: A Systematic Review and Meta-Analysis. Nutrients 15 , doi:10.3390/nu15194191 (2023). Wang, M. et al. Molecular crosstalk between articular cartilage, meniscus, synovium, and subchondral bone in osteoarthritis. Bone & Joint Research 11 , 862-872, doi:10.1302/2046-3758.1112.Bjr-2022-0215.R1 (2022). Tchetina, E. V. et al. Metabolic Dysregulation and Its Role in Postoperative Pain among Knee Osteoarthritis Patients. International Journal of Molecular Sciences 25 , doi:10.3390/ijms25073857 (2024). Maruotti, N., Corrado, A. & Cantatore, F. P. Osteoblast role in osteoarthritis pathogenesis. Journal of Cellular Physiology 232 , 2957-2963, doi:10.1002/jcp.25969 (2017). Ge, M. et al. Multi-omics analysis of synovial tissue and fluid reveals differentially expressed proteins and metabolites in osteoarthritis. Journal of Translational Medicine 23 , doi:10.1186/s12967-025-06310-y (2025). Hochberg, M. C. et al. Guidelines for the medical management of osteoarthritis. Part I. Osteoarthritis of the hip. American College of Rheumatology. Arthritis and rheumatism 38 , 1535-1540, doi:10.1002/art.1780381103 (1995). Poplawski, A. & Binder, H. Feasibility of sample size calculation for RNA-seq studies. Briefings in Bioinformatics 19 , 713-720, doi:10.1093/bib/bbw144 (2018). Surowiec, I. et al. Multivariate strategy for the sample selection and integration of multi-batch data in metabolomics. Metabolomics 13 , doi:10.1007/s11306-017-1248-1 (2017). Vinaixa, M. et al. A Guideline to Univariate Statistical Analysis for LC/MS-Based Untargeted Metabolomics-Derived Data. Metabolites 2 , 775-795, doi:10.3390/metabo2040775 (2012). Di Guida, R. et al. Non-targeted UHPLC-MS metabolomic data processing methods: a comparative investigation of normalisation, missing value imputation, transformation and scaling. Metabolomics 12 , doi:10.1007/s11306-016-1030-9 (2016). Bro, R. & Smilde, A. K. Principal component analysis. Analytical Methods 6 , 2812-2831, doi:10.1039/c3ay41907j (2014). Trygg, J. & Wold, S. Orthogonal projections to latent structures (O-PLS). Journal of Chemometrics 16 , 119-128, doi:10.1002/cem.695 (2002). Pang, Z. et al. MetaboAnalyst 6.0: towards a unified platform for metabolomics data processing, analysis and interpretation. Nucleic Acids Research 52 , W398-W406, doi:10.1093/nar/gkae253 (2024). Ranstam, J. & Cook, J. A. LASSO regression. British Journal of Surgery 105 , 1348-1348, doi:10.1002/bjs.10895 (2018). Leek, J. T., Johnson, W. E., Parker, H. S., Jaffe, A. E. & Storey, J. D. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics 28 , 882-883, doi:10.1093/bioinformatics/bts034 (2012). Pertea, M. et al. StringTie enables improved reconstruction of a transcriptome from RNA-seq reads. Nature Biotechnology 33 , 290-+, doi:10.1038/nbt.3122 (2015). Langmead, B., Trapnell, C., Pop, M. & Salzberg, S. L. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biology 10 , doi:10.1186/gb-2009-10-3-r25 (2009). Wen, M., Shen, Y., Shi, S. & Tang, T. miREvo: an integrative microRNA evolutionary analysis platform for next-generation sequencing experiments. Bmc Bioinformatics 13 , doi:10.1186/1471-2105-13-140 (2012). Friedlaender, M. R., Mackowiak, S. D., Li, N., Chen, W. & Rajewsky, N. miRDeep2 accurately identifies known and hundreds of novel microRNA genes in seven animal clades. Nucleic Acids Research 40 , 37-52, doi:10.1093/nar/gkr688 (2012). Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology 15 , doi:10.1186/s13059-014-0550-8 (2014). Wu, T. et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innovation (New York, N.Y.) 2 , 100141-100141, doi:10.1016/j.xinn.2021.100141 (2021). Liberzon, A. et al. The Molecular Signatures Database Hallmark Gene Set Collection. Cell Systems 1 , 417-425, doi:10.1016/j.cels.2015.12.004 (2015). Szklarczyk, D. et al. STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Research 47 , D607-D613, doi:10.1093/nar/gky1131 (2019). Shannon, P. et al. Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Research 13 , 2498-2504, doi:10.1101/gr.1239303 (2003). Warde-Farley, D. et al. The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function. Nucleic Acids Research 38 , W214-W220, doi:10.1093/nar/gkq537 (2010). Salmena, L., Poliseno, L., Tay, Y., Kats, L. & Pandolfi, P. P. A ceRNA Hypothesis: The Rosetta Stone of a Hidden RNA Language? Cell 146 , 353-358, doi:10.1016/j.cell.2011.07.014 (2011). van der Kraan, P. M. & van den Berg, W. B. Osteophytes: relevance and biology. Osteoarthritis and Cartilage 15 , 237-244, doi:10.1016/j.joca.2006.11.006 (2007). Mobasheri, A. et al. The role of metabolism in the pathogenesis of osteoarthritis. Nature Reviews Rheumatology 13 , 302-311, doi:10.1038/nrrheum.2017.50 (2017). Katsoula, G., Kreitmaier, P. & Zeggini, E. Insights into the molecular landscape of osteoarthritis in human tissues. Current Opinion in Rheumatology 34 , 79-90, doi:10.1097/bor.0000000000000853 (2022). Kumm, J., Tamm, A., Lintrop, M. & Tamm, A. Diagnostic and prognostic value of bone biomarkers in progressive knee osteoarthritis: a 6-year follow-up study in middle-aged subjects. Osteoarthritis and Cartilage 21 , 815-822, doi:10.1016/j.joca.2013.03.008 (2013). Martel-Pelletier, J., Pelletier, J. P. & Fahmi, H. Cyclooxygenase-2 and prostaglandins in articular tissues. Seminars in Arthritis and Rheumatism 33 , 155-167, doi:10.1016/s0049-0172(03)00134-3 (2003). Goggs, R., Carter, S. D., Schulze-Tanzil, G., Shakibaei, M. & Mobasheri, A. Apoptosis and the loss of chondrocyte survival signals contribute to articular cartilage degradation in osteoarthritis. Veterinary Journal 166 , 140-158, doi:10.1016/s1090-0233(02)00331-3 (2003). Pang, H. et al. Low back pain and osteoarthritis pain: a perspective of estrogen. Bone Research 11 , doi:10.1038/s41413-023-00280-x (2023). Gilmer, G. et al. Uncovering the "riddle of femininity" in osteoarthritis: a systematic review and meta-analysis of menopausal animal models and mathematical modeling of estrogen treatment. Osteoarthritis and Cartilage 31 , 447-457, doi:10.1016/j.joca.2022.12.009 (2023). Aziz, A., Nathan, K. G., Kamarul, T., Mobasheri, A. & Sharifi, A. The interplay between dysregulated metabolites and signaling pathway alterations involved in osteoarthritis: a systematic review. Therapeutic Advances in Musculoskeletal Disease 16 , doi:10.1177/1759720x241299535 (2024). Zeng, Z. et al. Pelargonidin ameliorates inflammatory response and cartilage degeneration in osteoarthritis via suppressing the NF-κB pathway. Archives of Biochemistry and Biophysics 743 , doi:10.1016/j.abb.2023.109668 (2023). Lu, L. & Tian, L. Postmenopausal osteoporosis coexisting with sarcopenia: the role and mechanisms of estrogen. Journal of Endocrinology 259 , doi:10.1530/joe-23-0116 (2023). Mimpen, J. Y. et al. Interleukin-17A Causes Osteoarthritis-Like Transcriptional Changes in Human Osteoarthritis-Derived Chondrocytes and Synovial Fibroblasts In Vitro. Frontiers in Immunology 12 , doi:10.3389/fimmu.2021.676173 (2021). Shi, S. et al. Identification of key regulators responsible for dysregulated networks in osteoarthritis by large-scale expression analysis. Journal of Orthopaedic Surgery and Research 16 , doi:10.1186/s13018-021-02402-9 (2021). Wang, Z. et al. RANKL expression of primary osteoblasts is enhanced by an IL-17-mediated JAK2/STAT3 pathway through autophagy suppression. Connective Tissue Research 62 , 411-426, doi:10.1080/03008207.2020.1759562 (2021). Rai, M. F. et al. Three decades of advancements in osteoarthritis research: insights from transcriptomic, proteomic, and metabolomic studies. Osteoarthritis and Cartilage 32 , 385-397, doi:10.1016/j.joca.2023.11.019 (2024). Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterial.pdf Supplementary Information Supplementary Material 1 Supplementary Material 2 IntegrationDMsDEGsPathways.csv 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-6359791","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":453991745,"identity":"14efe150-6251-4125-91d8-0b21555c7847","order_by":0,"name":"Yang Lu","email":"","orcid":"","institution":"Chongqing Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yang","middleName":"","lastName":"Lu","suffix":""},{"id":453991746,"identity":"1af4be79-3fef-4067-80a1-fba6af18fce7","order_by":1,"name":"Shasha Wang","email":"","orcid":"","institution":"Chongqing Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Shasha","middleName":"","lastName":"Wang","suffix":""},{"id":453991747,"identity":"acd7477e-d7bf-49fb-b58a-49845304bd7e","order_by":2,"name":"Jinkun Li","email":"","orcid":"","institution":"Chongqing Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Jinkun","middleName":"","lastName":"Li","suffix":""},{"id":453991748,"identity":"8822d000-ad1e-4b09-be5c-dba77d8e120a","order_by":3,"name":"Jianpin Gan","email":"","orcid":"","institution":"Chongqing Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Jianpin","middleName":"","lastName":"Gan","suffix":""},{"id":453991749,"identity":"7dbac68a-d8e4-4cb2-9cab-a98f2c97bc15","order_by":4,"name":"Fenglin Zhu","email":"","orcid":"","institution":"Chongqing Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Fenglin","middleName":"","lastName":"Zhu","suffix":""},{"id":453991753,"identity":"09a64ad6-1321-4249-afb6-033b56442b6b","order_by":5,"name":"Qin Shao","email":"","orcid":"","institution":"Chongqing Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Qin","middleName":"","lastName":"Shao","suffix":""},{"id":453991754,"identity":"3ff0d3df-fdbb-42e4-bff6-b83d4fa540e1","order_by":6,"name":"Wenfu Cao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+klEQVRIie3QMUvDUBDA8RcC7RJ96xVr+xVOAp0y+FHuISRL3Tu2CJclxbUufgxx88IDp/cBClkyObkUlxRUjKuYWLcO7z/fj+NOKZ/veJsoHYYiFMFkMMzlEBKrUc5G6nESn0aODiPoXFzWSWru4RJ7Z3Vun2H3iGa5JRSa24xBkWoWD50EXJqO7hya1YZIyNlrPltJULiqk6Caz6oTfjc3QCKmaMlYKAy4h+jXWfXBaBjMUsynzQZA2E+g3RK0pIisap+c0p8Eti/Zfs0Yb4b8TZILbp9c9t2ib68s7hnPn6x+2zURTKd5XtbNopv8nvxz3ufz+Xw/+gLy4mDzBx1dIQAAAABJRU5ErkJggg==","orcid":"","institution":"Chongqing Medical University","correspondingAuthor":true,"prefix":"","firstName":"Wenfu","middleName":"","lastName":"Cao","suffix":""},{"id":453991755,"identity":"698e0c86-4f7f-4997-b4ff-055ec5a8e51c","order_by":7,"name":"Bin Wu","email":"","orcid":"","institution":"Chongqing Medical University","correspondingAuthor":false,"prefix":"","firstName":"Bin","middleName":"","lastName":"Wu","suffix":""}],"badges":[],"createdAt":"2025-04-02 09:38:21","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6359791/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6359791/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":82560197,"identity":"0e5d8330-3cbe-4b3e-b237-00901f2e9c7b","added_by":"auto","created_at":"2025-05-13 01:33:08","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1093706,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of DMs. \u003cstrong\u003eA\u003c/strong\u003e Pie chart showing the distribution of identified metabolites by chemical classification. \u003cstrong\u003eB\u003c/strong\u003e Volcanic plot showing DMs between the OA and control groups. Red dots denote highly increased metabolites, blue dots denote highly decreased metabolites, and gray dots denote metabolites with no notable variations. \u003cstrong\u003eC\u003c/strong\u003e OPLS-DA score plot demonstrating clear separation between the OA red triangles and control blue circles groups, indicating distinct metabolic profiles between the two groups. \u003cstrong\u003eD\u003c/strong\u003e Permutation test n=200 validating the OPLS-DA model reliability, with the Q2 blue and R2Y red values indicating model quality and predictability. \u003cstrong\u003eE\u003c/strong\u003e Lollipop chart showing the top 20 significantly altered metabolites between the OA and control groups.\u003cstrong\u003eF\u003c/strong\u003e Metabolic pathway enrichment analysis bubble plot. The x-axis represents the enrichment ratio, while the y-axis shows the names of metabolic pathways. Bubble size indicates the number of enriched metabolites, and color represents significance (P-value).\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6359791/v1/c26e6c65034435a47f85f07c.png"},{"id":82561918,"identity":"6a37fafc-2632-489f-8a19-ced6fb633a9f","added_by":"auto","created_at":"2025-05-13 01:41:08","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":851823,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification and validation of metabolic characteristics associated with bone homeostasis in OA. \u003cstrong\u003eA\u003c/strong\u003e Hierarchical clustering heatmap showing that the DMs significantly correlated with bone turnover markers. \u003cstrong\u003eB\u003c/strong\u003e The distribution patterns of significantly altered metabolites between the control group orange and the OA group blue. \u003cstrong\u003eC\u003c/strong\u003e LASSO regression trajectory plot. \u003cstrong\u003eD\u003c/strong\u003eCross-validation optimization plot for LASSO model selection. \u003cstrong\u003eE\u003c/strong\u003e ROC curve analysis showing the diagnostic efficacy of a combination of six metabolites.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6359791/v1/458bb678fefc9a13c46fff75.png"},{"id":82559165,"identity":"16584105-de96-48dc-bb73-1a130d7944b4","added_by":"auto","created_at":"2025-05-13 01:25:08","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1366863,"visible":true,"origin":"","legend":"\u003cp\u003eAnalysis of the transcriptome. \u003cstrong\u003eA\u003c/strong\u003e The differential expression patterns of three different RNA types mRNAs, miRNAs, and lncRNAs are shown in volcano plots. \u003cstrong\u003eB\u003c/strong\u003e Bar plots of the results of the GO enrichment analysis showing the functional enrichment outcomes of genes with variable expression. \u003cstrong\u003eC\u003c/strong\u003e The top 20 significantly enriched signaling pathways are depicted in a bubble plot. \u003cstrong\u003eD\u003c/strong\u003eResults of the GSEA.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6359791/v1/b23a92661234153be8c5ff71.png"},{"id":82560199,"identity":"575e4d16-a6dc-4d4c-aa49-9f76aec61dec","added_by":"auto","created_at":"2025-05-13 01:33:08","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2023477,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of Hub Genes and PPI Networks in OA. \u003cstrong\u003eA\u003c/strong\u003e Global PPI network of DEGs constructed via the STRING database. \u003cstrong\u003eB\u003c/strong\u003e High-density topological module Cluster 1 identified by the MCODE algorithm score = 14.286, containing 29 nodes and 200 edges. \u003cstrong\u003eC\u003c/strong\u003e Core regulatory network identified through cytoHubba plugin analysis via the MCC algorithm. \u003cstrong\u003eD\u003c/strong\u003e Venn diagram illustrating the intersection of the hub genes identified by MCODE and cytoHubba analyses. \u003cstrong\u003eE\u003c/strong\u003eIntegrated functional network of the hub genes generated via the GeneMANIA database, which revealed significant downregulation of critical nodes.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6359791/v1/9a5d67f3b47fe6639eedeedc.png"},{"id":82560200,"identity":"fe22a933-6d18-4fda-ba1c-53f5cffdab64","added_by":"auto","created_at":"2025-05-13 01:33:08","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":766405,"visible":true,"origin":"","legend":"\u003cp\u003eceRNA regulatory network built on the basis of differentially expressed RNAs and their core modules. \u003cstrong\u003eA\u003c/strong\u003e The complete regulatory network mediated by lncRNAs is displayed, with node shapes representing different RNA types diamonds for miRNAs, ellipses for lncRNAs, and polygons for mRNAs. The gray connecting lines indicate lncRNA‒miRNA-mRNA interactions. \u003cstrong\u003eB\u003c/strong\u003e The core regulatory module identified by the MCC algorithm in cytoHubba, including key miRNAs and regulatory lncRNAs.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6359791/v1/87c4183a21cf17bc9998b01b.png"},{"id":82560201,"identity":"71095142-38c6-4342-8c66-3dc1c4601a97","added_by":"auto","created_at":"2025-05-13 01:33:08","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1437297,"visible":true,"origin":"","legend":"\u003cp\u003eIntegrative analysis of the transcriptome and metabolome in OA. \u003cstrong\u003eA\u003c/strong\u003e Spearman correlation heatmap between DEGs and DMs. \u003cstrong\u003eB\u003c/strong\u003e DEG‒DM interaction network. Node shapes: circles denote genes, diamonds represent metabolites; node colors: red indicates upregulation, and green signifies downregulation. \u003cstrong\u003eC\u003c/strong\u003eHistidine metabolic pathway enrichment analysis. The left side shows the pathway map, displaying the locations and interactions of key metabolites and genes. \u003cstrong\u003eD\u003c/strong\u003e Bubble plot of joint pathway analysis for DEGs-DMs. The bubble size denotes the number of enriched genes. \u003cstrong\u003eE\u003c/strong\u003e Expression heatmap of key genes in the IL-17 signaling pathway.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-6359791/v1/effdc2c55dcfc14e8759175d.png"},{"id":82559171,"identity":"2397690b-046e-4cc0-85b8-55ee429e09fe","added_by":"auto","created_at":"2025-05-13 01:25:08","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":749312,"visible":true,"origin":"","legend":"\u003cp\u003eValidation of five specifically expressed hub genes \u003cstrong\u003eA\u003c/strong\u003e-\u003cstrong\u003eB\u003c/strong\u003e Violin plots displaying the expression of specifically expressed hub genes via the validation dataset. \u003cstrong\u003eC\u003c/strong\u003eQuantification of IL1B, JUND, FOS, IL6, and JUNB mRNA levels in PBMCs. *P\u0026lt;0.05; **P\u0026lt;0.01; ***P\u0026lt;0.001.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-6359791/v1/9b5209a774865627d4f92980.png"},{"id":89451388,"identity":"cea9ace4-150c-4275-af45-e9a20507c2cf","added_by":"auto","created_at":"2025-08-20 06:16:55","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":8627269,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6359791/v1/7928ba2d-fad3-49d4-a305-2b2653a12df2.pdf"},{"id":82559162,"identity":"27f9d7c7-b874-46dd-86f6-12d088ff5457","added_by":"auto","created_at":"2025-05-13 01:25:08","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":382834,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Information\u003cbr\u003e\n Supplementary Material 1\u003cbr\u003e\nSupplementary Material 2\u003c/p\u003e","description":"","filename":"SupplementaryMaterial.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6359791/v1/503ef425716b3f24340caccd.pdf"},{"id":82559158,"identity":"992893cd-1e7e-4c07-af63-616dd120c17f","added_by":"auto","created_at":"2025-05-13 01:25:08","extension":"csv","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":3822,"visible":true,"origin":"","legend":"","description":"","filename":"IntegrationDMsDEGsPathways.csv","url":"https://assets-eu.researchsquare.com/files/rs-6359791/v1/34fc92b4e5510b53db21eb63.csv"}],"financialInterests":"No competing interests reported.","formattedTitle":"Integrated Metabolomic and Transcriptomic Analysis Identifies Candidate Regulatory Networks in Bone Metabolic Dysregulation of Osteoarthritis","fulltext":[{"header":"Introduction","content":"\u003cp\u003eOA constitutes a frequently encountered deteriorative disorder of the locomotor system that typically manifests with articular nociception, compromised joint function, and architectural derangements of affected articulations\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Recent Global Burden of Disease studies indicate that over the last 30 years (1990\u0026ndash;2020), the number of cases of OA has increased by 339\u0026nbsp;million (132%), currently affecting approximately 7.6% of the world's population, constituting a major public health challenge. Females are more likely than males to have this disorder, and its prevalence increases notably with age\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Multiple risk factors including aging, joint injuries, metabolic imbalances, changes in bone density, obesity, and sex variations collectively contribute to OA development\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. From a pathological standpoint, OA distinguishes itself through several hallmark tissue alterations: aberrant osseous projections at joint peripheries, pathological hardening of bone beneath cartilage surfaces, inflammatory transformation of synovial linings, and structural compromise of articular cartilage, all functioning within a sophisticated network of reciprocal pathological interactions\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWhile OA has long been seen as a disease of cartilage, growing evidence shows that abnormal bone metabolism serves as a critical determinant in OA progression\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Clinical studies have found that bone metabolic markers in OA patients differ significantly from healthy individuals, including changes in standardized uptake values (SUV) and bone metabolic kinetic parameter K\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Some OA patients show increased levels of bone sialoprotein, osteocalcin, and abnormal RANKL/OPG ratios\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Notably, these bone changes often occur before cartilage damage and are closely related to disease severity\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. However, despite these significant clinical observations, the specific metabolic pathways and gene regulatory networks that drive these bone metabolic abnormalities remain incompletely elucidated, impeding the development of targeted therapeutic interventions.\u003c/p\u003e \u003cp\u003eTo explore the molecular basis of these bone metabolic abnormalities, metabolomics as a sophisticated analytical technique that systematically quantifies small molecule metabolites in biological systems and identifies relationships between metabolic profiles and disease states\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e, offers a new perspective for OA research. Previous metabolomic investigations have revealed distinct metabolic signatures in OA patients, particularly involving altered lipid metabolism, amino acid processing, and energy metabolism pathways\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. However, these studies have primarily examined synovial fluid or cartilage tissue\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, while systemic metabolic changes that may drive bone metabolic abnormalities remain poorly characterized. Concurrently, with advancements in transcriptomics, analysis of gene expression regulation in OA has gained significant attention. These studies have identified candidate biomarkers and implicated several signaling cascades in OA pathophysiology\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. However, conventional transcriptomic approaches are often restricted to localized tissue examination, limiting their ability to detect how systemic metabolic alterations affect bone homeostasis. Thus, single-omics approaches struggle to comprehensively reveal the complex mechanisms underlying OA bone metabolic abnormalities.\u003c/p\u003e \u003cp\u003eGiven this research gap, the integration of metabolomic and transcriptomic datasets provides a powerful research paradigm. This approach allows for concurrent examination of both metabolite fluctuations and gene expression changes. It utilizes the complementary nature of these omics platforms to thoroughly assess the interrelationship between transcriptional regulation and metabolic dysfunction. This systems biology approach can uncover mechanistic insights not apparent through single-omics analysis. In particular, such multi-omics integration, through cross-validation, can establish causal relationships between metabolism, gene expression, and phenotype, providing more robust evidence regarding the biological mechanisms underlying bone metabolic disturbances in OA.\u003c/p\u003e \u003cp\u003eIn this study, we investigated the metabolic and transcriptional drivers of OA bone metabolic abnormalities by analyzing peripheral blood samples. Peripheral blood as a research subject offers multiple advantages, not only providing a minimally invasive sampling method but also comprehensively reflecting systemic metabolic changes\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, while containing circulating factors directly involved in bone remodeling processes\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, and has been validated by multiple studies to effectively capture OA-specific molecular signatures\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. We conducted untargeted metabolomic analysis and whole-transcriptome sequencing on a screened study cohort. Using a comprehensive data analysis approach, we first categorized genes by their biological functions, then identified important biochemical pathways, measured relationships between different molecular components, and created visual maps of gene and metabolite interactions. This multi-step analysis helped us uncover the underlying molecular processes and disease patterns involved in bone metabolism disorders. In summary, this study provides a comprehensive multi-omics analysis of OA bone metabolic abnormalities, with potential to establish blood-based diagnostic and prognostic biomarker systems, identify molecular targets driving bone metabolic abnormalities, offer new perspectives on the pathogenesis of this disease, and contribute to the development of personalized medical approaches for OA patients.\u003c/p\u003e"},{"header":"Materials \u0026 methods","content":"\u003ch2\u003eParticipants\u003c/h2\u003e\n\u003cp\u003eThis exploratory pilot study recruited 60 participants, 30 OA patients and 30 Controls, from the Department of Rheumatology and Immunology at Chongqing Hospital of Traditional Chinese Medicine. Participants were selected for the study based on several criteria: (1) with the primary demographic requirement limiting participation to adults aged 40-70 years; (2) BMI between 18.5 and 30 kg/m²; and (3) a diagnosis of OA on the basis of the revised criteria of the American College of Rheumatology (ACR)\u003csup\u003e18\u003c/sup\u003e. Study exclusion was implemented for individuals who met any of these disqualifying criteria: (1) comorbid bone or joint diseases, such as osteoporosis, rheumatoid arthritis, or traumatic arthritis; (2) metabolic disorders, including type 2 diabetes; (3) use of oral or intra-articular glucocorticoids within the past 3 months; and (4) joint replacement surgery within the past 6 months. Given the exploratory nature of this multiomics study, the sample size was determined based on similar pilot studies published in the field\u003csup\u003e19,20\u003c/sup\u003e and practical resource constraints. Existing research has demonstrated that 20-30 samples per group can effectively identify major metabolic alterations\u003csup\u003e21\u003c/sup\u003e. When evaluated retrospectively, our participant numbers demonstrated adequate sensitivity (70% power) for detecting pronounced effect magnitudes (Cohen's d ≥ 0.8) while maintaining standard type I error rates of 0.05, which is sufficient for identifying candidate pathways and generating hypotheses for future large-scale validation studies.\u0026nbsp;Demographic characteristics (sex, age, and BMI) were comparable between the two groups, with no statistically significant differences identified (p \u0026gt; 0.05). All subjects provided written informed consent prior to study participation. This investigation was conducted in accordance with the ethical standards delineated in the Declaration of Helsinki and received approval from the Ethics Committee of Chongqing Hospital of Traditional Chinese Medicine (approval number: 2022-KY-YJS-LY).\u003c/p\u003e\n\u003ch2\u003eSample Collection and Marker Measurement\u003c/h2\u003e\n\u003cp\u003eTen milliliters of fasting venous blood was collected from each participant. Five milliliters were transferred to procoagulant tubes and centrifuged (3,000 rpm, 10 min, 4°C) for serum separation. The obtained serum was stored at -80°C for subsequent biochemical and metabolomic analyses. The remaining 5 mL was collected in EDTA tubes for peripheral blood mononuclear cell (PBMC) isolation using Lymphoprep™ density gradient centrifugation (1.077 g/mL; STEMCELL Technologies, Canada). Isolated PBMCs were washed twice with sterile phosphate-buffered saline (PBS) at 4°C and pelleted by centrifugation (1,000 rpm, 5 min). For long-term preservation, the cell pellets were resuspended in fetal bovine serum containing 10% dimethyl sulfoxide (DMSO), gradually cooled (1°C/min) to -80°C overnight, and then transferred to liquid nitrogen until RNA extraction and transcriptomic analysis.\u003c/p\u003e\n\u003cp\u003eSerum levels of PINP, N-MID, and β-CTX were measured via enzyme-linked immunosorbent assay (ELISA) kits (Wuhan Elabscience Biotechnology Co., Ltd.) in accordance with the manufacturer’s instructions. An automatic microplate reader (RT-6100, Rayto, USA) was used to quantify the results, enabling an assessment of bone metabolic markers.\u003c/p\u003e\n\u003ch2\u003eUntargeted Metabolite Extraction and Analysis\u003c/h2\u003e\n\u003cp\u003eAfter the serum samples were thawed on ice, 400 μL of extraction solution (methanol:water, 8:2, v/v) was combined with 100 μL aliquots of serum or plasma. After two minutes of vortexing, the mixture was centrifuged for 10 min at 4°C at 12,000 rpm. The supernatants were subsequently collected and placed into liquid chromatography–mass spectrometry (LC‒MS) vials for further examination. Equal quantities from each of the ten samples were combined to create a quality control (QC) sample, which was used to guarantee the dependability of the analytical platform. Blank samples (containing pure solvent) were also included to avoid contamination during sample preparation.\u003c/p\u003e\n\u003cp\u003eMetabolomic analysis was performed using ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) on a Vanquish UHPLC system coupled with a Q Exactive™ HF-X mass spectrometer (Thermo Fisher Scientific, Germany) at Novogene Co., Ltd. (Beijing, China). Chromatographic separation was achieved using an ACQUITY UPLC BEH amide column (2.1 mm × 100 mm, 1.7 μm; Waters). The acquired LC-MS data were processed using Compound Discoverer v3.1 software. Metabolite identification and quantification were performed following peak alignment across samples, applying stringent filtering criteria including a retention time deviation tolerance of 0.2 min and a mass accuracy threshold of 5 ppm\u003csup\u003e22\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eTarget peaks were identified on the basis of reference standards and databases. Compounds were accurately extracted by refining parameters such as mass deviation, signal intensity variation, and signal-to-noise ratio thresholds. The peak areas, which represent the relative abundances of the compounds, were quantified through established quantitative analysis methods. Molecular formulas were predicted according to molecular ion peaks as well as fragment ion data, and compounds were identified by comparison with databases such as mzCloud, mzVault, and MassList. Background ion interference was minimized via the use of blank samples, and the quantitative results were standardized to calculate relative peak areas. Compounds with a coefficient of variation (CV) exceeding 30% in the QC samples were excluded to ensure reliable metabolite quantification.\u003c/p\u003e\n\u003cp\u003ePrincipal component analysis (PCA) and orthogonal projections to latent structures discriminant analysis (OPLS-DA) were employed to examine metabolite variations between the two groups\u003csup\u003e23,24\u003c/sup\u003e. The OPLS-DA model's robustness was confirmed by permutation testing, and the ideal number of principal components was ascertained using 10-fold cross-validation. P values \u0026lt; 0.05 and variable importance in projection (VIP) scores \u0026gt; 1 were used to identify differential metabolites (DMs). Kyoto Encyclopedia of Genes and Genomes (KEGG), the Human Metabolome Database (HMDB), and the LIPID Metabolites and Pathways Strategy (LIPID MAPS) were used to annotate these metabolites. For visualization of the differences in metabolites, a thorough analysis of metabolite differences and sample groups was performed via the pheatmap tool for clustering heatmaps and the ggplot2 R package for creating volcano plots. The significantly changed metabolites were subjected to Z score normalization, and the pathways most affected by these alterations were identified using pathway enrichment analysis using MetaboAnalyst 6.0\u003csup\u003e25\u003c/sup\u003e. R was also used to perform correlation analysis of the DMs.\u003c/p\u003e\n\u003ch2\u003eCorrelation of DMs with OA Clinical Markers\u003c/h2\u003e\n\u003cp\u003eSpearman correlation coefficients were calculated to investigate the connection between DMs and OA clinical markers; values greater than 0.7 indicate strong relationships. The glmnet package in R was used to identify metabolites that could effectively differentiate OA patients from controls via least absolute shrinkage and selection operator (LASSO) regression, a regularization strategy intended to increase prediction accuracy\u003csup\u003e26\u003c/sup\u003e. For analysis of the diagnostic potential for important metabolites, receiver operating characteristic (ROC) analysis was conducted via the pROC package in R. The classification efficacy was measured using the area under the curve (AUC), and a P value \u0026lt; 0.05 was considered statistically significant.\u003c/p\u003e\n\u003ch2\u003eWhole-Transcriptome Sequencing and Analysis\u003c/h2\u003e\n\u003cp\u003eFollowing the manufacturer's instructions, total RNA was extracted using the TRIzol reagent (Invitrogen, Carlsbad, California, USA) and then purified. A NanoDrop spectrophotometer (Thermo Fisher Scientific, USA) was used to determine the concentration and purity of the RNA, while electrophoresis on 1% agarose gels was used to evaluate RNA integrity as well as contamination. An Agilent 2100 Bioanalyzer system (Agilent Technologies, USA) was used to further evaluate the quality of the RNA. For library preparation, only samples with a 28S/18S ratio \u0026gt; 1.8 and an RNA integrity number (RIN) \u0026gt; 7.5 were employed to ensure high-quality starting material.\u003c/p\u003e\n\u003cp\u003eThe NEBNext® Ultra™ II RNA Library Prep Kit for Illumina® (New England Biolabs [NEB], Ipswich, Massachusetts, USA) was used to construct long noncoding RNA (lncRNA) libraries, whereas the NEBNext® Multiplex Small RNA Library Prep Set for Illumina® (NEB) was used to create small RNA libraries. Each sample was given a unique index code for sequence identification after ribosomal RNA (rRNA) was eliminated. The AMPure XP system (Beckman Coulter, Brea, California, USA) was harnessed to purify the library.To minimize technical variation, all samples were processed in parallel using the same reagent lots. The Illumina NovaSeq 6000 platform was harnessed for sequencing, producing paired-end reads of 150 bp. The target fragment size and effective concentration were used to pool the libraries. Clean reads were utilized for further analysis after adapter sequences and low-quality reads were removed from the raw data. Library preparation and sequencing were conducted by Novogene Bioinformatics Technology Co., Ltd. (Beijing, China).\u003c/p\u003e\n\u003cp\u003eHigh-throughput sequencing data (Illumina PE150/SE50) were processed with CASAVA base calling, converting the raw image files into FASTQ format. FastQC software (version 0.11.5) was used for QC as well as data filtering. Reads with a Phred quality score \u0026lt; 20 or lengths \u0026lt; 50 nucleotides were discarded. This included assessing sample-to-sample correlation matrices and PCA to identify potential outliers or batch effects. To address potential experimental batch effects, we employed the ComBat algorithm implemented in the sva R package\u003csup\u003e27\u003c/sup\u003e, which uses an empirical Bayes framework to adjust for known batch effects while preserving biological variation of interest. The reference genome index was constructed via HISAT2 (v2.0.5), which was also employed to align filtered paired-end reads to the reference genome. Transcript assembly was carried out via StringTie software (version 1.3.3b), and transcript abundance was quantified as fragments per kilobase of transcript per million mapped reads (FPKM)\u003csup\u003e28\u003c/sup\u003e. For small RNA analysis, length-filtered reads were aligned to the reference sequence via Bowtie (version 1.0.1)\u003csup\u003e29\u003c/sup\u003e, and matches were compared with miRBase to identify small RNAs. Novel miRNAs were predicted using miRNA detection tools such as miREvo\u003csup\u003e30\u003c/sup\u003e and miRDeep2\u003csup\u003e31\u003c/sup\u003e. The expression levels of known and novel miRNAs were normalized and quantified as transcripts per million (TPM) values.\u003c/p\u003e\n\u003cp\u003eDifferential expression analysis of RNAs was conducted via DESeq2, which applies a negative binomial distribution model\u003csup\u003e32\u003c/sup\u003e. Differentially expressed mRNAs, lncRNAs, and miRNAs (DEGs, DE-lncRNAs, and DE-miRNAs) were identified via a Benjamini‒Hochberg adjusted P value (padj) ≤ 0.05 and a |log2-fold change (FC)| ≥ 1 as thresholds. Heatmaps were generated with the heatmap R package, and volcano plots were created via the ggplot2 R package.\u003c/p\u003e\n\u003cp\u003eGene Ontology (GO) and KEGG enrichment analyses were conducted to study the functional relevance of the DEGs. KEGG included information on pathways, illnesses, and medications, whereas GO analysis covered three primary categories: biological process (BP), molecular function (MF), and cellular component (CC). The clusterProfiler R package\u003csup\u003e33\u003c/sup\u003e was used for both analyses, and a P value \u0026lt; 0.05 was deemed significant.\u003c/p\u003e\n\u003cp\u003eFurther research into the BPs associated with OA bone metabolism was conducted using gene set enrichment analysis (GSEA). The \"symbols.gmt\" gene set from the Molecular Signatures Database (MSigDB, v7.5.1)\u003csup\u003e34\u003c/sup\u003e was employed using GSEA software (v4.2.0). Gene sets were considered highly enriched if their P value was less than 0.05 and if their false discovery rate (FDR) was less than 0.25 after 1,000 permutations.\u003c/p\u003e\n\u003cp\u003eProtein‒protein interaction (PPI) networks were constructed with the STRING database\u003csup\u003e35\u003c/sup\u003e and Cytoscape software\u003csup\u003e36\u003c/sup\u003e. Hub genes were identified via the CytoHubba and Molecular Complex Detection (MCODE) plugins. With the MCODE settings set to degree cutoff = 2, node score cutoff = 0.2, k-core = 2, and max depth = 100, DEGs were included in the hub gene networks. The maximum clique centrality (MCC) technique was utilized by the CytoHubba plugin to rank the hub genes, and the PPI network was visualized via Cytoscape (v3.9.0).\u003c/p\u003e\n\u003cp\u003eFor prediction of additional PPIs, the GeneMANIA\u003csup\u003e37\u003c/sup\u003e web tool was utilized. This tool incorporates bioinformatic methods, including physical interaction, coexpression, genetic interaction, gene enrichment, and colocalization, to construct comprehensive interaction networks. In this study, GeneMANIA was used to predict interaction networks for the hub genes and proteins.\u003c/p\u003e\n\u003ch2\u003eceRNA Network Construction\u003c/h2\u003e\n\u003cp\u003eBy binding microRNAs (miRNAs), competing endogenous RNAs (ceRNAs) control post-transcriptional gene expression. lncRNAs function as miRNA sponges in this situation, binding to miRNAs and modifying the expression of downstream target mRNAs. In light of current database limitations, we attempted an initial exploration using correlation-based ceRNA network analysis in this research\u003csup\u003e38\u003c/sup\u003e. First, differential expression analysis was employed to identify mRNAs, miRNAs, and lncRNAs that were differentially expressed. Considerably negative correlations between miRNAs and their target RNAs (mRNAs and lncRNAs) and considerably positive correlations between RNA pairs that share the same miRNA (e.g., lncRNA and mRNA) were required for the calculation of expression correlations between these three RNA types. These connections led to the construction of ceRNA ternary regulatory modules using lncRNA‒miRNA-mRNA pairs that shared the same miRNA. Finally, all significant ternary regulatory units were integrated into a ceRNA regulatory network, and key nodes and critical regulatory modules were identified through network topology analysis. The entire network was constructed and visualized via Cytoscape software, which may suggest possible regulatory relationships that could potentially inform future functional validation studies.\u003c/p\u003e\n\u003ch2\u003eJoint Metabolomic and Transcriptomic Analysis\u003c/h2\u003e\n\u003cp\u003eDMs and DEGs linked to markers of bone metabolism were chosen according to predetermined standards to conduct an integrated analysis of metabolomic and transcriptomic data. P \u0026lt; 0.05 and an absolute log2-fold change (|log2FC|) \u0026gt; 1 were used to indicate DEGs, whereas a VIP score \u0026gt; 1 and P \u0026lt; 0.05 were employed to identify DMs. Spearman's correlation analysis was used to study the connections among bone metabolic markers, DEGs, and DMs. Only those that were deemed substantially associated and included in the integrated analysis had an absolute correlation coefficient (|r|) \u0026gt; 0.7 and P \u0026lt; 0.05. Heatmaps were created to show the relationships between genes and metabolites. These selected DMs and DEGs were further used to construct networks via Cytoscape software. This visualization enabled a detailed exploration of metabolite‒gene relationships and provided an overview of the composite network. MetaboAnalyst 6.0 was used to identify enriched metabolic pathways and KEGG pathways represented in both the metabolomic and transcriptomic data. Additionally, the summarized data facilitated the classification of DEGs and metabolites, providing a comprehensive view of the integrated analysis.\u003c/p\u003e\n\u003ch2\u003eReal-Time Quantitative PCR (RT-qPCR)\u003c/h2\u003e\n\u003cp\u003eRNA extraction and gene expression validation were performed on PBMCs obtained from 20 subjects (10 OA patients and 10 controls). Total RNA was isolated using a TRIzol-based protocol. Briefly, samples in 1.5-mL Eppendorf tubes were treated with 1 mL TRIzol reagent and 200 µL chloroform, followed by ice incubation (10 min) and centrifugation (12,000 ×g, 15 min). The aqueous phase (400 µL) was precipitated with an equivalent volume of isopropanol at ambient temperature for 10 min. Following centrifugation (12,000 ×g, 10 min), the resulting RNA precipitate underwent dual washing procedures with 75% ice-cold ethanol (1 mL) and was subsequently solubilized in RNase-free water.\u003c/p\u003e\n\u003cp\u003eRNA integrity was evaluated using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific), with A260/A280 ratios between 1.8-2.0 considered acceptable. Reverse transcription was conducted using 2× SPARKscript II All-in-One RT SuperMix (SparkJade; Dongsi Kejie Biotechnology Co., Ltd., China).\u003c/p\u003e\n\u003cp\u003eQuantitative PCR was performed using SYBR Green qPCR Master Mix (without ROX; Applied Biosystems) under the following thermal parameters: initial denaturation (95°C, 30 s) followed by 60 amplification cycles (95°C for 10 s, 62°C for 30 s). Gene expression was normalized to GAPDH using the comparative threshold (2−ΔΔCt) method. Target gene primers (Table S1) were synthesized by Shanghai Bioengineering Technology Service Company (Shanghai, China). All analyses were performed in triplicate with GAPDH as the reference standard.\u003c/p\u003e\n\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\n\u003cp\u003eData analysis was conducted using SPSS (version 24.0; IBM, Armonk, NY, USA), GraphPad Prism (version 9.0; GraphPad Software, San Diego, CA, USA), and R (version 4.3.0; R Foundation for Statistical Computing, Vienna, Austria). Statistical summaries included tabular formats, frequency distributions, and graphical illustrations. The Shapiro-Wilk method was employed to evaluate distribution normality of continuous variables. For normally distributed parameters, results were presented as mean ± standard deviation and analyzed via independent-samples t tests. Non-parametric data were expressed as median with interquartile range (25th–75th percentile) and evaluated using the Mann-Whitney U test. Categorical variables were reported as frequency counts with percentages, with between-group comparisons performed using either Pearson's chi-square analysis or Fisher's exact test when appropriate. Statistical significance threshold was established at P \u0026lt; 0.05.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eClinical Characteristics\u003c/p\u003e\n\u003cp\u003eThis study enrolled 30 patients with primary osteoarthritis (OA) and 30 age- and sex-matched controls. To mitigate potential confounding factors, we excluded subjects with inflammatory arthritis, secondary OA, disorders affecting bone metabolism, or those receiving bone-active medications. As demonstrated in Table 1, no statistically significant differences were observed between groups regarding demographic characteristics (p \u0026gt; 0.05). The distribution of comorbidities (hypertension, diabetes mellitus, dyslipidemia) and medication use (non-steroidal anti-inflammatory drugs, analgesics) was relatively balanced between OA patients and controls, with no significant differences detected. To rigorously control for potential confounding influences, we constructed multivariate linear regression models with bone turnover markers as dependent variables and OA status as the primary independent variable, while adjusting for demographic characteristics, comorbidities, and medication use as covariates. The adjusted analyses (Table S2) revealed that the significant association between OA and elevated bone turnover marker levels persisted after controlling for these confounding factors (p \u0026lt; 0.001), with negligible attenuation of effect sizes, indicating that this association reflects a genuine biological relationship between OA and altered bone metabolism rather than an artifact of differences in comorbidity patterns or medication effects.\u003c/p\u003e\n\u003cp\u003eTable 1. Baseline Characteristics of Study Participants\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"99%\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOsteoarthritis (n = 30)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eControl (n = 30)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eP-value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eDemographics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eSex, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e0.792\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e19 (63.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e17 (56.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e11 (36.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e13 (43.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eAge, years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e56.7 \u0026plusmn; 6.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e54.7 \u0026plusmn; 4.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e0.211\u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eBMI, kg/m\u0026sup2;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e22.1 \u0026plusmn; 2.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e21.9 \u0026plusmn; 2.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e0.749\u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eBone turnover markers\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003e\u0026beta;-CTX, pg/mL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e88.2 [75.0-96.1]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e64.1 [49.6-69.2]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003ec\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eN-MID, ng/mL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e1.88 \u0026plusmn; 0.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e1.30 \u0026plusmn; 0.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003ePINP, ng/mL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e1.89 \u0026plusmn; 0.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e1.24 \u0026plusmn; 0.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eComorbidities, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eHypertension\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e12 (40.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e8 (26.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e0.273\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eDiabetes mellitus\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e11 (36.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e9 (30.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e0.584\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eDyslipidemia\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e14 (46.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e11 (36.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e0.432\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eCurrent medications, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eNSAIDs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e7 (23.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e10 (33.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e0.390\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 35px;\"\u003e\n \u003cp\u003eAnalgesics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31px;\"\u003e\n \u003cp\u003e9 (30.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e6 (20.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e0.371\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eData are presented as mean \u0026plusmn; standard deviation, median [interquartile range], or number (percentage).\u003cbr\u003e\u0026nbsp;BMI: body mass index; \u0026beta;-CTX: \u0026beta;-isomerized C-terminal telopeptide of type I collagen; N-MID: N-terminal mid-fragment osteocalcin; PINP: procollagen type I N-terminal propeptide; NSAIDs: non-steroidal anti-inflammatory drugs.\u003cbr\u003e\u0026nbsp;\u003csup\u003ea\u003c/sup\u003eChi-square test or Fisher\u0026apos;s exact test;\u0026nbsp;\u003csup\u003eb\u003c/sup\u003eIndependent t-test;\u0026nbsp;\u003csup\u003ec\u003c/sup\u003eMann-Whitney U test.\u003c/p\u003e\n\u003cp\u003eUntargeted Metabolomic Analysis\u003c/p\u003e\n\u003cp\u003eWe investigated metabolic alterations between OA patients and controls through untargeted metabolomic analysis. LC-MS/MS analysis was performed on serum samples from 60 participants (30 OA patients and 30 age-sex matched controls). Following data processing, we identified 488 metabolic features using HDBM and KEGG annotation databases, with 263 detected in positive ion mode and 225 in negative ion mode. The database-annotated metabolites were categorized into chemical groups by the integration of positive and negative ion modes. The largest fractions were composed of lipids as well as lipid-like molecules, organic acids and their derivatives, organoheterocyclic compounds, benzoenoids, and other groups (Fig. 1A). Differences in metabolites between the positive and negative ion modes were detected via univariate analysis, and a volcano plot representing typical DMs (VIP \u0026gt; 1.0, P \u0026lt; 0.05) is displayed (Fig. 1B). OPLS-DA modeling revealed clear separation between OA patients and controls (R2Y=0.94, Q2Y=0.88) (Fig. 1C), with permutation testing validating model reliability (Fig. 1D). The matchstick plot displayed the top 20 upregulated and downregulated metabolites in OA patients (Fig. 1E). Notably, benzoic acid derivatives, fatty acids, and glycerophosphocholines were significantly upregulated, while amino acids, peptides, bile acids, and arachidonic acid derivatives were markedly downregulated.\u003c/p\u003e\n\u003cp\u003eKEGG pathway enrichment analysis revealed significant involvement of multiple metabolic pathways in OA (Fig. 1F). Upregulated metabolites were enriched in retinol metabolism, ubiquinone biosynthesis, and arachidonic acid metabolism pathways, while downregulated metabolites were primarily associated with histidine metabolism, steroid hormone biosynthesis, and bile acid production. Further analysis of specific metabolites revealed elevated inflammatory mediators (prostaglandin G2) and decreased chondroprotective compounds (estradiol) in OA patients.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFinally, we examined metabolites that play biological roles in eukaryotes using the HDMB database. Prostaglandin G2 (PGG2), hydroquinone, vitamin A, 3,4-dihydroxycinnamic acid, hydroxybenzoic acid, gonadotropin, and phenylacetaldehyde were among the representative metabolites whose levels substantially increased in the OA group. Metabolite levels in the OA group, in contrast, were considerably lower for estradiol, histidine, arginine, hydroxydodecanoic acid, cholic acid, lysine, glycochenodeoxycholic acid, 6-keto-prostaglandin, glutamic acid, lithocholic acid, and phosphatidylcholine. Overall, our findings indicate that OA patients have metabolic traits distinct from those of Controls.Our findings indicate that OA is associated with significant metabolic reprogramming, characterized by enhanced inflammatory pathways and suppressed bone anabolic processes, providing a potential strategy for early diagnosis and targeted treatment of OA.\u003c/p\u003e\n\u003cp\u003eCorrelation Analysis of Metabolites and Clinical Markers\u003c/p\u003e\n\u003cp\u003eTo screen for high-confidence metabolic features related to bone homeostasis, we selected significant differentially expressed metabolites (P \u0026lt; 0.05, VIP \u0026gt; 1) and analyzed their associations with bone metabolic markers using Spearman correlation analysis. Results are shown in Fig. 2A. Among upregulated metabolites, phenylglyoxylic acid, tropine, and PGG2 correlated with \u0026beta;-CTX levels, tropine correlated with PINP levels, and phenylglyoxylic acid and tropine correlated with N-MID levels. Among downregulated metabolites, multiple compounds showed significant correlations with bone metabolic markers, with 4-nitrophenol, stearamide, 6-hydroxymelatonin, lithocholic acid, and anserine correlating with all three bone metabolic markers.\u003c/p\u003e\n\u003cp\u003eComparing bone homeostasis markers, taurolithocholic acid was associated with bone formation markers PINP and N-MID, while 3-hydroxy-3-methylbutanoic acid and propylparaben were associated with the bone resorption marker \u0026beta;-CTX. These metabolic profile variations demonstrated the characteristics of abnormal bone homeostasis in OA patients.\u003c/p\u003e\n\u003cp\u003eUsing bone metabolism-related differential metabolites as markers, we assessed their predictive performance for OA through ROC analysis. Serum biomarkers with AUC \u0026ge; 80%, including seven metabolites such as 1-methylguanosine, were screened (Fig. 2B). After LASSO algorithm feature selection (Fig. 2C and D), a model comprising six metabolites demonstrated excellent diagnostic efficacy (Fig. 2E, AUC = 0.991). These metabolites\u0026mdash;1-methylguanosine, 4-nitrophenol, anserine, tropine, 20-carboxy-leukotriene B4, and thymopentin\u0026mdash;show potential as effective diagnostic markers for OA.\u003c/p\u003e\n\u003cp\u003eThese differential metabolites align with our previous pathway analysis results. For instance, PGG2 upregulation supports the role of inflammatory pathways in OA, bile acids reflect lipid metabolism disorders, and changes in amino acid-related molecules confirm the importance of protein metabolism in OA pathology. By integrating metabolomic and clinical phenotype data, we established a metabolite-clinical feature association network, providing multi-level evidence for the molecular pathological mechanisms of OA.\u003c/p\u003e\n\u003cp\u003eTranscriptomic Analysis\u003c/p\u003e\n\u003cp\u003eBased on our metabolomics cohort, we selected 20 matched pairs for transcriptomic analysis to investigate transcriptional variations in OA. Through RNA-seq and bioinformatic analysis, we identified 694 DEGs, 53 DE-miRNAs, and 36 DE-lncRNAs in PBMCs. Volcano plots displayed the distribution of differentially expressed genes (Fig. 3A), with 482 downregulated and 212 upregulated mRNAs, 1 upregulated and 35 downregulated lncRNAs, and 35 upregulated and 18 downregulated miRNAs in the OA group.\u003c/p\u003e\n\u003cp\u003eGO enrichment analysis revealed DEGs primarily enriched in T cell activation and differentiation pathways, mRNA splicing, and RNA polymerase II-related processes (Fig. 3B, left). DE-miRNAs were enriched in cytoskeletal organization, chromatin assembly, and signaling receptor binding (Fig. 3B, middle), while DE-lncRNAs were enriched in cytokine-mediated signaling, receptor complexes, and apoptotic processes (Fig. 3B, right).\u003c/p\u003e\n\u003cp\u003eKEGG pathway analysis revealed DEGs were enriched in MAPK signaling, Spliceosome, and IL-17 signaling pathways (Fig. 3C), indicating inflammatory and RNA processing mechanisms in OA. DE-miRNAs were enriched in viral infection, cytokine-receptor interaction, and Huntington disease pathways (Fig. S1A), suggesting roles in immune regulation. DE-lncRNAs showed significant enrichment in cancer pathways, neurodegeneration, and Alzheimer disease (Fig. S1B), pointing to involvement in cellular stress responses.\u003c/p\u003e\n\u003cp\u003eGSEA demonstrated multiple downregulated pathways in OA (Fig. 3D), including apoptosis, inflammatory response, and TNF-\u0026alpha; signaling via NF-\u0026kappa;B, with the latter showing particularly pronounced downregulation. Notably, cytokine-cytokine receptor interaction was enriched across all RNA types, suggesting this represents a core mechanism in OA pathophysiology regulated at multiple transcriptional levels.\u003c/p\u003e\n\u003cp\u003ePPI Network and Hub Gene Identification\u003c/p\u003e\n\u003cp\u003eTo deeply investigate functional interactions among DEGs, we constructed protein-protein interaction (PPI) networks. Using the STRING database, we created a PPI network with 103 nodes and 524 edges, demonstrating functional associations between DEGs (Fig. 4A). Using the MCODE method in Cytoscape software, we identified the top-ranked module Cluster 1 with a score of 14.286 (29 nodes and 200 edges) (Fig. 4B). Cluster 1 comprised 14 genes, including 2 upregulated genes (CCR2 and CXCR6) and 12 downregulated genes (FOSL1, CCL20, JUND, JUNB, JUN, CXCR2, CXCL8, IL6, IL1B, IFNG, FOSB, and FOS). Further analysis through the MCC algorithm of the cytoHubba plugin identified the top 20 core genes (Fig. 4C). The intersection of nodes ranked by MCODE and cytoHubba analyses determined the final set of hub genes (Fig. 4D). Among these genes, the expression levels of FOS, IL1B, IL6, and JUN were significantly reduced in OA samples (P \u0026lt; 0.05) (Fig. S1C), highlighting their potential as biomarkers for OA. Additionally, we employed the GeneMANIA database to predict interaction networks of these hub genes, revealing that most DEGs were closely related to the hub genes (Fig. 4E). These core genes form regulatory modules that play crucial roles in OA-related inflammation and cartilage degradation processes, influencing OA development by regulating downstream gene expression.\u003c/p\u003e\n\u003cp\u003eNoncoding RNA Regulatory Network\u003c/p\u003e\n\u003cp\u003eLncRNAs can function as upstream regulators of miRNAs to influence gene expression. To reveal the regulatory roles of non-coding RNAs in OA, we constructed a ceRNA regulatory network based on expression profile data. First, we processed all RNA-seq data uniformly to reduce the effect of sample-to-sample sequencing depth. Strict screening criteria (Spearman correlation coefficient threshold |r| \u0026gt; 0.7, FDR \u0026lt; 0.05) were used to identify miRNA-mRNA interaction pairs, and transcripts with unknown functions prefixed with LOC and XLOC were disqualified. With the same correlation criterion of |r| \u0026gt; 0.7, miRNA-lncRNA interaction pairs were screened. Finally, we utilized Cytoscape to create and visualize the lncRNA-miRNA-mRNA ceRNA network according to expression patterns of miRNAs, lncRNAs, and mRNAs from OA patients and controls. According to network topology analysis, the network exhibited a notable core-periphery structure and typical scale-free properties. The core nodes were composed primarily of highly connected miRNAs that might play important regulatory roles in OA onset and progression (Fig. 5A).\u003c/p\u003e\n\u003cp\u003eWe used the cytoHubba plugin in Cytoscape software to perform topological analysis of the complete network, identifying major regulatory modules in the ceRNA network. We then used the MCC technique to extract the core subnetwork, consisting of the top 20 most important hub genes. For example, HLA-E, the most important node in this subnetwork, directly regulates miRNA molecules including hsa-miR-136-3p, hsa-miR-487b-3p, hsa-miR-299-5p, and hsa-miR-410-3p. The subnetwork contained 11 miRNA nodes (shown by diamonds) and 7 target gene and lncRNA nodes (shown by ellipses). Notably, lncRNA molecules, including LINC00880, NUTM2G, and RUSC1-AS1, may exert critical synergistic functions in the regulatory network by competitively interacting with specific miRNAs (Fig. 5B). This ceRNA network reveals the complex roles of lncRNA-miRNA-mRNA three-layer regulatory relationships in OA pathogenesis, providing new perspectives for understanding non-coding RNA functions in OA.\u003c/p\u003e\n\u003ch2\u003eIntegration and Crosstalk between the Metabolome and Transcriptome\u003c/h2\u003e\n\u003cp\u003eTo further explore the pathogenic mechanisms of bone homeostasis imbalance in OA, we performed a multi-dimensional integrated analysis of whole transcriptome and metabolome data from the same biological samples. First, we employed GSEA using the C2 and C7 collections from the MSigDB to identify key pathways related to inflammation and metabolism. Subsequently, using Spearman correlation analysis, we screened differentially expressed genes and DMs associated with bone metabolism markers, establishing the connection between OA DEGs and metabolites. We visualized the overall relationships between metabolites and phenotypes using a heatmap (Fig. 6A). Based on stringent selection criteria (|cor| \u0026gt; 0.7, P \u0026lt; 0.05), we constructed a completely linked network of DMs and DEGs (Fig. 6B), clearly identifying transcription-metabolism co-regulatory modules.Within this network, we identified key co-regulatory feature clusters, including genes linked to inflammation, such as TNFAIP3; cell signaling pathways, such as PRDM3B and SORL1; and receptors, such as ROR and PPX. Notably, genes including TNFAIP3, LINC00905, and CKB6 exhibited significant positive or negative correlations with metabolites such as isoleucylglycerol, propylparaben, and panaxanthine.\u003c/p\u003e\n\u003cp\u003eNext, we used the MetaboAnalyst 6.0 platform combined with GSEA methodology to perform a joint pathway analysis of DEGs and DMs to identify metabolic pathways co-enriched with genes and metabolites (raw data provided in Integration_DMs_DEGs_Pathways.csv). The results showed significant enrichment (P \u0026lt; 0.05) in metabolic pathways of histidine, caffeine, \u0026beta;-alanine, taurine, hypotaurine, glycerolipid, and aminoacyl-tRNA biosynthesis. Among these, histidine metabolism emerged as the most significantly enriched metabolic pathway, showing notable regulatory abnormalities in the gene-metabolite network analysis (Fig. 6C).\u003c/p\u003e\n\u003cp\u003eAs shown in Fig. 6C, the histidine metabolic pathway forms a complex metabolic network involving multiple key metabolites and enzymes. L-histidine can be metabolized through three main pathways: (1) entering the histidine catabolic pathway (blue box in the lower right) via catalysis by HAL (histidine ammonia-lyase), ultimately producing L-glutamate; (2) conversion to histamine via HDC (histidine decarboxylase), which subsequently enters the methylation process (blue box in the middle) under the action of HNMT (histamine N-methyltransferase), generating methylimidazole acetic acid; (3) participation in the carnosine metabolic cycle (circle in the upper right), forming anserine through catalysis by CARNS1 and CARNMT1. Additionally, this network connects to aspartate metabolism (left side) through enzymes such as ASPA and ALDH2. In OA patients, all three main metabolic pathways exhibit significant abnormalities, particularly the methylation process of histamine (indicated by the red arrow) showing an upregulation trend, which may lead to inflammation and bone metabolic disorders.\u003c/p\u003e\n\u003cp\u003eWe also examined the regulatory pathways associated with DEGs and metabolites using combined analysis (Fig. 6D). Based on analysis results from MSigDB\u0026apos;s C2 and C7 collections, the modulation of OA bone metabolism involves signaling pathways such as the TNF signaling pathway, the MAPK signaling pathway, the legionellosis signaling pathway, and the IL-17 signaling pathway (P \u0026lt; 0.05). In particular, we deeply explored the key genes in the IL-17 signaling pathway and found that the previously identified hub genes were key nodes in this pathway. For example, genes such as JUND, CCL20, FOSL1, IL1B, JUN, FOS, CXCR2, IL6, JUNB, CXCL8, and CCR2 were significantly differentially expressed (Fig. 6E). Through analysis of MSigDB\u0026apos;s C3 collection, we further discovered highly overlapping binding sites for STAT3 and NF-\u0026kappa;B transcription factors in these pathways, indicating their synergistic role in transcription-metabolism co-regulation. Our multi-omics integrated analysis revealed abnormalities in the transcription-metabolism co-regulatory network in OA, particularly the coordinated dysregulation of histidine metabolism and IL-17 signaling pathways may constitute a key link in the pathogenic mechanism of OA. The close connection between histidine metabolism pathways and inflammatory responses provides potential new targets for precision treatment of OA.\u003c/p\u003e\n\u003ch2\u003eValidation of Key Genes in Independent Datasets and Clinical Samples\u003c/h2\u003e\n\u003cp\u003eTo enhance the reliability of our findings, we employed a comprehensive cross-validation strategy using both public datasets and independent clinical samples. The mRNA expression levels of DEGs linked to bone metabolism in OA were validated by analyzing two independent RNA-seq datasets from the Gene Expression Omnibus (GEO) database: GSE48556 (106 OA samples and 33 normal controls) and GSE114007 (20 OA samples and 18 normal controls). In the GSE48556 dataset, we observed significant differential expression of IL1B, JUND, FOS, IL6, and JUNB between OA and normal samples (Fig. 7A). The validation in the GSE114007 dataset showed consistent downregulation of JUND, FOS, IL6, and JUNB in OA samples, though IL1B did not reach statistical significance in this cohort (Fig. 7B).\u003c/p\u003e\n\u003cp\u003eTo further validate these findings in clinical specimens, we performed RT-qPCR analysis on peripheral blood samples from an independent cohort of OA patients and controls (n = 10 per group). This analysis confirmed significantly reduced expression of all five genes in the OA group compared to the normal control group. The most pronounced differences were observed for IL1B, JUND, and JUNB (P \u0026lt; 0.001), while FOS (P \u0026lt; 0.05) and IL6 (P \u0026lt; 0.01) also showed significant downregulation (Fig. 7C).\u003c/p\u003e\n\u003cp\u003eThese consistent results across multiple independent datasets and clinical samples provide robust validation of our multi-omics analysis. They confirm that the dysregulation of inflammatory genes plays a critical role in OA pathogenesis. The convergent evidence from computational analysis, public datasets, and clinical specimens strengthens our conclusion that these inflammation-related genes are central to regulating bone metabolism homeostasis in OA and influencing disease progression. Moreover, the differential expression patterns observed across these validation approaches align with our findings from the histidine metabolism and IL-17 signaling pathway analyses, supporting the proposed transcription-metabolism co-regulatory network as a key mechanism in OA development.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOA is a common degenerative joint disorder characterized by cartilage degradation and osteophyte formation. These pathological changes lead to significant functional impairment and disability, reducing patients' quality of life and adversely affecting long-term prognosis\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. Despite extensive research, the exact pathomechanism of OA remains incompletely understood. Current evidence indicates that typical histopathological features of OA include cartilage matrix degradation, abnormal subchondral bone remodeling, and synovial inflammation\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e,\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. Several theories attempt to explain these pathological changes, including mechanical stress theory, inflammatory cytokine theory, and metabolic dysregulation hypothesis\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e,\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. As the disease progresses, articular cartilage is gradually eroded, subchondral bone undergoes sclerosis and cystic changes, and osteophytes form, ultimately resulting in joint pain, stiffness, and limited mobility\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Existing research suggests that OA results from the combined effects of genetic factors, lifestyle, and environmental influences, with certain OA phenotypes potentially attributable to specific metabolic and transcriptional alterations\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e,\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eBone metabolic imbalance plays a central role in OA pathogenesis. Under physiological conditions, osteoblasts synthesize and secrete bone matrix, while osteoclasts mediate bone resorption and degradation, maintaining a dynamic equilibrium within the local microenvironment. However, factors such as aging and altered estrogen levels can disrupt this balance, leading to abnormal bone metabolism and accelerated disease progression. Therefore, this study employed an integrated multi-omics approach to elucidate the metabolic-transcriptional regulatory networks involved in OA bone homeostasis dysregulation and explore potential biomarkers.\u003c/p\u003e \u003cp\u003eIn this study, we collected 60 peripheral blood samples (30 from OA patients and 30 from controls) and subsequently performed untargeted metabolomic analysis and whole transcriptome sequencing, identifying 127 DMs and 694 DEGs. Concurrently, we observed significantly elevated bone turnover markers (N-MID, PINP, and β-CTX) in OA patients, confirming the importance of bone metabolic imbalance in OA. Through metabolomic analysis, we found that proinflammatory metabolites (such as PGG2) were significantly elevated in OA patients, while protective metabolites (such as estriol and 3-indolepropionic acid) were markedly reduced. These findings align with previous reports\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. PGG2, as an intermediate in the prostaglandin synthesis pathway, not only participates in inflammatory responses but also directly affects cartilage degradation by regulating matrix metalloproteinase expression\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. The decreased estriol levels may reflect weakened bone protective mechanisms, consistent with previous clinical observations regarding estrogen replacement therapy delaying OA progression in women\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. These metabolite alterations are not isolated phenomena but represent coordinated dysregulation across multiple metabolic pathways, potentially forming the metabolic foundation of OA pathogenesis.\u003c/p\u003e \u003cp\u003eIn pathway enrichment analysis, we identified multiple metabolic and signaling pathways associated with OA. Metabolomic analysis revealed that upregulated metabolites were primarily enriched in retinol metabolism, ubiquinone biosynthesis, and arachidonic acid metabolism pathways, while downregulated metabolites were mainly associated with histidine metabolism, steroid hormone biosynthesis, and bile acid production. Transcriptomic analysis revealed that DEGs were primarily enriched in DNA repair pathways, cell junctions, and actin cytoskeleton-related processes. Notably, the IL-17 signaling pathway was significantly dysregulated in OA, consistent with previous studies\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. Through PPI network analysis, we identified core genes including IL1B, JUND, FOS, IL6, and JUNB, which have been implicated in OA pathogenesis in multiple studies\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. The activation of the IL-17 signaling pathway involves not only the release of inflammatory mediators but may also directly regulate bone metabolism by influencing the differentiation balance between osteoblasts and osteoclasts\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. In particular, core genes JUND and FOS, as transcription factors, can simultaneously regulate multiple downstream targets, forming complex regulatory networks. These genes are not simply effector molecules but key nodes integrating multiple signal inputs, potentially playing coordinating roles in the pathological cascade of OA.\u003c/p\u003e \u003cp\u003eFurthermore, by integrating metabolomic and transcriptomic data, we constructed a transcription-metabolism co-regulatory network and discovered that the interaction between histidine metabolism and IL-17 signaling pathways might be a key driver of bone metabolic imbalance in OA. Specifically, the histidine metabolism pathway participates in OA pathogenesis through three main routes: (1) entering the histidine degradation pathway via HAL catalysis, ultimately producing L-glutamate; (2) conversion to histamine via HDC, subsequently entering methylation processes under HNMT action; and (3) participation in the carnosine metabolism cycle. In OA patients, all three major metabolic pathways exhibited significant abnormalities, particularly an upward trend in histamine methylation, potentially leading to inflammation and bone metabolic disorders.\u003c/p\u003e \u003cp\u003eOur study has several limitations. First, our use of peripheral blood rather than joint tissues may not accurately reflect local OA pathological processes. Muhammad et al.\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e demonstrated significant molecular differences between systemic circulation and joint tissues in OA, highlighting the need for integrated multi-omics approaches across biological compartments. Second, our small sample size may have limited detection of subtle metabolic and transcriptional changes. Third, our correlation-based analysis lacks functional validation, preventing causal inference, with potential confounding factors possibly influencing results. Future research should include tissue-specific comparisons between matched blood and joint samples, larger longitudinal studies, functional validation experiments, and exploration of therapeutic approaches targeting histidine metabolism and IL-17 signaling pathways.\u003c/p\u003e \u003cp\u003eIn conclusion, our exploratory multi-omics study provides preliminary evidence of metabolic and transcriptional signatures associated with bone metabolic imbalance in OA, identifying potential interactions between histidine metabolism and IL-17 signaling pathways. While these findings require further validation in larger studies and functional experiments, they offer new perspectives on the molecular mechanisms of bone metabolism abnormalities in OA, potentially contributing to the development of novel diagnostic markers and therapeutic strategies.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn this study, we explored the metabolic-transcriptional regulatory networks involved in bone metabolic imbalance in OA through an integrated multi-omics approach. We analyzed peripheral blood samples from OA patients and identified differentially expressed metabolites and genes, observing significantly elevated bone turnover markers (N-MID, PINP, and β-CTX) in OA patients, which validated the importance of bone metabolic imbalance in OA pathogenesis. Our findings revealed potential interactions between histidine metabolism and IL-17 signaling pathways in OA, with abnormal histidine metabolism potentially playing a crucial role in OA pathomechanisms. The study identified core genes including IL1B, JUND, FOS, IL6, and JUNB, which may serve coordinating functions in the pathological cascade of OA. Our results suggest that increased proinflammatory metabolites coupled with decreased protective metabolites constitute the metabolic signature of OA, potentially providing molecular targets for early diagnosis and treatment of this condition. Nevertheless, larger-scale studies and functional validation experiments are still needed to further elucidate the clinical significance of these findings and their exact roles in OA pathophysiology.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eAUC\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Area Under the Curve\u003c/p\u003e\n\u003cp\u003eβ-CTX \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Beta-CrossLaps of Type I Collagen\u003c/p\u003e\n\u003cp\u003eceRNA\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Competing Endogenous RNA\u003c/p\u003e\n\u003cp\u003eDEGs\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Differentially Expressed Genes\u003c/p\u003e\n\u003cp\u003eDMs\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Differential Metabolites\u003c/p\u003e\n\u003cp\u003eELISA\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Enzyme-Linked Immunosorbent Assay\u003c/p\u003e\n\u003cp\u003eFPKM\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Fragments Per Kilobase per Million Mapped Reads\u003c/p\u003e\n\u003cp\u003eGO\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Gene Ontology\u003c/p\u003e\n\u003cp\u003eGSEA\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Gene Set Enrichment Analysis\u003c/p\u003e\n\u003cp\u003eCON\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Controls\u003c/p\u003e\n\u003cp\u003eIL-17\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Interleukin-17\u003c/p\u003e\n\u003cp\u003eKEGG\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Kyoto Encyclopedia of Genes and Genomes\u003c/p\u003e\n\u003cp\u003eLASSO\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Least Absolute Shrinkage and Selection Operator\u003c/p\u003e\n\u003cp\u003eMAPK\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Mitogen-Activated Protein Kinase\u003c/p\u003e\n\u003cp\u003eN-MID\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Mid-Regional Fragment of Osteocalcin\u003c/p\u003e\n\u003cp\u003eOA\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Osteoarthritis\u003c/p\u003e\n\u003cp\u003ePBMCs\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Peripheral Blood Mononuclear Cells\u003c/p\u003e\n\u003cp\u003ePINP\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Procollagen I N-Terminal Propeptide\u003c/p\u003e\n\u003cp\u003ePPI\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Protein-Protein Interaction\u003c/p\u003e\n\u003cp\u003eQC\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Quality Control\u003c/p\u003e\n\u003cp\u003eRIN\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;RNA Integrity Number\u003c/p\u003e\n\u003cp\u003eROC\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Receiver Operating Characteristic\u003c/p\u003e\n\u003cp\u003eTNF\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Tumor Necrosis Factor\u003c/p\u003e\n\u003cp\u003eUHPLC-MS\u0026nbsp; \u0026nbsp;\u0026nbsp;Ultrahigh-Performance Liquid Chromatography-Tandem Mass Spectrometry\u003c/p\u003e\n\u003cp\u003eVIP \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Variable Importance in Projection\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe extend our gratitude to all study participants for their invaluable contributions to this research.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYL, BW, and\u0026nbsp;WC\u0026nbsp;designed the study. YL, JG, SW, QS, and FZ conducted patient recruitment and data collection. YL and JL performed experiments. YL, JL, and SW analyzed data. YL wrote the manuscript. BW and\u0026nbsp;WC\u0026nbsp;supervised the project and acquired funding. All authors critically reviewed and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets analyzed during this study are available in the NCBI GEO repository under accession numbers GSE48556 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE48556) and GSE114007 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE114007). The datasets generated during this study, including integrated gene-metabolite pathway analysis results, are provided in Additional file 3: Integration_DMs_DEGs_Pathways.csv.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication \u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll participants provided written informed consent for the publication of anonymized data.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research received no specific grant from funding agencies in the public, commercial, or not-for-profit sectors.Declarations\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Chongqing Hospital of Traditional Chinese Medicine (Approval No. 2022-KY-YJS-LY). Written informed consent was obtained from all participants or their legal guardians.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor details\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e1\u003c/sup\u003eCollege of Traditional Chinese Medicine, Chongqing Medical University, Chongqing 400016, China. \u003csup\u003e2\u003c/sup\u003eDepartment of Rheumatology, Chongqing Hospital of Traditional Chinese Medicine, Chongqing 400021, China. \u003csup\u003e3\u003c/sup\u003eDepartment of Nephrology and Rheumatology, Chongqing Jiangjin District Hospital of Traditional Chinese Medicine,Chongqing 402260, China. \u003csup\u003e4\u003c/sup\u003eChongqing Key Laboratory of Traditional Chinese Medicine to Prevent and Treat Autoimmune Diseases, Chongqing, 400021, China.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBijlsma, J. W. J., Berenbaum, F. \u0026amp; Lafeber, F. P. J. G. Osteoarthritis: an update with relevance for clinical practice. \u003cem\u003eLancet\u003c/em\u003e \u003cstrong\u003e377\u003c/strong\u003e, 2115-2126, doi:10.1016/s0140-6736(11)60243-2 (2011).\u003c/li\u003e\n\u003cli\u003eSteinmetz, J. D.\u003cem\u003e et al.\u003c/em\u003e Global, regional, and national burden of osteoarthritis, 1990-2020 and projections to 2050: a systematic analysis for the Global Burden of Disease Study 2021. \u003cem\u003eLancet Rheumatology\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, e508-e522, doi:10.1016/s2665-9913(23)00163-7 (2023).\u003c/li\u003e\n\u003cli\u003eLong, H.\u003cem\u003e et al.\u003c/em\u003e Prevalence Trends of Site-Specific Osteoarthritis From 1990 to 2019: Findings From the Global Burden of Disease Study 2019. \u003cem\u003eArthritis \u0026amp; Rheumatology\u003c/em\u003e \u003cstrong\u003e74\u003c/strong\u003e, 1172-1183, doi:10.1002/art.42089 (2022).\u003c/li\u003e\n\u003cli\u003eHe, Y.\u003cem\u003e et al.\u003c/em\u003e Pathogenesis of Osteoarthritis: Risk Factors, Regulatory Pathways in Chondrocytes, and Experimental Models. \u003cem\u003eBiology-Basel\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, doi:10.3390/biology9080194 (2020).\u003c/li\u003e\n\u003cli\u003eMacfarlane, E., Seibel, M. J. \u0026amp; Zhou, H. Arthritis and the role of endogenous glucocorticoids. \u003cem\u003eBone research\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 33-33, doi:10.1038/s41413-020-00112-2 (2020).\u003c/li\u003e\n\u003cli\u003eLoeser, R. F., Goldring, S. R., Scanzello, C. R. \u0026amp; Goldring, M. B. Osteoarthritis: A disease of the joint as an organ. \u003cem\u003eArthritis and Rheumatism\u003c/em\u003e \u003cstrong\u003e64\u003c/strong\u003e, 1697-1707, doi:10.1002/art.34453 (2012).\u003c/li\u003e\n\u003cli\u003eWatkins, L.\u003cem\u003e et al.\u003c/em\u003e Assessment of quantitative \u0026lt;SUP\u0026gt;18\u0026lt;/SUP\u0026gt;F Sodium fluoride PET measures of knee subchondral bone perfusion and mineralization in osteoarthritic and healthy subjects. \u003cem\u003eOsteoarthritis and Cartilage\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 849-858, doi:10.1016/j.joca.2021.02.563 (2021).\u003c/li\u003e\n\u003cli\u003eHwang, J. S., Lee, H. S. \u0026amp; Gong, H. S. Three-Dimensional Analysis of the Trapezium Subchondral Bone and its Association with Trapeziometacarpal Joint Osteoarthritis. \u003cem\u003eCalcified Tissue International\u003c/em\u003e \u003cstrong\u003e112\u003c/strong\u003e, 320-327, doi:10.1007/s00223-022-01040-5 (2023).\u003c/li\u003e\n\u003cli\u003eMuratovic, D., Atkins, G. J. \u0026amp; Findlay, D. M. Is RANKL a potential molecular target in osteoarthritis? \u003cem\u003eOsteoarthritis and Cartilage\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 493-500, doi:10.1016/j.joca.2023.10.010 (2024).\u003c/li\u003e\n\u003cli\u003eSahu, S., Naik, S., Bandyopadhyay, D. \u0026amp; Tripathy, S. Serum levels of osteoprotegerin, RANK-L \u0026amp; vitamin D in different stages of osteoarthritis of the knee. \u003cem\u003eIndian Journal of Medical Research\u003c/em\u003e \u003cstrong\u003e154\u003c/strong\u003e, 491-496, doi:10.4103/ijmr.IJMR_873_19 (2021).\u003c/li\u003e\n\u003cli\u003eZhou, J. \u0026amp; Zhong, L. Applications of liquid chromatography-mass spectrometry based metabolomics in predictive and personalized medicine. \u003cem\u003eFrontiers in Molecular Biosciences\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, doi:10.3389/fmolb.2022.1049016 (2022).\u003c/li\u003e\n\u003cli\u003eVan Pevenage, P. M., Birchmier, J. T. \u0026amp; June, R. K. Utilizing metabolomics to identify potential biomarkers and perturbed metabolic pathways in osteoarthritis: A systematic review. \u003cem\u003eSeminars in Arthritis and Rheumatism\u003c/em\u003e \u003cstrong\u003e59\u003c/strong\u003e, doi:10.1016/j.semarthrit.2023.152163 (2023).\u003c/li\u003e\n\u003cli\u003eLiao, Z.\u003cem\u003e et al.\u003c/em\u003e Differential Metabolites in Osteoarthritis: A Systematic Review and Meta-Analysis. \u003cem\u003eNutrients\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, doi:10.3390/nu15194191 (2023).\u003c/li\u003e\n\u003cli\u003eWang, M.\u003cem\u003e et al.\u003c/em\u003e Molecular crosstalk between articular cartilage, meniscus, synovium, and subchondral bone in osteoarthritis. \u003cem\u003eBone \u0026amp; Joint Research\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 862-872, doi:10.1302/2046-3758.1112.Bjr-2022-0215.R1 (2022).\u003c/li\u003e\n\u003cli\u003eTchetina, E. V.\u003cem\u003e et al.\u003c/em\u003e Metabolic Dysregulation and Its Role in Postoperative Pain among Knee Osteoarthritis Patients. \u003cem\u003eInternational Journal of Molecular Sciences\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, doi:10.3390/ijms25073857 (2024).\u003c/li\u003e\n\u003cli\u003eMaruotti, N., Corrado, A. \u0026amp; Cantatore, F. P. Osteoblast role in osteoarthritis pathogenesis. \u003cem\u003eJournal of Cellular Physiology\u003c/em\u003e \u003cstrong\u003e232\u003c/strong\u003e, 2957-2963, doi:10.1002/jcp.25969 (2017).\u003c/li\u003e\n\u003cli\u003eGe, M.\u003cem\u003e et al.\u003c/em\u003e Multi-omics analysis of synovial tissue and fluid reveals differentially expressed proteins and metabolites in osteoarthritis. \u003cem\u003eJournal of Translational Medicine\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, doi:10.1186/s12967-025-06310-y (2025).\u003c/li\u003e\n\u003cli\u003eHochberg, M. C.\u003cem\u003e et al.\u003c/em\u003e Guidelines for the medical management of osteoarthritis. Part I. Osteoarthritis of the hip. American College of Rheumatology. \u003cem\u003eArthritis and rheumatism\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, 1535-1540, doi:10.1002/art.1780381103 (1995).\u003c/li\u003e\n\u003cli\u003ePoplawski, A. \u0026amp; Binder, H. Feasibility of sample size calculation for RNA-seq studies. \u003cem\u003eBriefings in Bioinformatics\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 713-720, doi:10.1093/bib/bbw144 (2018).\u003c/li\u003e\n\u003cli\u003eSurowiec, I.\u003cem\u003e et al.\u003c/em\u003e Multivariate strategy for the sample selection and integration of multi-batch data in metabolomics. \u003cem\u003eMetabolomics\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, doi:10.1007/s11306-017-1248-1 (2017).\u003c/li\u003e\n\u003cli\u003eVinaixa, M.\u003cem\u003e et al.\u003c/em\u003e A Guideline to Univariate Statistical Analysis for LC/MS-Based Untargeted Metabolomics-Derived Data. \u003cem\u003eMetabolites\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 775-795, doi:10.3390/metabo2040775 (2012).\u003c/li\u003e\n\u003cli\u003eDi Guida, R.\u003cem\u003e et al.\u003c/em\u003e Non-targeted UHPLC-MS metabolomic data processing methods: a comparative investigation of normalisation, missing value imputation, transformation and scaling. \u003cem\u003eMetabolomics\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, doi:10.1007/s11306-016-1030-9 (2016).\u003c/li\u003e\n\u003cli\u003eBro, R. \u0026amp; Smilde, A. K. Principal component analysis. \u003cem\u003eAnalytical Methods\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 2812-2831, doi:10.1039/c3ay41907j (2014).\u003c/li\u003e\n\u003cli\u003eTrygg, J. \u0026amp; Wold, S. Orthogonal projections to latent structures (O-PLS). \u003cem\u003eJournal of Chemometrics\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 119-128, doi:10.1002/cem.695 (2002).\u003c/li\u003e\n\u003cli\u003ePang, Z.\u003cem\u003e et al.\u003c/em\u003e MetaboAnalyst 6.0: towards a unified platform for metabolomics data processing, analysis and interpretation. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, W398-W406, doi:10.1093/nar/gkae253 (2024).\u003c/li\u003e\n\u003cli\u003eRanstam, J. \u0026amp; Cook, J. A. LASSO regression. \u003cem\u003eBritish Journal of Surgery\u003c/em\u003e \u003cstrong\u003e105\u003c/strong\u003e, 1348-1348, doi:10.1002/bjs.10895 (2018).\u003c/li\u003e\n\u003cli\u003eLeek, J. T., Johnson, W. E., Parker, H. S., Jaffe, A. E. \u0026amp; Storey, J. D. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 882-883, doi:10.1093/bioinformatics/bts034 (2012).\u003c/li\u003e\n\u003cli\u003ePertea, M.\u003cem\u003e et al.\u003c/em\u003e StringTie enables improved reconstruction of a transcriptome from RNA-seq reads. \u003cem\u003eNature Biotechnology\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 290-+, doi:10.1038/nbt.3122 (2015).\u003c/li\u003e\n\u003cli\u003eLangmead, B., Trapnell, C., Pop, M. \u0026amp; Salzberg, S. L. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. \u003cem\u003eGenome Biology\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, doi:10.1186/gb-2009-10-3-r25 (2009).\u003c/li\u003e\n\u003cli\u003eWen, M., Shen, Y., Shi, S. \u0026amp; Tang, T. miREvo: an integrative microRNA evolutionary analysis platform for next-generation sequencing experiments. \u003cem\u003eBmc Bioinformatics\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, doi:10.1186/1471-2105-13-140 (2012).\u003c/li\u003e\n\u003cli\u003eFriedlaender, M. R., Mackowiak, S. D., Li, N., Chen, W. \u0026amp; Rajewsky, N. miRDeep2 accurately identifies known and hundreds of novel microRNA genes in seven animal clades. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e40\u003c/strong\u003e, 37-52, doi:10.1093/nar/gkr688 (2012).\u003c/li\u003e\n\u003cli\u003eLove, M. I., Huber, W. \u0026amp; Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. \u003cem\u003eGenome Biology\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, doi:10.1186/s13059-014-0550-8 (2014).\u003c/li\u003e\n\u003cli\u003eWu, T.\u003cem\u003e et al.\u003c/em\u003e clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. \u003cem\u003eInnovation (New York, N.Y.)\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 100141-100141, doi:10.1016/j.xinn.2021.100141 (2021).\u003c/li\u003e\n\u003cli\u003eLiberzon, A.\u003cem\u003e et al.\u003c/em\u003e The Molecular Signatures Database Hallmark Gene Set Collection. \u003cem\u003eCell Systems\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, 417-425, doi:10.1016/j.cels.2015.12.004 (2015).\u003c/li\u003e\n\u003cli\u003eSzklarczyk, D.\u003cem\u003e et al.\u003c/em\u003e STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, D607-D613, doi:10.1093/nar/gky1131 (2019).\u003c/li\u003e\n\u003cli\u003eShannon, P.\u003cem\u003e et al.\u003c/em\u003e Cytoscape: A software environment for integrated models of biomolecular interaction networks. \u003cem\u003eGenome Research\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 2498-2504, doi:10.1101/gr.1239303 (2003).\u003c/li\u003e\n\u003cli\u003eWarde-Farley, D.\u003cem\u003e et al.\u003c/em\u003e The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, W214-W220, doi:10.1093/nar/gkq537 (2010).\u003c/li\u003e\n\u003cli\u003eSalmena, L., Poliseno, L., Tay, Y., Kats, L. \u0026amp; Pandolfi, P. P. A \u0026lt;i\u0026gt;ceRNA\u0026lt;/i\u0026gt; Hypothesis: The Rosetta Stone of a Hidden RNA Language? \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e146\u003c/strong\u003e, 353-358, doi:10.1016/j.cell.2011.07.014 (2011).\u003c/li\u003e\n\u003cli\u003evan der Kraan, P. M. \u0026amp; van den Berg, W. B. Osteophytes: relevance and biology. \u003cem\u003eOsteoarthritis and Cartilage\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 237-244, doi:10.1016/j.joca.2006.11.006 (2007).\u003c/li\u003e\n\u003cli\u003eMobasheri, A.\u003cem\u003e et al.\u003c/em\u003e The role of metabolism in the pathogenesis of osteoarthritis. \u003cem\u003eNature Reviews Rheumatology\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 302-311, doi:10.1038/nrrheum.2017.50 (2017).\u003c/li\u003e\n\u003cli\u003eKatsoula, G., Kreitmaier, P. \u0026amp; Zeggini, E. Insights into the molecular landscape of osteoarthritis in human tissues. \u003cem\u003eCurrent Opinion in Rheumatology\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 79-90, doi:10.1097/bor.0000000000000853 (2022).\u003c/li\u003e\n\u003cli\u003eKumm, J., Tamm, A., Lintrop, M. \u0026amp; Tamm, A. Diagnostic and prognostic value of bone biomarkers in progressive knee osteoarthritis: a 6-year follow-up study in middle-aged subjects. \u003cem\u003eOsteoarthritis and Cartilage\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 815-822, doi:10.1016/j.joca.2013.03.008 (2013).\u003c/li\u003e\n\u003cli\u003eMartel-Pelletier, J., Pelletier, J. P. \u0026amp; Fahmi, H. Cyclooxygenase-2 and prostaglandins in articular tissues. \u003cem\u003eSeminars in Arthritis and Rheumatism\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 155-167, doi:10.1016/s0049-0172(03)00134-3 (2003).\u003c/li\u003e\n\u003cli\u003eGoggs, R., Carter, S. D., Schulze-Tanzil, G., Shakibaei, M. \u0026amp; Mobasheri, A. Apoptosis and the loss of chondrocyte survival signals contribute to articular cartilage degradation in osteoarthritis. \u003cem\u003eVeterinary Journal\u003c/em\u003e \u003cstrong\u003e166\u003c/strong\u003e, 140-158, doi:10.1016/s1090-0233(02)00331-3 (2003).\u003c/li\u003e\n\u003cli\u003ePang, H.\u003cem\u003e et al.\u003c/em\u003e Low back pain and osteoarthritis pain: a perspective of estrogen. \u003cem\u003eBone Research\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, doi:10.1038/s41413-023-00280-x (2023).\u003c/li\u003e\n\u003cli\u003eGilmer, G.\u003cem\u003e et al.\u003c/em\u003e Uncovering the \u0026quot;riddle of femininity\u0026quot; in osteoarthritis: a systematic review and meta-analysis of menopausal animal models and mathematical modeling of estrogen treatment. \u003cem\u003eOsteoarthritis and Cartilage\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 447-457, doi:10.1016/j.joca.2022.12.009 (2023).\u003c/li\u003e\n\u003cli\u003eAziz, A., Nathan, K. G., Kamarul, T., Mobasheri, A. \u0026amp; Sharifi, A. The interplay between dysregulated metabolites and signaling pathway alterations involved in osteoarthritis: a systematic review. \u003cem\u003eTherapeutic Advances in Musculoskeletal Disease\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, doi:10.1177/1759720x241299535 (2024).\u003c/li\u003e\n\u003cli\u003eZeng, Z.\u003cem\u003e et al.\u003c/em\u003e Pelargonidin ameliorates inflammatory response and cartilage degeneration in osteoarthritis via suppressing the NF-\u0026kappa;B pathway. \u003cem\u003eArchives of Biochemistry and Biophysics\u003c/em\u003e \u003cstrong\u003e743\u003c/strong\u003e, doi:10.1016/j.abb.2023.109668 (2023).\u003c/li\u003e\n\u003cli\u003eLu, L. \u0026amp; Tian, L. Postmenopausal osteoporosis coexisting with sarcopenia: the role and mechanisms of estrogen. \u003cem\u003eJournal of Endocrinology\u003c/em\u003e \u003cstrong\u003e259\u003c/strong\u003e, doi:10.1530/joe-23-0116 (2023).\u003c/li\u003e\n\u003cli\u003eMimpen, J. Y.\u003cem\u003e et al.\u003c/em\u003e Interleukin-17A Causes Osteoarthritis-Like Transcriptional Changes in Human Osteoarthritis-Derived Chondrocytes and Synovial Fibroblasts \u0026lt;i\u0026gt;In Vitro\u0026lt;/i\u0026gt;. \u003cem\u003eFrontiers in Immunology\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, doi:10.3389/fimmu.2021.676173 (2021).\u003c/li\u003e\n\u003cli\u003eShi, S.\u003cem\u003e et al.\u003c/em\u003e Identification of key regulators responsible for dysregulated networks in osteoarthritis by large-scale expression analysis. \u003cem\u003eJournal of Orthopaedic Surgery and Research\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, doi:10.1186/s13018-021-02402-9 (2021).\u003c/li\u003e\n\u003cli\u003eWang, Z.\u003cem\u003e et al.\u003c/em\u003e RANKL expression of primary osteoblasts is enhanced by an IL-17-mediated JAK2/STAT3 pathway through autophagy suppression. \u003cem\u003eConnective Tissue Research\u003c/em\u003e \u003cstrong\u003e62\u003c/strong\u003e, 411-426, doi:10.1080/03008207.2020.1759562 (2021).\u003c/li\u003e\n\u003cli\u003eRai, M. F.\u003cem\u003e et al.\u003c/em\u003e Three decades of advancements in osteoarthritis research: insights from transcriptomic, proteomic, and metabolomic studies. \u003cem\u003eOsteoarthritis and Cartilage\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 385-397, doi:10.1016/j.joca.2023.11.019 (2024).\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":"Osteoarthritis, Bone homeostasis, Metabolomics, Transcriptomics, Regulatory networks","lastPublishedDoi":"10.21203/rs.3.rs-6359791/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6359791/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eOsteoarthritis (OA) represents a prevalent articular condition characterized by significant disturbances in bone metabolic balance, yet the molecular pathways governing these skeletal alterations remain inadequately defined.This study conducted metabolomic and transcriptomic analyses of 30 OA patients and 30 controls, revealing significant alterations in purine, lipid, and amino acid metabolism in OA patients. Retinol metabolism and ubiquinone biosynthesis showed increased activity, while histidine metabolism decreased. Key metabolites correlated with bone turnover markers (β-CTX, PINP, N-MID). Transcriptomic analysis identified differentially expressed mRNAs (DEGs) involved in apoptosis, inflammation, and lipid metabolism. Integrated analysis revealed candidate regulators connecting IL-17, MAPK, histidine, and fatty acid signaling pathways, providing potential intervention targets for OA bone metabolic abnormalities.\u003c/p\u003e","manuscriptTitle":"Integrated Metabolomic and Transcriptomic Analysis Identifies Candidate Regulatory Networks in Bone Metabolic Dysregulation of Osteoarthritis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-13 01:25:03","doi":"10.21203/rs.3.rs-6359791/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":"bb34974f-a9f4-45f0-9a3f-bedb11d013c0","owner":[],"postedDate":"May 13th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":48279214,"name":"Health sciences/Rheumatology/Osteoimmunology"},{"id":48279215,"name":"Health sciences/Biomarkers"},{"id":48279216,"name":"Health sciences/Pathogenesis"},{"id":48279217,"name":"Health sciences/Rheumatology"}],"tags":[],"updatedAt":"2025-08-20T06:08:37+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-13 01:25:03","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6359791","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6359791","identity":"rs-6359791","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}