Table_1_Characterizing mitochondrial features in osteoarthritis through integrative multi-omics and machine learning analysis.docx

PurposeOsteoarthritis (OA) stands as the most prevalent joint disorder. Mitochondrial dysfunction has been linked to the pathogenesis of OA. The main goal of this study is to uncover the pivotal role of mitochondria in the mechanisms driving OA development.Materials and methodsWe acquired seven bulk RNA-seq datasets from the Gene Expression Omnibus (GEO) database and examined the expression levels of differentially expressed genes related to mitochondria in OA. We utilized single-sample gene set enrichment analysis (ssGSEA), gene set enrichment analysis (GSEA), and weighted gene co-expression network analysis (WGCNA) analyses to explore the functional mechanisms associated with these genes. Seven machine learning algorithms were utilized to identify hub mitochondria-related genes and develop a predictive model. Further analyses included pathway enrichment, immune infiltration, gene-disease relationships, and mRNA-miRNA network construction based on these hub mitochondria-related genes. genome-wide association studies (GWAS) analysis was performed using the Gene Atlas database. GSEA, gene set variation analysis (GSVA), protein pathway analysis, and WGCNA were employed to investigate relevant pathways in subtypes. The Harmonizome database was employed to analyze the expression of hub mitochondria-related genes across various human tissues. Single-cell data analysis was conducted to examine patterns of gene expression distribution and pseudo-temporal changes. Additionally, The real-time polymerase chain reaction (RT-PCR) was used to validate the expression of these hub mitochondria-related genes.ResultsIn OA, the mitochondria-related pathway was significantly activated. Nine hub mitochondria-related genes (SIRT4, DNAJC15, NFS1, FKBP8, SLC25A37, CARS2, MTHFD2, ETFDH, and PDK4) were identified. They constructed predictive models with good ability to predict OA. These genes are primarily associated with macrophages. Unsupervised consensus clustering identified two mitochondria-associated isoforms that are primarily associated with metabolism. Single-cell analysis showed that they were all expressed in single cells and varied with cell differentiation. RT-PCR showed that they were all significantly expressed in OA.ConclusionSIRT4, DNAJC15, NFS1, FKBP8, SLC25A37, CARS2, MTHFD2, ETFDH, and PDK4 are potential mitochondrial target genes for studying OA. The classification of mitochondria-associated isoforms could help to personalize treatment for OA patients.

目的：骨关节炎（OA）作为最常见的关节疾病，其发病机制与线粒体功能障碍密切相关。本研究的主要目标是揭示线粒体在驱动OA发展机制中的关键作用。材料与方法：我们从基因表达综合数据库（GEO）中获取了七个大规模RNA测序数据集，并检测了与OA相关的线粒体差异表达基因的表达水平。我们运用单样本基因集富集分析（ssGSEA）、基因集富集分析（GSEA）和加权基因共表达网络分析（WGCNA）等方法，探究与这些基因相关的功能机制。采用七个机器学习算法识别关键线粒体相关基因，并构建预测模型。进一步的分析包括通路富集、免疫浸润、基因-疾病关系以及基于这些关键线粒体相关基因的mRNA-miRNA网络构建。利用基因图谱数据库进行全基因组关联研究（GWAS）。通过GSEA、基因集变异分析（GSVA）、蛋白质通路分析和WGCNA等方法，研究了亚型中的相关通路。Harmonizome数据库用于分析关键线粒体相关基因在各种人类组织中的表达。通过单细胞数据分析，考察基因表达分布模式和伪时间变化的规律。此外，实时聚合酶链反应（RT-PCR）用于验证这些关键线粒体相关基因的表达。结果：在OA中，线粒体相关通路显著激活。鉴定出九个关键线粒体相关基因（SIRT4、DNAJC15、NFS1、FKBP8、SLC25A37、CARS2、MTHFD2、ETFDH和PDK4），它们构建的预测模型具有良好的预测OA的能力。这些基因主要与巨噬细胞相关。无监督一致性聚类识别出两种与代谢相关的线粒体关联异构体。单细胞分析显示，它们均表达在单个细胞中，并随细胞分化而变化。RT-PCR显示，它们在OA中均显著表达。结论：SIRT4、DNAJC15、NFS1、FKBP8、SLC25A37、CARS2、MTHFD2、ETFDH和PDK4是研究OA的潜在线粒体靶基因。线粒体关联异构体的分类有助于个性化治疗OA患者。