基于多组学分析探讨陈皮改善膝骨关节炎滑膜纤维化的作用机制

Exploration of the Mechanism of Citri Reticulatae Pericarpium in Improving Synovial Fibrosis of Knee Osteoarthritis Based on Multi-Omics Analysis

摘要:

目的：探究陈皮对脂质代谢重编程的调控作用及其在改善膝骨关节炎（KOA） 滑膜纤维化增生中的潜在作用机制。方法：通过SymMap、TCMID、TCMSP、TCM-ID等多数据库检索陈皮的潜在活性成分及其作用靶点，并运用DO/GO/KEGG功能富集分析明确其在KOA治疗中的潜在作用机制。从GEO数据库下载与KOA滑膜病变相关的基因芯片数据，筛选差异表达基因，构建加权基因共表达网络（WGCNA），识别与KOA滑膜病变相关的核心基因。进一步从MSigDB数据库获取脂质代谢相关的基因，联合差异表达分析和WGCNA分析鉴定陈皮靶向调控脂质代谢重编程的作用靶点。利用Lasso回归模型构建诊断模型，并通过外部数据集进行验证以期明确其核心作用靶点。获取KOA滑膜病变相关的单细胞测序数据，确定核心基因在不同细胞亚群中的分布。最后，开展虚拟对接筛选和分子动力学模拟，从动态和静态多维验证陈皮潜在作用成分及其与核心蛋白的结合稳定性。结果：鉴定出114种陈皮活性成分，其中40个潜在靶点在骨关节炎、脂质代谢和炎症通路中富集。转录组分析鉴定得到2 060个KOA滑膜病变相关的差异表达基因，联合WGCNA分析的576个枢纽基因，最终得到四个核心基因（PTGS2、RELA、HMGCR、TNF）。利用外部验证集构建的诊断模型具有良好的准确性和敏感度，且核心基因在KOA滑膜组织中显著上调。单细胞测序显示成纤维细胞在KOA滑膜组织占比最高且核心基因异常上调。虚拟活性成分筛选和动力学模拟结果显示活性成分豆甾醇、橙皮素、β-谷甾醇与核心蛋白PTGS2、RELA、HMGCR、TNF稳定结合。结论：陈皮活性成分可经PTGS2-RELA-HMGCR-TNF轴调控滑膜成纤维细胞脂质代谢重编程，进而改善KOA滑膜纤维化。

Abstract:

Abstract： Objective： To investigate the regulatory effect of Citri Reticulatae Pericarpium on lipid metabolism reprogramming and its potential mechanism in improving synovial fibrosis in knee osteoarthritis （KOA）. Methods： Potential active components of Citri Reticulatae Pericarpium and their target genes were retrieved from multiple databases， including SymMap， TCMID， TCMSP， and TCM-ID. Functional enrichment analysis （DO/GO/KEGG） was conducted to clarify the potential mechanisms of Citri Reticulatae Pericarpium in KOA treatment. Microarray data related to KOA synovial lesions were downloaded from the GEO database， and differential expression genes （DEGs） were screened. Weighted gene co-expression network analysis （WGCNA） was performed to identify core genes associated with KOA synovial lesions. Lipid metabolism-related gene sets were obtained from the MSigDB database. By combining DEG analysis and WGCNA， target genes of Citri Reticulatae Pericarpium involved in regulating lipid metabolism reprogramming were identified. A diagnostic model was constructed using the Lasso regression model and validated with external datasets to confirm core targets. Single-cell sequencing data of KOA synovial lesions were analyzed to determine the distribution of core genes across different cell subsets. Virtual docking screening and molecular dynamics simulations were performed to verify the potential active components of Citri Reticulatae Pericarpium and their binding stability with core proteins from both dynamic and static perspectives. Results： A total of 114 active components of Citri Reticulatae Pericarpium were identified，with 40 potential targets enriched in pathways related to osteoarthritis， lipid metabolism， and inflammation. Transcriptomic analysis identified 2 060 DEGs associated with KOA synovial lesions. Combined with 576 hub genes from WGCNA， four core genes （PTGS2， RELA， HMGCR， TNF） were identified. The diagnostic model constructed with an external validation set demonstrated good accuracy and sensitivity. Core genes were significantly upregulated in KOA synovial tissues. Single-cell sequencing showed that fibroblasts constituted the highest proportion of cell types in KOA synovial tissues， with core genes abnormally upregulated. Virtual screening and molecular dynamics simulations revealed that active components， such as stigmasterol， hesperetin， and β -sitosterol， stably bound to core proteins PTGS2， RELA， HMGCR， and TNF. Conclusion：The active components of Citri Reticulatae Pericarpium can regulate lipid metabolism reprogramming in synovial fibroblasts through the PTGS2-RELA-HMGCR-TNF axis，thereby improving synovial fibrosis in KOA.