激素性股骨头坏死衰老关键基因的生物信息学鉴定和验证  

Bioinformatics identification and validation of aging key genes in hormonal osteonecrosis of the femoral head

作　　者：丘博元 刘飞 童思文 欧志学[2] 王伟伟 Qiu Boyuan;Liu Fei;Tong Siwen;Ou Zhixue;Wang Weiwei(Graduate School of Guangxi University of Chinese Medicine,Nanning 530000,Guangxi Zhuang Autonomous Region,China;Department of Joint and Sports Medicine,Guilin Traditional Chinese Medicine Hospital,Guilin 541000,Guangxi Zhuang Autonomous Region,China)

机构地区：[1]广西中医药大学研究生院,广西壮族自治区南宁市530000 [2]桂林市中医医院关节与运动医学科,广西壮族自治区桂林市541000

出　　处：《中国组织工程研究》2025年第26期5608-5620,共13页Chinese Journal of Tissue Engineering Research

基　　金：广西中医药大学研究生教育创新计划项目(YCSY2023067),项目负责人:丘博元;广西自然科学基金项目(2023GXNSFAA026412),项目负责人:欧志学;广西壮族自治区中医药管理局自筹经费科研课题(GXZYZ20210372),项目负责人:欧志学。

摘　　要：背景:激素性股骨头坏死与衰老紧密相连,但调控靶点和机制尚不明确。通过生物信息学联合机器学习分析并加以实验验证,确定细胞衰老介导激素性股骨头坏死发生发展的关键基因,将为激素性股骨头坏死的防治提供新思路。目的:使用生物信息学分析筛选激素性股骨头坏死的衰老核心基因并进行实验验证,以探讨其作用机制。方法:从GEO数据库的GPL15207平台获取了GSE123568数据集,其中包含了30例激素性股骨头坏死患者与10名健康对照的外周血清样本基因表达谱;从CellAge数据库获取了279个细胞衰老相关基因的数据。对激素性股骨头坏死基因谱进行差异分析及加权基因共表达网络(WGCNA)分析,两者与衰老相关基因取交集再取并集得到激素性股骨头坏死衰老潜在基因,并进行GO和KEGG富集分析。机器学习方法筛出枢纽基因,构建Nomogram模型,进行共识聚类及免疫浸润分析。最后收集临床股骨头样本通过qPCR及Westernblot检测方法进行验证。结果与结论:①共获得41个潜在基因,主要富集在衰老及氧化应激反应等生物过程,以及FoxO及肿瘤坏死因子信号通路中。②机器学习鉴定后得到枢纽基因过氧化氢酶、结缔组织生长因子、叉头框蛋白O3、胰岛素受体底物2和丝裂原活化蛋白激酶激酶11,Nomogram模型预测能力良好。③共识聚类分析将患者分为a,b和c共3组,过氧化氢酶、叉头框蛋白O3、胰岛素受体底物2和丝裂原活化蛋白激酶激酶11在3个分子亚型之间表达存在差异(P<0.05),免疫浸润结果显示活化CD4+T细胞、活化CD8+T细胞、嗜酸性粒细胞等免疫细胞的丰度在3个分子亚类中存在差异(P<0.05)。④qPCR和Western blot结果显示,激素性股骨头坏死组过氧化氢酶、结缔组织生长因子、叉头框蛋白O3和丝裂原活化蛋白激酶激酶11比对照组表达降低(P<0.05),胰岛素受体底物2的表达升高(P<0.05)。⑤上述结果BACKGROUND:Hormonal osteonecrosis of the femoral head is strongly associated with aging,but the regulatory targets and mechanisms are still unclear.Through bioinformatics combined with machine learning analysis and experimental verification,the key genes of hormonal osteonecrosis of the femoral head mediated by cell senescence will be identified,which will provide new ideas for the prevention and treatment of hormonal osteonecrosis of the femoral head.OBJECTIVE:To screen and validate the senescence core genes of hormonal osteonecrosis of the femoral head using bioinformatics analysis to explore its mechanism of action.METHODS:The GSE123568 dataset was obtained from the GPL15207 platform of the GEO database,which contained the gene expression profiles of peripheral serum samples of 30 hormonal osteonecrosis of the femoral head patients and 10 healthy controls.Data on 279 cellular senescence-related genes were obtained from the CellAge database.Differential analysis and weighted correlation network analysis(WGCNA)were performed on hormonal osteonecrosis of the femoral head gene profiles,and both were intersected with senescence-related genes and then concatenated to obtain hormonal osteonecrosis of the femoral head senescence potential genes,and GO and KEGG analyses were performed.The machine learning method screened out the pivotal genes,constructed nomogram model,and performed consensus clustering and immune infiltration analysis.Finally,clinical femoral samples were collected for validation by qPCR and western blot assay.RESULTS AND CONCLUSION:(1)41 potential genes were obtained,which were mainly enriched in biological processes such as aging and oxidative stress response,as well as FoxO and tumor necrosis factor signaling pathways.(2)The pivotal genes catalase,connective tissue growth factor,forkhead box protein O3,insulin receptor substrate 2,and mitogen-activated protein kinase kinase 11 were obtained after machine learning identification,and the predictive ability of nomogram model was good.(3)The patients w

关 键 词：激素性股骨头坏死 衰老 WGCNA分析 机器学习 免疫浸润分析 实验验证 

分 类 号：R459.9[医药卫生—治疗学] R364[医药卫生—临床医学] R681.8