中国组织工程研究 ›› 2026, Vol. 30 ›› Issue (36): 9413-9422.doi: 10.12307/2026.910

• 骨组织构建 bone tissue construction • 上一篇    下一篇

多种机器学习鉴定抗菌肽作为骨关节炎钠死亡关键治疗靶点:细胞学验证

王槐旌1,郭锦荣2,万东平1,梅其杰2,袁景钊1,徐文飞2,曾  超2,郑海军2,袁长深2,段  戡2   

  1. 1广西中医药大学研究生院,广西壮族自治区南宁市  530200;2广西中医药大学第一附属医院,广西壮族自治区南宁市  530023
  • 收稿日期:2025-11-13 修回日期:2026-03-25 出版日期:2026-12-28 发布日期:2026-05-20
  • 通讯作者: 袁长深,硕士,主任医师,硕士生导师,广西中医药大学第一附属医院,广西壮族自治区南宁市 530023 共同通讯作者:段戡,博士,主任医师,博士生导师,广西中医药大学第一附属医院,广西壮族自治区南宁市 530023
  • 作者简介:第一作者:王槐旌,男,1996年生,云南省昆明市人,汉族,博士,主要从事骨关节炎的基础与临床研究。 共同第一作者:郭锦荣,男,1985年生,甘肃省天水市人,汉族,硕士,副主任医师,主要从事骨关节炎的基础与临床研究。
  • 基金资助:
    国家自然科学基金资助项目(82160912),项目负责人:段戡;广西自然科学基金项目(面上项目:中医药壮瑶医药联合专项)(2023GXNSFAA026051),项目负责人:袁长深;广西中医药大学研究生教育创新计划项目(YCBZ2024152),项目负责人:王槐旌;广西中医药大学研究生教育创新计划项目(YCBZ2025192),项目负责人:万东平

Identification of antimicrobial peptides as key therapeutic targets for necrosis by sodium overload in osteoarthritis using multiple machine learning approaches: cytological validation

Wang Huaijing1, Guo Jinrong2, Wan Dongping1, Mei Qijie2, Yuan Jingzhao1, Xu Wenfei2, Zeng Chao2, Zheng Haijun2, Yuan Changshen2, #br# Duan Kan2#br#   

  1. 1Graduate School of Guangxi University of Chinese Medicine, Nanning 530200, Guangxi Zhuang Autonomous Region, China; 2The First Affiliated Hospital of Guangxi University of Chinese Medicine, Nanning 530023, Guangxi Zhuang Autonomous Region, China
  • Received:2025-11-13 Revised:2026-03-25 Online:2026-12-28 Published:2026-05-20
  • Contact: Yuang Changshen, MS, Chief physician, Master’s supervisor, The First Affiliated Hospital of Guangxi University of Chinese Medicine, Nanning 530023, Guangxi Zhuang Autonomous Region, China Co-corresponding author: Duan Kan, PhD, Chief physician, Doctoral supervisor, The First Affiliated Hospital of Guangxi University of Chinese Medicine, Nanning 530023, Guangxi Zhuang Autonomous Region, China
  • About author:Wang Huaijing, PhD, Graduate School of Guangxi University of Chinese Medicine, Nanning 530200, Guangxi Zhuang Autonomous Region, China Guo Jinrong, MS, Associate chief physician, The First Affiliated Hospital of Guangxi University of Chinese Medicine, Nanning 530023, Guangxi Zhuang Autonomous Region, China Wang Huaijing and Guo Jinrong contributed equally to this work.
  • Supported by:
    National Natural Science Foundation of China, No. 82160912 (to DK); Guangxi Natural Science Foundation Project (General Project: Traditional Chinese Medicine Zhuang Yao Medicine Joint Special Project), No. 2023GXNSFAA026051 (to YCS); Innovation Project of Guangxi Graduate Education of GXUCM, No. YCBZ2024152 (to WHJ); Innovation Project of Guangxi Graduate Education of GXUCM, No. YCB2025192 (to WDP)

摘要:



文题释义:
抗菌肽:由37个螺旋状氨基酸组成,具有广谱抗菌活性,能调节炎症,诱导免疫细胞到达受伤或感染部位,结合和中和脂多糖,促进上皮化和修复损伤。
钠死亡:以钠过载为特征,通过非选择性单价阳离子通道瞬时受体电位M通道成员4促进钠离子内流引起细胞坏死,钠超载可加速软骨细胞破坏,加剧炎症反应,而抑制钠过载可有效抑制炎症反应,延缓疾病进展。

背景:钠死亡在心血管疾病中发挥重要作用,但在骨关节炎中的作用机制尚不明确。
目的:通过生物信息学筛选骨关节炎钠死亡关键基因,探究关键基因在骨关节炎中的作用机制。
方法:检索GEO数据库获取骨关节炎芯片数据GSE117999和GSE169077,将差异表达基因与钠死亡相关基因取交集获取骨关节炎钠死亡差异基因,对骨关节炎钠死亡差异基因进行GO和KEGG富集分析,并进行蛋白互作网络分析获得Hub基因。对骨关节炎钠死亡差异基因进行免疫细胞浸润分析及加权基因共表达网络分析,进一步筛选骨关节炎钠死亡差异表达基因及免疫相关差异表达基因。基于最小绝对收缩和选择算子、极限梯度提升算法与随机森林方法筛选骨关节炎钠死亡关键基因构建诊断性模型。Pearson相关性分析探究骨关节炎钠死亡关键基因与诊断性基因的相关性,然后预测在骨关节炎发病过程中与诊断性基因相关的药物。最后利用细胞实验进行验证。
结果与结论:①筛选获得9个骨关节炎钠死亡差异基因,主要富集于细胞丝氨酸水解酶活性、丝氨酸型肽酶活性、丝氨酸型内肽酶活性等生物学过程,以及白细胞介素17信号通路;②免疫浸润分析发现浆细胞、静态树突状细胞差异表达,可推测这2种免疫细胞在骨关节炎发病中起一定作用;③加权基因共表达网络分析获得3个基因:乳铁蛋白、抗菌肽、S100钙结合蛋白A8,结合免疫浸润分析获得1个基因:α-S1-酪蛋白;④蛋白互作分析获得5个Hub基因:基质金属蛋白酶3、载脂蛋白D、抗菌肽、S100钙结合蛋白A8、乳铁蛋白;⑤机器学习筛选出1个诊断性基因:抗菌肽,且与Hub基因具有相关性;⑥最后预测得到可调控诊断基因的药物共计5种:聚乙二醇缀合Toll样受体7/8激动剂NKTR-262、罗哌卡肽、17-丁酸氯倍他索、表达干扰素β和酪氨酸酶相关蛋白1的重组水疱性口炎病毒、吉洛拉单抗;⑦细胞实验结果显示:与空白组相比,经白细胞介素1β处理的软骨细胞中抗菌肽蛋白表达量显著升高。经生物信息学分析与实验验证,抗菌肽基因对骨关节炎具有诊断价值,钠死亡相关基因在骨关节炎发病中起一定作用。
https://orcid.org/0009-0004-8593-9638(王槐旌)


中国组织工程研究杂志出版内容重点:干细胞;骨髓干细胞;造血干细胞;脂肪干细胞;肿瘤干细胞;胚胎干细胞;脐带脐血干细胞;干细胞诱导;干细胞分化;组织工程

关键词: 骨关节炎, 钠死亡, CAMP, 生物信息学, 机器学习, 诊断模型, 免疫浸润, 细胞实验

Abstract: BACKGROUND: Necrosis by sodium overload (NECSO) plays a significant role in cardiovascular diseases, but its mechanism in osteoarthritis remains unclear.
OBJECTIVE: To screen key genes related to NECSO in osteoarthritis through bioinformatics and to explore their mechanisms in osteoarthritis.
METHODS: Osteoarthritis microarray datasets GSE117999 and GSE169077 were obtained from the GEO database. Differentially expressed genes were intersected with NECSO-related genes to identify osteoarthritis-NECSO differentially expressed genes. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses and protein-protein interaction network analysis were performed to obtain Hub genes. Immune cell infiltration analysis and weighted gene co-expression network analysis were conducted to further screen NECSO-related and immune-related differentially expressed genes in osteoarthritis. Key genes were screened using least absolute shrinkage and selection operator, extreme gradient boosting, and random forest methods to construct and validate a diagnostic model. The correlation between key NECSO genes and diagnostic genes was analyzed using the Pearson’s correlation analysis, and potential drugs targeting diagnostic genes were predicted. Finally, cell experiments were conducted for validation.
RESULTS AND CONCLUSION: (1) Nine osteoarthritis-NECSO differentially expressed genes were identified, mainly enriched in biological processes such as cellular serine hydrolase activity, serine-type peptidase activity, and serine-type endopeptidase activity, as well as the interleukin-17 signaling pathway. (2) Immune infiltration analysis revealed differential expression of plasma cells and resting dendritic cells, suggesting these two immune cells play certain roles in osteoarthritis pathogenesis. (3) Weighted gene co-expression network analysis identified three genes: lactoferrin, cathelicidin antimicrobial peptide, and S100 calcium-binding protein A8. Combined with immune infiltration analysis, one gene, alpha-S1-casein, was identified. (4) Protein-protein interaction analysis identified five Hub genes: matrix metalloproteinase 3, apolipoprotein D, cathelicidin antimicrobial peptide, S100 calcium-binding protein A8, and lactoferrin. (5) Machine learning screened out one diagnostic gene, cathelicidin antimicrobial peptide, which was correlated with Hub genes. (6) Five drugs targeting the diagnostic gene were predicted: PEG-conjugated Toll-like receptor 7/8 agonist NKTR-262, ropocamptide, clobetasol 17-butyrate, recombinant vesicular stomatitis virus expressing interferon-beta and tyrosinase-related protein 1, and giloralimab. (7) Cell experiments showed significantly higher expression of cathelicidin antimicrobial peptide protein in chondrocytes treated with interleukin-1β compared with the blank group. These findings suggest that the cathelicidin antimicrobial peptide gene has diagnostic value for osteoarthritis, and NECSO-related genes play a role in its pathogenesis.

Key words: osteoarthritis, necrosis by sodium overload, CAMP, bioinformatics, machine learning, diagnostic model, immune infiltration, cell experiment

中图分类号: