[关键词]
[摘要]
目的 通过网络药理学、机器学习和分子对接技术探讨丹参治疗急性脑梗死的作用机制。方法 通过中药系统药理数据库与分析平台(TCMSP)与Swiss Target Prediction数据库,预测并筛选丹参的主要活性成分及其作用靶点;同时,通过OMIM、GeneCards等相关数据库获取脑梗死相关靶点,经R软件韦恩图分析确定交集靶点;借助STRING构建蛋白相互作用(PPI)网络,通过Cytoscape构建“丹参-活性成分-靶点-通路”网络,初筛核心靶点;整合GEO数据集GSE58294,经limma包差异分析和WGCNA共表达网络分析筛选候选靶点,通过R包进行基因本体(GO)和京都基因与基因组百科全书(KEGG)富集分析;运用随机森林、支持向量机、逻辑回归3种机器学习模型及SHAP分析、接受者操作特征(ROC)曲线筛选并验证核心基因;最后经AutoDock Tools进行分子对接,验证核心成分与靶点的结合稳定性。结果 对丹参的610个活性成分相关靶点与脑梗死相关的11 520个靶点进行交集分析,共识别出610个交集靶点,核心成分包括丹参酮ⅡA等。GO富集分析共得到747个显著条目,BP主要涉及肌肉系统过程、肌肉收缩、平滑肌收缩及其正调控,以及对肽激素的反应等;CC显著富集于顶端质膜、神经元投射末端、内质网腔、膜微区及膜筏等结构;MF主要包括蛋白酪氨酸激酶活性、组蛋白H2AXY142与H3Y41激酶活性,以及氧化还原酶活性等。表明目标基因在肌肉功能、细胞信号转导及表观遗传调控中发挥重要作用。KEGG富集分析鉴定出24条相关通路,涉及色氨酸代谢、破骨细胞分化、核因子-κB(NF-κB)信号通路、造血细胞谱系及急性髓系白血病等。交集靶点分别为B淋巴细胞瘤-2相关蛋白A1(BCL2A1)、钙调蛋白依赖性蛋白激酶1(CAMK1)、组织蛋白酶K(CTSK)、鞘氨醇1磷酸酯受体3(S1PR3)、丝裂原激活蛋白激酶激酶6(MAP2K6)、类固醇硫酸酯酶(STS)、钾电压阀门通道1(KCNE1)。分子对接证实丹参酮ⅡA与核心靶点结合稳定。结论 丹参可通过丹参酮ⅡA等核心成分靶向调控CTSK、S1PR3和BCL2A1等关键靶点,干预NF-κB信号通路、色氨酸代谢等核心通路,发挥治疗急性脑梗死的作用。
[Key word]
[Abstract]
Objective To explore the mechanism of Salvia miltiorrhiza in treatment of acute ischemic stroke based on network pharmacology, machine learning and molecular docking technology. Methods The main active components and their targets of S. miltiorrhiza were predicted and screened by TCMSP and Swiss Target Prediction database. At the same time, the relevant targets of cerebral infarction were obtained through OMIM, Gene Cards and other related databases, and the intersection targets were determined by Wayne diagram analysis of R software. The protein-protein interaction (PPI) network was constructed by STRING, and the “S. miltiorrhiza-active ingredient-target-pathway” network was constructed by Cytoscape to screen the core targets. The GEO dataset GSE58294 was integrated, and candidate targets were screened by limma package difference analysis and WGCNA co-expression network analysis. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were performed by R package. Three machine learning models of random forest, support vector machine and logistic regression, SHAP analysis and receiver operating characteristic (ROC) curve were used to screen and verify the core genes. Finally, molecular docking was performed by AutoDock Tools to verify the binding stability of the core components and the target. Results A total of 610 intersection targets were identified by intersection analysis of 610 active components related targets and 11 520 targets related to ischemic stroke, and the core components included tanshinone ⅡA. A total of 747 significant items were obtained by GO enrichment analysis. BP mainly involved muscle system processes, muscle contraction, smooth muscle contraction and its positive regulation, and response to peptide hormones. CC was significantly enriched in the apical plasma membrane, neuronal projection end, endoplasmic reticulum cavity, membrane microdomain and membrane raft. MF mainly included protein tyrosine kinase activity, histone H2AXY142 and H3Y41 kinase activity, and oxidoreductase activity. It showed that the target gene plays an important role in muscle function, cell signal transduction and epigenetic regulation. KEGG enrichment analysis identified 24 related pathways, involving tryptophan metabolism, osteoclast differentiation, nuclear factor-κB (NF-κB) signaling pathway, hematopoietic cell lineage and acute myeloid leukemia. The intersection targets were B-cell lymphoma-2-associated protein A1 (BCL2A1), calmodulin-dependent protein kinase 1 (CAMK1), cathepsin K (CTSK), sphingosine 1 phosphate receptor 3 (S1PR3), mitogen-activated protein kinase kinase 6 (MAP2K6), steroid sulfatase (STS), potassium voltage valve channel 1 (KCNE1). Molecular docking confirmed that tanshinone ⅡA binded stably to the core target. Conclusion S. miltiorrhiza can target and regulate key targets such as CTSK, S1PR3 and BCL2A1 through core components such as tanshinone ⅡA, interfere with core pathways such as NF-κB signaling pathway and tryptophan metabolism, and play a role in the treatment of acute ischemic stroke.
[中图分类号]
R285;R286.1
[基金项目]
国家自然科学基金资助项目(82260259)