目的 通过网络药理学构建三七合剂的活性成分-肝缺血再灌注损伤（hepatic ischemia-reperfusion injury，HIRI）靶点-生物通路的网络关系，探讨三七合剂干预HIRI的关键靶点及作用机制。方法 通过国内外文献调研和中药系统药理学分析平台（TCMSP）及Pharm Mapper、Swiss Target Prediction等服务器，以口服利用度（OB）和类药性（DL）作为限定条件筛选并收集三七合剂干预HIRI的相关靶点；利用OMIM数据库筛选并整理HIRI相关的基因和蛋白靶点。利用Excel表合并、整理成分与疾病靶点的交集，通过Cytoscape3.7.2软件的插件Network Analyzer分析，以拓扑参数度（degree）≥5（平均自由度为4.5）为筛选条件找到核心靶点，并将交集靶点导入STRING服务器，并以置信度得分（Confidence Score）≥0.85为筛选条件构建核心蛋白互作（Hub-PPI）网络；将交集靶点导入FunRich 3.0软件进行生物学过程和生物学通路分析，利用Cytoscape3.7.2构建中药-活性成分-HIRI靶点-生物学通路网络。结果 三七合剂能降低HIRI小鼠天门冬氨酸转氨酶（AST）及谷氨酸转氨酶（ALT）的表达（P<0.01）；三七合剂经筛选得到45个活性成分，对应靶点3 273个，主要化合物包括熊果酸、齐墩果酸、马钱子苷、槲皮素、人参皂苷F₂、芍药苷等；整理HIRI得到196个靶点，其与成分的交集靶点为46个，主要包括11-β-羟类固醇脱氢酶（HSD11B1）、腺苷受体A3（ADORA3）、环氧化酶2（PTGS2）、腺苷受体A1（ADORA1）、蛋白激酶C-ε（PKCε）等；经STRING服务器设定Confidence Score≥0.85的限定条件，得到高置信度PPI网络并经cluster处理将其聚为3类；经FunRich软件分析得到蛋白质代谢作用、信号传导、酶活性的负调控、炎症反应、跨膜受体蛋白酪氨酸激酶信号通路共5条生物学过程（P<0.05）；整合素连接激酶信号、TNF受体信号通路、p38丝裂原活化蛋白激酶信号通路、TRAIL信号通路等16条生物通路（P<0.01）。结论 初步探讨了三七合剂干预HIRI是通过多成分和多靶点的相互作用，及调控多条生物通路及生物学过程共同完成的，但关键的核心靶标和具体的调控机制尚待进一步的实验验证。

Objective Through network pharmacology, the network relationship between the active component of Sanqi Mixture, the target of hepatic ischemia-reperfusion injury(HIRI), and biological pathway was constructed to explore the key target and mechanism of effect of Sanqi Mixture on HIRI. Method Through literature research at home and abroad, Traditional Chinese Medicine Systems Pharmacology (TCMSP) platform, Pharm Mapper, Swiss Target Prediction and other servers, oral availability (OB) and drug-likeness (DL) were selected as the limited conditions to collect the relevant targets for Sanqi Mixture for intervention in HIRI. The OMIM database was used to screen and collate HIRI related genes and protein targets. Excel table was used to merge and sort the intersection between disease and targets through Cytoscape3.7.2 software plug-ins Network Analyzer, with topological parameters (degree) ≥ 5 (average degrees of freedom 4.5) for the filter to find the core targets; And the intersection targets were imported to the server STRING, and with Confidence Score of 0.85 or higher for the filter conditions to build the core protein interactions (Hub-PPI) network. The intersection target was introduced into FunRich 3.0 software for biological process and biological pathway analysis, and Cytoscape3.7.2 was used to construct the network of "traditional Chinese medicine-active ingredient-HIRI target-biological pathway". Result Sanqi mixture could reduce the expression of Aspartate aminotransferase (AST) and glutamate transaminase (ALT) in HIRI mice (P < 0.01). After screening, 45 active components of Sanqi Mixture were obtained, corresponding to 3 273 targets, and the main compounds included ursolic acid, oleanolic acid, brucine, quercetin, ginsenoside F₂, paeoniflorin, etc. Among the 196 targets obtained by HIRI, 46 targets were intersected with components, including 11-β-hydroxysteroid dehydrogenase (HSD11B1), adenosine receptor A3 (ADORA3), cyclooxygenase 2 (PTGS2), adenosine receptor A1 (ADORA1), protein kinase C-ε (PKC), etc. With the STRING server setting the qualified condition of Confidence Score ≥ 0.85, the PPI network with high Confidence was obtained and clustered into three categories through cluster processing. Five biological processes including protein metabolism, signal transduction, negative regulation of enzyme activity, inflammatory response and transmembrane receptor protein tyrosine kinase signal pathway were analyzed by FunRich software (P < 0.05). 16 biological pathways including integrin-linked kinase signal, TNF receptor signaling pathway, P38 mitogen-activated protein kinase signaling pathway, and TRAIL signaling pathway (P < 0.01). Conclusion It is preliminarily discussed that Sanqi Mixture intervenes HIRI through the interaction of multiple components and multiple targets, as well as the regulation of multiple biological pathways and biological processes. However, the key core targets and the specific regulation mechanism still need further experimental verification.

R285.5

天津市第一中心医院春风课题（2019CF15）