目的 基于超高效液相色谱-四极杆-静电场轨道阱组合质谱联用系统（UHPLC-Q Exactive）、网络药理 学、转录组学及分子对接技术探讨刺梨治疗动脉粥样硬化（AS）的潜在作用机制。方法 将 12 只雄性 ApoE-/-小 鼠随机分为模型组（6 只）、刺梨原液组（6 只，500 mg·kg-1 ），给予高脂饲料喂养 10 周诱导 AS 小鼠模型；造模 成功后，按照相应剂量灌胃给药，每日 1 次，持续给药 12 周。采用 UHPLC-Q Exactive 技术分析刺梨原液的化 学成分；采用 UHPLC-Q Exactive HF-X 技术分析模型组与刺梨原液组小鼠血清的化学成分，通过 OPLS-DA 模 型得到的变量权重值（VIP）及 t 检验的 P 值来确定显著差异代谢物；将筛选出的显著差异代谢物与刺梨原液的 化学成分逐一对比，筛选出刺梨入血成分。使用 Swiss Target Prediction 数据库预测刺梨原液入血成分的作用靶 点；通过 GeneCards、OMIM、DisGeNET 数据库检索 AS 疾病相关靶点；对刺梨原液入血成分作用靶点与 AS 疾 病相关靶点取交集，筛选出刺梨原液治疗 AS 的潜在作用靶点。利用 STRING 数据库构建潜在作用靶点蛋白互 作（PPI）网络，筛选出核心靶点；通过 Cytoscape 3.10.0 软件构建“成分-靶点-疾病”网络，筛选出核心成分； 通过 DAVID 数据库对潜在作用靶点进行 GO 功能及 KEGG 通路富集分析。对小鼠主动脉组织总 RNA 进行转录 组学测序，筛选差异表达基因（DEGs）；使用 Venny 2.1 平台对 DEGs 与刺梨原液治疗 AS 的潜在作用靶点取交 集，对交集靶点进行 PPI 网络分析，得到刺梨干预 AS 的关键调控靶点。将关键调控靶点分别与刺梨干预 AS 的核心成分进行分子对接验证。结果 共鉴定出刺梨化学成分 468 种，模型组 vs 刺梨原液组血清差异代谢 物 423 种，筛选出刺梨原液入血成分 65 种。得到入血成分对应的主要作用靶点 137 个，AS 疾病相关靶点 5 000 个，刺梨原液-AS 交集靶点 96 个。对 137 个刺梨入血成分作用靶点与 5 000 个 AS 疾病相关靶点取交 集，共获得刺梨治疗 AS 的潜在作用靶点 96 个。筛选出刺梨干预 AS 的核心靶点包括 AKT1、STAT3、EGFR、 PTGS2、HSP90AA1、TLR4、MMP9、ESR1、ACE、MAPK1 等；筛选出核心成分包括苯丙氨酸-脯氨酸二肽 （Phe-Pro）、酪氨酸-谷氨酸二肽（Tyr-Glu）、11-（2-羟基-3，4-二甲基-5-氧代呋喃-2-基）十一烷酸[11-（2- hydroxy-3，4-dimethyl-5-oxofuran-2-yl） undecanoic acid]、γ-谷氨酰-酪氨酸二肽（gamma-Glu-Tyr）和 2-丙基 戊二酸（2-Propylglutaric acid）等。潜在作用靶点主要涉及炎症反应调控、脂质代谢过程调节及细胞增殖调控等 生物过程，以及神经活性配体-受体相互作用、癌症、雌激素、化学致癌作用受体和钙离子等 KEGG 信号通 路。共鉴定出模型组 vs 刺梨原液组的 DEGs 共有 1 385 个，与 96 个潜在作用靶点取交集，得到 13 个交集靶 点，通过 PPI 分析最终得到 7 个刺梨干预 AS 的关键调控靶点：MMP9、CASP1、CCR1、CCR5、PTPN1、 MMP13、ANPEP。分子对接显示，苯丙氨酸-脯氨酸二肽、11-（2-羟基-3，4-二甲基-5-氧代呋喃-2-基）十一 烷酸分别与 MMP9、CASP1、CCR1、CCR5、MMP13、ANPEP 结合稳定。结论 刺梨可能通过苯丙氨酸-脯氨 酸二肽、11-（2-羟基-3，4-二甲基-5-氧代呋喃-2-基）十一烷酸等核心入血成分，作用于 MMP9、MMP13、 CASP1、ANPEP 等关键靶点，调控细胞外基质降解通路、NLRP3 炎症小体和血管活性肽代谢等通路，发挥治 疗 AS 的作用。

Objective To investigate the potential mechanism of Rosa roxburghii Tratt（. RRT） in treating atherosclerosis （AS） using ultra-high performance liquid chromatography-quadrupole-electrostatic field orbitrap mass spectrometry （UHPLC-Q Exactive），network pharmacology，transcriptomics and molecular docking. Methods Twelve male ApoE-/- mice were randomly divided into model group （n=6） and RRT crude extract group （n=6，500 mg·kg-1 ），fed with highfat diet for 10 weeks to establish AS model. After successful modeling，intragastric administration was performed once daily for 12 weeks. UHPLC-Q Exactive was used to analyze chemical components of RRT crude extract. UHPLC-Q Exactive HF-X identified serum metabolites in both groups，with significant differential metabolites screened via OPLSDA-derived variable importance in projection （VIP） values and t-test P-values. Blood-absorbed RRT components were identified by comparing these metabolites with RRT’s chemical profile. Swiss Target Prediction predicted targets of absorbed components， while GeneCards/OMIM/DisGeNET provided AS-related targets. The intersection between the target proteins of blood-absorbed RRT components and AS-related disease targets was analyzed to identify potential therapeutic targets for AS treatment. STRING database constructed protein-protein interaction （PPI） networks to identify core targets. Cytoscape 3.10.0 built “component-target-disease” networks to screen core components. Potential therapeutic targets were subjected to Gene Ontology（GO） functional annotation and Kyoto Encyclopedia of Genes and Genomes （KEGG）pathway enrichment analysis using the DAVID database. Total RNA from mouse aortic tissues was sequenced for transcriptomic analysis to identify differentially expressed genes （DEGs）. The intersection between DEGs and potential therapeutic targets of RRT crude extract for AS was analyzed using Venny 2.1 platform. PPI network analysis of the overlapping targets revealed key regulatory targets for RRT intervention in AS. Molecular docking validation was subsequently performed between these key targets and core components of RRT. Results A total of 468 RRT chemical components and 423 serum differential metabolites （model group versus RRT crude extract group） were identified，65 blood-absorbed RRT components were screened. These components corresponded to 137 targets， which intersected with 5 000 AS-related targets to yield 96 potential therapeutic targets. Core targets included AKT1， STAT3， EGFR， PTGS2， HSP90AA1， TLR4， MMP9， ESR1， ACE and MAPK1 were screened out. Key components comprised Phe-Pro dipeptide， Tyr-Glu dipeptide， 11-（2-hydroxy-3， 4-dimethyl-5-oxofuran-2-yl） undecanoic acid， gamma-Glu-Tyr dipeptide and 2-propylglutaric acid. Potential targets primarily involved inflammatory response regulation， lipid metabolism modulation， cell proliferation control， and KEGG signaling pathways like neuroactive ligand-receptor interaction， cancer， estrogen signaling， chemical carcinogenesis and calcium. There were 1 385 DEGs （model group versus RRT crude extract group），intersecting with 96 potential targets to yield 13 overlapping targets. PPI analysis revealed 7 key regulatory targets： MMP9， CASP1， CCR1， CCR5， PTPN1，MMP13 and ANPEP. Molecular docking demonstrated stable binding between Phe-Pro/11- （2-hydroxy-3，4- dimethyl-5-oxofuran-2-yl） undecanoic acid and MMP9，CASP1，CCR1，CCR5，MMP13，ANPEP. Conclusion RRT may exert anti-AS effects through core blood-absorbed components （e.g.，Phe-Pro dipeptide and 11-（2-hydroxy- 3，4-dimethyl-5-oxofuran-2-yl） undecanoic acid） acting on key targets （MMP9，MMP13，CASP1，ANPEP） to regulate extracellular matrix degradation，NLRP3 inflammasome and vasoactive peptide metabolism pathways.

R285.5

贵州省科技计划项目（黔科合基础-ZK[2024]一般354）；中药民族药贵州省科技创新领军人才工作站（黔科合平台KXJZ[2024]034）