目的 建立鸡骨草Abrus cantoniensis和毛鸡骨草A.mollis的HPLC指纹图谱，比较不同批次鸡骨草和毛鸡骨草样品的化学成分差异，并对维采宁-2和夏佛塔苷进行定量测定，结合网络药理学和体外实验验证，初步探讨鸡骨草和毛鸡骨草的同用合理性及治疗肝癌的潜在作用机制。方法 采用HPLC方法，Ultimate^Ⓡ AQ-C₁₈色谱柱（250 mm×4.6 mm，5 μm），流动相为甲醇-水溶液，梯度洗脱，体积流量1.0 mL·min^-1，检测波长270 nm，柱温30℃。建立12批鸡骨草和6批毛鸡骨草的指纹图谱并进行相似度评价和特征峰匹配。结合聚类分析（CA）、主成分分析（PCA）及正交偏最小二乘法-判别分析（OPLS-DA）确定鸡骨草和毛鸡骨草的差异性特征成分，并对其中2种成分进行定量测定。通过中药系统药理学数据库与分析平台（TCMSP）筛选鸡骨草活性成分及其潜在靶点，并利用GeneCards数据库筛选肝癌相关靶点，结合韦恩图分析获取共有靶点。基于共有靶点，进行基因本体（GO）和京都基因与基因组百科全书（KEGG）分析，揭示鸡骨草治疗肝癌的关键通路。使用STRING 11.5数据库构建靶点蛋白质与蛋白质相互作用（PPI）网络，并构建药物-成分-疾病-靶点-通路网络，确定主要活性成分。采用MTT法确定鸡骨草醇提物（ACEE）对HepG2细胞的最佳用药剂量，并应用实时荧光定量PCR（qRTPCR）实验检测排名前5的mRNA表达量。结果 建立了12批鸡骨草和6批毛鸡骨草样品的指纹图谱，标定10个共有峰，共指认其中5个主要特征峰：峰1为相思子碱、峰2为刺桐碱、峰5为维采宁-2、峰8为夏佛塔苷、峰9为异夏佛塔苷。通过化学模式识别筛选得到峰3、峰5（维采宁-2）、峰6、峰7、峰8（夏佛塔苷）、峰9（异夏佛塔苷）所代表的成分是区分不同批次样品的差异性标志物，其中维采宁-2和夏佛塔苷的质量分数分别为0.05～4.36 mg·g^-1、0.10～4.34 mg·g^-1，不同批次间差异较大。网络药理学结合指纹图谱指认共确定9个活性成分，265个潜在靶点，与肝癌共有靶点130个。KEGG富集分析共得到112条信号通路，主要涉及癌症通路、脂质与动脉粥样硬化通路等。根据活性成分-靶点-通路网络筛选出5个肝癌关键靶点。根据MTT结果，最终选择质量浓度0.25 mg·mL-1的ACEE进行qRT-PCR实验，与对照组比较，关键靶点AKT1、PIK3CA、STAT3、BCL2、GSK3B的mRNA表达水平显著下降（P＜0.001）。结论 建立的指纹图谱及含量测定方法简便可行，网络药理学筛选出的关键靶点经体外实验证明与鸡骨草抗肝癌的作用机制密切相关，为鸡骨草质量的控制和药效机制的研究提供参考。

Objective To establish high-performance liquid chromatography (HPLC) Fingerprints for Abrus cantoniensis and A. mollis, comparison of chemical composition differences among different batches of A. cantoniensis and A. mollis samples, quantitative determination of vicenin-2 and schaftoside, combined with network pharmacology and in vitro experiments to preliminary explore the rationality of co-use of A. cantoniensis and A. mollis and the potential mechanism of anti-hepatocellular carcinoma.Methods The HPLC method was used with an Ultimate^Ⓡ AQ-C₁₈ column (250 mm×4.6 mm, 5 μm), a mobile phase of methanol-water solution, gradient elution; flow rate of 1.0 mL·min^-1; detection at 270 nm, 30 ℃. Fingerprints for 12 batches of A. cantoniensis and six batches of A. mollis were established and evaluated for similarity and characteristic peak matching. Combined with Cluster Analysis (CA), principal component analysis (PCA), and orthogonal partial least squares-discriminant analysis (OPLS-DA), the differential characteristic components of A. cantoniensis and A. mollis were determined, and two of these components were quantitatively measured. Active components of A. cantoniensis and their potential targets were screened through the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP), and liver cancer-related targets were screened using the GeneCards database, with common targets obtained through venn diagram analysis. Based on common targets, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were conducted to reveal key pathways of A. cantoniensis in treating liver cancer. The STRING 11.5 database was used to construct a protein-protein interaction (PPI) network of target proteins, and an active component-target-pathway network was built to identify main active components. The MTT method was used to determine the optimal dosage of A. cantoniensis ethanol extract (ACEE) for HepG2 cells, and real-time fluorescence quantitative PCR (qRT-PCR) was used to detect the expression levels of the top five mRNA.Results Fingerprints for 12 batches of A. cantoniensis and 6 batches of A. mollis samples were established, with ten common peaks calibrated and five main characteristic peaks identified: peak 1 as abrine, peak 2 as hypaphorine, peak 5 as vicenin-2, peak 8 as schaftoside, and peak 9 as isoschaftoside. Chemical pattern recognition identified components represented by peaks 3, 5 (vicenin-2), 6, 7, 8 (schaftoside), and 9 (isoschaftoside) as differential markers distinguishing different batch samples, with mass fractions of vicenin-2 and schaftoside ranging from 0.05—4.36 mg·g^-1 and 0.10—4.34 mg·g^-1, respectively, showing significant differences among batches. Network pharmacology combined with fingerprint identification confirmed nine active components and 265 potential targets, with 130 common targets related to liver cancer. KEGG yielded 112 signaling pathways, mainly involving cancer pathways, lipid and atherosclerosis pathways, etc. Five key liver cancer targets were screened based on the active component-target-pathway network. According to MTT results, 0.25 mg·mL^-1 ACEE was selected for qRT-PCR, compared with the control group, the mRNA expression levels of key targets AKT1, PIK3CA, STAT3, BCL2 and GSK3B were significantly decreased (P＜0.001).Conclusion The established HPLC fingerprint and content determination method are simple and feasible. The key targets screened by network pharmacology have been verified by in vitro experiments to be closely related to the effect of A. cantoniensis in the treatment of liver cancer, providing a reference for the quality control and the study of its pharmacological mechanisms.

R286.2

国家自然科学基金委区域创新发展联合基金重点支持项目（U23A20521）；广西科技重大专项研究项目（桂科AA22096029）