目的 建立经典名方薏苡附子败酱散（Yiyi Fuzi Baijiang Powder，YFBP）基准样品的HPLC指纹图谱，对其指标性成分（原儿茶酸、绿原酸、咖啡酸、异绿原酸A、异绿原酸C、苯甲酰新乌头原碱、亚油酸）进行定量测定，研究YFBP的量值传递规律，并结合化学模式识别评价其质量。方法 通过建立15批YFBP基准样品的HPLC指纹图谱，明确其特征峰归属及出膏率范围，并对指标性成分进行定量测定。同时采用层次聚类分析（hierarchical cluster analysis，HCA）、主成分分析（principal component analysis，PCA）与正交偏最小二乘法-判别分析（orthogonal partial least squares-discriminant analysis，OPLS-DA），寻找影响YFBP质量的主要化学成分。结果 15批YFBP基准样品HPLC指纹图谱相似度＞0.817；共归属30个特征峰，并指认出8个特征峰信息，其中30号峰（亚油酸）为薏苡仁专属峰；2（去甲猪毛菜碱）、9、22（苯甲酰新乌头原碱）号峰为附子专属峰；3（原儿茶酸）、4～6、7（绿原酸）、8、10（咖啡酸）、11、13、18、19、20（异绿原酸A）、21（异绿原酸C）、23～26号峰为败酱草专属峰；1、12、14～17、27～29号峰为薏苡仁和败酱草共有特征峰；15批基准样品出膏率为20.06%～33.22%；指标性成分原儿茶酸、绿原酸、咖啡酸、异绿原酸A、异绿原酸C、苯甲酰新乌头原碱、亚油酸的质量分数分别为0.012～0.076、0.025～0.308、0.011～0.047、0.011～0.049、0.014～0.044、0.031～0.111、0.031～0.067 mg/g，转移率分别为5.05%～30.86%、5.24%～21.31%、3.57%～43.22%、4.59%～18.69%、6.97%～23.13%、9.68%～62.32%、6.60%～19.95%；化学模式识别分析将15批样品分为2类，OPLS-DA筛选出了导致质量差异的14个特征峰，其中包括已被识别的峰3（原儿茶酸）、10（咖啡酸）、20（异绿原酸A）。结论 结合化学模式识别，建立了经典名方YFBP基准样品的HPLC指纹图谱及多指标性成分定量测定方法，并对其量值传递过程进行分析，为YFBP基准样品的质量控制及复方制剂开发提供了科学参考依据。

Objective The HPLC fingerprints of the benchmark samples of the classic prescription Yiyi Fuzi Baijiang Powder (薏苡附子败酱散, YFBP) was established, and the contents of its index components (protocatechuic acid, chlorogenic acid, caffeic acid, isochlorogenic acid A, isochlorogenic acid C, benzoylmesaconine and linoleic acid) were determined to study the quantitative value transfer pattern of YFBP and to evaluate its quality by combining with chemical pattern recognition to assess its quality. Methods By HPLC fingerprinting of 15 batches of YFBP benchmark samples, the common peaks and extraction yield range were identified, and the marker components were quantitatively determined. Simultaneously, hierarchical cluster analysis (HCA), principal component analysis (PCA), and orthogonal partial least squares-discriminant analysis (OPLS-DA) were employed to identify the key chemical components influencing the quality of YFBP. Results In 15 batches of YFBP benchmark samples, the HPLC fingerprint similarity was greater than 0.817. A total of 30 common peaks were identified, and eight common peaks were recognised, with specific attributions as follows: Peak 30 (linoleic acid) was specific to Yiyiren (Coicis Semen). Peaks 2 (salsolinol), 9, and 22 (benzoylmesaconine) were specific to Fuzi (Aconiti Lateralis Radix Praeparata). Peaks 3 (protocatechuic acid), 4—6, 7 (chlorogenic acid), 8, 10 (caffeic acid), 11, 13, 18, 19, 20 (isochlorogenic acid A), 21 (isochlorogenic acid C), 23—26 were specific to Baijiangcao (Patriniae Herba). Peaks 1, 12, 14—17, 27—29 were common to both Coicis Semen and Patriniae Herba. The extraction rate of the 15 batches ranged from 20.06% to 33.22%. The mass fraction of protocatechuic acid, chlorogenic acid, caffeic acid, isochlorogenic acid A, isochlorogenic acid C, benzoylmesaconine and linoleic acid were 0.012—0.076, 0.025—0.308, 0.011—0.047, 0.011—0.049, 0.014—0.044, 0.031—0.111, and 0.031—0.067 mg/g. Transfer rate were 5.05%—30.86%, 5.24%—21.31%, 3.57%—43.22%, 4.59%—18.69%, 6.97%—23.13%, 9.68%—62.32%, and 6.60%—19.95%. The chemometric analysis classified the 15 batches of samples into two groups, and OPLS-DA identified 14 common peaks responsible for quality differences. Among these, three peaks have been identified: peak 3 (protocatechuic acid), peak 10 (caffeic acid), and peak 20 (isochlorogenic acid A). Conclusion Based on chemical pattern recognition, an HPLC fingerprint and a multi-index component content determination method were established for the YFBP benchmark samples of a classical prescription. The process of quantitative transfer was analyzed, providing a scientific reference for quality control and the development of compound preparations based on YFBP standard samples.

R283.6

国家自然科学基金资助项目(82260847);贵州省科学技术厅科技创新人才团队(黔科合平台人才-CXTD[2023]020);贵州省教育厅高校科技创新团队(黔教技[2023]069号);贵州省科技计划项目(黔科合基础-ZK[2023]重点046);国家中医药管理局高水平中医药重点学科建设项目(zyyzdxk-2023185);贵州省基础研究自然科学项目(黔科合基础-ZK[2022]一般483)