目的 建立不同产地批次大青叶Isatidis Folium的HPLC指纹图谱及指标成分含量测定方法，结合化学模式识别法评价不同产地大青叶质量。方法 建立大青叶HPLC指纹图谱，应用中药色谱指纹图谱相似度评价系统（2012.130723版本）、IBM SPSS Statistics 27.0、SIMCA14.1软件，结合聚类分析（hierarchical cluster analysis，HCA）、主成分分析（principal component analysis，PCA）、正交偏最小二乘法-判别分析（orthogonal partial least squares discrimination analysis，OPLS-DA）对不同产地大青叶质量进行评价，根据变量重要性投影（variable importance for projection，VIP）值，对差异性成分异牡荆素、靛玉红进行定量分析，基于含量测定结果建立熵权逼近理想排序法（technique for order preference by similarity to ideal solution，TOPSIS）-灰色关联度（grey relation analysis，GRA）融合模型进一步分析评价。结果 建立的大青叶HPLC指纹图谱中，识别共有峰19个，指认峰6个，分别为峰4（丁香酸）、峰9（阿魏酸）、峰12（异牡荆素）、峰16（色胺酮）、峰18（靛蓝）、峰19（靛玉红）。指纹图谱相似度在0.871～0.989；HCA将15批大青叶分为2类；PCA共提取4个成分，累积方差贡献率为91.745%；OPLS-DA筛选出异牡荆素、靛玉红等6个差异性成分。大青叶中异牡荆素、靛玉红的质量分数在0.781～4.385 mg/g、0.134～0.921 mg/g，熵权TOPSIS-GRA融合模型结果显示15批不同产地大青叶样品含量存在差异，编号为S1、S2、S3的样品质量较优。结论 建立的HPLC指纹图谱和含量测定结果可为大青叶的质量控制和整体性评价提供参考。

Objective The HPLC fingerprint and the index component content determination method were established to evaluate the quality of Daqingye (Isatidis Folium) from different producing areas with chemical pattern recognition methods. Methods Establish HPLC fingerprint of Isatidis Folium, apply the TCM chromatographic fingerprint similarity evaluation system (2012.130723 version), IBM SPSS Statistics 27.0 and SIMCA14.1 software. Combining hierarchical cluster analysis (HCA) and principal component analysis (PCA) and orthogonal partial least squares discrimination analysis (OPLS-DA) to evaluate the quality of Isatidis Folium from different producing areas. Quantitative analysis of the differential components, isovitexin and indirubin, was conducted based on the variable importance for projection (VIP) values. Based on the content measurement results, the technique for order preference by similarity to ideal solution (TOPSIS) -grey relation analysis (GRA) fusion model was established to further analyze and evaluate. Results In the established HPLC fingerprint of Isatidis Folium, 19 common peaks were identified, with six peaks assigned, which are peak 4 (syringic acid), peak 9 (ferulic acid), peak 12 (isovitexin), peak 16 (tryptanthrin), peak 18 (indigo), and peak 19 (indirubin). The similarity of the fingerprint spectra ranged from 0.871 to 0.989; HCA classifies 15 batches of Isatidis Folium into two categories; PCA extracts four components with a cumulative variance contribution rate of 91.745%; OPLS-DA identifies six differential components including isovitexin and indirubin. The average mass fraction of isovitexin and indirubin in Isatidis Folium were 0.781—4.385 mg/g and 0.134—0.921 mg/g, respectively. The entropy weight TOPSIS-GRA fusion model shows that there are differences in the content of 15 batches of Isatidis Folium samples from different origins, with samples numbered S1, S2, and S3 having superior quality. Conclusion The established HPLC fingerprint and content determination results can provide a reference for the quality control and holistic evaluation of Isatidis Folium.

R286.2

国家自然科学基金面上项目（82474099）；大青叶㕮咀饮片研究横向课题（14207848）