目的 基于HPLC指纹图谱、化学模式识别以及多成分含量测定相结合的方法，评价不同产区艾纳香药材的质量。方法 建立4个不同产区34批艾纳香HPLC指纹图谱，确定共有峰，根据前期提取分离出的艾纳香单体成分指认了其中5个色谱峰并测定样品中含量，结合相似度分析、聚类分析（hierarchicalcluster analysis，HCA）、主成分分析（principal component analysis，PCA）和正交偏最小二乘法-判别分析（partial least squares discrimination analysis，OPLS-DA）等化学模式识别方法以及药材综合质量评分寻找不同产区艾纳香药材中的差异。结果 化学指纹图谱共标定15个共有峰，各批艾纳香样品相似度在0.433～0.977，4个产区艾纳香样品在化学组分和含量上存在差异。4个产区的艾纳香药材基本上可以分为4类，每个产区大部分样品各自分布于不同的象限，不同产区之间有交叉样品，样品之间的离散程度较大，每个产区的样品本身质量就存在较大差异，尤其是海南和广西样品。通过变量权重重要性排序（variable importance in projection，VIP）值图确定了5个产区差异性指标成分，综合质量评分中3份海南琼中产样品评分排在前3位。测定5个黄酮类成分含量中3，3'，5，7-四羟基-4'-甲氧基二氢黄酮、圣草酚、3，3'，5-三羟基-4'，7-二甲氧基二氢黄酮和sakuranrtin均以贵州产整体含量较高，艾纳香素以海南产整体含量较高。结论 结合HPLC指纹图谱、HCA、PCA及多指标定性定量等分析方法可以更全面地评价艾纳香药材质量，为艾纳香药材质量控制和优良种质资源选育提供参考依据。

Objective The quality of Blumea balsamifera from different regions was evaluated based on the method of HPLC fingerprint, multi-component quantification and chemical pattern recognition. Methods The HPLC fingerprints of 34 batches of B. balsamifera from different regions in four provinces were established to determine the common peaks. Five chemical components were identified and determined according to the monomer composition extracted from B. balsamifera at early stage. The differences in the herbs from different regions were searched by integrating cluster analysis (HCA), principal component analysis (PCA), orthogonal partial least squares discriminant analysis (OPLS-DA) and comprehensive quality score. Results Fifteen common peaks were selected in the chemical fingerprint. Similarity of all the B. balsamifera samples were ranged from 0.433 to 0.977, suggesting that there are differences in chemical composition and content of B. balsamifera samples from four different regions. The samples from four regions were divided into four groups basically and most samples of each region were distributed in different quadrants. There were overlapping samples in different regions and the degree of dispersion among samples was large, which showed that the quality of the samples from each origin varies greatly, especially for Hainan and Guangxi. Five components were confirmed as the main compositions which caused the differences of chemical constituents in the samples from different regions through the VIP value map. Three samples from Qiongzhong county of Hainan ranked the top three in the comprehensive quality scores. In the determined five chemical components, the overall contents of 3,3',5,7-tetrahydro-4'-methoxy dihydroflavones, eriodictyol, 3,3',5-trihydroxy-4',7-dimethoxy dihydroflavones and sakuranrtin were higher in Guizhou than other origins, while the overall content of blumeatin is higher in Hainan than other origins. Conclusion The quality of B. balsamifera can be evaluated more comprehensively combined with fingerprint, cluster analysis and principal component analysis, which can provide reference for the quality control and breeding of elite germplasm resources.

R282.6

海南省自然科学基金资助项目（2019CXTD414）；海南省自然科学基金资助项目（319QN312）；农业农村部部门预算项目子任务（RZJP202006）；国家热带植物种质资源库项目