聚类分析结合主成分分析评价不同产地枳壳药材的一致性与差异性

摘要

Objective:To evaluate the consistency and difference and the influence of processing and harvest period on quality of Fructus Aurantii from different regions,in order to provide theoretical support for its quality control.Methods:HPLC was used to acquire characteristic fingerprint of 37batches Fructus Aurantii from different regions.The traditional Chinese medicine fingerprint similarity evaluation system(2012version)was used to establish common peak pattern and analyze the similarity.The principal component analysis(PCA)was carried on by using SPSS 17.0software.The hierarchical cluster analysis(HCA)was applicated by SIM-CA 13.0software with common peak area as variable.Results:A set of HPLC fingerprint containing 19common peak was obtained. Six flavonoids common peak were identified,which on behalf of narirutin,naringin,hesperidin,neohesperidin,hesperetin and nobiletin. The similarity coefficients weregreater than 0.90of samples,except those of S1,S5,S8and S18.PCA results showed that the first five principal components represented 86.435％of the total variability on the basis of the total variance explained and the screen plot,which could be chosen for comprehensive evaluation of Fructus Aurantii from different regions.HCA classified the 37batches sample into three main categories on the basis of the measurement parameters,which was consistent with the results of PCA and the content determination.Conclusion:The chemical composition of Fructus Aurantii from different regions was not unified,even in the same area.The qualities of Fructus Aurantii from Ruanjiang in Hunan province and Xingan in Jiangxi province were best.The processing of stir-baking with bran could reduce the content of flavonoids in Fructus Aurantii.The method is simple and reproducible, which can provide a certain basis for quality control and evaluation standard. Keywords:Fructus Aurantii;Fingerprint;Cluster Analysis;Principal Component Analysis;Quality Control%目的:评价不同产地枳壳药材的一致性与差异性,以及炮制和采收期对枳壳品质的影响,为其质量控制提供理论支持.方法:采用HPLC对37批次不同产地枳壳药材进行特征指纹图谱采集;使用中药指纹图谱相似度评价系统(2012版)建立共有峰模式,并对其进行相似度分析;利用SPSS17.0软件进行主成分分析(PCA);应用SIMCA 13.0软件,以共有峰峰面积为变量进行层序聚类分析(HCA).结果:建立了枳壳HPLC指纹图谱,匹配了19个共有峰,鉴定了其中6个黄酮类共有峰,分别为芸香柚皮苷、柚皮苷、橙皮苷、新橙皮苷、橙皮素和川陈皮素;37批枳壳药材中,除了S1、S5、S8和S18,其余样本的相似度均大于0.90;根据PCA主成分解释总方差和碎石图,前5个主成分可反映原来变量86.435％的信息,可用于对不同产地的枳壳进行综合评分.聚类分析将37批次样本分为三类,其结果与主成分综合评分排序及含测结果基本一致.结论:不同产地的枳壳药材的化学成分存在差异,甚至同一地区的枳壳药材质量也并不统一,其中湖南沅江和江西新干产地的枳壳质量最佳.麸炒炮制会降低枳壳中黄酮类成分的含量.建立的方法简单、重复性好,可为枳壳的质量控制和评价标准制定提供一定依据.