基于多指标成分定量联合多元统计分析评价不同产地鹅不食草药材质量

Quality evaluation of Centipedae Herba from different origins based on multi-component quantification combined with multivariate statistical analysis

基于多指标成分定量联合多元统计分析综合评价不同产地鹅不食草质量差异,提高鹅不食草药材的整体质量控制水平。以Waters XTerra MS C18柱为色谱柱,乙腈-0.2%磷酸为流动相,采用HPLC法同时测定鹅不食草中绿原酸、异绿原酸A、异绿原酸C、槲皮素、山柰酚、3-甲氧基槲皮素、山金车内酯D、山金车内酯C、小堆心菊素C、短叶老鹳草素A、羽扇豆醇、β-谷甾醇和蒲公英甾醇含量。对16批鹅不食草多指标成分定量检测结果进行聚类分析(cluster analysis,CA)、主成分分析(principal component analysis,PCA)和因子分析(factor analysis,FA),对不同产地鹅不食草药材进行分组和综合质量评价。利用正交偏最小二乘法-判别分析(orthogonal partial least squares-discriminant analysis,OPLS-DA)分析挖掘影响产品质量的差异性标志物。方法学验证符合中华人民共和国药典要求,13种成分在各自范围内线性关系良好(R^(2)>0.999),平均加样回收率(n=9)在96.88%~100.1%之间(RSD<2.0%)。CA结果显示16批鹅不食草聚为3类。PCA结果显示前2个主成分特征值分别为10.187和2.059,方差贡献率分别为78.361%和15.839%,表明前2个主成分代表鹅不食草94.200%信息量,对鹅不食草质量评价具有很好的代表性。FA结果显示16批鹅不食草样品主成分综合得分在-1.451~1.344,其中浙江和江苏产鹅不食草综合得分较高,排名居于前5位,湖北、湖南和广东产鹅不食草综合得分排名居中,贵州和四川产鹅不食草综合得分排名靠后。OPLS-DA结果显示短叶老鹳草素A、异绿原酸A、小堆心菊素C、槲皮素、山金车内酯C是影响鹅不食草产品质量的差异性标志物。所建立的HPLC法操作便捷,结果准确,可用于鹅不食草多指标成分定量控制;多元统计分析可用于不同产地鹅不食草的整体质量评价。

This study aims to comprehensively evaluate the quality differences of Centipedae Herba in different producing areas by multi-component quantification combined with multivariate statistical analysis,and to improve the overall quality control level of Centipedae Herba.The Waters XTerra MS C 18 column was used,with acetonitrile-0.2%phosphoric acid serving as the mobile phase.The contents of chlorogenic acid,isochlorogenic acid A,isochlorogenic acid C,quercetin,kaempferol,3-O-methylquercetin,arnicolide D,arnicolide C,microhelenin C,brevilin A,lupeol,β-sitosterol and taraxasterol in Centipedae Herba were simultaneously determined by HPLC.Cluster analysis(CA),principal component analysis(PCA)and factor analysis(FA)were performed on the quantitative detection results of multi-components components of 16 batches of Centipedae Herba,and the differential markers affecting product quality were excavated by OPLS-DA.The methodology was verified in accordance with the requirements of Chinese Pharmacopoeia.The 13 components showed good linear relationships in their respective ranges(R^(2)>0.999),and the average recovery rate(n=9)were 96.88%-100.1%(RSD<2.0%).The CA results showed that 16 batches of Centipedae Herba were clustered into 3 categories.The PCA results showed that the eigenvalues of the first two principal components were 10.187 and 2.059,respectively,and the variance contribution rates were 78.361%and 15.839%,respectively,indicating that the first two principal components represented 94.200%of the information content of Centipedae Herba,and had a good representation for the quality evaluation of Centipedae Herba.The results of FA showed that the comprehensive scores of principal components of 16 batches of Centipedae Herba samples were-1.451-1.344.Among them,the comprehensive scores of Centipedae Herba in Zhejiang and Jiangsu were higher,ranking in the top five,the comprehensive scores of Centipedae Herba in Hubei,Hunan and Guangdong were in the middle,and the comprehensive scores of Centipedae Herba in Guizhou and Sichuan were lower.The results of OPLS-DA showed that short-leaved geraniin A,isochlorogenic acid A,inulin C,quercetin,and seneciolide C were the differential markers affecting the quality of Centipedae Herba.The established HPLC method is convenient and accurate,and can be used for quantitative control of multi-index components in Centipedae Herba.Multivariate statistical analysis can be used for the overall quality evaluation of Centipedae Herba in different origins.