Chromatographic fingerprint coupled with hierarchical clustering analysis and principal component analysis for quality evaluation and original discrimination of rhizomes of Helleborus thibetanus Franch by HPLC-DAD

Abstract

We developed an HPLC-based fractionation procedure coupled with hierarchical clustering and principal component statistical analysis to evaluate the quality and identify the authenticity of rhizomes of Hellebours thibetanus Franch. Professional software, namely Analysis System for Chromatographic Fingerprint and Data of Traditional Chinese Medicine (Version 2004), was used to establish a reference fingerprint of rhizomes of Hellebours thibetanus Franch. 17 raw samples and 1 processed sample of Hellebours thibetanus Franch were collected from various sources. 16 common peaks were determined in all the samples and the relative standard deviations of the relative retention times and relative peak areas were less than 1.0% when quercetin was used as the reference substance. The quality of the samples was classified into three levels by comparing the degree of similarity, these were high, general and inferior levels, respectively. Although there was slight discrimination among the samples which came from different locations, the similarity was very good between samples from the same level. The relative peak areas of the 16 common peaks of the samples were selected as input data for hierarchical clustering analysis and principal component analysis. 18 batches of samples were sorted into five clusters by the pattern recognition methods. The raw samples and the processed sample could also be distinguished by these methods. We suggest that this HPLC-based fingerprint method could be useful for the authentication and quality evaluation of H. thibetanus Franch.