Discrimination of Different Part of Curcuma longa by HPLC Fingerprints Combined with Multivariate Statistical Analysis

Abstract

Aim: The purpose of this study is to develop a method to explore the difference between the rhizomes and tuberous roots of Curcuma longa, and screen out the difference markers. Materials and Methods: The quantitative analysis and fingerprints of rhizomes and tuberous roots were established by HPLC, rhizomes and tuberous roots of Curcuma longa were clearly discriminated by Hierarchical Cluster Analysis (HCA), Similarity Analysis (SA) and Principal Component Analysis (PCA). The difference markers were screened out by Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA). Results: The contents of curcumin, bisdemethoxycurcumin and demethoxycurcumin in all rhizomes were higher than those in tuberous roots. Multivariate statistical analysis shown that the samples of rhizomes were grouped into the same categories and samples of tuberous root were grouped another group in each analysis mode. And the OPLS-DA model had a good productivity and good fit indicated by the value of R2Y=0.981, Q2=0.946and R2X=0.816. The important markers for discrimination samples were the peak 14, peak 10 (demethoxycurcumin) and peak 11 (curcumin). Conclusion: The fingerprinting combination of multivariate statistical analysis can be applied to distinguish the rhizomes and tuberous roots of Curcuma longa.