Virtual Separation of Phytochemical Constituents by Their Adduct-Ion Patterns in Full Mass Spectra.

Yufeng Zhang,Jianyang Pan,Jinghua Zhong,Yi Wang,Xiaohui Fan,Yiyu Cheng

DOI: https://doi.org/10.1016/j.chroma.2012.01.006

2012-01-01

Abstract:In the present study, a tool called classifier for traditional Chinese medicine (CTCM) was developed to facilitate the discrimination of phytochemical constituents in two-dimensional datasets of liquid chromatography/mass spectrometry (LC/MS). Based on the full mass spectral characteristics of components in a mixture, particularly their adduct-ion patterns, an entire LC/MS dataset can be separated into several sub-datasets, each corresponding to one or several types of natural products. CTCM has been verified using 24 standard compounds and successfully applied in two previously reported LC/MS datasets, which confirmed the capability of proposed tool to extract adduct-ion patterns from LC/MS datasets. Moreover, the LC/MS dataset of a Wei-Fu-Chun (WFC) tablet, a prescription drug consisting of three crude herbs used for the treatment of enteric diseases, was analyzed using CTCM. The analysis indicated that the compounds in WFC could be split into three groups, with the main constituents including saponins from Radix Ginseng Rubra, flavonoids from Fructus aurantii, and phenolic compounds from Isodon amethystoides. The major compounds in the three groups were either positively identified or tentatively characterized by multi-stage and high resolution MS. The proposed tool provides a novel approach for processing the LC/MS datasets of complex samples, such as traditional Chinese medicine and botanical drugs.