Metabolite profiling to discriminate different species and genus from thistles in Korea using liquid chromatography with quadrupole time-of-flight mass spectrometry.

This study was designed to classify and identify closely related thistle species in the genus Cirsium, as well as Carduus and Cephalonoplos species, which are also thistles. The comprehensive and untargeted metabolite profiles of nine Korean thistles were determined using ultra high performance liquid chromatography combined with hybrid quadrupole time-of-flight mass spectrometry. The difference in metabolite profiles among species was explored using principal component analysis and hierarchical clustering analysis. The significantly different metabolites (Bonferroni-corrected P-value < 0.001) were used to construct a partial least squares discriminant analysis model to predict the species of thistle. Nine species were successfully classified using a partial least squares discriminant analysis model and confirmed using a cross-validation method. Species with similar features were grouped based on unique patterns in variable clusters. The present study suggests that liquid chromatography with quadrupole time-of-flight mass spectrometry untargeted metabolomic profiling with chemometric analysis is an efficient and powerful tool for discriminating between different species of medicinal herbs.