Discovery of Quality Markers of Nucleobases, Nucleosides, Nucleotides and Amino Acids for Chrysanthemi Flos From Different Geographical Origins Using UPLC–MS/MS Combined With Multivariate Statistical Analysis

Nucleobases, nucleosides, nucleotides and amino acids, as crucial nutrient compositions, play essential roles in determining the flavor, function and quality of Chrysanthemi Flos. The quality of Chrysanthemi Flos from different geographical origins is uneven, but there have been no reports about the screening of their quality markers based on nutritional ingredients. Here, we developed a comprehensive strategy integrating ultra performance liquid chromatography coupled with triple-quadrupole linear ion-trap tandem mass spectrometry (UPLC–MS/MS) and multivariate statistical analysis to explore quality markers of Chrysanthemi Flos from different geographical origins and conduct quality evaluation and discrimination of them. Firstly, a fast, sensitive, and reliable UPLC–MS/MS method was established for simultaneous quantification 28 nucleobases, nucleosides, nucleotides and amino acids of Chrysanthemi Flos from nine different regions in China. The results demonstrated that Chrysanthemi Flos from nine different cultivation regions were rich in the above 28 nutritional contents and their contents were obvious different; however, correlation analysis showed that altitude was not the main factor for these differences, which required further investigation. Subsequently, eight crucial quality markers for nine different geographical origins of Chrysanthemi Flos, namely, 2′-deoxyadenosine, guanosine, adenosine 3′,5′-cyclic phosphate (cAMP), guanosine 3′,5′-cyclic monophosphate (cGMP), arginine, proline, glutamate and tryptophan, were screened for the first time using partial least squares discriminant analysis (PLS-DA) and the plot of variable importance for projection (VIP). Moreover, a hierarchical clustering analysis heat map was employed to intuitively clarify the distribution of eight quality markers in the nine different regions of Chrysanthemi Flos. Finally, based on the contents of selected eight quality markers, support vector machines (SVM) model was established to predict the geographical origins of Chrysanthemi Flos, which yielded excellent prediction performance with an average prediction accuracy of 100%. Taken together, the proposed strategy was suitable to discover the quality markers of Chrysanthemi Flos and could be used to discriminate its geographical origin.