Journal of Functional Foods, vol. 64, pp. 103684-103684
Rational and efficient approaches to research and development are becoming more popular for increasing value enhancement in bioactive materials. In this study, metabolite profiles of 17 different Lagerstroemia species from different countries were performed using UPLC-QTof-MS combined with chemometric tools to identify potential metabolite predictors of antidiabetic activity. PLS-DA of the Lagerstroemia samples indicated two large, distinct clusters in the score plot. On the loading scatter plot, significant changes in metabolites of antidiabetic candidates were found between species, and two flavonol glycosides and corosolic acid were evaluated as key markers among the 25 analyzed metabolites. Based on the screening results, samples high in two triterpenic acids (corosolic acid and asiatic acid) showed a tendency to inhibit glucose uptake, which could have contributed to the correlation of active and inactive extracts. Therefore, these comprehensive results support our biochemometric models and contribute to the rational and efficient studies of pharmacologically active metabolites.