Metabolomics and chemoinformatics approaches to identify natural compounds of interest from Asteraceae

The generation of extracts and pure compounds libraries has been shown to be an important strategy to obtain compounds of interest. Many strategies have been used to generate natural products libraries, especially considering the Lipinski rule. However, the combination of powerful analytical methods, chemical structure databases and chemoinformatics tools can certainly provide additional information. We describe herein the generation of an extract library consisting of ca. 300 extracts from species from the family Asteraceae and ca. 150 pure compounds that have previously been isolated from these species as well as chemoinformatics strategies used for dereplication of medicinal plants extracts [3]. The extracts were obtained using an optimized and standardized method. Afterwards, the extracts and the pure compounds were analyzed by UHPLC-UV-HRMS. The same chromatographic system (Kinetex C18, 1.7 µm, MeCN-H₂O) was used to obtain metabolic profiles of selected extracts whose dereplication would be performed. Two-dimensional chemical structures and their analytical data (accurate masses and retention times) were used to build a QSRR model for retention time prediction of compounds that were not included in the library. Using PCA it was possible to verify matches of chemical structures with those present in chromatograms of medicinal plants extracts. As a result, several compound matches were identified; for example, compounds from Lycnophora ericoides (a medicinal plant known as “Brazilian arnica”) were dereplicated, like the flavonoid chrysin 6,8-di-C-glucopyranoside and the sesquiterpene lactones 4β,5-dihydro-15-deoxygoyazensolide and 4β,5-dihydrolychnopholide, which were identified with 80 – 90% of confidence. Based on the results, we conclude that focused libraries combined with chemoinformatics tools are important to accelerate the identification of known compounds and also to improve the possibility of finding new compounds.