Plant metabolites are among the most studied natural products, with thousands of reports describing new chemical structures and valuable biological activities. However, the study of phylogenetic diversification and distribution patterns of plant secondary metabolites is still in its early steps, and several botanical families have not been studied in this context. Highly sensitive hyphenated detection techniques, such as Liquid Chromatography-Mass Spectrometry (LC-MS), along with bioinformatics tools for the classification of metabolites, have proven to be of great value in chemosystematics investigations. In this work, our main goal was to perform a chemosystematic study of the Malpighiaceae botanical family using MS/MS untargeted metabolomics, spectral searches, and recently described in silico classification tools. Leaves of a unique collection of 137 Malpighiaceae species (comprising all the major phylogenetic clades currently accepted for this family) were used to prepare hydroethanolic extracts (EtOH–H₂O, 4:1), which were further analyzed by UHPLC-ESI-MS/MS in an untargeted method in positive and negative ionization mode, separately. The data was processed using MZmine2 and subjected to the Feature-Based Molecular Networking (FBMN) workflow in the GNPS platform. In addition, a chemical hierarchy analysis was performed using Qemistree, and the in silico classifications were retrieved using the CANOPUS tool. Several spectral matches were retrieved from the public spectral libraries in the platform, and the several molecular families were clade or genera-specific. The in silico classification tools allowed to expand the chemical knowledge regarding this dataset and highlighted the different results positive and negative ionization modes can generate. Finally, the in silico results were used to perform ancestral character reconstructions aiming to map the evolutionary history of the chemical classes retrieved using the maximum likelihood criteria in the Malpighiaceae molecular phylogeny. All the major phylogenetic clades contained at least one homoplastic (i.e., non-exclusive) or synapomorphic (i.e., exclusive) class of metabolites supporting them. In conclusion, we expect that our findings can guide future studies in Malpighiaceae and that the workflow followed in this study will be used in future investigations in several fields, such as chemotaxonomy, metabolomics, chemical ecology, and for the discovery of new natural products.

Paula Carolina Pires Bueno

Appearances