RESUMEN
A computer-aided, 13C NMR-based dereplication method is presented for the chemical profiling of natural extracts without any fractionation. An algorithm was developed in order to compare the 13C NMR chemical shifts obtained from a single routine spectrum with a set of predicted NMR data stored in a natural metabolite database. The algorithm evaluates the quality of the matching between experimental and predicted data by calculating a score function and returns the list of metabolites that are most likely to be present in the studied extract. The proof of principle of the method is demonstrated on a crude alkaloid extract obtained from the leaves of Peumus boldus, resulting in the identification of eight alkaloids, including isocorydine, rogersine, boldine, reticuline, coclaurine, laurotetanine, N-methylcoclaurine, and norisocorydine, as well as three monoterpenes, namely, p-cymene, eucalyptol, and α-terpinene. The results were compared to those obtained with other methods, either involving a fractionation step before the chemical profiling process or using mass spectrometry detection in the infusion mode or coupled to gas chromatography.
Asunto(s)
Alcaloides/análisis , Aporfinas/química , Espectroscopía de Resonancia Magnética con Carbono-13/métodos , Monoterpenos/análisis , Monoterpenos/química , Peumus/química , Hojas de la Planta/química , Alcaloides/química , Monoterpenos Ciclohexánicos , Cimenos , Espectrometría de Masas , Estructura Molecular , Extractos Vegetales/análisis , Extractos Vegetales/químicaRESUMEN
Natural product chemistry began in Reims, France, in a pharmacognosy research laboratory whose main emphasis was the isolation and identification of bioactive molecules, following the guidelines of chemotaxonomy. The structure elucidation of new compounds of steadily increasing complexity favored the emergence of methodological work in nuclear magnetic resonance. As a result, our group was the first to report the use of proton-detected heteronuclear chemical shift correlation spectra for the computer-assisted structure elucidation of small organic molecules driven by atom proximity relationships and without relying on databases. The early detection of known compounds appeared as a necessity in order to deal more efficiently with complex plant extracts. This goal was reached by an original combination of mixture fractionation by centrifugal partition chromatography, analysis by 13 C NMR, digital data reduction and alignment, hierarchical data clustering, and computer database search.
Asunto(s)
Inteligencia Artificial , Productos Biológicos/química , Extractos Vegetales/química , Cromatografía Liquida , Francia , Espectroscopía de Resonancia Magnética , Estructura MolecularRESUMEN
Wood residues produced from forestry activities represent an interesting source of biologically active, high value-added secondary metabolites. In this study, 30 extracts from 10 barks of deciduous and coniferous tree species were investigated for their potential dermo-cosmetic use. The extracts were obtained from Fagus sylvatica, Quercus robur, Alnus glutinosa, Prunus avium, Acer pseudoplatanus, Fraxinus excelsior, Populus robusta, Larix decidua, Picea abies, and Populus tremula after three successive solid/liquid extractions of the barks with n-heptane, methanol, and methanol/water. All extracts were evaluated for their radical scavenging capacity, for their elastase, collagenase, and tyrosinase inhibitory activities, as well as for their antibacterial activity against gram-positive Staphylococcus aureus. In parallel, the global metabolite profiles of all extracts were established by 1D and 2D NMR and related to their biological activity. The results showed that the methanol extracts of Q. robur, A. glutinosa, L. decidua, and P. abies barks exhibit particularly high activities on most bioassays, suggesting their promising use as active ingredients in the dermo-cosmetic industry.