Deciphering Steroidal and Aporphine Alkaloids as Antileukemic Agents by Approaches of Molecular Networking and...

The chemodiversity of plants is a valuable resource for drug discovery, and its combination with modern approaches can reduce the time consumption for bioactive metabolite discovery. This study aimed to evaluate the chemical constituents from 18 plant species of different families against leukemia cancer cells and the application of statistical analysis from metabolomic data and molecular networking for the prediction of bioactive metabolites. The samples, extracted by an accelerated solvent extractor using ethanol and water 7:3 (v/v), were analyzed by LC-DAD-MS and evaluated against leukemia cancer cells (Kasumi-1, KG-1, and K-562). Chemical data were aligned, analyzed by statistics, and applied to create the molecular network. Sesbania virgata, Aeschynomene denticulata, Erythroxylum angiufugum, Psidium guineense, Astronium fraxinifolium, Coccoloba ochreolata, Solanum glaucophyllum (S. glaucophyllum), and Paullinia pinnata inhibited K-562 leukemia cancer cell viability by approximately 70% at 100 μg/mL, while Ocotea diospyrifolia showed 35% inhibition for the KG-1 lineage. Alkaloid fractions from S. glaucophyllum and O. diospyrifolia revealed EC₅₀ values ranging from 13.9 to 6.4 μg/mL for K-562 and KG-1 cell lines, effectively inducing cell death with apoptotic characteristics, membrane integrity loss, and signs of late apoptosis. The molecular networking was essential and crucial to complement the statistical analysis, which was performed from 430 features and targeted steroidal and aporphine alkaloids. Boldine revealed EC₅₀ values of 46, 116, and 145 μM for Kasumi, KG-1, and K-562 cancer cell lines, respectively. The findings marked the relevance of a broader chemical data analysis to predict bioactive compounds, emphasizing potential benefits in the search for metabolites against leukemia cancer cells, particularly steroidal and aporphine alkaloids.