Abstract
Background: Leishmaniasis is a neglected disease that affects 15 million people around the world. Many limitations are associated to the treatment as high cost and toxicity. Several classes of natural substances with proven leishmanicidal activity were reported in the literature. Phytochemsitry study of Anaxagorea dolichocarpa (Annonacea) reported the isolation of aporphine alkaloids.
Methods: In this study, we evaluate the potential activity of the azaphenanthrene alkaloids eupolaramine, imbiline 1, imbiline 4, sampangine, 3-metoxisampangine and 4- metoxisampangine, isolated from A. dolichocarpa, together with a homemade databank of 142 aporphynic alkaloids isolated from Annonaceae, through ligand-based and structurebased virtual screening (VS) against Leishmania donovani. A diverse set selected from CHEMBL databank of 1397 structures, with tested antileishmanial activity against promastigote L. donovani, were classified according pIC50 values in order to generate and validate Random Forest model that show higher statistical indices values. The structure of six different L. donovani enzymes were downloaded from PDB databank and alkaloids structures were submitted to molecular docking.
Results: From the six azaphenanthrene alkaloids, sampangine, 3-methoxy, and 4-methoxy were indicated as potential actives by the RF model. Docking results gave similar values for all six azaphenanthrene alkaloids. So, we performed in vitro tests with sampangine, imbiline 1, imbiline 4, and eupolaramine, which are available in our laboratory, and that show significant values of pIC50 (> 5.26).
Conclusion: Combined approach of VS allowed us to select that aporphynic alkaloid xyloguyelline as potential multitarget compound for leishmanial treatment, presenting activity against five strategic enzymes to treatment with probability of activity over 60% by RF model.
Keywords: Alkaloids, Anaxagorea dolichocarpa, Virtual Screening, Molecular Docking, Random Forest, Leishmania donovani, antileishmania activity.