A 3D-QSAR Study on the Antitrypanosomal and Cytotoxic Activities of Steroid Alkaloids by Comparative Molecular Field Analysis

As part of our research for new leads against human African trypanosomiasis (HAT), we report on a 3D-QSAR study for antitrypanosomal activity and cytotoxicity of aminosteroid-type alkaloids recently isolated from the African medicinal plant A. DC. (Apocynaceae), some of which are strong trypanocides against ( ), with low toxicity against mammalian cells. Fully optimized 3D molecular models of seventeen congeneric alkaloids were subjected to a comparative molecular field analysis (CoMFA). CoMFA models were obtained for both, the anti- and cytotoxic activity data. Model performance was assessed in terms of statistical characteristics ( ², ², and ² for partial least squares (PLS) regression, internal cross-validation (leave-one-out), and external predictions (test set), respectively, as well as the corresponding standard deviation error in prediction (SDEP) and -values). With ² = 0.99, ² = 0.83 and ² = 0.79 for anti- activity and ² = 0.94, ² = 0.64, ² = 0.59 for cytotoxicity against L6 rat skeletal myoblasts, both models were of good internal and external predictive power. The regression coefficients of the models representing the most prominent steric and electrostatic effects on anti- and for L6 cytotoxic activity were translated into contour maps and analyzed visually, allowing suggestions for possible modification of the aminosteroids to further increase the antitrypanosomal potency and selectivity. Very interestingly, the 3D-QSAR model established with the alkaloids also applied to the antitrypanosomal activity of two aminocycloartane-type compounds recently isolated by our group from L. (Buxaceae), which indicates that these structurally similar natural products share a common structure⁻activity relationship (SAR) and, possibly, mechanism of action with the steroids. This 3D-QSAR study has thus resulted in plausible structural explanations of the antitrypanosomal activity and selectivity of aminosteroid- and aminocycloartane-type alkaloids as an interesting new class of trypanocides and may represent a starting point for lead optimization.