A feature based pharmacophore for Candida albicans MyristoylCoA: protein N-myristoyltransferase inhibitors

Abstract

A three-dimensional pharmacophore model has been generated for Candida albicans MyristoylCoA: protein N-myristoyltransferase (NMT) inhibitors, using the software program catalyst. The in vitro NMT inhibitory activity of a series of peptidic inhibitors was used for pharmacophore generation. The effect of altering the control parameters and feature selection was studied to arrive at the pharmacophore model. The selection of the best hypothesis model was based on the total cost, predictive ability, difference in the cost from the null hypothesis and alignment of the training set compounds on to the hypothesis. The pharmacophore model selected has four features; one hydrophobic, two hydrogen bond acceptor and one positive ionisable function. Groups identified as necessary by scanning alanine mutagenesis studies of the peptidic substrate of C. albicans NMT, have been identified as pharmacophore features. Comparison of the ligand binding with the enzyme in the crystal structure of NMT and that proposed by the phamacophore is consistent. The pharmacophore thus generated can be used as a template for designing non-peptidic inhibitors of NMT.