A Comparison of 3D Conformations of Endothelin-1 Analogs to Find the Pharmacophore Model Required for Endothelin Receptor Ligand Activity

Abstract

Endothelin receptors (ETRs) are GPCRs whose activation results in the increase of intracellular calcium levels, and are ultimately involved in vasoconstriction and vasodilation. Because control of ETR activity has important medical implications, it is beneficial to rapidly discover effective synthetic ETR ligands by virtual drug screening methods. Despite its biological importance, little is known about ETR structure, since there are no reported X-ray structures to date. However, structural information is available for Endothelin-1, the natural ligand to ETR and the most potent vasoconstrictor yet identified. Thus, our work takes the ligand-based approach instead of receptor-based approach to ETR ligand pharmacophore discovery. Two well-studied ETR ligands, BQ-123 and bosentan, were chosen to be reference compounds in addition to Endothelin-1, and 30 known synthetic ligands were defined as the sample active compounds set. Conformers of the sample set were prepared by genetic algorithm-based approach using OpenBabel, while conformers of the reference set were prepared by molecular dynamics, using myPresto/cosgene. For molecular superimpositions, we built and used a program called 'BMSIP,' which is an algorithmic extension of the ROCS program. While it performs rigid-body superimpositions using a volume overlap score function based on spherical Gaussian descriptions of atomic shape, BMSIP also introduces a scores-weighting matrix to bias atom-pairings by atom-type correspondence. Superimposition of the sample set against the reference set was performed to obtain a density map that shows the frequency of atom overlaps between the superimposed compounds and reference compounds. This map can indicate the essential chemical features (the pharmacophore) required for interaction with ETRs. The results suggest that nonpolar groups alone, in particular aromatic groups, are important for ETR binding affinity and provide a good initial guess for directed ETR ligand design.