3D pharmacophore models for thromboxane A(2) receptor antagonists.

Thromboxane A(2) (TXA(2)) is an endogenous arachidonic acid derivative closely correlated to thrombosis and other cardiovascular diseases. The action of TXA(2) can be effectively inhibited with TXA(2) receptor antagonists (TXRAs). Previous studies have attempted to describe the interactions between the TXA(2) receptor and its ligands, but their conclusions are still controversial. In this study, ligand-based computational drug design is used as a new and effective way to investigate the structure-activity relationship of TXRAs. Three-dimensional pharmacophore models of TXRAs were built with HypoGenRefine and HipHop modules in CATALYST software. The optimal HypoGenRefine model was developed on the basis of 25 TXRAs. It consists of two hydrophobic groups, one aromatic ring, one hydrogen-bond acceptor and four excluded volumes. The optimal HipHop model contains two hydrophobic groups and two hydrogen-bond acceptors. These models describe the key structure-activity relationship of TXRAs, can predict their activities, and can thus be used to design novel antagonists.