Modelling drug-receptor interactions in an average binding site for NK2

A tentative procedure applied to the search for a new antagonist of neurokinin A (NKA) is presented. In parallel a tentative 3-D model of the NK2 receptor was created, using bacteriorhodopsin (BRD) as a template. The residue substitutions were performed in BRD to obtain the sequence for NK2R_H and the seven a-helical segments were optimized forcing the a-helical backbone to match the corresponding aligned parts of BRD, while the arrangements of the side chains were model built based on available site-directed mutagenesis studies. Constrained MM and molecular dynamics simulations were carried out H-bonding a low energy conformer of the known drugs to residues in the receptor site, allowing both the receptor site and drugs to relax. The Connolly surface for each ligand allowed to determine an "average" binding site in which all the low energy conformers of known and prospective drugs were docked and classified according to a statistical index. The whole procedure was repeated exploiting the lately published structure of an actual G protein coupled receptor as a better template, thus producing a cavity in the binding site to directly dock the drugs. Corollary validations of the force fields used are also mentioned. In addition intra- and intermolecular interactions suitable to produce more active drugs were evaluated.