Computer-aided molecular modeling of a D2-agonist dopamine pharmacophore.

Using computer-aided molecular modeling techniques to analyze models recently proposed for the receptor binding sites of dopaminergic agonists, we superimposed the chemical structures of various compounds that mimic the pharmacological behavior of dopamine, as well as inactive enantiomers, on a postulated three-dimensional frame of reference. We analyzed the vector directionalities of the lone pairs of the nitrogen common to these molecules, and the acidic hydrogen of phenols (in aminoindanes, aminotetralins, apomorphines, p-phenol-piperazines, octahydrobenzo(g)quinolines, octahydrobenzo(f)quinolines, and benzazepines) or of nitrogen (in ergoline-type compounds and related structures). This model, when expressed as distances from that of the reference compound pergolide, correlates with the dopaminergic binding affinity observed in compounds previously reported to act on the dopaniergic system in the central nervous system (CNS). The regression analysis of log KD with respect to the distances of the vectors of the acidic hydrogen support the hypothesis that these compounds bind to the receptor as donors in hydrogen bond formation.