Pharmacophore and pseudoreceptor modelling of class Ib antiarrhythmic and local anaesthetic lidocaine analogues.

A molecular modelling study was carried out in order to investigate the molecular binding behaviour of antiarrhythmically and local anaesthetically active aminoacylanilide derivatives from the lidocaine type at their specific sodium channel binding site. An examination of relevant X-ray structures and of results derived from systematic and random search conformational analyses yielded information about the spatial requirements of these sodium channel blocking compounds. Common structural elements in combination with their non-covalent interaction potentials were used to generate a rational pharmacophore model. To further support and refine this model an atomistic pseudoreceptor of the Na+ channel binding site was constructed using a training set of eight well-defined lidocaine homologues. With the final pseudoreceptor, composed of tyrosine, phenylalanine, serine, valine and three isoleucine residues, it was possible to correlate experimental versus calculated dissociation constants of the training set with a correlation coefficient of 0.98. To test the accuracy of this model, the affinities of three additional compounds, not used for pseudoreceptor modelling, were predicted. After free relaxation within the binding cavity using a Monte-Carlo minimization the test set yielded a RMS error in the prediction of 0.039 kcal/mol corresponding to an uncertainty factor of 1.06. In addition, this hypothetical receptor model provides evidence for an exceptional binding mode of the lidocaine metabolite glycinexylidide (GX) which could explain its low binding affinity and thereby possibly the minor physiological effects with respect to lidocaine.