Theoretical studies applicable to the design of novel anticonvulsants: an AM1 molecular orbital structure–activity study of dihydropyridine calcium channel antagonists

The development and synthesis of anticonvulsant new chemical entities that are distinct from the cyclic ureides in current clinical use is a continuing neuropharmacologic priority. The design of neuronal specific dihydropyridines active at the L-type calcium channel protein represents a rational approach to this design problem. To provide the structural data required for the design of anticonvulsant dihydropyridines, an AM1 semi-empirical molecular orbital study has been undertaken. Forty-six dihydropyridine calcium channel antagonists have been fully optimized at the AM1 level. Each of the 46 analogues was considered in six conformations to provide a systematic evaluation of changes in ester and phenyl ring orientation. The calculational validity of the AM1 Hamiltonian when applied to dihydropyridines was demonstrated by comparing AM1-optimized structures to experimental (X-ray crystallographic) and abinitio molecular orbital (STO 3G basis set) geometries. For each dihydropyridine, 79 AM1-derived geometric and electronic descriptors were obtained. The resulting descriptor matrix comparing structural descriptors with biological activity was statistically reduced to provide regression and discriminant structure–activity models for dihydropyridine calcium channel antagonism.