Computational studies of anticonvulsants and their mechanisms of action

Epilepsy is a common neurological condition that affects an estimated 50 million people worldwide. Since currently available anticonvulsant drugs are effective in less than 65% of patients, there is an ongoing need for new, more effective treatments. Today's drug design efforts are assisted by computational methods that permit a deeper understanding of biological phenomena at the molecular level. In this thesis, several of these computational approaches were used to gain insight into anticonvulsants and their mechanisms of action.; The succinimide derivatives are an older, yet understudied class of anticonvulsants. Quantitative structure-activity relationship (QSAR) methods were used to construct models for predicting their anticonvulsant activity. An extensive in silico library of succinimide analogues was screened using the resulting QSAR models. The screening study provided insight into the structural features that lead to improved bioactivity.; The neuroactive peptide FMRFamide has demonstrated anticonvulsant activity in the same seizure model as 3-methyl-3-phenylsuccinimide (MPS). Since these compounds may bind at the same active site, their structural similarities were explored. A possible bioactive conformation of the C-terminus of FMRFamide was proposed, based on results from rigid and flexible fit alignments with MPS.; The gamma-aminobutyric acid type A (GABAA) receptor is a ligand-gated ion channel that mediates the effects of the inhibitory neurotransmitter GABA. A comparative model of its extracellular and transmembrane domains was constructed using a recently published structure of the nicotinic acetylcholine receptor as a template. The ligand-binding domains and channel pore of the final model were in general agreement with results obtained from site-directed mutagenesis and ligand binding studies.; Knowing whether a compound will cross the blood-brain barrier (BBB) is important when designing anticonvulsont drugs. QSAR methods were used to construct a model for the prediction of BBB permeability. A set of novel descriptors, which estimates the surface area contributions of hydrogen-bond donors and acceptors, was developed and subsequently incorporated into the QSAR study. Molecular size, hydrophobicity and H-bonding ability were identified as important predictors of BBB permeability.