Molecular mapping of drug interaction domains on voltage-gated calcium channels

Recent studies provide convincing evidence that the pharmacologic block of so-called non-L-type Ca2+-channels in mammalian brain confers important therapeutic benefits, such as neuroprotection and analgesia. Non L-type Ca2+-channel blockers are thus considered promising new tools to treat stroke and severe pain and may also exert therapeutic actions in migraine. However, non-peptidergic compounds suitable for clinical use have not yet been reported. The aim of the proposed project is the structural analysis of two distinct domain on the alpha1-subunit of non L- type Ca2+-channels that can act as binding sites for such drugs. The first domain corresponds to the region which is known to bind dihydropyridine Ca2+-channel modulators on L- type Ca2+-channel alpha1-subunits. Employing site-directed mutagenesis and molecular modeling the molecular architecture of this binding domain will be analyzed and a detailed structure-activity relationship of the dihydropyridine -alpha1 -subunit complex will be derived. This structural information will then be used to predict strucutres of non-peptidergic compounds selectively acting on the correspondeing regions of non L-type Ca2+- channel alpha1 -subunits. The second domain represents the cytoplasmic linker involved in the high affinity interaction with the channels beta-subunit. Drugs that prevent beta-subunit association are predicted to act as Ca2+-channel blockers. Alternatively, drugs mimicking beta-subunit binding could serve as channel activators. Essentially irreversible beta-subunit binding to the alpha1 -subunit is mediated by an 18-50 amino acid binding motif (AID) within this linker and a conserved stretch of 50-200 amino acids (BID) on the beta-subunit. As a first step towards the rational design of drugs interfering with AID - BID interaction the three-dimensional solution structure of the recombinant high affinity AID - BID complex will be determined using NMR spectroscopy. Screening of random peptide libraries using the phage display system will be performed in order to isolate short peptides with high affinity for AID or BID. The peptides will be investigated for their potency as blockers in electrophysiological and biochemical experiments. The three-dimensional structure of the bound peptides as determined by NMR can then be used to predict and search for non-peptidergic structures selectively modulating non L-type channel function. Taken together a broad experimental approach ranging from site-directed mutagenesis, NMR analysis, molecular modeling, random peptide library screening and electrohphysilogical studies will be employed in this project to provide a solid structural basis for the development of therapeutically interesting non L-type Ca2+-channel modulators.

In electrically excitable cells (such as muscle or neuronal cells) voltage-gated calcium channels have the important function to convert electrical excitation into an intracellular signal. This signal is generated by these channels by allowing influx of calcium ions into the cell. This calcium signal then triggers appropriate cellular responses, such as muscle contraction or neurotransmitter release. Therefore voltage gated calcium channels control important body functions, including control of blood pressure, cardiac activity and all brain functions. In this project we could prove that mutations in calcium channel proteins can change their functional properties so that they cause human diseases. Mutations in a1A subunits, a calcium channel subunit which forms neuronal P/Q- type channels, cause a rare hereditary form of migraine, so-called Familial Hemiplegic Migraine (FHM). This provides an excellent disease model in which we could test the consequences of these mutations on channel function. The observed functional changes allowed important insight into the pathophysiological mechanisms underlying a migraine attack and the potential use a1A modulators as novel anti-migraine drugs. Our analysis revealed that FHM mutations change calcium channel function thus allowing activation of these channels, calcium influx and neurotransmitter release even when neurons are only excited weakly. Such weak excitation would normally not lead to neurotransmitter release. This hyperexcitability proved that migraine is a disease primarily affecting neurons and that modulators of P/Q-type channel blockers should have antimigraine properties. This hypothesis cannot be tested directly because at present no selective modulators of P/Q-type channels with high brain penetration exist. In a separate part of the project we have therefore analyzed which molecular properties small organic compounds should have in order to bind to P/Q-type channel a1A subunits. We determined the molecular basis why these subunits cannot bind drugs which recognize structurally highly related calcium channel subunits with very high affinity. We found that a three amino acid difference between a1A and drug-sensitive subunits explains most of the difference. These amino acid residues were found to either obstruct the access of drugs to a potential binding site on a1A or not to provide the appropriate interaction sites for the drug. We therefore postulated that small and more hydrophobic compounds structurally related to dihydropyridines could provide promising lead compounds for new P/Q-type calcium channel modulators. This hypothesis needs to be tested in future studies. All study results were published in highly ranked international journals. They represent an important basis for our understanding of migraine pathophysiology and the development of new anti-migraine drugs.