Ion channels in drug discovery : focus on biological assays

Abstract

Ion channels are well characterised drug targets. However, the techniques used to study ion channel pharmacology have not been particularly applicable to modern drug discovery. The aim of this thesis was to examine the usefulness of both established and novel methods used in the discovery and pharmacological characterisation of drugs interacting with ion channels. Three different classes of ion channels have been studied; the GABAA receptor representing the classical ligand-gated ion channel, the voltage-gated sodium channel (NaV-1.7) and the newly discovered non-selective cation channel, Transient Receptor Potential Melastatin-8 (TRPM8).

Due to the difficulty in expression and purification of membrane proteins, no crystal structures of GABAA receptors are currently available and structure-based drug design is not directly applicable. Using pharmacophore modelling and site-directed mutagenesis, ligand interactions were studied at a structural level. A pharmacophore model for the benzodiazepine binding site (BzR) on the GABAA receptor, based on structure-activity relationship studies for 136 different ligands from 10 structurally different classes was used to model flavonoids binding to the GABAA receptor. By synthesising a series of flavonoids, the structure activity relationship (SAR) was investigated using 3H-flumazenil binding to rat cortical membranes in-vitro. The results demonstrate that flavonoids with high affinity for the BzR spanning the whole efficacy range from agonists to inverse agonists can be synthesised and the receptor binding properties of the flavonoids can successfully be fitted in a comprehensive pharmacophore model.

In order to investigate which arginine residues potentially contribute to the formation of the GABA binding pocket, six arginines conserved in all human GABAA receptor alpha subunits as well as two non-conserved arginines in the extracellular, N-terminal segment of the alpha5 subunit, were substituted by lysines. The individual alpha5 subunit mutants were co-expressed with human beta2 and gamma2s GABAA receptor subunits in CHO cells using transient transfection. Electrophysiological whole-cell patch-clamp recordings showed that, of the eight arginine residues tested, only the two arginines, at position 70 and 123, appear to be essential for GABA-gated chloride current, since the EC50 values of the two mutant constructs increase more than 100-fold as compared to wild-type alpha5beta2gamma2s GABAA receptors.

Ion channels are challenging targets particularly in the early phases of the drug discovery process, due to the lack of technologies available to screen large numbers of compounds in functionally relevant assays. The human sodium channel NaV-1.7 was stably expressed in HEK293 cells and 3 high throughput screening (HTS) assays were compared. 1) A Li+ flux atomic absorption spectroscopy (AAS) assay, 2) a Fluorometric Imaging Plate Reader (FLIPR) membrane potential assay 3) a Fluorescent Energy Transfer (FRET) based membrane potential assay. These 3 assays were then compared to an automated electrophysiological assay (the Ionworks-HT platform) to characterize eleven known sodium channel inhibitors.

The results demonstrated that all 3 HTS assays are suitable for the identification of NaV-1.7 inhibitors, but for an HTS assay the Li+-flux assay was more robust than the FLIPR and FRET based membrane potential assays. Furthermore, there was a better correlation between the Ionworks assay and the Li+-flux assay regarding the IC50 values of the sodium channel inhibitors investigated.

Human TRPM8 channels stably expressed in HEK293 were used to develop a FLIPR based HTS assay using the calcium sensitive dye Fluo-4. The two known TRPM8 agonists menthol and icilin induced [Ca2+]i increases with EC50 values comparable to electrophysiological measurements. Screening of a compound library identified 15 novel antagonists, which were characterized using both the FLIPR assay and twoelectrode voltage clamp electrophysiology in Xenopus oocytes expressing TRPM8. The antagonists were additionally tested in a FLIPR assay using cold as the agonist, which additionally demonstrated concentration dependent antagonism. These results show that menthol, icilin and cold can all be used as agonists when searching for antagonists and demonstrate that a cellular HTS assay for investigating the pharmacological activity of TRPM8 ligands can be developed.

The work in this thesis illustrates the validity of multiple methods and technologies when studying ion channel pharmacology. Traditional ligand binding assays in combination with molecular modelling are important tools for structure activity relationship studies, but functional assays are needed to determine ligand efficacy. The functional cellular assays that have become available recently to study ion channels have shown great potential in terms of screening throughput and delivering quality data. These assays are capable of determine both the pharmacological properties of ion channel ligands and perform SAR studies. With these developments ion channels may become more tractable targets for the pharmaceutical industry, leading to new improved drugs for the patients.