Modeling Block of the Kv1.5 Channel by Cationophilic and Cationic Ligands: Implications for General Mechanisms of the Inner Pore Block in P-Loop Channels

Abstract

Many small molecules of intriguingly different chemical structures bind in the inner pore of potassium channels. Previous mutational, electrophysiological, and ligand-binding experiments revealed common and diverse characteristics of action of different ligands including ligand-channel stoichiometries, voltage- and use-dependencies, and patterns of ligand-sensing residues. However, generally accepted structural interpretations for most of the available data are lacking. Here we used energy calculations with experimentally based constraints to dock flecainide, ICAGEN-4, benzocaine, vernakalant, and AVE0118 into the inner pore of the Kv1.5 channel. We arrived at ligand-binding models that for the first time explain: (i) the different Hill coefficients, (ii) opposite voltage dependencies of cationic and electroneutral cationophilic ligands, and (iii) effects of mutations of residues, which do not face the pore, on the ligand action. Two concepts were crucial for the modeling. First, the inner-pore block of a potassium channel requires a cationic “blocking particle”. A ligand that lacks a positively charged group, blocks the channels not per se, but in a complex with a permeant cation. Second, flexible hydrophobic moieties of ligands have a tendency to escape from the aqueous pore environment into subunit interfaces. Previously these concepts have allowed to explain action of so different ligands of potassium channels as correolide, propafenone, phenylalkylamines, and PAP-1-like immunosuppressants. The same concepts have been successfully used to analyze structure-function relationships of phenylalkylamines, benzothiazepines and dihydropyridines in calcium channels and local anesthetics in sodium channels. Supported by RFBR to DBT and NSERC to BSZ.