Abstract

hKv1.3 channels in lymphocytes are targets for chemotherapy of autoimmune diseases. Phenylalkylamines block Kv1.3 channels by poorly understood mechanisms. In the inactivation-reduced mutant H399T, the second mutation A413C in S6 substantially decreases potency of phenylalkylamines with a para-methoxy group at the phenylethylamine end, whereas potency of phenylalkylamines lacking this group is less affected. Intriguingly, completely demethoxylated emopamil blocks mutant H399T/A413C with a 2:1 stoichiometry. Here we generated a triple mutant H399T/C412A/A413C and found that its emopamil-binding properties are similar to those of the double mutant. These data rule out disulfide bonding Cys412-Cys413, which would substantially deform the inner-helix, suggest a clash of Cys413 with the para-methoxy group, and provide a distance constraint to dock phenylalkylamines in a Kv1.2-based homology model. Monte Carlo-minimizations predict that the verapamil ammonium group donates an H-bond to the backbone carbonyl of Thr391 at the P-loop turn, the pentanenitrilephenyl moiety occludes the pore, while the phenylethylamine meta- and para-methoxy substituents approach, respectively, the sidechains of Met390 and Ala413. In the double-mutant model, the Cys413 sidechains accept H-bonds from two emopamil molecules whose phenyl rings fit in the hydrophobic inter-subunit interfaces, while the pentanenitrilephenyl moieties occlude the pore. Since these interfaces are unattractive for a methoxylated phenyl ring, the ammonium group of respective phenylalkylamines cannot approach the Cys413 sidechain and binds at the focus of P-helices, while the para-methoxy group clashes with Cys413. Our study proposes an atomistic mechanism of Kv1.3 block by phenylalkylamines and highlights the intra- and inter-subunit interfaces as ligand-binding loci.