Molecular Determinants of D and L-Sotalol Stereoselective Proarrhythmia Procilivties

Abstract

Cardiac arrhythmias can be caused by drug induced blockade of cardiac potassium channel KV11.1 encoded by the human Ether-à-go-go-Related Gene (hERG), being a major obstacle for the development of new pharmaceuticals. Inhibition of this protein activity leads to an increased cardiac action potential duration manifesting as a prolonged QT interval on the surface ECG, which can cause potentially lethal arrhythmias. However, not all hERG channel blocking and QT prolonging drugs are arrhythmogenic leading to possible attrition of safe and effective pharmaceuticals. Previously we developed a multi-scale in silico pipeline to predict arrhythmogenicity of two hERG blocking drugs, dofetilide and moxifloxacin, from their chemical structures using a combination of atomistic structural and functional kinetic modeling approaches (Yang et al. Circ. Res. 2020, 126: 947). Here we applied such computational pipeline backed by an enantiomeric separation and electrophysiology recordings to predict the pro-arrhythmia risks of two enantiomers of sotalol, an anti-arrhythmic drug with known stereoselective beta-blocking properties. We determined similar hERG channel binding affinities of d and l-sotalol from both patch-clamp recordings and umbrella sampling all-atom molecular dynamics (MD) simulations. MD derived hERG channel - drug on and off rates were used as functional model parameters. In cardiac tissue simulations our model predicted substantially higher emergent arrhythmogenicity of d-sotalol compared to the l-stereoisomer due to a more potent beta adrenergic receptor blockade by the latter and also provided good agreement with clinical QT interval prolongation for the racemic mixture. Our study helped to elucidate molecular mechanisms of stereoselective arrhythmogenicity for dl-sotalol allowing to predict cardiac safe and dangerous drug formulations. This work has been made possible by NIH NHLBI and Common Fund as well as American Heart Association funding, usage of XSEDE, PSC Anton 2, TACC Frontera, and Oracle for Research computational resources.