In Silico Prediction of the Chemical Block of Human Ether-a-Go-Go-Related Gene (hERG) K⁺ Current

Abstract

A variety of compounds with different chemical properties directly interact with the cardiac repolarizing K⁺ channel encoded by the human ether-a-go-go-related gene (hERG). This causes acquired forms of QT prolongation, which can result in lethal cardiac arrhythmias, including torsades de pointes one of the most serious adverse effects of various therapeutic agents. Prediction of this phenomenon will improve the safety of pharmacological therapy and also facilitate the process of drug development. Here we propose a strategy for the development of an in silico system to predict the potency of chemical compounds to block hERG. The system consists of two sequential processes. The first process is a ligand-based prediction to estimate half-maximal concentrations for the block of compounds inhibiting hERG current using the relationship between chemical features and activities of compounds. The second process is a protein-based prediction that comprises homology modeling of hERG, docking simulation of chemical-channel interaction, analysis of the shape of the channel pore cavity, and Brownian dynamics simulation to estimate hERG currents in the presence and absence of chemical blockers. Since each process is a combination of various calculations, the criterion for assessment at each calculation and the strategy to integrate these steps are significant for the construction of the system to predict a chemical’s block of hERG current and also to predict the risk of inducing cardiac arrhythmias from the chemical information. The principles and criteria of elemental computations along this strategy are described.