Modeling the hERG potassium channel in a phospholipid bilayer: Molecular dynamics and drug docking studies

Abstract

The hERG potassium channel has recently been a matter of extensive studies both at experimental and computational levels, because of its possible involvement in the potentially lethal drug-induced long QT syndrome. In this context, in the absence of an experimentally determined 3D structure, to acquire a detailed description of the channel pore and an understanding of atomic determinants for drug binding is of enormous interest at both academic and industrial levels. To contribute to this aim, we first built the open and closed states of the channel by homology modeling techniques, and then submitted both channel models to ns-time-scale MD simulations in explicit membrane. An in-depth analysis of the dynamical behavior of the channel pore with particular attention to the cavity volume was carried out. Finally, using hERG conformations coming from MD simulations, docking experiments were performed to identify a possible binding mode for the most potent hERG channel blocker so far known, the antihistaminic drug astemizole. We show that the combined use of MD and docking is suitable to identify a possible binding mode of drugs in a fairly good agreement with experiments. Moreover, the exploitation of MD snapshots in the docking experiments allowed us to capture some induced-fit effects related to the side chain conformations of Tyr652 and Phe656, which are residues playing a pivotal role in the hERG drug binding.

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

“…Farid et al (2006) reported similar compound-hERG associations in docking simulations with terfenadine, cisapride, MK-499, s-ibtilide, clofilium, and sertindole and its analogs [47]. On the other hand, if no constraints are given for the docking space for nifekalant during computation, compounds tended to be distributed along the ion conduction pathway as reported by Rajamani et al (2005) [49] and Masetti et al (2007) [48].…”

“…An alternative would be to take into account the target protein. A structure-based screening method with homology models of the ion conduction unit of hERG to predict inhibitory activity of compounds is discussed below [47][48][49][50][51][52].…”

“…So although hERG possesses large N and C termini [2,89], the crystal structure of a K + channel with an open pore is a suitable reference structure to generate the 3D model of an ion conduction unit of hERG [47,48,50,[90][91][92]]. …”

“…In the homology model, all amino acids responsible for drug binding face the central cavity. Masetti et al (2007) performed an MD simulation for the open conformation model of hERG based on the KvAP structure to examine the stability and the quality of its conformation [48]. They showed that the root mean squared fluctuation of the Cα positions was small during the simulation, indicating that the open conformation model is stable and potentially reflects one aspect of the open state of hERG.…”

“…Docking of ligands to the hERG homology model. To predict compound positions on hERG, several studies have docked hERG blockers in the central cavity of the pore [47][48][49][50][51][52]. We used a protein-ligand fl exible docking program, Sievgene [62].…”