A structure-based computational workflow to predict liability and binding modes of small molecules to hERG

Abstract

Off-target interactions of drugs with the human ether-à-go-go related gene 1 (hERG1) channel have been associated with severe cardiotoxic conditions leading to the withdrawal of many drugs from the market over the last decades. Consequently, predicting drug-induced hERG-liability is now a prerequisite in any drug discovery campaign. Understanding the atomic level interactions of drug with the channel is essential to guide the efficient development of safe drugs. Here we utilize the recent cryo-EM structure of the hERG channel and describe an integrated computational workflow to characterize different drug-hERG interactions. The workflow employs various structure-based approaches and provides qualitative and quantitative insights into drug binding to hERG. Our protocol accurately differentiated the strong blockers from weak and revealed three potential anchoring sites in hERG. Drugs engaging in all these sites tend to have high affinity towards hERG. Our results were cross-validated using a fluorescence polarization kit binding assay and with electrophysiology measurements on the wild-type (WT-hERG) and on the two hERG mutants (Y652A-hERG and F656A-hERG), using the patch clamp technique on HEK293 cells. Finally, our analyses show that drugs binding to hERG disrupt and hijack certain native—structural networks in the channel, thereby, gaining more affinity towards hERG.

Assessing hERG1 Blockade from Bayesian Machine-Learning-Optimized Site Identification by Ligand Competitive Saturation Simulations

“…One approach to address this challenge is to perform long atomistic simulations followed by identification of ensembles of conformational states for docking from those simulations ,, using clustering algorithms or experimentally inspired structural insights. However, assigning statistical weights for the selected rotameric states of specific residues such as those in the aromatic cassette is challenging and their contributions to protein–ligand affinity, therefore, are somewhat arbitrary.…”

“…The main purpose of the GLIDE-XP docking studies was to evaluate the usefulness of the receptor state captured in the cryo-EM for high-throughput screening of large ligand libraries. Several recent studies utilizing large-scale MD simulations pointed out that the conformational arrangement of the aromatic cassette (F557, F656, and Y652) captured in the cryo-EM structure may inhibit high-affinity binding to the intracavitary site. ,,, GLIDE outputs Gscore, which is calculated in kcal/mol and accounts for the hydrophobic interactions, π–π stacking interactions between aromatic rings, RMSD, desolvation, protein–ligand interaction, and hydrogen bond formation . For the purposes of our study, we only show the results of the best docking score of all ligands binding to the high-affinity site present the central cavity of the hERG1.…”

Structure-Based Prediction of hERG-Related Cardiotoxicity: A Benchmark Study

“…It is worth noting that most of these structure-based investigations were performed employing homology models based on the crystal structure of other K + channels, − as the first near-atomic resolution structure of hERG was determined only recently through single-particle cryoelectron microscopy. In particular, among the different models deposited by the authors, the one provided with the best resolution (3.7 Å PDB code: 5VA1) is today emerging as the structure of choice to perform molecular docking simulations, as highlighted by the recent literature. ,− Despite providing insights into the molecular determinants of drug binding, all of these studies focus on small data sets of compounds already proved to be (or potentially be) hERG binders. In other words, they do not provide any useful model for discerning hERG binders from safe compounds.…”

Molecular Dynamics-Derived Pharmacophore Model Explaining the Nonselective Aspect of KV10.1 Pore Blockers

“…The rotation of Y464 for MK-499 showed significantly less interference with the π-π interactions of residues F359 and F468, compared to the rotation of F468. Such π-π networking was also observed in the hERG channel with residues F619, F557, Y652, and F656, and there was disruption of the π-π stacking between F557 and Y652 upon binding of cisapride [62]. The π-π network of K V 10.1 lacks an additional aromatic ring that is present in hERG, as F619, which is instead M431 in K V 10.1.…”