Improving Small-Molecule Force Field Parameters in Ligand Binding Studies.

Stefano Raniolo,Vittorio Limongelli

doi:10.3389/fmolb.2021.760283

Abstract

Small molecules are major players of many chemical processes in diverse fields, from material science to biology. They are made by a combination of carbon and heteroatoms typically organized in system-specific structures of different complexity. This peculiarity hampers the application of standard force field parameters and their in silico study by means of atomistic simulations. Here, we combine quantum-mechanics and atomistic free-energy calculations to achieve an improved parametrization of the ligand torsion angles with respect to the state-of-the-art force fields in the paradigmatic molecular binding system benzamidine/trypsin. Funnel-Metadynamics calculations with the new parameters greatly reproduced the high-resolution crystallographic ligand binding mode and allowed a more accurate description of the binding mechanism, when the ligand might assume specific conformations to cross energy barriers. Our study impacts on future drug design investigations considering that the vast majority of marketed drugs are small-molecules.

Highlights

Small molecules are organic compounds of relatively low molecular weight which are responsible of specific chemical reactions
Benzamidine is a small molecule acting as an inhibitor of trypsin and it interacts mainly through an amidine group contacting an aspartate in the binding pocket of the protein (ASP189) (Figure 1)
The results obtained with a new technique can be compared with those achieved with already established methods and with experimental values obtained through isothermal titration calorimetry (ITC) (Katz et al, 2001; Talhout and Engberts, 2001; Doudou et al, 2009; Buch et al, 2011; Söderhjelm et al, 2012; Limongelli et al, 2013; Takahashi et al, 2014)

Summary

Introduction

Small molecules are organic compounds of relatively low molecular weight which are responsible of specific chemical reactions. Few of them are successful and in even fewer cases the ligand/protein bound poses are experimentally resolved by X-ray crystallography, providing important insights at atomistic level on the possible interaction mechanism and the binding site in the target molecule. Such information is invaluable for a de novo drug design campaign on that target or to modify the ligand structure in order to improve its activity and toxicity, but it is available merely for a selection of drugs. The development of reliable and relatively fast computational techniques capable of inferring mechanisms of binding in ligand/protein complexes has taken the scene

Methods

Results

Conclusion