Halogen bonds: how do we model them?
Halogenated compounds have a prominent role in drug discovery, often aiming at improving drug-like properties such as membrane permeability or pharmacokinetic stability. Additionally, they often bind to their biomolecular targets via halogen bonds, which are non-covalent interactions of the type R–X∙∙∙B (X = Cl, Br, I; B = Lewis base; R = substituent). This behaviour is explained by the existence of a positive region on the electrostatic potential of X, named σ-hole, and it was soon evident that halogen bonds were present in biological molecules, inspiring several applications ranging from protein-ligand recognition to anion transport across membranes by synthetic receptors.
Tackling halogen bonds in biomolecular simulations is not trivial since the anisotropy of the halogen is not properly represented by a single charge used to electrostatically describe the halogen. Strategies and examples of how to tackle this interaction in the context of biomolecular simulations will be shown, in particular, how the conformational space of halogen-bonded protein-ligand systems is improved in molecular dynamics simulations by the introduction of an extra-point of charge at a given distance from the halogen. Additionally, the optimization of parameters that lead to improved calculated solvation energies of halogenated species, which are required for the calculation of binding energies, is going to be addressed. The importance of these methods in computer-aided drug design and discovery is going to be highlighted, aiming at establishing bridges and triggering new collaborations within BioISI.