Benchmarking biomolecular force field‐based Zn2+ for mono‐ and bimetallic ligand binding sites. Issue 8 (9th December 2022)

• Record Type: Journal Article
• Authors: Melse, Okke; Antes, Iris; Kaila, Ville R. I.; Zacharias, Martin
• Is Part Of: Journal of computational chemistry. Volume 44:Issue 8(2023)
• Page Start: 912
• Page End: 926
• Publication Date: 2022-12-09
• Subjects: binding site -- dummy atom model -- metal ion model -- Metalloproteins -- zinc; Chemistry -- Data processing -- Periodicals; 542.85
• DOI: 10.1002/jcc.27052
• Languages: English
• ISSNs: 0192-8651

Abstract

Abstract: Zn 2+ is one of the most versatile biologically available metal ions, but accurate modeling of Zn 2+ ‐containing metalloproteins at the biomolecular force field level can be challenging. Since most Zn 2+ models are parameterized in bulk solvent, in‐depth knowledge about their performance in a protein environment is limited. Thus, we systematically investigate here the behavior of non‐polarizable Zn 2+ models for their ability to reproduce experimentally determined metal coordination and ligand binding in metalloproteins. The benchmarking is performed in challenging environments, including mono‐ (carbonic anhydrase II) and bimetallic (metallo‐β‐lactamase VIM‐2) ligand binding sites. We identify key differences in the performance between the Zn 2+ models with regard to the preferred ligating atoms (charged/non‐charged), attraction of water molecules, and the preferred coordination geometry. Based on these results, we suggest suitable simulation conditions for varying Zn 2+ site geometries that could guide the further development of biomolecular Zn 2+ models. Abstract : This benchmarking study reveals key performance differences and limitations of widely‐used non‐polarizable Zn 2+ models in metalloproteins. The characterization of these models regarding their preferred coordination geometry, preferred ligating atoms, and ligand binding stability during molecular dynamics simulations allowed us to suggest suitable simulation conditions for varying Zn 2+ site geometries, which can guide further development of these models.