Back to Search Start Over

Development of new zinc ion parameters suitable for classical MD simulations of zinc metallo-peptidases

Authors :
Tomić, Antonija
Horvat, Gordan
Ramek, Michael
Agić, Dejan
Brkić, Hrvoje
Tomić, Sanja
Publication Year :
2019

Abstract

Reliable representation of the systems with transition metals is still a weak point of the force field based molecular modeling methods. We tested different strategies for modeling the zinc ion in classical molecular dynamics simulations based on nonpolarizable (nonbonded, hybrid bonded/nonbonded and cationic dummy atom models) as well as on polarizable (Drude-2013) potentials using human dipeptidyl peptidase III (DPP III) as a template. In the experimentally determined human DPP III structure the zinc ion is mostly tetrahedrally coordinated by two histidines, glutamate and one water or the substrate/inhibitor molecule. The quantum mechanics - molecular mechanics (QMMM) calculations showed an exchange of the four- and five-coordinated zinc ion during the reaction. Since neither of the strategies based on the nonpolarizable potential showed good agreement with experimental findings and the results of QMMM calculations, and the simulations utilizing the polarizable potential turned out to be extremely expensive, we modified the approach of Yang et al. [1] by extending the region used in quantum mechanical (QM) calculations based parametrization procedure. Namely, the zinc coordinated water molecule and the residues constituting the second metal ion coordination sphere were included in QM calculations resulting with a new set of the zinc ion parameters, so called 4-ligand-extended parameters, to be used within the AMBER force fields ff12SB and ff14SB. These parameters enabled realistic modeling of the active site configuration in different DPP III orthologues as well as in distant peptidases like thermolysin, neprilysin and aminopeptidase N.

Details

Language :
English
Database :
OpenAIRE
Accession number :
edsair.57a035e5b1ae..90673b400c1fbe5124568182c0511ddf