A Transferable Non-bonded Pairwise Force Field to Model Zinc Interactions in Metalloproteins.

Ruibo Wu,Zhenyu Lu,Zexing Cao,Yingkai Zhang
DOI: https://doi.org/10.1021/ct100525r
2010-01-01
Journal of Chemical Theory and Computation
Abstract:Herein we introduce a novel practical strategy to overcome the well-known challenge of modeling the divalent zinc cation in metalloproteins. The main idea is to design short-long effective functions (SLEF) to describe charge interactions between the zinc ion and all other atoms. This SLEF approach has the following desired features: (1). It is pairwise, additive and compatible with widely used atomic pair-wise force fields for modeling biomolecules; (2). It only changes interactions between the zinc ion and other atoms, and does not affect force field parameters that model other interactions in the system; (3). It is a non-bonded model that is inherently capable to describe different zinc ligands and coordination modes. By optimizing two SLEF parameters as well as zinc vdW parameters through force matching based on Born-Oppenheimer ab initio QM/MM molecular dynamics simulations, we have successfully developed the first SLEF force field (SLEF1) to describe zinc interactions. Extensive molecular dynamics simulations of seven zinc enzyme systems with different coordination ligands and distinct chelation modes (4-,5-,6-fold), including the binuclear zinc active site, yielded zinc coordination numbers and binding distances in good agreement with the corresponding crystal structures as well as ab initio QM/MM MD results. This not only demonstrates the transferability and adequacy of the new SLEF1 force field in describing a variety of zinc proteins, but also indicates that this novel SLEF approach is a promising direction to explore for improving force field description of metal ion interactions.
What problem does this paper attempt to address?