Include Dispersion in Quantum Chemical Modeling of Enzymatic Reactions: the Case of Isoaspartyl Dipeptidase.

Hai-Mei Zhang,Shi-Lu Chen
DOI: https://doi.org/10.1021/acs.jctc.5b00246
2015-01-01
Journal of Chemical Theory and Computation
Abstract:The lack of dispersion in the B3LYP functional has been proposed to be the main origin of big errors in quantum chemical modeling of a few enzymes and transition metal complexes. In this work, the essential dispersion effects that affect quantum chemical modeling are investigated. With binuclear zinc isoaspartyl dipeptidase (IAD) as an example, dispersion is included in the modeling of enzymatic reactions by two different procedures, i.e., (i) geometry optimizations followed by single-point calculations of dispersion (approach I) and (ii) the inclusion of dispersion throughout geometry optimization and energy evaluation (approach II). Based on a 169-atom chemical model, the calculations show a qualitative consistency between approaches I and II in energetics and most key geometries, demonstrating that both approaches are available with the latter preferential since both geometry and energy are dispersion-corrected in approach II. When a smaller model without Arg233 (147 atoms) was used, an inconsistency was observed, indicating that the missing dispersion interactions are essentially responsible for determining equilibrium geometries. Other technical issues and mechanistic characteristics of IAD are also discussed, in particular with respect to the effects of Arg233.
What problem does this paper attempt to address?