Quantum Chemical Modeling Of Reaction Mechanism Of Isoaspartyl Dipeptidase

The lack of dispersion in the B3LYP functional has been proposed to be the mainorigin of big errors in quantum chemical modeling of a few enzymes and transition metalcomplexes. In this thesis, dispersion is included in the modeling of enzymatic reactions bytwo different procedures, i.e (i) geometry optimizations followed by single-pointcalculations of dispersion effects (approach I) and (ii) the inclusion of dispersionthroughout geometry optimization and energy evaluation (approach II). As an example, themechanism of binuclear zinc isoaspartyl dipeptidase (IAD) was investigated respectively byapproaches I and II, on the basis of a169-atom chemical model built from the X-ray crystalstructure. The calculations show an excellent consistency between approaches I and II, inboth key geometries and energetics. This not only demonstrates that both approaches areavailable for quantum chemical modeling of enzymatic reactions, but also gives a strongsupport to the previous results obtained by approach I. When a smaller model withoutArg233(147atoms) was used, an inconsistency between approaches I and II was observed,indicating that the construction of an accurate chemical model is to a great extent moreimportant than the choice between approaches I and II. The achievement of a consistencybetween approaches I and II is thus proposed to be a new criteria utilized to validate a builtactive-site model. The mechanistic characteristics of IAD are also discussed in detail, inparticular with respect to the effects of Arg233.