QM/MM studies of Ni–Fe hydrogenases: the effect of enzyme environment on the structure and energies of the inactive and active states

Abstract

The catalytically active (Ni–SI and Ni–R) and inactive states (Ni–A and Ni–B) of Ni–Fe hydrogenases have been studied using density functional theory (DFT) methods. Both isolated clusters and clusters embedded in the enzyme have been used to model the Ni–A, Ni–B, Ni–SI and Ni–R states. The BP86 and B3LYP functionals were employed, and hybrid quantum mechanical (QM)/molecular mechanical (MM) methods were used for the embedded calculations. The QM/MM studies, rather than the isolated cluster calculations, were generally found to give structures which correlated better with X-ray data. The structure of the unready state (Ni–A), was correctly predicted by the QM/MM, but not by the isolated cluster calculation. Comparison with the observed crystal structure favoured the catalytically active state, Ni–SI, to be the protonated (Ni–SI_II), rather than the unprotonated state (Ni–SI_I). In the QM/MM studies, the binding of H₂ to Ni–SI_II is preferred at the Ni (Ni–R(Ni)), rather than at the Fe centre (Ni–R(Fe)), in agreement with xenon binding studies, and in contrast to isolated cluster studies. These calculations cannot say with certainty which functional should be favoured, nor the preferred spin state of the catalytically active species. However, the lack of any predicted structure in which H₂ binds to the Fe centre, does favour a low spin state for Ni–SI_II, and the use of the BP86 functional. This is in agreement with recent high level ab initio calculations of a model of the Ni–SI_I state.