Modeling the active sites in metalloenzymes. 3. Density functional calculations on models for [Fe]-hydrogenase: structures and vibrational frequencies of the observed redox forms and the reaction mechanism at the Diiron Active Center.

Optimized structures for the redox species of the diiron active site in [Fe]-hydrogenase as observed by FTIR and for species in the catalytic cycle for the reversible H(2) oxidation have been determined by density-functional calculations on the active site model, [(L)(CO)(CN)Fe(mu-PDT)(mu-CO)Fe(CO)(CN)(L')](q)(L = H(2)O, CO, H(2), H(-); PDT = SCH(2)CH(2)CH(2)S, L' = CH(3)S(-), CH(3)SH; q = 0, 1-, 2-, 3-). Analytical DFT frequencies on model complexes (mu-PDT)Fe(2)(CO)(6) and [(mu-PDT)Fe(2)(CO)(4)(CN)(2)](2)(-) are used to calibrate the calculated CN(-) and CO frequencies against the measured FTIR bands in these model compounds. By comparing the predicted CN(-) and CO frequencies from DFT frequency calculations on the active site model with the observed bands of D. vulgaris [Fe]-hydrogenase under various conditions, the oxidation states and structures for the diiron active site are proposed. The fully oxidized, EPR-silent form is an Fe(II)-Fe(II) species. Coordination of H(2)O to the empty site in the enzyme's diiron active center results in an oxidized inactive form (H(2)O)Fe(II)-Fe(II). The calculations show that reduction of this inactive form releases the H(2)O to provide an open coordination site for H(2). The partially oxidized active state, which has an S = (1)/(2) EPR signal, is an Fe(I)-Fe(II) species. Fe(I)-Fe(I) species with and without bridging CO account for the fully reduced, EPR-silent state. For this fully reduced state, the species without the bridging CO is slightly more stable than the structure with the bridging CO. The correlation coefficient between the predicted CN(-) and CO frequencies for the proposed model species and the measured CN(-) and CO frequencies in the enzyme is 0.964. The proposed species are also consistent with the EPR, ENDOR, and Mössbauer spectroscopies for the enzyme states. Our results preclude the presence of Fe(III)-Fe(II) or Fe(III)-Fe(III) states among those observed by FTIR. A proposed reaction mechanism (catalytic cycle) based on the DFT calculations shows that heterolytic cleavage of H(2) can occur from (eta(2)-H(2))Fe(II)-Fe(II) via a proton transfer to "spectator" ligands. Proton transfer to a CN(-) ligand is thermodynamically favored but kinetically unfavorable over proton transfer to the bridging S of the PDT. Proton migration from a metal hydride to a base (S, CN, or basic protein site) results in a two-electron reduction at the metals and explains in part the active site's dimetal requirement and ligand framework which supports low-oxidation-state metals. The calculations also suggest that species with a protonated Fe-Fe bond could be involved if the protein could accommodate such species.