A model for dinitrogen binding in the E4 state of nitrogenase

Molybdenum nitrogenase is one of the most intriguing metalloenzymes in nature, featuring an exotic iron-molybdenum-sulfur cofactor, FeMoco, whose mode of action remains elusive. In particular, the molecular and electronic structure of the N-2-binding E-4 state is not known. In this study we present theoretical QM/MM calculations of new structural models of the E-4 state of molybdenum-dependent nitrogenase and compare to previously suggested models for this enigmatic redox state. We propose two models as possible candidates for the E-4 state. Both models feature two hydrides on the FeMo cofactor, bridging atoms Fe-2 and Fe-6 with a terminal sulfhydryl group on either Fe-2 or Fe-6 (derived from the S2B bridge) and the change in coordination results in local lower-spin electronic structure at Fe-2 and Fe-6. These structures appear consistent with the bridging hydride proposal put forward from ENDOR studies and are calculated to be lower in energy than other proposed models for E-4 at the TPSSh-QM/MM level of theory. We critically analyze the DFT method dependency in calculations of FeMoco that has resulted in strikingly different proposals for this state. Importantly, dinitrogen binds exothermically to either Fe-2 or Fe-6 in our models, contrary to others, an effect rationalized via the unique ligand field (from the hydrides) at the Fe with an empty coordination site. A low-spin Fe site is proposed as being important to N-2 binding. Furthermore, the geometries of these states suggest a feasible reductive elimination step that could follow, as experiments indicate. Via this step, two electrons are released, reducing the cofactor to yield a distorted 4-coordinate Fe-2 or Fe-6 that partially activates N-2. We speculate that stabilization of an N-2-bound Fe(i) at Fe-6 (not found for Fe-2 model) via reductive elimination is a crucial part of N-2 activation in nitrogenases, possibly aided by the apical heterometal ion (Mo or V). By using protons from the sulfhydryl group (to regenerate the sulfide bridge between Fe-2 and Fe-6) and the nearby homocitrate hydroxy group, we calculate a plausible route to yield a diazene intermediate. This is found to be more favorable with the Fe-6-bound model than the Fe-2-bound model; however, this protonation is uphill in energy, suggesting protonation of N-2 might occur later in the catalytic cycle or via another mechanism.