Structural evolution of nitrogenase states under alkaline turnover

Abstract

Biological nitrogen fixation, performed by the enzyme nitrogenase, supplies nearly 50% of the bioavailable nitrogen pool on Earth, yet the structural nature of the enzyme intermediates involved in this cycle remains ambiguous. Here we present four high resolution cryoEM structures of the nitrogenase MoFe-protein, sampled along a time course of alkaline reaction mixtures under an acetylene atmosphere. This series of structures reveals a sequence of salient changes including perturbations to the inorganic framework of the FeMo-cofactor; depletion of the homocitrate moiety; diminished density around the S2B belt sulfur of the FeMo-cofactor; rearrangements of cluster-adjacent side chains; and the asymmetric displacement of the FeMo-cofactor. We further demonstrate that the nitrogenase associated factor T protein can recognize and bind an alkaline inactivated MoFe-protein in vitro. These time-resolved structures provide experimental support for the displacement of S2B and distortions of the FeMo-cofactor at the E₀-E₃ intermediates of the substrate reduction mechanism, prior to nitrogen binding, highlighting cluster rearrangements potentially relevant to nitrogen fixation by biological and synthetic clusters.

Fig. 1. Time sequence cryoEM structures of the nitrogenase MoFe-protein under alkaline turnover.

A CryoEM density maps of MoFe-protein from acetylene reaction mixtures at pH 9.5 at 20 s (yellow; MoFe^{Alkaline-20sec}), 5 min (green; MoFe^{Alkaline-5min}), 20 min (blue; MoFe^{Alkaline-20min}), and 60 min (purple; MoFe^{Alkaline-60min}). The coloring convention of each time point will be maintained throughout the manuscript. B Structural models of the adjacent cryoEM maps. α- and β-subunits are highlighted in the figure panel. C Left axis: Specific activity assays of acetylene turnover to ethylene under alkaline conditions (solid symbols; purple line is fit according to a two phase exponential decay model for illustrative purposes only; n = 2 technical replicates). Right axis: Total ethylene formation assays under alkaline (pH 9.5) conditions (striped symbols; green curve is fit according to equation 1 in Yang et al.; n = 3 technical replicates). Assay conditions contained ~1 nmol of MoFe-protein, or ~2 nmol of active site. Source data are provided as a Source Data file. D CryoEM density for the ordered active site FeMo-cofactor in the MoFe^{Alkaline-5min} time point. E CryoEM density for the P-cluster in the MoFe^{Alkaline-5min} time point. For D and E, atoms are colored according to the following scheme: green (carbon); red (oxygen); blue (nitrogen); yellow (sulfur); orange (iron); teal (molybdenum).

Fig. 2. Time-dependent sulfur S2B depletion in the ordered MoFe-protein active sites.

A Overlay of cryoEM density (12 σ) surrounding the ordered FeMo-cofactor and homocitrate (HCA) for the MoFe^{Alkaline-20sec} (yellow), MoFe^{Alkaline-5min} (green), and MoFe^{Alkaline-20min} (blue) time points. The belt sulfurs of FeMo-cofactor S3A, S5A, and S2B are labeled. B–D Structures of the MoFe^{Alkaline-20sec}, MoFe^{Alkaline-5min}, and MoFe^{Alkaline-20min} times points, which each contain one ordered active site (boxed). E–G Volumes of cryoEM density around the FeMo-cofactor belt sulfur atoms were calculated at 15 σ within a 1.25 Å spherical radius; values are shown in boxes adjacent to the corresponding belt sulfur.

Fig. 3. Time-dependent FeMo-cofactor rearrangement in the disordered MoFe-protein active sites.

A Size exclusion chromatography (S.E.C.) of reaction mixtures stopped at the designated time points with addition of excess MgADP. Dotted lines indicate the peak elution time for the MoFe-protein from each reaction mixture: 65.8 min (20 sec; yellow), 65.6 min (5 min; green), 65.2 min (20 min; blue), 64.9 min (60 min; purple). Source data are provided as a Source Data file. B–E Structure models of the MoFe^{Alkaline-20sec} (yellow), MoFe^{Alkaline-5min} (green), MoFe^{Alkaline-20min} (blue), and MoFe^{Alkaline-60min} (purple) time points with disordered active sites indicated with boxes. These panels are rotated 180° relative to the orientation depicted in Fig. 2B–D. F–J CryoEM density surrounding the disordered FeMo-cofactor for the MoFe^{Alkaline-20sec}, MoFe^{Alkaline-5min}, MoFe^{Alkaline-20min}, and MoFe^{Alkaline-60min} time points. The belt sulfurs of FeMo-cofactor S3A, S5A, and S2B are labeled.

Fig. 4. Rearrangements in the α-Gln191 side chain.

A Previously determined crystal structure of the VFe nitrogenase for reference. Residue α-Q176 in the VFe nitrogenase is the equivalent of α-Q191 in the MoFe nitrogenase. O/N oxygen or nitrogen, H₂S hydrogen sulfide, CO₃²⁻ carbonate. B–E CryoEM density surrounding the α-Q191 residue within the active sites of the alkaline turnover structures. Top panels represent ordered active sites, bottom panels represent disordered active sites.

Fig. 5. Structure of a turnover-dependent cofactor-displaced MoFe-protein state.

A–D Left panel: CryoEM density maps of the MoFe^{Alkaline-20sec} (yellow), MoFe^{Alkaline-5min} (green), MoFe^{Alkaline-20min} (blue), and MoFe^{Alkaline-60min} (purple) time points with C2 symmetry imposed; resolution is indicated in Ångstroms. Right panel: After symmetry expansion, three dimensional variability analysis (3DVA) was carried out and clustered into 5 groups, shown from left to right in order of decreasing number of particles. The boxed groups B and D of the 5 min time point correspond to (E–G). E Groups B and D of the MoFe^{Alkaline-5min} time point were subjected to local refinement with a mask covering the disordered dimer (top panel) resulting in 1.97 Å and 2.13 Å resolution maps, respectively (middle panel). Densities for the FeMo-cofactor for each map are highlighted in the bottom panel which is related to the middle panel by ~45° rotations around both the X and Y axes. F CryoEM density for residues in the loop housing α-Cys275 and the FeMo-cofactor in the α-subunit of the Group D (Cofactor in) map. Residues capping the active site (355 to 358) are highlighted in yellow. G CryoEM density for residues in the loop housing α-Cys275 and the FeMo-cofactor in the α-subunit of the Group B (Cofactor out) map.

Fig. 6. Structural evolution of the nitrogenase active site under alkaline turnover.

Changes observed within the MoFe-protein FeMo-cofactor in cryoEM time points are diagrammed and shown from left to right as a function of time. Upper panel: Percent activity is calculated as percent remaining specific activity of acetylene reduction to ethylene at pH 7.8 from the data shown in Fig. 1C. Lower panel: Proceeding from the resting state (E₀), the first observed changes are depletions in the density surrounding the belt sulfur S2B (over the 20 sec to 20 min time points in the ordered active sites), which could represent positional displacement, or replacement, of this atom. Subsequently, depletions in the density at the HCA site and cofactor distortions occur at first asymmetrically between the αβ-dimers (in the 20 sec through 20 min time points) and then symmetrically (60 min time point). Within the 5 min time point, two distinct subclasses are seen either with the cofactor density displaced towards the MoFe-protein surface lacking identifiable HCA density, or in the resting state position with HCA density present. In the final time point, both cofactors occupy the resting state position, but lack identifiable HCA density and the α-His442 loop has rearranged so that the α-His442 participates in a Histidine quartet with residues α-His274, α-His362, and α-His451, while α-Trp444 instead lies adjacent to the Mo atom of the cofactor. Dashed lines between Fe7 and adjacent atoms indicate flexibility of these bonds. Color coding is as follows: Blue (cartoon representation of changing cryoEM density); yellow (sulfur atoms); orange (iron atoms); teal (molybdenum atoms); gray (carbon atoms).