Electron Transfer in Nitrogenase

Abstract

Nitrogenase is the only enzyme capable of reducing N₂ to NH₃. This challenging reaction requires the coordinated transfer of multiple electrons from the reductase, Fe-protein, to the catalytic component, MoFe-protein, in an ATP-dependent fashion. In the last two decades, there have been significant advances in our understanding of how nitrogenase orchestrates electron transfer (ET) from the Fe-protein to the catalytic site of MoFe-protein and how energy from ATP hydrolysis transduces the ET processes. In this review, we summarize these advances, with focus on the structural and thermodynamic redox properties of nitrogenase component proteins and their complexes, as well as on new insights regarding the mechanism of ET reactions during catalysis and how they are coupled to ATP hydrolysis. We also discuss recently developed chemical, photochemical, and electrochemical methods for uncoupling substrate reduction from ATP hydrolysis, which may provide new avenues for studying the catalytic mechanism of nitrogenase.

Figure 1.

One half of the ADP.AlF₄⁻ stabilized nitrogenase complex (PDB ID: 1M34). FeP is shown in green, and MoFeP is red (α-subunit) and blue (β-subunit). Metalloclusters are depicted as spheres, and the nucleotides are red sticks.

Figure 2.

MgADP bound FeP (PDB ID: 6N4L). The Walker A (P-loop) motif is shown in red (γ9 - 16), Switch I region in magenta (γ39 - 69), and Switch II region in blue (γ125 - 132). MgADP is depicted is as sticks, and the [4Fe-4S] cluster as spheres colored by element.

Figure 3.

Dithionite reduced [4Fe-4S]¹⁺ cluster of FeP. The FeS cluster is colored by element (sulfides are yellow, irons are orange), and hydrogen-bonding networks are depicted by dashed gray lines. (a) FeP in the nucleotide-free state (PDB ID: 6N4K). (b) FeP in the MgADP-bound state (PDB ID: 6N4L). Reprinted in part with permission from ref . Copyright 2019 John Wiley and Sons.

Figure 4.

MoFeP (PDB ID: 3U7Q). MoFeP is a dimer of αβ-dimers (one dimer pictured is opaque and the other is transparent). The α- and β-subunits are shown in red and blue, respectively. Each αβ-dimer contains one P-cluster and one FeMoco, represented as spheres colored by element (sulfurs are yellow, irons are orange, molybdenum is teal, and carbon is black).

Figure 5.

Redox-dependent structural changes of the Av nitrogenase P-cluster. (a) The dithionite-reduced P-cluster (P^N) is ligated by six Cys residues, and the central S1 sulfide is coordinated by six Fe’s (PDB ID: 3MIN). (b) Upon one-electron oxidation, the P-cluster (P¹⁺) gains a Ser ligand βSer188) to Fe6, which dissociates from the S1 sulfide (PDB ID: 6CDK). (c) The two-electron, indigo disulfonate oxidized P-cluster (P²⁺) involves additional ligation of the backbone amide of the bridging αCys88 to Fe5, which also dissociates from S1 (PDB ID: 2MIN).

Figure 6.

Reductases of FeP in Av. (a) Av flavodoxin II (NifF) with the flavin mononucleotide (FMN) shown as orange sticks. Residues 56-60 are important in modulating the midpoint potential of FMN (purple loop), and residues 64-71 are hypothesized to be necessary for FeP binding specificity (green) (PDB ID: 5K9B). (b) Av ferredoxin I (FdI) contains a [3Fe-4S] and a [4Fe-4S] cluster (PDB ID: 6FDR).

Figure 7.

Nucleotide-dependent docking geometries (DGs) of the nitrogenase complex. FeP is shown in green, and MoFeP in red (α-subunit) and blue (β-subunit). Metalloclusters are shown as spheres colored by element, and nucleotides as red sticks. (a) The nucleotide-free FeP-MoFeP structure (nf) is in DG1, with FeP primarily in contact with the β-subunit of MoFeP (PDB ID: 2AFH). (b) MgAMPPCP-FeP-MoFeP complex (pcp) structure is in DG2, with FeP situated almost equally over both subunits MoFeP (PDB ID: 4WZB). (c) MgADP-FeP-MoFeP structure (adp) is in DG3 with FeP mostly in contact with the α-subunit of MoFeP (PDB ID: 2AFI).

Figure 8.

Nucleotide-dependent conformational changes in FeP. FeP and the γ100s helices from the alf structure (DG2) are shown in light gray and light green (PDB ID: 1M34) and from the adp structure (DG3) in dark gray and dark green (PDB ID: 2AFI). Nucleotides and the [4Fe-4S] clusters are shown as sticks and spheres, respectively. (a) Structural alignment of the left subunit of FeP from alf and adp demonstrating (top) hinge-like movement of the right subunit about a pivot point near the [4Fe-4S] cluster resulting in af having a flatter binding surface, as seen by the γ100s helices, and (bottom) depicting sliding motion of the subunits. (b) The flattening of the surface in alf poises residues γLys10 and γAsp129 across the subunit interface in a position favorable for ATP hydrolysis. (c) alf has a more surface-exposed [4Fe-4S] cluster than adp, as shown in FeP cross-sections.

Figure 9.

Overlay of aligned MoFeP structures observed in alf (light gray, PDB ID: 1M34) and adp (dark gray, PDB ID: 2AFI) complexes (RMSD over all α-C’s = 0.317 Å). The P-cluster and FeMoco are depicted as spheres.

Figure 10.

Nucleotide-dependent distances (center-to-center) between the P-cluster and the FeP [4Fe-4S] cluster in the nf pcp, alf and adp nitrogenase complex structures. Reprinted in part with permission from ref . Copyright 2005 AAAS.

Figure 11.

Asymmetric nucleotide binding in the nitrogenase complex (pcp/adp) (PDB ID: 4WZA). (a) FeP is green, and MoFeP is red (α-subunit) and blue (β-subunit). MgADP is located over the α-subunit and MgAMPPCP over the β-subunit. (b) The 2F_o–F_c electron density map (1.0 σ) around the nucleotides is shown as a black mesh. The nucleotides are colored by element, and Mg²⁺ ions are depicted as spheres.

Figure 12.

Docking models of FeP reductases with FeP in different nucleotide-bound states. (a) MgAMPPCP-FeP (top, PDB ID: 4WZB) and MgADP-FeP (bottom, PDB ID: 1FP6) bound to NifF (PDB ID: 1YOB). The MgADP-FeP-NifF structure from these simulations places the redox pair in closer proximity (6.4 Å) than the MgAMPPCP-FeP-NifF model (9.4 Å). Adapted from ref under the Creative Commons Attribution License (https://creativecommons.org/licenses/bync/3.0/lgalcode). Copyright 2017 ASBMB. (b) MgAMPPCP-FeP (PDB ID: 4WZB) bound to NifF (PDB ID:5K9B) and (c) FdI (PDB ID: 6FDR). Both NifF and FdxA in these models share the same binding surface on FeP. The MgAMPPCP-FeP-NifF model (b) from these simulations places the redox pair in closer proximity (~5 Å) than the model in (a). Both (b) and (c) are adapted with permission from ref . Copyright 2017 John Wiley and Sons.

Figure 13.

Shethna II is a homodimeric [2Fe-2S] ferredoxin. (a) Reduced Shethna II dimer (PDB ID: 5FFI) is in a closed conformation that likely cannot bind the nitrogenase complex. (b) Upon oxidation, Shethna II occupies a more open conformation that is thought to form a complex with FeP and MoFeP (PDB ID: 5FFI). (b) Docking model of Shethna II forming an oxygen-protected 1:1:1 complex with FeP and MoFeP. Shethna II binds the cleft between FeP and the a-subunit of MoFeP. (c) is adapted from ref .

Figure 14.

Thorneley-Lowe (TL) FeP cycle. (a) The original TL FeP cycle. Adapted from ref . (b) Adapted TL FeP cycle with proposed order of ET events, encounter complex, and transduction complex. Adapted from ref .

Figure 15.

Functionally relevant encounter complex mediated by electrostatic interactions. (a) In DG1 (PDB ID: 2AFH), EDC-crosslinkable residues γGlu112 and βLys400 are observed within H-binding distance (2.9 Å). (b) βLys400Glu MoFeP exhibits 28% less C₂H₂ reduction activity than wild-type MoFeP. (c) βLys400Glu MoFeP activity is more sensitive to [NaC1] than wild-type (WT) MoFeP. (b) The βLys400Glu MoFeP-FeP complex is more susceptible to iron chelation than the WT MoFeP-FeP complex. Panels (b), (c) and (d) are reprinted in part from ref .

Figure 16.

The updated FeP cycle. FeP is light blue and MoFeP is dark blue. ATP and ADP are denoted by T and D, respectively. FeP forms an ensemble of transient encounter complexes with MoFeP mediated by electrostatics before reaching the metastable DG1 complex. The DG2 complexes marked with * have been proposed, but not experimentally observed. Adapted with permission from ref . Copyright 2016 John Wiley and Sons.

Figure 17.

Evidence for conformational gating of electron transfer. (a) Stopped-flow experiments monitoring the oxidation of FeP at 430 nm. The rate of ET decreases from 160 s⁻¹ to 14 s⁻¹ as sucrose molality (m) increases from 0 to 2. (b) Logarithmic plot of rate of ET (k₂) vs m of different osmolytes. (a) and (b) are reprinted from ref .

Figure 18.

Evidence for the deficit spending model: stopped-flow experiments monitoring the oxidation of FeP by MoFeP βSer188Cys and wild-type) at 430 nm. ~65% of reduced βSer188Cys MoFeP is in the P¹⁺ state. (a) ATP-dependent oxidation of FeP by wild-type (WT) MoFeP (green) occurs at approximately the same rate as βSer188Cys MoFeP (blue), but unlike WT MoFeP, ~65% of FeP is oxidized by βSer188Cys MoFeP in the instrument dead time. No oxidation of FeP by βSer188Cys MoFeP occurs with MgADP (red) or without nucleotides (yellow). (b) AMPPCP enables oxidation of FeP by βSer188Cys MoFeP, unlike WT MoFeP. (c) Oxidation of FeP by βSer188Cys MoFeP occurs without FeMoco (apo), but not for WT MoFeP. Adapted from ref .

Figure 19.

Covariance of Ser and Tyr ligation to the oxidized P-cluster. Residue numbers correspond to Gluconocetabacter diazotrophicus (Gd). (a) The DT-reduced P-cluster of Gd, which contains alanine in place of serine at position β187 (β188 by Av numbering) (PDB ID: 5KOH). (b) IDS-oxidation of the P-cluster from Gd results in ligation of βTyr98 (residue β99 by Av numbering) to Fe8 and the backbone amide of αCys104 (residue α88 by Av numbering) to Fe6 (PDB ID: 5KOJ). (c) A sample of sequences demonstrating covariance of residues β99 and β188 (Av numbering) such that organisms have either a Tyr or a Ser in on of those positions, respectively. For a more complete list, see supplemental information of reference .

Figure 20.

Av MoFeP P-cluster primary coordination sphere mutants in the IDS-oxidized state. The anomalous electron density difference maps (near the Fe K-edge are shown in black mesh. (a) Av βSer188Ala MoFeP contains no oxygenic ligand. This mutant has two redox-labile iron centers (Fe1 and Fe5) whose positions are indicated with dashed circles. (PDB ID: 6O7S) (b) The Gd-like Av βPhe99Tyr/βSer188Ala MoFeP contains a tyrosine ligand. Upon oxidation, three irons (Fe1, Fe5 and Fe8, indicated with black circles) are partially occupied (~67% occupied) such that there is one redox-labile iron per P-cluster on average. (PDB ID: 6O7N)

Figure 21.

Evidence for ET preceding ATP hydrolysis. The time-course of ET (red), ATP hydrolysis (blue), P_i release (cyan), and FeP-MoFeP dissociation (olive). Reprinted with permission from ref . Copyright 2013 NAS.

Figure 22.

Light-induced catalysis by MoFeP with a Ru-photosensitizer. (a) The chemical structure of the [Ru(bpy)2(phen)]²⁺ photosensitizer. (b) Model of RuBP binding to aLeu158Cys in the cleft above the P-cluster. The α-subunit and β-subunit of MoFeP are black and gray, respectively. Reprinted from ref . (c) The Ru-labeled MoFeP proposed catalytic cycle. The P-cluster is depicted as cycling between P^N and P¹⁺, however it is also possible that it is cycling between P^N and a super-reduced state. FeMoco^N denotes the initial reduction state FeMoco. X indicates the number of ET events required for substrate reduction (X = 2, 2, and 6 for C2H2, H⁺, and HCN reduction, respectively). Adapted from ref . (d) Graph of total CH4 produced after illumination using the complete photoreduction system and using negative controls in which components were omitted. Adapted from ref . (e) CO inhibition of C2H2 reduction but not of H+ reduction. Reprinted with permission from ref . Copyright 2013 John Wiley and Sons.

Figure 23.

Light induced catalysis by MoFeP with CdS nanorods. (a) The proposed reaction scheme for the photoreduction of N₂ to NH₃ by CdS nanorods. (b) Production of NH₃ over time. Adapted with permission from ref . Copyright 2016 AAAS.

Figure 24.

Catalysis by MoFeP mutants using Eu^II reductants. (a) The time-course production of NH₃ from hydrazine by ATP- uncoupled catalysis. Negative controls (no protein in solution) are plotted as empty symbols. Reprinted from ref . (b) The crystal structure of βTyr98His demonstrates differences in the position of the solvent molecules around FeMoco that may contribute to the ability to ATP-uncoupled ET (wild-type PDB ID: 3U7Q, βTyr98His PDB: 4XPI). The native βTyr98 is depicted as blue sticks, and the mutation to His is depicted as gray sticks.

Figure 25.

Bioelectrocatalysis with MoFeP. (a) The proposed scheme of bioelectrocatalytic substrate reduction using the redox mediator cobaltocene. Adapted from ref under the Creative Commons Attribution License (https://creativecommons.org/licenses/by-nc/3.0/legalcode). (b) The proposed scheme of bioelectrocatalytic substrate reduction using LPEI-pyrene hydrogels. (c) Cyclic voltammogram of MoFeP embedded in LPEI-pyrene hydrogel after 5 min (red) and 10 min (black) N₂. Ar control shown as dashed line. (b) and (c) are adapted and reptrinted, respectively, from ref under the Creative Commons Attribution License (https://creativecommons.org/licenses/by-nc/3.0/legalcode).