Does the crystal structure of vanadium nitrogenase contain a reaction intermediate? Evidence from quantum refinement

On the Shoulders of Giants-Reaching for Nitrogenase

Only a single enzyme system-nitrogenase-carries out the conversion of atmospheric N2 into bioavailable ammonium, an essential prerequisite for all organismic life. The reduction of this inert substrate at ambient conditions poses unique catalytic challenges that strain our mechanistic understanding even after decades of intense research. Structural biology has added its part to this greater tapestry, and in this review, I provide a personal (and highly biased) summary of the parts of the story to which I had the privilege to contribute. It focuses on the crystallographic analysis of the three isoforms of nitrogenases at high resolution and the binding of ligands and inhibitors to the active-site cofactors of the enzyme. In conjunction with the wealth of available biochemical, biophysical, and spectroscopic data on the protein, this has led us to a mechanistic hypothesis based on an elementary mechanism of repetitive hydride formation and insertion.

Catalysis and structure of nitrogenases

In providing bioavailable nitrogen as building blocks for all classes of biomacromolecules, biological nitrogen fixation is an essential process for all organismic life. Only a single enzyme, nitrogenase, performs this task at ambient conditions and with ATP as an energy source. The assembly of the complex iron-sulfur enzyme nitrogenase and its catalytic mechanism remains a matter of intense study. Recent progress in the structural analysis of the three known isoforms of nitrogenase-differentiated primarily by the heterometal in their active site cofactor-has revealed a degree of structural plasticity of these clusters that suggest two distinct binding sites for substrates and reaction intermediates. A mechanistic proposal based on this finding integrates most of the available experimental data. Furthermore, the first applications of high-resolution cryo-electron microscopy have highlighted further dynamic conformational changes. Structures obtained under turnover conditions support the proposed alternating half-site reactivity in the C2-symmetric nitrogenase complex.

The binding of reducible N2 in the reaction domain of nitrogenase

The binding of N₂ to FeMo-co, the catalytic site of the enzyme nitrogenase, is central to the conversion to NH₃, but also has a separate role in promoting the N₂-dependent HD reaction (D₂ + 2H⁺ + 2e^- → 2HD). The protein surrounding FeMo-co contains a clear channel for ingress of N₂, directly towards the exo-coordination position of Fe2, a position which is outside the catalytic reaction domain. This led to the hypothesis [I. Dance, Dalton Trans., 2022, 51, 12717] of 'promotional' N₂ bound at exo-Fe2, and a second 'reducible' N₂ bound in the reaction domain, specifically the endo-coordination position of Fe2 or Fe6. The range of possibilities for the binding of reducible N₂ in the presence of bound promotional N₂ is described here, using density functional simulations with a 486 atom model of the active site and surrounding protein. The pathway for ingress of the second N₂ through protein, past the first N₂ at exo-Fe2, and tumbling into the binding domain between Fe2 and Fe6, is described. The calculations explore 24 structures involving 6 different forms of hydrogenated FeMo-co, including structures with S2BH unhooked from Fe2 but tethered to Fe6. The calculations use the most probable electronic states. End-on (η¹) binding of N₂ at the endo position of either Fe2 or Fe6 is almost invariably exothermic, with binding potential energies ranging up to -18 kcal mol^-1. Many structures have binding energies in the range -6 to -14 kcal mol^-1. The relevant entropic penalty for N₂ binding from a diffusible position within the protein is estimated to be 4 kcal mol^-1, and so the binding free energies for reducible N₂ are suitably negative. N₂ binding at endo-Fe2 is stronger than at endo-Fe6 in three of the six structure categories. In many cases the reaction domain containing reducible N₂ is expanded. These results inform computational simulation of the subsequent steps in which surrounding H atoms transfer to reducible N₂.

The HD Reaction of Nitrogenase: a Detailed Mechanism

Nitrogenase is the enzyme that converts N₂ to NH₃ under ambient conditions. The chemical mechanism of this catalysis at the active site FeMo-co [Fe₇ S₉ CMo(homocitrate)] is unknown. An obligatory co-product is H₂ , while exogenous H₂ is a competitive inhibitor. Isotopic substitution using exogenous D₂ revealed the N₂ -dependent reaction D₂ +2H⁺ +2e^- →2HD (the 'HD reaction'), together with a collection of additional experimental characteristics and requirements. This paper describes a detailed mechanism for the HD reaction, developed and elaborated using density functional simulations with a 486-atom model of the active site and surrounding protein. First D₂ binds at one Fe atom (endo-Fe6 coordination position), where it is flanked by H-Fe6 (exo position) and H-Fe2 (endo position). Then there is synchronous transfer of these two H atoms to bound D₂ , forming one HD bound to Fe2 and a second HD bound to Fe6. These two HD dissociate sequentially. The final phase is recovery of the two flanking H atoms. These H atoms are generated, sequentially, by translocation of a proton from the protein surface to S3B of FeMo-co and combination with introduced electrons. The first H atom migrates from S3B to exo-Fe6 and the second from S3B to endo-Fe2. Reaction energies and kinetic barriers are reported for all steps. This mechanism accounts for the experimental data: (a) stoichiometry; (b) the N₂ -dependence results from promotional N₂ bound at exo-Fe2; (c) different N₂ binding K_m for the HD reaction and the NH₃ formation reaction results from involvement of two different sites; (d) inhibition by CO; (e) the non-occurrence of 2HD→H₂ +D₂ results from the synchronicity of the two transfers of H to D₂ ; (f) inhibition of HD production at high pN₂ is by competitive binding of N₂ at endo-Fe6; (g) the non-leakage of D to solvent follows from the hydrophobic environment and irreversibility of proton introduction.

Understanding the tethered unhooking and rehooking of S2B in the reaction domain of FeMo-co, the active site of nitrogenase

The active site of the nitrogen fixing enzyme nitrogenase is an Fe₇MoS₉C cluster, and investigations of the enigmatic chemical mechanism of the enzyme have focussed on a pair of Fe atoms, Fe2 and Fe6, and the S2B atom that bridges them. There are three proposals for the status of the Fe2-S2B-Fe6 bridge during the catalytic cycle: one that it remains intact, another that it is completely labile and absent during catalysis, and a third that S2B is hemilabile, unhooking one of its bonds to Fe2 or Fe6. This report examines the tethered unhooking of S2B and factors that affect it, using DFT calculations of 50 geometric/electronic possibilities with a 485 atom model including all relevant parts of surrounding protein. The outcomes are: (a) unhooking the S2B-Fe2 bond is feasible and favourable, but alternative unhooking of the S2B-Fe6 bond is unlikely for steric reasons, (b) energy differences between hooked and unhooked isomers are generally <10 kcal mol^-1, usually with unhooked more stable, (c) ligation at the exo-Fe6 position inhibits unhooking, (d) unhooking of hydrogenated S2B is more favourable than that of bare S2B, (e) hydrogen bonding from the NεH function of His195 to S2B occurs in hooked and unhooked forms, and possibly stabilises unhooking, (f) unhooking is reversible with kinetic barriers ranging 10-13 kcal mol^-1. The conclusion is that energetically accessible reversible unhooking of S2B or S2BH, as an intrinsic property of FeMo-co, needs to be considered in the formulation of mechanisms for the reactions of nitrogenase.

Calculating the chemical mechanism of nitrogenase: new working hypotheses

The enzyme nitrogenase converts N₂ to NH₃ with stoichiometry N₂ + 8H⁺ + 8e^- → 2NH₃ + H₂. The mechanism is chemically complex with multiple steps that must be consistent with much accumulated experimental information, including exchange of H₂ and N₂ and the N₂-dependent hydrogenation of D₂ to HD. Previous investigations have developed a collection of working hypotheses that guide ongoing density functional investigations of mechanistic steps and sequences. These include (i) hypotheses about the serial provision of protons and their conversion to H atoms bonded to S and Fe atoms of the FeMo-co catalytic site, (ii) the migration of H atoms over the surface of FeMo-co, (iii) the roles of His195, (iv) identification of three protein channels, one for the ingress of N₂, a separate pathway for the passage of exogenous H₂ (D₂) and product H₂ (HD), and a hydrophilic pathway for egress of product NH₃. Two additional working hypotheses are described in this paper. N₂ passing along the N₂ channel approaches and binds end-on to the exo coordination position of Fe2, with favourable energetics when FeMo-co is pre-hydrogenated. This exo-Fe2-N₂ is apparently not reduced but has a promotional role by expanding the reaction zone. A second N₂ can enter via the N₂ ingress channel and bind at the endo-Fe6 position, where it is surrounded by H atom donors suitable for the N₂ → NH₃ conversion. It is proposed that this endo-Fe6 position is also the binding site for H₂ (generated or exogenous), accounting for the competitive inhibition of N₂ reduction by H₂. The HD reaction occurs at the endo-Fe6 site, promoted by N₂ at the exo-Fe2 site. The second hypothesis concerns the most stable electronic states of FeMo-co with ligands bound at Fe2 and Fe6, and provides a protocol for management of electronic states in mechanism calculations.

Computationally enhanced X-ray diffraction analysis of a gold(III) complex interacting with the human telomeric DNA G-quadruplex. Unravelling non-unique ligand positioning

The crystal structure of the human telomeric DNA Tel24 G-quadruplex (Tel24 = TAG₃(T₂AG₃)₃T) in complex with the novel [AuL] species (with L = 2,4,6-tris(2-pyrimidyl)-1,3,5-triazine - TPymT-α) was solved by a novel joint molecular mechanical (MM)/quantum mechanical (QM) innovative approach. The quantum-refinement crystallographic method (crystallographic refinement enhanced with quantum mechanical calculation) was adapted to treat the [AuL]/G-quadruplex structure, where each gold complex in the binding site was found spread over four equally occupied positions. The four positions were first determined by docking restrained to the crystallographically determined metal ions' coordinates. Then, the quantum refinement method was used to resolve the poorly defined density around the ligands and improve the crystallographic determination, revealing that the binding preferences of this metallodrug toward Tel24 G-quadruplex arise from a combined effect of pyrimidine stacking, metal-guanine interactions and charge-charge neutralizing action of the π-acid triazine. The occurrence of interaction in solution with the Tel24 G-quadruplex DNA was further proved through DNA melting experiments, which showed a slight destabilisation of the quadruplex upon adduct formation.

Thermodynamically Favourable States in the Reaction of Nitrogenase without Dissociation of any Sulfide Ligand

We have used combined quantum mechanical and molecular mechanical (QM/MM) calculations to study the reaction mechanism of nitrogenase, assuming that none of the sulfide ligands dissociates. To avoid the problem that there is no consensus regarding the structure and protonation of the E₄ state, we start from a state where N₂ is bound to the cluster and is protonated to N₂H₂, after dissociation of H₂. We show that the reaction follows an alternating mechanism with HNNH (possibly protonated to HNNH₂) and H₂NNH₂ as intermediates and the two NH₃ products dissociate at the E₇ and E₈ levels. For all intermediates, coordination to Fe6 is preferred, but for the E₄ and E₈ intermediates, binding to Fe2 is competitive. For the E₄, E₅ and E₇ intermediates we find that the substrate may abstract a proton from the hydroxy group of the homocitrate ligand of the FeMo cluster, thereby forming HNNH₂, H₂NNH₂ and NH₃ intermediates. This may explain why homocitrate is a mandatory component of nitrogenase. All steps in the suggested reaction mechanism are thermodynamically favourable compared to protonation of the nearby His‐195 group and in all cases, protonation of the NE2 atom of the latter group is preferred.

A Mechanism for Nitrogenase Including Loss of a Sulfide

Nitrogenase is the only enzyme in nature that can fix N₂ from the air. The active cofactor of the leading form of this enzyme contains seven irons and one molybdenum connected by sulfide bridges. In several recent experimental studies, it has been suggested that the cofactor is very flexible, and might lose one of its sulfides during catalysis. In this study, the possible loss of a sulfide has been investigated by model calculations. In previous studies, we have shown that there should be four activation steps before catalysis starts, and this study is based on that finding. It was found here that, after the four reductions in the activation steps, a sulfide will become very loosely bound and can be released in a quite exergonic step with a low barrier. The binding of N₂ has no part in that release. In our previous studies, we suggested that the central carbide should be protonated three times after the four activation steps. With the new finding, there will instead be a loss of a sulfide, as the barrier for the loss is much lower than the ones for protonating the carbide. Still, it is suggested here that the carbide will be protonated anyway, but only with one proton, in the E₃ to E₄ step. A very complicated transition state for H₂ formation involving a large structural change was obtained. The combined step, with a loss of H₂ and binding of N₂ , is calculated to be endergonic by +2.3 kcal mol^-1 ; this is in excellent agreement with experiments in which an easily reversible step has been found.

Structures and reaction dynamics of N2 and H2 binding at FeMo-co, the active site of nitrogenase

The chemical reactions occurring at the Fe₇MoS₉C(homocitrate) cluster, FeMo-co, the active site of the enzyme nitrogenase (N₂ → NH₃), are enigmatic. Experimental information collected over a long period reveals aspects of the roles of N₂ and H₂, each with more than one type of reactivity. This paper reports investigations of the binding of H₂ and N₂ at intact FeMo-co, using density functional simulations of a large 486 atom relevant portion of the protein, resulting in 27 new structures containing H₂ and/or N₂ bound at the exo and endo coordination sites of the participating Fe atoms, Fe2 and Fe6. Binding energies and transition states for association/dissociation are determined, and trajectories for the approach, binding and separation of H₂/N₂ are described, including diffusion of these small molecules through proximal protein. Influences of surrounding amino acids are identified. FeMo-co deforms geometrically when binding H₂ or N₂, and a procedure for calculating the energy cost involved, the adaptation energy, is introduced here. Adaptation energies, which range from 7 to 36 kcal mol^-1 for the reported structures, are influenced by the protonation state of the His195 side chain. Seven N₂ structures and three H₂ structures have negative binding free energies, which include the estimated entropy penalties for binding of N₂, H₂ from proximal protein. These favoured structures have N₂ bound end-on at exo-Fe2, exo-Fe6 and endo-Fe2 positions of FeMo-co, and H₂ bound at the endo-Fe2 position. Various postulated structures with N₂ bridging Fe2 and Fe6 revert to end-on-N₂ at endo positions. The structures are also assessed via the calculated potential energy barriers for association and dissociation. Barriers to the binding of H₂ range from 1 to 20 kcal mol^-1 and barriers to dissociation of H₂ range from 3 to 18 kcal mol^-1. Barriers to the binding of N₂, in either side-on or end-on mode, range from 2 to 18 kcal mol^-1, while dissociation of bound N₂ encounters barriers of 3 to 8 kcal mol^-1 for side-on bonding and 7 to 18 kcal mol^-1 for end-on bonding. These results allow formulation of mechanisms for the H₂/N₂ exchange reaction, and three feasible mechanisms for associative exchange and three for dissociative exchange are identified. Consistent electronic structures and potential energy surfaces are maintained throughout. Changes in the spin populations of Fe2 and Fe6 connected with cluster deformation and with metal-ligand bond formation are identified.

Combining crystallography with quantum mechanics

In standard crystallographic refinement of biomacromolecules, the crystallographic raw data are supplemented by empirical restraints that ensure that the structure makes chemical sense. These restraints are typically accurate for amino acids and nucleic acids, but less so for cofactors, substrates, inhibitors, ligands and metal sites. In quantum refinement, this potential is replaced by more accurate quantum mechanical (QM) calculations. Several implementations have been presented, differing in the level of QM and whether it is used for the entire structure or only for a site of particular interest. It has been shown that the method can improve and correct errors in crystal structures and that it can be used to determine protonation and tautomeric states of various ligands and to decide what is really seen in the structure by refining different interpretations and using standard crystallographic and QM quality measures to decide which fits the structure best.

Critical evaluation of a crystal structure of nitrogenase with bound N2 ligands

Recently, a 1.83 Å crystallographic structure of nitrogenase was suggested to show N₂-derived ligands at three sites in the catalytic FeMo cluster, replacing the three [Formula: see text] bridging sulfide ligands (two in one subunit and the third in the other subunit) (Kang et al. in Science 368: 1381-1385, 2020). Naturally, such a structure is sensational, having strong bearings on the reaction mechanism of the enzyme. Therefore, it is highly important to ensure that the interpretation of the structure is correct. Here, we use standard crystallographic refinement and quantum refinement to evaluate the structure. We show that the original crystallographic raw data are strongly anisotropic, with a much lower resolution in certain directions than others. This, together with the questionable use of anisotropic B factors, give atoms an elongated shape, which may look like diatomic atoms. In terms of standard electron-density maps and real-space Z scores, a resting-state structure with no dissociated sulfide ligands fits the raw data better than the interpretation suggested by the crystallographers. The anomalous electron density at 7100 eV is weaker for the putative N₂ ligands, but not lower than for several of the [Formula: see text] bridging sulfide ions and not lower than what can be expected from a statistical analysis of the densities. Therefore, we find no convincing evidence for any N₂ binding to the FeMo cluster. Instead, a standard resting state without any dissociated ligands seems to be the most likely interpretation of the structure. Likewise, we find no support that the homocitrate ligand should show monodentate binding.

Response to Comment on "Structural evidence for a dynamic metallocofactor during N2 reduction by Mo-nitrogenase"

Peters et al comment on our report of the dynamic structure of the nitrogenase metallocofactor during N₂ reduction. Their claim that their independent structural refinement and consideration of biochemical data contradict our finding is incorrect and is strongly refuted by our biochemical and structural data that collectively and conclusively point to the binding of dinitrogen species to the nitrogenase cofactor.

Quantum refinement with multiple conformations: application to the P-cluster in nitrogenase

X-ray crystallography is the main source of atomistic information on the structure of proteins. Normal crystal structures are obtained as a compromise between the X-ray scattering data and a set of empirical restraints that ensure chemically reasonable bond lengths and angles. However, such restraints are not always available or accurate for nonstandard parts of the structure, for example substrates, inhibitors and metal sites. The method of quantum refinement, in which these empirical restraints are replaced by quantum-mechanical (QM) calculations, has previously been suggested for small but interesting parts of the protein. Here, this approach is extended to allow for multiple conformations in the QM region by performing separate QM calculations for each conformation. This approach is shown to work properly and leads to improved structures in terms of electron-density maps and real-space difference density Z-scores. It is also shown that the quality of the structures can be gauged using QM strain energies. The approach, called ComQumX-2QM, is applied to the P-cluster in two different crystal structures of the enzyme nitrogenase, i.e. an Fe8S7Cys6 cluster, used for electron transfer. One structure is at a very high resolution (1.0 Å) and shows a mixture of two different oxidation states, the fully reduced PN state (Fe82+, 20%) and the doubly oxidized P2+ state (80%). In the original crystal structure the coordinates differed for only two iron ions, but here it is shown that the two states also show differences in other atoms of up to 0.7 Å. The second structure is at a more modest resolution, 2.1 Å, and was originally suggested to show only the one-electron oxidized state, P1+. Here, it is shown that it is rather a 50/50% mixture of the P1+ and P2+ states and that many of the Fe-Fe and Fe-S distances in the original structure were quite inaccurate (by up to 0.8 Å). This shows that the new ComQumX-2QM approach can be used to sort out what is actually seen in crystal structures with dual conformations and to give locally improved coordinates.