P⁺ State of Nitrogenase P-Cluster Exhibits Electronic Structure of a [Fe₄S₄]⁺ Cluster

Abstract

Mo nitrogenase consists of two component proteins: the Fe protein, which contains a [Fe₄S₄] cluster, and the MoFe protein, which contains two different classes of metal cluster: P-cluster ([Fe₈S₇]) and FeMoco ([Mo-Fe₇S₉C•Homocitrate]). The P-cluster is believed to mediate the electron transfer between the Fe protein and the MoFe protein via inter-conversions between its various oxidation states, such as the all-ferrous state (P^N) and the one (P⁺)- and two (P²⁺)-electron oxidized states. While the structural and electronic properties of P^N and P²⁺ states have been well characterized, little is known about the electronic structure of the P⁺ state. Here, a mutant strain of Azotobacter vinelandii (DJ1193) was used to facilitate the characterization of the P⁺ state of P-cluster. This strain expresses a MoFe protein variant (designated ΔnifB β-188^Cys MoFe protein) that accumulates the P⁺ form of P-cluster in the resting state. MCD spectrum of the P-cluster in the oxidized ΔnifB β-188^Cys MoFe protein closely resembles that of the P²⁺ state in the oxidized wild-type MoFe protein, except for the absence of a major charge-transfer band centered at 823 nm. Moreover, magnetization curves of ΔnifB β-188^Cys and wild-type MoFe proteins suggest that the P²⁺ species in both proteins have the same spin state. MCD spectrum of the P⁺ state in the ΔnifB β-188^Cys MoFe protein, on the other hand, is associated with a classic [Fe₄S₄]⁺ cluster, suggesting that the P-cluster could be viewed as two coupled 4Fe clusters and that it could donate either one or two electrons to FeMoco by using one or both of its 4Fe halves. Such a mode of action of P-cluster could provide energetic and kinetic advantages to nitrogenase in the complex mechanism of N₂ reduction.

1. INTRODUCTION

Nitrogenase catalyzes the reduction of dinitrogen to ammonia. Mo nitrogenase, the most common form of this enzyme, consists of two protein components: the Fe protein and the MoFe protein. Both proteins are required for the enzymatic activity of nitrogenase, where the Fe protein serves as the obligate reductant of the MoFe protein.¹ The Fe protein has a γ₂ structure, and it contains a subunit-bridging [Fe₄S₄] cluster and one MgATP binding site in each subunit. The MoFe protein has an α₂β₂ structure, and it contains two metal clusters in each αβ-dimer: the P-cluster ([Fe₈S₇]) and the FeMo cofactor (or FeMoco, [MoFe₇S₉C•homocitrate]).^2–4 The P-cluster is bound at the α/β-subunit interface by six cysteine residues, three from each subunit;⁵ whereas the FeMoco resides in the α-subunit of MoFe protein.⁶ During catalysis, the Fe protein and the MoFe protein undergo repeated association/dissociation cycles, where electrons are transferred from the former to the latter concomitant with ATP hydrolysis, and substrate reduction eventually takes place at the active FeMoco site upon accumulation of a sufficient amount of electrons.

The P-cluster is believed to mediate the transfer of electrons from the [Fe₄S₄] cluster of the Fe protein to the FeMoco of the MoFe protein during nitrogenase catalysis.¹ In the resting state, the P-cluster exists in an all-ferrous form (P^N).⁷ It can also exist in three other stable, oxidized states (P⁺, P²⁺ and P³⁺).⁸ Formation of the P-cluster was investigated through the characterization of a so-called ΔnifH MoFe protein.⁹ Expressed in a genetic background where the nifH gene (encoding Fe protein) is deleted, the ΔnifH MoFe protein contains two pairs of [Fe₄S₄]-like clusters, one at each α/β-subunit interface that is normally occupied by the [Fe₈S₇] P-cluster. These [Fe₄S₄]-like cluster pairs can be reductively coupled into two mature P-clusters in a reaction involving the Fe protein and MgATP, suggesting that they are indeed the precursors to P-clusters.⁹

Two questions have arisen regarding the structural-functional relationship of the P-cluster: (1) what is the mechanistic role of the P-cluster in electron transfer from the Fe protein to the MoFe protein, and (2) why does the P-cluster possess a [Fe₈S₇] structure rather than the more common [Fe₄S₄] structure that is sufficient for electron transfer in most enzymes? The first question was tackled by a recent study,¹⁰ which suggested a slow one-electron electron transfer from the all-ferrous P-cluster (P^N) to FeMoco (Eq. 1), followed by a fast reduction of the one-equivalent-oxidized P-cluster (P⁺) by the reduced Fe protein (FeP_red) upon binding to the MoFe protein (Eq. 2).

$${\text{P}}^{\text{N}}+\text{FeMoco}\stackrel{\mathit{Slow}}{\to }{\text{P}}^{+}+{(\text{FeMoco})}^{-}$$ (1)

$${\text{FeP}}_{\text{red}}+{\text{P}}^{+}\stackrel{\mathit{Fast}}{\to }{\text{FeP}}_{\text{ox}}+{\text{P}}^{\text{N}}$$ (2)

The P⁺-state of P-cluster is obviously an important intermediate in this mechanism. However, unlike the resting (P^N, S = 0) and two-equivalent-oxidized (P²⁺, S = 3 or 4) states of P-cluster, where various spectroscopic,¹¹ crystallographic⁵ and theoretical¹² studies have been undertaken, little is known of the P⁺ state other than its EPR features.¹³ Here, we present a variable-temperature, variable-field (VTVH) magnetic circular dichroism (MCD) spectroscopic study of the P⁺ state of nitrogenase P-cluster.

2. EXPERIMENTAL SECTION

General Considerations

Unless noted otherwise, all chemicals and reagents were obtained from Fisher Scientific or Sigma-Aldrich.

Cell Growth and Purification

Wild-type (AvOP) and mutant (DJ1193)¹⁴ strains of A. vinelandii were were grown in 180 L batches in a 200 L New Brunswick fermentor (New Brunswick Scientific, Edison, NJ, USA) in Burke’s minimal medium supplemented with 2 mM ammonium acetate. The growth rate was measured by cell density at 436 nm using a Spectronic 20 Genesys (Spectronic Instruments, Westbury, NY). After the consumption of ammonia, the cells were derepressed for 3 h, followed by harvesting using a flow-through centrifugal harvester (Cepa, Lahr, Germany). The cell paste was washed with 50 mM Tris–HCl (pH 8.0). Published methods were used for the purification of wild-type Fe protein (from AvOP), wild-type MoFe protein (from AvOP), and ΔnifB β-188^Cys MoFe protein (from DJ1193) of A. vinelandii.¹⁵

Protein Sample Preparation

All MCD samples were prepared in an Ar-filled drybox (Vacuum Atmospheres, Hawthorne, CA) at an oxygen level of less than 4 ppm.¹⁶ Dithionite-reduced protein samples were in 25 mM Tris–HCl (pH 8.0), 10% glycerol, and 2 mM dithionite (Na₂S₂O₄). Indigodisulfonate (IDS)-oxidized protein samples were prepared by incubating samples with IDS for 5 min, followed by removal of excess IDS by a G25 size-exclusion column. Samples were subsequently concentrated in a Centricon-50 concentrator (Amicon) in anaerobic centrifuge tubes outside the drybox as described earlier.¹⁷ After concentration, these protein samples [50–100 mg mL⁻¹ in 25 mM Tris–HCl (pH 8.0) and 50% glycerol] were transferred to MCD sample cuvettes under anaerobic conditions and frozen in liquid nitrogen. MCD sample cells were constructed from optical-quality Spectrosil® quartz (170–2200 nm, 1 mm path length, Model BS-1-Q-1, Starna®, Model SUV R-1001 or FUV; Spectrocell, Oreland, PA). Each cuvette was cut into the appropriate dimensions to fit the sample holder (2.0 cm × 12.5mm), resulting in a sample volume of approximately 160 μL. All samples contained 50% glycerol to ensure the formation of an optical glass upon freezing and kept on dry ice in transit.

Spectroscopic Characterization

MCD spectra were recorded with a CD spectropolarimeter (Model J-710; Jasco, MD) interfaced with a superconducting magnet (Model Spectromag 400-7T; Oxford, U.K.) as previously described (S3). Sample temperatures were monitored with a thin film resistance temperature sensor (Model CX1050-Cu-1-4L; Lakeshore, Westerville, OH) positioned directly (1 mm) above the sample cuvette. The linearity of the magnetic field was monitored with a calibrated Hall generator (Model HGCA-3020; Lakeshore, Westerville, OH) placed directly outside the superconducting magnet.

MCD spectra were recorded at a rate of 50 nm min⁻¹ from 800 to 350 nm at a resolution of 2–10 nm. Since optical glasses formed at low temperatures often generate a strain-induced background CD spectrum, the CD spectrum was recorded in a zero magnetic field to determine whether the background signal was excessive. If so, the sample was replaced by a new sample. To further eliminate interference by this signal, the corrected MCD spectrum was obtained for each sample by first recording the spectrum with the magnetic field in the normal direction and subtracting from it the spectrum with the field in the reversed direction. All spectral intensities were quantified per αβ-dimer of the α₂β₂-tetrameric MoFe protein and corrected for path length, sample concentration and depolarization effects. The extent of depolarization was determined by placing a standard sample of nickel tartarate between the magnet cryostat and the detector. The CD spectrum was then recorded before and after light passed through the frozen protein sample in the magnet.

Analysis of Magnetization Data

Magnetization curves were recorded at a set wavelength and temperature, while the magnetic field was linearly varied from 0 to 6 T at a rate of 0.45 T min⁻¹ with a resolution of 2 s. MCD data were analyzed using a fit/simulation program created by Neese and Solomon.¹⁸ The program allows the calculation of best-fit saturation magnetization curves using experimental data as a basis set and is valid for any spin state, half-integer or integer, at any specified temperature.

Experimental data were analyzed by fitting the spin Hamiltonian parameters and the effective transition moment products. The effective transition moment products represent the planes of polarization that reflect the anisotropy of the g factors. Since the initial slope of the magnetization curve is dependent on the g factors, the transition polarizations relate the transition dipole to the g factor axes of a powder or randomly oriented sample. For S > ½ spin systems, the spin parameters, including the g factor (g), the axial zero-field splitting (D), and the rhombic distortion of the electronic environment (E/D), are determined based on the Hamiltonian (below), which is the expression for energy of the Zeeman interaction and the correction to the energy of the individual spin states arising from spin-orbit coupling.

$$\widehat{H}=\beta Bg\widehat{S}+D[{\widehat{S}}_z^2-(1/3)\widehat{S}(\widehat{S}+1)+(E/D)({\widehat{S}}_x^2-{\widehat{S}}_y^2)]$$ (3)

At low temperatures (~1.6 K), the lowest energy level is predominantly populated and dictates the behavior of the magnetization curve. As the temperature is raised, the spectral parameters of excited states become increasingly important in the profile of the magnetization curve. Best-fit simulations of the experimental data were initially performed at the lowest temperature to enable the determination of the spectral parameters. Subsequent simulations of high-temperature data facilitated the determination of the axial zero-field splitting, D. MCD spectral simulations were conducted using IgorPro® (WaveMetrics, Inc.). Major peaks were first identified assuming a Gaussian shape. Peak intensities and central wave-lengths were then adjusted and the baseline corrected to minimize the error between the simulated and empirical curves.

3. RESULTS

The lack of information on the P⁺ state is mainly due to the difficulty in generating significant concentrations of this state. The sequential oxidations of P^N → P⁺ and P⁺ → P²⁺ exhibit the same mid-point potential (−309 mV) at pH 8¹⁹ and, because of this reason, there is not a particular potential that permits the enrichment of the P⁺ state without the accumulation of high concentrations of P^N and P²⁺ states. To overcome this problem, a mutant strain of Azotobacter vinelandii (DJ1193) was employed in the current study.¹⁴ There are two major advantages of using DJ1193 for spectroscopic studies of the P⁺ state. First, DJ1193 contains a deletion of the nifB gene, which encodes an essential protein for the biosynthesis of FeMoco. As a result, the variant MoFe protein of DJ1193 lacks FeMoco, thus preventing the paramagnetic FeMoco (S = 3/2) from interfering with the MCD spectroscopic investigation of P-cluster. Second, DJ1193 contains a substitution of cysteine for serine at the β-188 position of the MoFe protein. The β-188^Ser residue is an important ligand in the structure of the P-cluster. Specifically, during the oxidation of P^N to P²⁺, β-188^Ser becomes a ligand of an Fe atom in the P²⁺ state (Figure 1). The substitution of cysteine for serine in DJ1193 stabilizes the paramagnetic P⁺ state in the as-isolated DJ1193 MoFe protein (designated the ΔnifB β-188^Cys MoFe protein), leading to an accumulation of 65% P-cluster in the P⁺ state and the remaining 35% in the normal, diamagnetic P^N state.¹⁴

Figure 1.

Structures of the P-cluster in the reduced (P^N, left) and two-equivalent oxidized (P²⁺, right) state. The positions of residues Cys^β153, Ser^β188 and Cys^α88 are shown. PYMOL was used to generate this figure using the PDB entries 3MIN and 2MIN.

Consistent with the earlier observation that the P-cluster in the wild-type MoFe protein can be oxidized to the paramagnetic P²⁺ state, the VTVH MCD spectrum of the oxidized ΔnifB β-188^Cys MoFe protein (Figure 2A) exhibits decreasing spectral intensity with increasing temperature, which is indicative of a paramagnetic ground state of the P-cluster.¹⁶ However, a closer examination of the MCD spectra of the oxidized ΔnifB β-188^Cys and wild-type MoFe proteins (Figure 2B) reveals both similarities and differences between them. Spectral simulations (Figure 3) of the transitions above 450 nm reveal positive bands at 486, 521 and 589 nm in the spectrum of the wild-type MoFe protein, which appear to be also present in the spectrum of the ΔnifB β-188^Cys MoFe protein, with bathochromic shifts to 527, 541 and 644 nm, respectively. The major difference in these spectra is the absence of a very intense, broad transition that is centered at > 800 nm (simulated at 823 nm) in the spectrum of the wild-type protein. In its place, there is a broad negative transition at 740 nm in the spectrum of the oxidized ΔnifB β-188^Cys MoFe protein. The ‘823’ nm transition is the largest spectral transition in the spectrum of the oxidized wild-type protein and has been used as an identifying feature of the P²⁺ state.^16,20,21 The absence of this transition implies either a loss or a large spectral shift of a major charge-transfer transition following the substitution of cysteine for serine at the β-188 position of MoFe protein.

Figure 2.

(A) Temperature-dependent MCD spectra of the oxidized ΔnifB β-188^Cys MoFe protein at 6.0 T and 1.56, 4.12, 9.3 and 21 K, respectively. Spectra were normalized for one αβ-dimer of the protein. Sharp inflection at 420 nm is due to a minor cytochrome impurity. (B) Comparison of the 1.6 K MCD spectra of P²⁺ in the wild-type MoFe protein (blue) and the ΔnifB β-188^Cys MoFe protein (red) in the high-wavelength region (450–800 nm). The spectrum of the wild-type protein was multiplied by 1/3 to allow a better visualization. Arrows represent the proposed spectral shifts of transitions in the wild-type MoFe protein relative to those in the ΔnifB β-188^Cys MoFe protein, as determined by the spectral simulation (see Figure 3). Concentrations of the wild-type and ΔnifB β-188^Cys MoFe proteins were 18.0 and 74.5 mg/mL, respectively.

Figure 3.

MCD spectra of the oxidized wild-type (blue, multiplied by 1/3) and ΔnifB β-188^cys (red) MoFe proteins. Simulations dashed blue and dashed red, respectively) were made using Gaussian curves (thin blue and thin red, respectively). Arrows indicate the shifts of the main transitions in the wildtype MoFe protein to those of the ΔnifB β-188^cys MoFe protein.

EPR, MCD and Mössbauer data^8,21,22 suggest that the P²⁺ state in the wild-type MoFe protein is an S = 3 or 4 integer spin state. The 1.6 K magnetization curve of the oxidized ΔnifB β-188^Cys MoFe protein mimics that of the wild-type MoFe protein (Figure 4) in that it also exhibits a sharp initial slope, suggesting that the P-cluster of this variant MoFe protein is present in a similar, high spin state.^20,21 Such similarities between both the MCD spectra and the magnetization curves of the two MoFe proteins imply that the P²⁺ state of the ΔnifB β-188^Cys MoFe protein is essentially the same as that of the wild-type protein and suggest a great deal of resemblance between the electronic states of their P²⁺-clusters.

Figure 4.

Comparison of the magnetization curves of the wildtype MoFe protein (blue) and the ΔnifB β-188^Cys MoFe protein (red) at 1.6 K and 4.2 K. Data for both proteins were recorded at 770 nm.

While the MCD spectrum of the P²⁺ state reveals interesting electronic properties of this cluster, the MCD spectrum of the P⁺ state is more pertinent to the enzymatic mechanism of nitrogenase (Eq. 1). The EPR spectrum of the P⁺-cluster in the as-isolated ΔnifB β-188^Cys MoFe protein is virtually identical to that of the P⁺-cluster in the wild-type MoFe protein (Table 1), showing the same mixed S = 1/2, (2×)5/2 states that are indicative of a nearly identical electronic structure of their P-cluster species.^13,14 Consistent with this observation, the VTVH MCD spectrum of the P⁺-cluster in the ΔnifB β-188^Cys MoFe protein (Figure 5A) clearly illustrates that it is present in a paramagnetic ground state. Moreover, magnetization curves of the ΔnifB β-188^Cys MoFe protein exhibit nesting (a fanning of the curves with temperature), implying the presence of a high-spin state component (S > 1/2) in this protein (Figure 5B). These curves are best simulated using a mixed paramagnetic ground state (S = 1/2, 5/2) with average spectral parameters obtained from the EPR spectrum of the P⁺ state (Figure 4B).

Table 1.

EPR Parameters of p⁺

MoFe protein	S	gxa	gy	gz	D (cm−1)	E/D	%
Wild-type	1/2	2.06	1.95	1.81			11
	5/2b	6.7	5.3		−3.2	0.029	42
	5/2	7.3			−3.2	0.059	47
ΔnifBβ-188Cys	1/2	2.03	1.97	1.93			55
	5/2	6.7	5.3		−3.2	0.029	20
	5/2	7.7			−3.2	0.061	25

^aonly observed g-factors are listed;

^btwo different S = 5/2 signals are observed

Figure 5.

(A) Temperature-dependent MCD spectra of the reduced ΔnifB β-188^Cys MoFe protein at 1.59, 4.2 and 9.5 K. Spectra were normalized for one αβ-dimer of the protein. (B) Magnetization curves of the reduced ΔnifB β-188^Cys MoFe protein (red) recorded at 520 nm. Simulation of the magnetization curves (black) assuming two different spin (1/2 and 5.2) contributions and using the following parameters: 60% S = 1/2, 40% S = 5/2; for S = 5/2: D = −3 cm¹⁻, E/D = 0.003; for all curves: transition polarizations = 1.0 for the x-, y- and z-directions (i.e., equal contributions from all three principle axes). Protein concentration was 30.5 mg/mL.

4. DISCUSSION

It has recently been suggested that the P⁺ state of the P-cluster is a key player in the enzymatic mechanism of nitrogenase. While structures have been assigned to the P^N and P²⁺ states, the structure of P⁺ is unknown, although it has been suggested ²³ to be midway between those of P^N and P²⁺. The similarities of the MCD spectra and magnetization curves (Figure 2 and Figure 4) of P²⁺ in the wild-type protein and the ΔnifB β-188^Cys MoFe protein suggest a similarity between the electronic structures of P²⁺ in both proteins. Likewise, the great similarities in the EPR parameters of P⁺ in both proteins (Table 1) suggest that the structure of the P⁺ state is also conserved in both protein.

Most surprisingly, the MCD spectrum of the P⁺ state arises from a classic [Fe₄S₄]⁺-like cluster.^24,25 MCD spectra have long been used to characterize the structure and redox state of basic FeS clusters.^26,27 The spectrum of a [Fe₄S₄]⁺-like cluster exhibits a broad, derivative-like inflection that is centered around 600 nm, as well as a positive peak at ca. 520 nm and a negative trough at ca. 640 nm. Also, there is often a smaller positive peak at ca 740 nm. Figure 5 exemplifies these ‘finger-prints’, which are characteristic of a [Fe₄S₄]⁺-like cluster, are present in the MCD spectrum of the P⁺ state. This observation suggests that the wave function for the unpaired electron is localized on one 4Fe half of the P-cluster rather than delocalized over the entire 8Fe cluster. As such, it may be more correct to view the P-cluster as two linked [Fe₄S₄] clusters, where each 4Fe half retains the electronic characteristic of an isolated [Fe₄S₄] cluster. The suggestion of linked [Fe₄S₄] clusters was previously made to explain the intensity variation of the EPR spectrum during the stepwise oxidation of the P-cluster.^13,28 Similarly, theoretical calculations¹² have employed a split [Fe₄S₄] cluster model to explain the spin state of the P²⁺-cluster. Furthermore, the split cluster model is consistent with the identical mid-point potentials of the P^N → P⁺ and P⁺ → P²⁺ oxidations.¹⁹ Finally, it has been shown²⁹ that replacing the two cysteine ligands that bridge one 4Fe half of the P-cluster with the other retains good, albeit diminished catalytic activity. It is also interesting to note that the MCD spectrum of the P⁺-cluster in the ΔnifB β-188^Cys MoFe protein is very similar to that of the two different [Fe₄S₄]⁺-like clusters in the P-cluster precursor of the ΔnifH MoFe protein (Figure 6).^16,30 Such a similarity again leads to the question of why the [Fe₈S₇] P-cluster is used in place of the more common [Fe₄S₄] cluster for electron transfer in nitrogenase.

Figure 6.

Comparison of the MCD spectra of P⁺ in the reduced ΔnifB β-188^Cys MoFe protein (red) and the [Fe₄S₄]-type clusters in the reduced ΔnifH (green) MoFe proteins at a temperature of 1.6 K and a magnetic field of 6.0 T. Spectra were normalized for one αβ-dimer of the protein. The intensity of the ΔnifH MoFe protein spectrum was multiplied by 0.5 to compensate for the presence of two [Fe₄S₄]-type clusters in the protein, as compared to the presence of a single P-cluster in the ΔnifB β-188^Cys MoFe protein. The positive shift of the spectrum of ΔnifB β-188^Cys MoFe protein relative to that of the ΔnifH MoFe protein is likely caused by the presence of an S = 5/2 spin state in the former protein.

The answer to this question may be drawn indirectly from studies of proteins similar to the MoFe protein. DPOR (Dark Operative Protochlorophillide Reductase) and COR (Chlorophyllide a Reductase), two enzymes involved in bacteriochlorophyll synthesis,^31,32 are good examples of these MoFe protein homologs. Both DPOR and COR contain only a single [Fe₄S₄] cluster in an analogous position held by the P-cluster in the MoFe protein, and both are capable of catalyzing two-electron reduction of substrates: DPOR catalyzes the two-electron reduction of the D-ring of porphoryrin, and COR catalyzes the subsequent two-electron reduction of the B-ring of chlorin.^31,32 Similarly, NifEN, another functional homolog to the MoFe protein, contains a single [Fe₄S₄] cluster at a position that corresponds to the position of P-cluster in the MoFe protein.³³ Like DPOR and COR, NifEN can catalyze the two-electron reduction of certain substrates (i.e., C₂H₂ and N₃⁻), but it cannot reduce these substrates further, nor can it reduce other nitrogenase substrates that require more electrons (e.g., CN⁻, N₂H₄ and N₂).³⁴

The fact that all these MoFe protein homologs can only catalyze two-electron reduction via their respective [Fe₄S₄] cluster suggests that the more complex [Fe₈S₇] P-cluster is required specifically for the reduction of substrates that involve the transfer of more than two electrons, such as N₂. The Lowe-Thorneley (LT) model of N₂ reduction³⁰ involves eight one-electron transfer steps from the Fe protein to the MoFe protein (Figure 7), with E₀ representing the enzyme in the resting state and E_n representing the enzyme after receiving n electrons and n protons (where 8 ≥ n ≥ 0). The ‘electron inventory’³⁵ states that the n electrons are associated with the number transferred to the cofactor (m), the substrate (s) and the number oxidized at the P-cluster (p), such that n = m + s − p. Data obtained from previous studies suggest that the initial binding of N₂ occurs at E₃ or E₄ step and the reduction of N₂ occurs at higher E_n steps.³⁶

Figure 7.

Lowe-Thorneley (LT) model for N₂ reduction by nitrogenase. The resting state of one half of the MoFe protein is represented by E₀. State E_n represents the protein after having received n electrons from the Fe protein and n protons from the medium. The LT model proposes that the binding of N₂ occurs at E₃ or E₄ step, and the release of NH₃ occurs at E₅ or E₇ step.

A basic premise of the LT model is that each step involves the transfer of only one electron. This premise is centered on two assumptions. The first assumption is that the Fe protein can transfer only one electron to the MoFe protein during each step in the cycle (Figure 7). While it has been shown³⁷ that the FeS cluster in the Fe protein can be super reduced to the all ferrous [Fe₄S₄]⁰ state, thus making it a potential two-electron donor, there is no definitive proof that such a two-electron transfer process indeed occurs during enzymatic turnover. The second assumption is that the Fe protein first reduces the P-cluster, which subsequently reduces the FeMo cofactor. The recent proposal¹⁰ (Eqs. 1 and 2) seems to be just the opposite of this assumption and, if the P-cluster indeed donates an electron to FeMoco first (Eq. 1), then there is no longer a restriction on a strict one-electron transfer in this process.

The results presented herein suggest that the P-cluster may act as two coupled [Fe₄S₄] clusters, each capable of donating one electron to FeMoco. In this case, the P-cluster could undergo either a one- or two-electron transfer process as a complete unit. Over the past few decades, a vast amount of research has gone into the study of the mechanism and kinetics of nitrogenase. By far the majority of these studies focused on the early steps of the LT cycle, where the concept of one-electron transfer per step is favored.¹ Our hypothesis is consistent with this concept and the earlier proposal that the initial E₀ → E₁ mechanistic step involves one electron (Eq. 1). However, it can also be used to further reason that one or more of the steps later in the LT cycle may involve two electrons, followed by reduction of P²⁺ back to P^N.

In an intra MoFe protein two-electron transfer, n = 0 and p = 2, meaning m + s = 2. Recent spectroscopic data³⁸ suggests that the cofactor only cycles through two different states, the native state (M^N) and the one-electron reduced state (M^R). As such, a two-electron transfer would likely leave m unchanged resulting in both electrons residing on the substrate (s = 2). Since the two-electron transfer event is suggested to occur later in the cycle, where few, if any, experiments have focused on, there is currently no evidence to either support or refute this theory. Regardless, there could be an energy advantage to the two-electron transfer event. Calculations^39,40 show that the major energy barriers to N₂ reduction occur during the initial “uphill” reduction steps. The mechanistic step in the two-electron transfer process may be required to more efficiently surmount the energy barriers of this reaction. Given the two 4Fe halves in the 8Fe structure of P-cluster, it is possible that the MoFe protein utilizes two different electron transfer pathways from the P cluster to FeMoco, each originating from a different half of the P cluster.

There could also be a kinetic advantage to the two-electron transfer process. One of the side reactions (Eq. 3) in the LT model³⁵ is the slow, natural relaxation of En state to lower (E_n⁻²) states with the concomitant evolution of H₂.

$${\text{E}}_n+2{\text{H}}^{+}\stackrel{\mathit{Slow}}{\to }{\text{E}}_{n^{-}2}+{\text{H}}_2$$ (3)

It is obvious that this reaction decreases the ability of the enzyme to reach higher E_n state and, consequently, lowers the efficiency of the nitrogenase reaction. However, the two-electron transfer mechanism could counteract this inefficient side reaction by more rapidly “pushing” the enzyme to higher E_n states that are necessary for the binding and reduction of N₂.

In summary, the MCD spectrum of the P⁺-cluster in ΔnifB β-188^Cys MoFe protein has been identified as arising from a [Fe₄S₄]⁺-like cluster, suggesting that the P-cluster may function as two coupled [Fe₄S₄]-like clusters, where each cluster is capable of transferring electrons to FeMoco. Such a cluster arrangement of P-cluster is likely related to the mechanism of N₂ reduction and could serve as a focal point of future investigations of the structur-function relationship of nitrogenase.

ABBREVIATIONS

FeMoco

iron-molybdenum cofactor

VTVH

variable-temperature, variable-field

MCD

magnetic circular dichroism

IDS

indigodisulfonate