Decoding the nitrogenase mechanism: the homologue approach

Abstract

The (Mo)-nitrogenase is a complex metalloenzyme that catalyzes the key step in the global nitrogen cycle, the reduction of atmospheric dinitrogen (N(2)) to bioavailable ammonia (NH(3)), at the iron-molybdenum cofactor (FeMoco) site of its molybdenum-iron (MoFe) protein component. Despite the fundamental significance of biological nitrogen fixation and extensive studies over the past decades, the catalytic mechanism of nitrogenase has not been deciphered. One major challenge for the mechanistic study of nitrogenase is the redox versatility of its FeMoco center. The ability of FeMoco to shuttle between oxidation states in a rapid and unsynchronized manner results in a mixed oxidation state of the cofactor population during turnover. The substrate and the various intermediates can only interact with the FeMoco site in a transient manner, so it is extremely difficult to capture any substrate- or intermediate-bound form of nitrogenase for the direct examination of substrate-enzyme interactions during catalysis. In this Account, we describe the approach of identifying a partially "defective" nitrogenase homologue, one with a slower turnover rate, as a means of overcoming this problem. The NifEN protein complex serves as an ideal candidate for this purpose. It is an alpha(2)beta(2)-heterotetramer that contains cluster-binding sites homologous to those found in the MoFe protein: the "P-cluster site" at the interface of the alphabeta-subunit dimer, which accommodates a [Fe(4)S(4)]-type cluster; and the "FeMoco site" within the alpha-subunit, which houses an all-iron homologue to the FeMoco. Moreover, NifEN mimics the MoFe protein in catalysis: it is capable of reducing acetylene (C(2)H(2)) and azide (N(3)(-)) in an ATP- and iron (Fe) protein-dependent manner. However, NifEN is unable to reduce proton (H(+)) and N(2), and it is an inefficient enzyme with a restricted electron flux during the turnover. The extremely slow turnover rate of NifEN and the possible "synchronization" of its FeMoco homologue at a certain oxidation level permit the observation of a new S = 1/2 EPR signal upon turnover of C(2)H(2) by NifEN, which is analogous to the signal reported for a MoFe protein variant upon turnover of the same substrate. This result is exciting, because it suggests the possibility of naturally enriching a C(2)H(2)-bound form of NifEN for the successful crystallization of the first intermediate-bound nitrogenase homologue. On the other hand, the fact that NifEN represents a partially "defective" homologue of the MoFe protein makes it a promising mutational platform on which a functional MoFe protein equivalent may be reconstructed by introducing the missing features of MoFe protein step-by-step into NifEN. Such a strategy allows us to define the function of each feature and address questions such as the following: What is the function of P-cluster in catalysis? Are Mo and homocitrate the essential constituents of the cofactor in N(2) reduction? How does substrate accessibility affect the reactivity of the enzyme? This homologue approach could complement the mechanistic analysis of the nitrogenase MoFe protein, and information derived from both approaches will help achieve the ultimate goal of solving the riddle of biological nitrogen fixation.

Figure 1

Components of the electron transfer chains in the Fe protein/MoFe protein system (nitrogenase) and the Fe protein/NifEN system (nitrogenase homolog). (A) The crystal structure of the MgADP·AlF₄⁻-stabilized complex between the Fe protein and one αβ-dimer of the MoFe protein. (B) The electron transfer pathway between the Fe protein and the MoFe protein. (C) The hypothetical electron transfer pathway between the Fe protein and NifEN. It has been proposed that, during nitrogenase catalysis, electrons flow from the [Fe₄S₄] cluster of the Fe protein to the P-cluster and then the FeMoco of the MoFe protein. Likewise, electrons could flow from the [Fe₄S₄] cluster of the Fe protein to the [Fe₄S₄] cluster and then the FeMoco homolog of NifEN. The two subunits of the Fe protein are colored dark and light gray, and the α- and β-subunits of the MoFe protein are colored light purple and magenta. The atoms of the metal centers are color as follows: Mo, orange; Fe, purple; S, yellow; O, red; C, dark gray; X (C, N or O), light gray. These presentations are generated in PYMOL using 1N2C and 1M1N PDB coordinates.,

Figure 2

Substrate reducing activities of NifEN and MoFe protein. Shown are the activities of N₂ (①), H⁺ (②), C₂H₂ (③) and N₃⁻ (④) reduction by NifEN (A) and MoFe protein (B) of A. vinelandii. NifEN is unable to reduce N₂ and H⁺, but it is capable of catalyzing the two-electron reduction of C₂H₂ and N₃⁻ (A). The MoFe protein, on the other hand, is able to catalyze the concomitant reduction of N₂ and H⁺, the reduction of H⁺ that is independent of N₂, and the reduction of C₂H₂ and N₃⁻. In the MoFe protein-catalyzed reaction, N₃⁻ can be reduced by two, six and eight electrons to N₂+NH₃, N₂H₄+NH₃ and NH₃, respectively.

Figure 3

EPR properties of NifEN. Shown are the EPR spectra of NifEN under turnover (A) and non-turnover (B) conditions of C₂H₂, and under turnover (C) and non-turnover (D) conditions of N₃⁻. The turnover samples (A, C) contain ATP, which is absent from the non-turnover samples (B, D). The EPR spectra in the presence of C₂H₂ and N₃⁻ were recorded at 30 K and 6K, respectively.

Figure 4

Comparison between the primary sequences of NifEN and MoFe protein. (A) Partial sequence alignment of the α-subunits of MoFe protein (NifD) and NifEN (NifE) from A. vinelandii. (B) Partial sequence alignment of the β-subunits of MoFe protein (NifK) and NifEN (NifN) from A. vinelandii. In NifEN, one of the FeMoco ligands (Cys) is conserved; whereas the other (His) is replaced by an Asn. The homocitrate anchor (Lys) is replaced by an Arg. Additionally, the FeMoco “lock” (Trp) is replaced by a Glu. The P-cluster ligands (coordinating the P^N state) are also partially preserved in NifEN, with four of the six Cys ligands conserved and the other two replaced by Ser and Gly, respectively. The additional Ser ligand for the P-cluster (coordinating the P^OX state) is replaced by an Asp. The proton gating residue (His^α195) is replaced by an Asn in NifEN.

Figure 5

Comparison between the “P” centers in NifEN and MoFe protein. Consistent with the presence of four Cys ligands (Cys^α37, Cys^α62, Cys^α81 and Cys^α124), the “P” center in NifEN is a [Fe₄S₄]-type cluster (A). In contrast, the “P” center in the MoFe protein is a [Fe₈S₇] cluster (B). In the reduced state (P^N), it is bridged by six Cys ligands, three from the α-subunit (Cys^α62, Cys^α88 and Cys^α154) and three from the β-subunit (Cys^β70, Cys^β95 and Cys^β153) (B). Upon a two-electron oxidation, the P-cluster can be converted to the oxidized state (P^OX), in which half of the cluster is “opened” up concomitantly with the additional coordination from Cys^α88 and Ser^β188 (B).

Figure 6

Comparison between the “M” centers in NifEN and MoFe protein. The “M” center in NifEN is a Mo- and homocitrate-free homolog of the FeMoco, and it is likely ligated by Asn^α418 and Cys^α250, respectively, at the opposite ends of the cluster (A). In contrast, the “M” center in the MoFe protein is a fully-complemented FeMoco, and it is ligated by His^α442 and Cys^α275, respectively, at the Mo-end and the opposite Fe-end of the cluster (B). The homocitrate entity at the Mo-end of the cluster is further coordinated by Lys^α426 (B). Additionally, two highly conserved residues, Trp^α444 and Ser^α443, have been proposed to be part of a “lock” mechanism that secures the FeMoco in place (B).

Figure 7

Comparison between NifEN and MoFe protein in the “proton gating” mechanism. In NifEN, an essential “proton gating” residue is replaced by another residue (Asn^α164) near the “M” center (A). In the MoFe protein, this “proton gating” residue (His^α195) is likely involved in the delivery of H⁺ to the active “M” center for the concomitant reduction of H⁺ and N₂ (B).