|Home | About | Journals | Submit | Contact Us | Français|
The final step of FeMo cofactor (FeMoco) assembly involves the insertion of FeMoco into its binding site in the molybdenum-iron (MoFe) protein of nitrogenase. Here we examine the role of His α274 and His α451 of Azotobacter vinelandii MoFe protein in this process. Our results from combined metal, activity, EPR, stability and insertion analyses show that mutations of His α274 and/or His α451, two of the histidines that belong to a so-called His triad, to small uncharged Ala specifically reduce the accumulation of FeMoco in MoFe protein. This observation indicates that the enrichment of histidines at the His triad is important for FeMoco insertion and that the His triad potentially serves as an intermediate docking point for FeMoco through transitory ligand coordination and/or electrostatic interaction.
Nitrogenase catalyzes the nucleotide-dependent reduction of the inert atmospheric dinitrogen to the bioavailable form of ammonia [for recent reviews see refs. 1–5]. The molybdenum (Mo) nitrogenase of Azotobacter vinelandii is a two-component system, comprising the iron (Fe) protein and the molybdenum iron (MoFe) protein. The Fe protein (also designated Av2) is a homodimer (encoded by nifH), with one ATP binding site per subunit and a single [4Fe-4S] cluster bridged between the subunits. The MoFe protein (also designated Av1) is an α2β2-tetramer (encoded by nifD and nifK), containing two unique metal clusters per αβ subunit pair: the [8Fe-7S] P-cluster , which is bridged between α- and β-subunits and ligated to six Cys ligands; and the [Mo-7Fe-9S-X-homocitrate]1 FeMo cofactor (FeMoco), which is located within the α-subunit and bound to the protein by Cys α275 at the terminal Fe and His α442 at the opposite Mo that is further coordinated by an endogenous homocitrate. Nitrogenase catalysis involves complex association/dissociation between the Fe protein and the MoFe protein, during which process electrons, concomitant with ATP hydrolysis by the Fe protein, are sequentially transferred from the [4Fe-4S] cluster in the Fe protein through the P-cluster to the FeMoco in the MoFe protein, where substrate reduction takes place.
The biosynthesis of nitrogenase component proteins and complex metal clusters are controlled by the nitrogen fixation (nif) genes . The assembly of MoFe protein occurs in two steps. First, a P-cluster-replete, yet FeMoco-depleted form is synthesized in a process that involves the expression of nifD and nifK—the structural genes encoding the α- and β-subunits of the MoFe protein, and the formation of P-cluster—likely through fusion of its substructural, [4Fe-4S] cluster units at the target location (in situ) in the protein [8–10]. Then, the FeMoco—assembled outside of the MoFe protein (ex situ) in a process that involves the participation of nifB, nifE, nifN, nifH, nifV and nifQ, is inserted into its binding site, resulting in the formation of an active, holo form of MoFe protein . The ex situ assembly of FeMoco is presumably initiated with the formation of an Fe/S core on NifB, followed by the relocation and rearrangement of the core on the NifEN complex [11–14]. Recently, we identified a Mo-free, NifEN-bound FeMoco precursor that closely resembled the core structure of the mature FeMoco [11, 12]. We also showed that, coupled to the hydrolysis of ATP, the Fe protein inserted Mo and homocitrate into the precursor while it was still bound to NifEN, resulting in a fully complemented cluster that could be inserted into the MoFe protein through direct protein-protein interaction [13, 14]. These recent developments on ex situ FeMoco biogenesis not only clarify certain aspects of the cluster assembly in nitrogenase, they may also provide useful insights into the biosynthetic strategy of complex metal clusters in other systems [15–17].
On the other hand, the mechanism of the final step of FeMoco assembly—namely, how FeMoco is inserted into its target binding site in the MoFe protein, remains largely unknown. Much of the current understanding in this regard comes from the crystallographic studies of a ΔnifB MoFe protein from a nifB-deletion strain of A. vinelandii . In agreement with the genetically-based theory that nifB is the starting point of FeMoco biosynthesis, the ΔnifB MoFe protein is FeMoco-deficient, yet it contains normal P-clusters . More importantly, the ΔnifB MoFe protein can be readily reconstituted, in vitro, into an active holo protein by addition of isolated FeMoco, suggesting that it represents a physiologically relevant snapshot of MoFe protein right before the final insertion of FeMoco . In contrast to the wild-type MoFe protein, the ΔnifB MoFe protein contains a positively charged funnel that could allow the entry and navigation of the negatively charged FeMoco toward its target location . Such a FeMoco insertion funnel is created as a result of major structural rearrangement of the αIII domain of the ΔnifB MoFe protein, which involves the significant repositioning of a number of residues. Some of these residues belong to a stretch of polypeptide ranging from α353 to α364 which, upon repositioning by distances up to 20 Å, forms a loop at the entrance of the funnel. Among them are the positively charged Arg α359 and Arg α361, as well as the His α362, which could serve as the first contact point for FeMoco during the insertion process. Moving into the funnel, the Cα of His α442 is relocated by a distance of approximately 5 Å and joins His α274 and His α451 to form a striking His triad (Fig. 1), which could help guide the negatively charged FeMoco to the destined location. The movement of His α442 also leads to a switch in position of this residue with Trp α444, which could help secure FeMoco at its final location. Additionally, Lys α426 is shifted by roughly 5 Å, which could serve as another anchoring point for FeMoco during insertion. These observations have led to the proposed mechanism of FeMoco insertion, which involves the initial docking of FeMoco at the entrance loop of the funnel, the continued cruising of FeMoco along the funnel under the guidance of the His triad, the anchoring of the Mo/homocitrate end of FeMoco by His α442 and Lys α426 at the bottom of the funnel, the swapping of positions of Trp α444 and His α442 that results in the capture of the FeMoco, and finally the closing up of the funnel by the movement of the entrance loop.
To test the structurally-based hypothesis of FeMoco insertion, we have constructed a series of MoFe protein variants with site-directed mutations of the key residues for this process and subjected them to extensive biochemical, biophysical and spectroscopic analyses. Recently, we showed that substitutions of Trp α444 for small residues, or, His α362 for uncharged or negatively charged residues drastically decreased the level of FeMoco accumulation [19, 20]. In this study we examine the role of two His triad residues, His α274 and His α451, in the process of FeMoco insertion. Our data clearly indicate that mutations of His α274 and/or His α451 to small uncharged Ala result in dramatically reduced levels of FeMoco incorporation, indicating that these residues are important for FeMoco insertion and that they likely interact with FeMoco through transitory ligand formation and/or electrostatic attraction during the insertion process.
Unless noted otherwise, all chemicals and reagents were obtained from Fisher Scientific, Baxter Scientific, or Sigma.
Table 1 summarizes the A. vinelandii strains used in this study. From this point on, the MoFe protein is designated as Av1, whereas the Fe protein is referred to as Av2. The mutations of the Av1 proteins are indicated by the appropriate superscripts. Note that the Av2 purified from the same strain as the Av1 variant are given the same superscript; however, the mutation is located in Av1, not Av2. The wild-type strain AvYM13Awild-type (expressing His-tagged Av1wild-type and non tagged Av2wild-type) was constructed as described elsewhere . The variant strains AvYM14AαH274A (expressing His-tagged Av1αH274A and non tagged Av2αH274A), AvYM22AαH451A (expressing His-tagged Av1αH451A and non tagged Av2αH451A) and AvYM23AαH274A/αH451A (expressing His-tagged Av1αH274A/αH451A and non tagged Av2αH274A/αH451A) were constructed as follows. First, plasmid pHR30 was constructed, which contained the chromosomal fragment of nifD and nifK genes of A. vinelandii. Then, two oligos were used to create desired site-directed mutations of the nifD gene carried on pHR30, following the procedure of the commercially available GeneEditor in vitro Site-Directed Mutagenesis System (Promega, Madison, WI). The oligos used for mutations were (i) H274A, 5′-AAGCTTAACCTGGTTGCCTGCTACCGCTCGATG-3′ and (ii) H451A, 5′-TATTCGGGCCCCTACGCCGGCTTCGATGGCTTC-3′. Oligos (i) and (ii) were either used alone to create single mutation or together to create double mutations. The resulting plasmids were pHR38 (αH274A), pHR39 (αH451A) and pHR40 (αH274A/αH451A). Finally, pHR38, pHR39 and pHR40 were transformed into AvYM13A using a previously described method [21, 22]. The resulting variant strains were AvYM14AαH274A, AvYM22AαH451A and AvYM23AαH274A/αH451A, respectively, with site-directed mutations of nifD gene on the chromosomal DNA. Additionally, the nifB gene in AvYM14AαH274A, AvYM22AαH451A or AvYM23AαH274A/αH451A was deleted as described elsewhere . The resulting strains were AvYM24AΔnifB/αH274A (expressing His-tagged Av1ΔnifB/αH274A), AvYM32AΔnifB/αH451A (expressing His-tagged Av1ΔnifB/αH451A) and AvYM33AΔnifB/αH274A/αH451A (expressing His-tagged Av1ΔnifB/αH274A/αH451A), respectively, with the deletion of nifB gene in addition to the site-directed mutations of nifD gene on the chromosomal DNA. The site-directed mutations of nifD were subsequently confirmed by DNA sequencing.
All A. vinelandii strains were grown in 180 L batches in a 200 L New Brunswick fermentor (New Brunswick Scientific, Edison, NJ) in Burke’s minimal medium supplemented with 2 mM ammonium acetate. The growth rate was measured by cell density at 436 nm using Spectronic 20 Genesys (Spectronic Instruments, Westbury, NY). After the consumption of ammonia, cells were de-repressed for 3 h followed by harvesting using a flow through centrifugal harvester (Cepa, Lahr/Schwarzwald, Germany). The cell paste was washed with 50 mM Tris·HCl (pH 8.0). Published methods were used for the purification of all Av2 proteins  and His-tagged Av1 proteins [11, 21]. FeMoco was isolated as described elsewhere .
All EPR samples were prepared in a Vacuum Atmospheres dry box (Hawthorne, CA) with an oxygen level of less than 4 ppm. Unless noted otherwise, all samples were in 25 mM Tris·HCl (pH 8.0), 10% glycerol and 2 mM Na2S2O4. Av1 protein samples were oxidized by incubation with excess indigo disulfonate (IDS) for 30 min. Subsequently, IDS was removed by a single passage over an anion-exchange column as described elsewhere . All perpendicular and parallel mode EPR spectra were recorded using a Bruker ESP 300 Ez spectrophotometer (Bruker, Billerica, MA), interfaced with an Oxford Instruments ESR-9002 liquid helium continuous flow cryostat (Oxford Instruments, Oxon, U.K.). Unless noted otherwise, all spectra were recorded at 10 K using a microwave power of 50 mW, a gain of 5 × 104, a modulation frequency of 100 kHz, and a modulation amplitude of 5 G. The microwave frequencies of 9.62 and 9.39 GHz were used for the perpendicular (10 scans) and parallel (20 scans) mode EPR spectra, respectively. Spin quantitation of EPR signals was carried out as described in detail earlier .
All nitrogenase activity assays and FeMoco insertion assays were carried out as described previously [21, 24]. The products H2 and C2H4 were analyzed as published elsewhere . Ammonium was determined by a high performance liquid chromatography fluorescence method . FeMoco maturation assays were carried out as published earlier . Molybdenum  and iron  were determined as published elsewhere. The iron chelation assays were performed as described elsewhere [30, 31] using an Fe protein concentration of 0.2 mg/ml.
Two approaches were used to determine the stability of the purified Av1 proteins: (i) heat treatment and (ii) prolonged storage at room temperature. A total amount of 100 mg purified Av1 [in 25 mM Tris·HCl (pH 8.0), 10% glycerol, 250 mM NaCl and 2 mM Na2S2O4] was incubated in a crimped anaerobic vial (volume, 8.7 ml; gas atmosphere, 100% Ar) at 56°C for 30 s (i) or stored at room temperature for 8 h (ii). Precipitated protein was subsequently removed in both cases by centrifugation at 10,000 rpm for 10 min (Biofuge fresco, Heraeus, Germany) in a Vacuum Atmospheres dry box. Concentrations of Av1 proteins were subsequently determined by Bio-Rad protein assays (Bio-Rad Laboratories, Hercules, CA). Enzymatic activities were determined as described above.
His α274 and His α451 are highly conserved (both ~80%) among the currently known MoFe protein primary sequences2. In Av1ΔnifB, the snapshot of Av1 prior to the final step of FeMoco insertion, His α274 and His α451 are joined by His α442 to form a remarkable His triad half way down the insertion funnel (Fig. 1) . It has been hypothesized that such a histidine-rich core provides a midway docking point for FeMoco along its insertion pathway, during which process the Mo end of FeMoco makes the first contact with one of its eventual ligands, His α442 . Thus, the His triad is crucial for the on-target cruising of FeMoco toward its destination and disruption of the histidine enrichment at this particular point may prevent the FeMoco from being properly inserted into its target location. Based on this theory, a series of A. vinelandii strains expressing Av1 protein variants with site-directed mutations at the His triad was constructed. To minimize the mutational effect that may result in the instability of the protein, the FeMoco ligand His α442 was left untouched; whereas His α274 and His α451 were mutated, either singularly or simultaneously, to small uncharged Ala. These A. vinelandii variant strains (Table 1), designated AvYM14AαH274A (expressing Av1αH274A and Av2αH274A), AvYM22AαH451A (expressing Av1αH451A and Av2αH451A) and AvYM23AαH274A/αH451A (expressing Av1αH274A/αH451A and Av2αH274A/αH451A), respectively, were used in studies described in sections 3.1, 3.2 and 3.3.1–3.3.4. Additionally, a series of nifB-deletion A. vinelandii strains expressing the respective FeMoco-deficient forms of the His triad Av1 protein variants were constructed. These variant strains (Table 1), designated AvYM24AΔnifB/αH274A (expressing Av1ΔnifB/αH274A), AvYM32AΔnifB/αH451A (expressing Av1ΔnifB/αH451A) and AvYM33AΔnifB/αH274A/αH451A (expressing Av1ΔnifB/αH274A/αH451A), respectively, were used in studies described in section 3.3.5.
Under N2-fixing conditions3, the doubling time of AvYM13Awild-type is approximately 4.6 h; while those of AvYM14AαH274A, AvYM22AαH451A and AvYM23AαH274A/αH451A are 4.9 h, 5.3 h and 5.9 h, respectively, all of which are longer than that of AvYM13Awild-type. Considering that cell growth under N2-fixing conditions reflects the amount of active nitrogenase, these results indicate that mutations of the His triad residues to small uncharged Ala result in decreased nitrogenase activities. As expected, double mutations of His α274 and His α451 present a more serious problem to the protein than a single mutation of either His residue, consistent with a greater extent of disruption of the histidine-rich environment at this particular location when both His residues are replaced by Ala.
Reduced growth rates of the His triad variant strains do not originate from altered expression and properties of Av24. The Av2 proteins of AvYM14AαH274A, AvYM22AαH451A and AvYM23AαH274A/αH451A are expressed at nearly the same level as Av2wild-type, as shown by Western blot analysis under non-saturated conditions (Fig. 2A and 2B). In addition, an approximate amount of 400 mg non-tagged Av2 was purified from 200 g cells of AvYM14AαH274A, AvYM22AαH451A or AvYM23AαH274A/αH451A, as was from AvYM13Awild-type (data not shown)5, again confirming that Av2 expression is unperturbed in these variant strains. The monomer of Av2αH274A, Av2αH451A or Av2αH274A/αH451A, like that of Av2wild-type, is ~30 kDa (Fig. 3A). The molecular masses of all of these Av2 proteins are ~60 kDa based on their elution profiles on gel filtration columns (data not shown), suggesting that they all have the same, homodimeric subunit composition as Av2wild-type. Compared to Av2wild-type, all Av2 proteins of the His triad variant strains have the same metal content of approximately 4 mol Fe/mol protein (Table 2) and display, in the dithionite-reduced state, the same characteristic S = 1/2 EPR signal of rhombic line shape in the g = 2 region (Fig. 4). These data suggest that, like Av2wild-type, all of these Av2 proteins have, per dimer, one [4Fe-4S] cluster that can be reduced by dithionite to an oxidation state of +1. All Av2 proteins of the His triad variant strains show the same substrate reducing activities as Av2wild-type, with regard to C2H4 formation under C2H2/Ar, H2 formation under Ar, NH3 formation under N2, or H2 formation under N2 (data not shown), suggesting that they are as efficient as Av2wild-type in nitrogenase catalysis. In addition, all of them have the same capacity as Av2wild-type in FeMoco maturation assay (data not shown), indicating that they are as functional as Av2wild-type in FeMoco assembly.
Mutations at the His triad do not interfere with the accumulation of a certain amount of Av1 in cells, nor do they affect the structural integrity of Av1. The Av1 proteins of the His triad variant strains are expressed at approximately the same amount as Av1wild-type, as evidenced by Western blot analysis under non-saturated conditions (Fig. 2A and 2C). Additionally, nearly 600 mg His-tagged Av1 was purified from 200 g cells of AvYM14AαH274A, AvYM22AαH451A or AvYM23AαH274A/αH451A, as was from AvYM13Awild-type (data not shown)5, further suggesting that Av1 expression is unaffected in these variant strains. Moreover, like Av1wild-type, the Av1 proteins of all these variant strains are composed of α (~56 kDa)- and β (~59 kDa)-subunits (Fig. 3B) and have the same, α2β2-tetrameric molecular mass of ~230 kDa based on their elution profiles on gel filtration columns (data not shown).
Nevertheless, disruptions of the His triad specifically affect FeMoco insertion, as shown by combined metal, activity, EPR, stability and insertion studies of the His triad Av1 variant proteins (see below).
The His triad Av1 variants contain reduced amounts of Fe and Mo (Table 2). The Mo contents of Av1αH274A, Av1αH451A and Av1αH274A/αH451A are 61%, 50% and 44%, respectively, of that of Av1wild-type (Table 2). Given that the Mo level reflects the FeMoco content, these results provide the first indication that mutations of the His triad residues result in decreased accumulation of FeMoco in the Av1 protein.
Consistent with the reduced Mo content, the His triad Av1 variants have reduced substrate reducing activities. Compared to Av1wild-type, the His triad Av1 variants show decreased activities of C2H4 formation under C2H2/Ar, H2 formation under Ar, and NH3 formation under N2 (Table 3). Further, the extent of decrease in activities, in the ascending order of Av1αH274A < Av1αH451A < Av1αH274A/αH451A, is consistent with the extent of decrease in growth rates of the strains expressing these proteins (see above). However, all three Av1 variants display unusually high capacities of H2 formation under N2—111%, 77% and 75%, respectively, of that of Av1wild-type, which do not correlate proportionally with their capacities of NH3 formation under the same conditions—51%, 46% and 36%, respectively, of that of Av1wild-type (Table 3). This discrepancy indicates a shift toward H2 evolution during the concomitant formation of NH3 and H2 under N2, a situation where neither NH3 nor H2 formation alone reflects the overall activity of the reaction. To obtain an accurate measure of the substrate reducing activity under N2, the total electron flux in the reaction was calculated on the basis of the electron pairs appearing in both products (NH3 and H2) and compared to those in the formation of C2H4 under C2H2/Ar and the formation of H2 under Ar (Table 4). These calculations reveal that, each Av1 variant shows consistently reduced activities with regard to all substrates, which are 59%–74% of Av1wild-type activity for Av1αH274A, 53%–61% of Av1wild-type activity for Av1αH451A, and 40%–48% of Av1wild-type activity for Av1αH274A/αH451A (Table 4). On average, Av1αH274A, Av1αH451A and Av1αH274A/αH451A have 67%, 56% and 43%, respectively, of the activity of Av1wild-type (Table 4). These activities correlate well with the Mo contents of the respective Av1 variant proteins (Table 2), strongly suggesting that the diminished activities of the His triad variants originate specifically from decreased levels of FeMoco in the Av1 proteins.
EPR analyses provide further evidence that disturbance of the His triad has significant impact on the FeMoco center while hardly affecting the P-cluster at all. In the IDS-oxidized state, Av1αH274A (Fig. 5A, trace 2), Av1αH451A (Fig. 5A, trace 3) and Av1αH274A/αH451A (Fig. 5A, trace 4) show the P-cluster (P2+ state) specific, g = 11.8, parallel mode EPR signals [34, 35], which integrate to 97%, 97% and 95%, respectively, of that of Av1wild-type (Fig. 5A, trace 1). This observation suggests the presence of largely intact P-clusters in these Av1 variant proteins6. In the dithionite-reduced state, however, although the FeMoco-specific, S = 3/2, perpendicular mode EPR signal  of Av1wild-type (Fig. 5B, trace 1) can be observed in the spectra of Av1αH274A (Fig. 5B, trace 2), Av1αH451A (Fig. 5B, trace 3) and Av1αH274A/αH451A (Fig. 5B, trace 4), the signal intensities in all three cases are significantly reduced, indicating a decreased FeMoco content in these Av1 variants. Indeed, the yields of FeMocos isolated from Av1αH274A (designated FeMocoαH274A), Av1αH451A (designated FeMocoαH451A) and Av1αH274A/αH451A (designated FeMocoαH274Aα/H451A) are 52%, 44% and 39%, respectively, of that from the same amount of Av1wild-type (designated FeMocowild-type) (Table 5). These numbers are consistent with the Mo contents (Table 2) and the average activities (Table 4) of these variant Av1 proteins, once again suggesting that the His triad mutations specifically affect the accumulation of FeMoco in Av1.
Although the FeMocos isolated from all three His triad Av1 variants are qualitatively indistinguishable from FeMocowild-type with regard to spectroscopic features (data not shown) and substrate reducing activities (Table 5), within the proteins—particularly those carrying the mutation of His α274 to Ala, they show EPR features that are altered from that of the wild-type (Fig. 5B). Compared to the S = 3/2 signal of Av1wild-type (g = 4.32, 3.65, 2.01), those of Av1αH274A (g = 4.47, 3.45, 1.98) and Av1αH274A/αH451A (g = 4.44, 3.45, 1.98) are broader in line shape (Fig. 5B, traces 1, 2 and 4). In addition, Av1αH274A shows extra features at g = 5.93, 5.26 and 4.65 (Fig. 5B, trace 2); whereas Av1αH274A/αH451A displays additional features at g = 5.93, 5.26 and 4.67 (Fig. 5B, trace 4). Av1αH451A (Fig. 5B, trace 3), on the other hand, exhibits the same S = 3/2 signal as Av1wild-type (Fig. 5B, trace 1), albeit at much reduced intensity.
The perturbed S = 3/2 signal in Av1αH274A or Av1αH274A/αH451A indicates a change in the immediate protein environment surrounding the FeMoco, rather than the presence of other EPR active species in the protein. Despite their altered shapes, the S = 3/2 signals of Av1αH274A (Fig. 6A and B, trace 2) and Av1αH274A/αH451A (Fig. 6A and B, trace 4) show the same temperature- and power-dependency as the normally shaped signals of Av1wild-type (Fig. 6A and B, trace 1) and Av1αH451A (Fig. 6A and B, trace 3). Upon variation of temperature or power, the intensities of the extra EPR features in the spectra of Av1αH274A (g = 5.93, 5.26 and 4.65) and Av1αH274A/αH451A (g = 5.93, 5.26 and 4.67) remain proportional to those of the major features of the S = 3/2 signal, reaching a maximum at 6 K (Fig. 6A) or 50 mW (Fig. 6B). Thus, these additional features are associated, exclusively, with the same cluster that gives rise to the main features of the S = 3/2 signal, i.e., the FeMoco center in Av1. Given that His α274 is positioned closer to Cys α275 (one of the two FeMoco ligands) than His α451 (Fig. 1), it is not surprising that mutation of His α274 has a more severe impact on the FeMoco-specific S = 3/2 signal than that of His α451.
Reduced accumulation of FeMoco in the His triad Av1 variants does not result from non-specific denaturation or incorrect folding of the proteins. Based on structural calculations by program MUMBO , mutations of His α274 and/or His α451 to Ala present no steric problems for the overall stability of Av1 (data not shown). This prediction is corroborated by the observation that Av1αH274A, Av1αH451A and Av1αH274A/αH451A exhibit the same stability as Av1wild-type upon heat treatment or prolonged storage at room temperature (data not shown). Furthermore, like Av1wild-type, Av1αH274A, Av1αH451A or Av1αH274A/αH451A can form a similarly stable complex with Av2wild-type in the presence of MgATP (Fig. 7). All these results indicate that mutations at the His triad do not alter the stability of the protein, and that the observed decrease in FeMoco accumulation is indeed a specific outcome of altering the His triad residues that are important for FeMoco insertion.
To further assess the effect of the His triad mutations, Av1αH274A, Av1αH451A and Av1αH274A/αH451A expressed in nifB-deletion background—namely, Av1ΔnifB/αH274A, Av1ΔnifB/αH451A and Av1ΔnifB/αH274A/αH451A, respectively (Table 1), were subjected to in vitro FeMoco insertion experiments. Like the Av1ΔnifB (containing the “wild-type” His triad), Av1ΔnifB/αH274A, Av1ΔnifB/αH451A and Av1ΔnifB/αH274A/αH451A are α2β2 tetramers that are FeMoco-deficient yet P-cluster-replete (data not shown). However, while Av1ΔnifB can be fully reconstituted by isolated FeMoco, Av1ΔnifB/αH274A, Av1ΔnifB/αH451A and Av1ΔnifB/αH274A/αH451A can only be reconstituted to reduced levels (Fig. 8). Furthermore, reconstitution reactions of Av1ΔnifB/αH274A, Av1ΔnifB/αH451A and Av1ΔnifB/αH274A/αH451A occur at significantly slower rates than Av1ΔnifB, especially during the initial phase (Fig. 8, inset). The levels of FeMoco incorporation under in vitro conditions—in the ascending order of Av1ΔnifB/αH274A/αH451A < Av1ΔnifB/αH451A < Av1ΔnifB/αH274A < Av1ΔnifB, reflect the amounts of FeMoco accumulation under in vivo conditions—in the increasing sequence of Av1αH274A/αH451A < Av1αH451A < Av1αH274A < Av1wild-type, convincingly making the case that the His triad variants are specifically defective in FeMoco insertion.
Histidine is capable of coordinating metals in the cluster-containing proteins, and there is no exception for that in the case of nitrogenase. In Av1wild-type, His α442 serves as one of the two protein ligands for FeMoco by ligating the Mo at the Mo/homocitrate end of the cofactor. In Av1ΔnifB, His α442 is grouped with two other histidines, His α274 and His α451, in forming a histidine-enriched cluster midway down the insertion funnel. It is very likely that, along with His α442, His α274 and His α451 provide an intermediate docking point for the FeMoco by transiently ligating to the Mo atom of the FeMoco. Meanwhile, His (pKa ≈ 6) is positively charged, at least partially, under physiological conditions (pH ≈ 6–7). Thus, in addition to coordinating the FeMoco through the Mo atom, the His triad could also provide a positively charged milieu that further attracts the FeMoco through its negatively charged homocitrate entity. Either way, our data clearly indicate that the histidine enrichment at this position is important for FeMoco incorporation. It is plausible that the temporary docking of FeMoco at the His triad allows His α442 to establish contact with the Mo atom of FeMoco and subsequently guide the FeMoco to its target location in a swinging action that covers a distance of around 5 Å (Fig. 1, arrow).
In addition to playing a pivotal role in the final step of FeMoco assembly, the His triad residues are, apparently, also closely involved in the process of substrate reduction. In contrast to the mutations of Trp α444 or His α362 (other key residues for FeMoco insertion) to Ala, which lead to proportionally reduced production of NH3 and H2 [19, 20], mutations of His α274 and/or His α451 to Ala cause a shift of N2 reduction toward H2 production (or proton reduction), indicating a re-distribution of available electrons in favor of H2 evolution (Table 3). Interestingly, early x-ray structural analysis of Av1wild-type has led to the hypothesis that His α274 and His α451 are potentially involved in the shuffling of protons to the active site (FeMoco), a process that is indispensable for N2 reduction . Since understanding the concurrent production of H2 during N2 reduction is central to elucidating the overall reaction mechanism of nitrogenase catalysis, the His triad residues could serve as ideal targets for future mechanistic studies of nitrogenase in this respect.
In summary, we show that His α274 and His α451 of Av1, part of a His triad, are specifically involved in FeMoco insertion. Mutations of His α274 and/or His α451 to small uncharged Ala result in significantly reduced levels of FeMoco accumulation, indicating that the His triad is important for FeMoco incorporation, potentially serving as the intermediate docking point for FeMoco through transitory ligand coordination and/or electrostatic interaction. Future studies will focus on structural analysis of the His triad Av1 variants so that the precise functions of these important residues in nitrogenase assembly and catalysis can be defined.
We wish to acknowledge Dr. Martin Stiebritz (Friedrich-Alexander-University Erlangen-Nürnberg, Germany) for his kind help on protein stability prediction using program MUMBO. This work was supported by NIH grant GM-67626 (M.W.R.). This paper is dedicated to the memory of Prof. Ed Stiefel.
1The identity of X is unknown but it is considered to be C, O or N .
2These numbers are based on a sequence comparison of known nitrogenase MoFe proteins (SWISS-PROT) using the program BLAST .
3Only organisms capable of expressing active nitrogenase enzymes are able to grow under the N2-fixing conditions, where no fixed or organic nitrogen (such as ammonia and urea) is present in the culture media.
4Although point mutations of Av1 have no direct effect on Av2, it has been well established that Av2 is involved in FeMoco biosynthesis . Since this work deals with FeMoco insertion, it is crucial to establish the catalytic and biosynthetic proficiency of Av2 in these variant strains.
5Note that for protein purification, all strains were grown in the presence of a limited amount of ammonia, so that a certain amount of cell mass could be accumulated regardless of the strain’s ability to fix nitrogen. After the consumption of ammonia, cells were de-repressed for 3 h, during which process the nitrogenase protein (active or inactive) was expressed. Under these conditions, a consistent yield of ~ 200 g cells per 180 L cell growth batch could be obtained for the wild-type and variant strains of A. vinelandii.
6In the dithionite-reduced state, Av1αH274A, Av1αH451A and Av1αH274A/αH451A show additional, minor S = 1/2 signals that integrate to roughly 0.04, 0.05 and 0.10 spin per protein, respectively (Fig. 5B, traces 2–4). This unique signal has been previously associated with a P-cluster analog, which is composed of two [4Fe-4S]-like fragments . It is likely, therefore, that mutations of the His triad residues near the FeMoco binding site somehow affect the P-cluster site and force a small portion of the substructural units of the [8Fe-7S] P-clusters slightly apart. Nevertheless, this portion of P-cluster species is negligible since the P-cluster (P2+ state) specific, parallel-mode EPR signal of all three Av1 variants integrate to nearly 100% of that of the wild type protein (Fig. 5A)
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.