Site-directed mutagenesis of the Klebsiella pneumoniae nitrogenase. Effects of modifying conserved cysteine residues in the alpha- and beta-subunits

Second and Outer Coordination Sphere Effects in Nitrogenase, Hydrogenase, Formate Dehydrogenase, and CO Dehydrogenase

Gases like H₂, N₂, CO₂, and CO are increasingly recognized as critical feedstock in "green" energy conversion and as sources of nitrogen and carbon for the agricultural and chemical sectors. However, the industrial transformation of N₂, CO₂, and CO and the production of H₂ require significant energy input, which renders processes like steam reforming and the Haber-Bosch reaction economically and environmentally unviable. Nature, on the other hand, performs similar tasks efficiently at ambient temperature and pressure, exploiting gas-processing metalloenzymes (GPMs) that bind low-valent metal cofactors based on iron, nickel, molybdenum, tungsten, and sulfur. Such systems are studied to understand the biocatalytic principles of gas conversion including N₂ fixation by nitrogenase and H₂ production by hydrogenase as well as CO₂ and CO conversion by formate dehydrogenase, carbon monoxide dehydrogenase, and nitrogenase. In this review, we emphasize the importance of the cofactor/protein interface, discussing how second and outer coordination sphere effects determine, modulate, and optimize the catalytic activity of GPMs. These may comprise ionic interactions in the second coordination sphere that shape the electron density distribution across the cofactor, hydrogen bonding changes, and allosteric effects. In the outer coordination sphere, proton transfer and electron transfer are discussed, alongside the role of hydrophobic substrate channels and protein structural changes. Combining the information gained from structural biology, enzyme kinetics, and various spectroscopic techniques, we aim toward a comprehensive understanding of catalysis beyond the first coordination sphere.

Biosynthesis of Nitrogenase Cofactors

Nitrogenase harbors three distinct metal prosthetic groups that are required for its activity. The simplest one is a [4Fe-4S] cluster located at the Fe protein nitrogenase component. The MoFe protein component carries an [8Fe-7S] group called P-cluster and a [7Fe-9S-C-Mo-R-homocitrate] group called FeMo-co. Formation of nitrogenase metalloclusters requires the participation of the structural nitrogenase components and many accessory proteins, and occurs both in situ, for the P-cluster, and in external assembly sites for FeMo-co. The biosynthesis of FeMo-co is performed stepwise and involves molecular scaffolds, metallochaperones, radical chemistry, and novel and unique biosynthetic intermediates. This review provides a critical overview of discoveries on nitrogenase cofactor structure, function, and activity over the last four decades.

MECHANISTIC FEATURES OF THE MO-CONTAINING NITROGENASE

Nitrogenase is the complex metalloenzyme responsible for biological dinitrogen reduction. This reaction represents the single largest contributor to the reductive portion of the global nitrogen cycle. Recent developments in understanding the mechanism of the Mo-based nitrogenase are reviewed. Topics include how nucleotide binding and hydrolysis are coupled to electron transfer and substrate reduction, how electrons are accumulated and transferred within the MoFe-protein, and how substrates bind and are reduced at the active site metal cluster.

New insights into structure-function relationships in nitrogenase: A 1.6 A resolution X-ray crystallographic study of Klebsiella pneumoniae MoFe-protein

The X-ray crystal structure of Klebsiella pneumoniae nitrogenase component 1 (Kp1) has been determined and refined to a resolution of 1.6 A, the highest resolution reported for any nitrogenase structure. Models derived from three 1.6 A resolution X-ray data sets are described; two represent distinct oxidation states, whilst the third appears to be a mixture of both oxidized and reduced states (or perhaps an intermediate state). The structures of the protein and the iron-molybdenum cofactor (FeMoco) appear to be largely unaffected by the redox status, although the movement of Ser beta90 and a surface helix in the beta subunit may be of functional significance. By contrast, the 8Fe-7S P-cluster undergoes discrete conformational changes involving the movement of two iron atoms. Comparisons with known component 1 structures reveal subtle differences in the FeMoco environment, which could account for the lower midpoint potential of this cluster in Kp1. Furthermore, a non-proline- cis peptide bond has been identified in the alpha subunit that may have a functional role. It is within 10 A of the FeMoco and may have been overlooked in other component 1 models. Finally, metal-metal and metal-sulphur distances within the metal clusters agree well with values derived from EXAFS studies, although they are generally longer than the values reported for the closely related protein from Azotobacter vinelandii. A number of bonds between the clusters and their ligands are distinctly longer than the EXAFS values, in particular, those involving the molybdenum atom of the FeMoco.

Isolation and characterization of nitrogenase MoFe protein from the mutant strain pHK17 of Klebsiella pneumoniae in which the two bridging cysteine residues of the P-clusters are replaced by the non-coordinating amino acid alanine

Nitrogenase MoFe protein (Kp1) from the mutant strain pHK17 or Klebsiella pneumoniae has been purified to give three catalytically active fractions. In this mutant, each of the two bridging cysteine ligands to the P-clusters, alpha-Cys-89 and beta-Cys-94, has been replaced by a non-coordinating residue, alanine. SDS/PAGE and earlier native gels showed that the three fractions retained the normal alpha 2 beta 2 tetrameric form of wild-type Kp1; therefore we conclude that in each of the fractions the subunits are folded differently, thus resulting in different surface charges and allowing separation of the fractions on ion-exchange chronatography. Earlier EPR and magnetic CD data had shown that the mutant fractions contain P-clusters, and thus the mutated residues are not as essential for maintaining the integrity of the P-clusters as they appear from the X-ray structure. The specific activity of each of the three fractions was less than that of wild-type Kp1, the most active fraction having only 50% of wild-type activity. No change in substrate specificity or in the relative distribution of electrons to various substrates was found. The relationship between ATP hydrolysis and substrate-reducing activity, the EPR spectra of the S = 3/2 spin state of the iron-molybdenum cofactor (FeMoco) and the pH profile of acetylene-reduction activities of the three fractions did not differ significantly from those exhibited by wild-type Kp1. The specific activities of the three mutant fractions and of wild-type Kp1 were linearly proportional to the intensity of the S = 3/2 EPR signal from the FeMoco centres. This implies that those molecules of the three mutant fractions and the wild-type protein that contain EPR-active FeMoco are fully active, i.e. that the Cys to Ala substitution of the P-cluster ligands does not affect the specific activity of the protein. This in turn implies that the P-clusters are not directly associated with the rate-limiting step in enzyme turnover. We conclude that the lower specific activities of the mutant fractions are observed because the fractions are mixtures of species containing a full complement of FeMoco and P-clusters and species lacking some or all of these clusters. On the basis of the Mo contents and EPR spectroscopy of the mutant fractions, we propose that the loss of the P-clusters causes (i) the physical loss or inhibition of binding of some FeMoco; (ii) the EPR and catalytic inactivation of some FeMoco; and/or (iii) the incorporation of a FeMoco-like species into the FeMoco site of the mutant molecules.

The genes encoding the delta subunits of dinitrogenases 2 and 3 are required for mo-independent diazotrophic growth by Azotobacter vinelandii

vnfG and anfG encode the delta subunits of alternative nitrogenases 2 and 3 in Azotobacter vinelandii, respectively. As a first step towards elucidating the role of these subunits, diazotrophic growth and acetylene reduction studies were conducted on mutants containing alterations in the genes encoding these subunits. Mutants containing a stop codon (C36stop) or an in-frame deletion in anfG were unable to grow in N-free, Mo-deficient medium (Anf-). Mutants in which cysteine 36 of AnfG (a residue conserved between VnfG and AnfG) was changed to Ala or Ser were Anf+. Thus, this conserved cysteine is not essential for the function of AnfG in dinitrogenase 3. A mutant with a stop codon in vnfG (C17stop) grew after a lag of 25 h in N-free, Mo-deficient medium containing V2O5. However, a Nif- Anf- strain with this mutation was unable to grow under these conditions. This shows that the vnfG gene product is required for nitrogenase 2-dependent growth. Strains with mutations in vnfG and anfG reduced acetylene to different degrees. This indicates that the delta subunits are not required for acetylene reduction by nitrogenases 2 and 3.

Selenol binds to iron in nitrogenase iron-molybdenum cofactor: an extended x-ray absorption fine structure study

The biological N2-fixation reaction is catalyzed by the enzyme nitrogenase. The metal cluster active site of this enzyme, the iron-molybdenum cofactor (FeMoco), can be studied either while bound within the MoFe protein component of nitrogenase or after it has been extracted into N-methylformamide. The two species are similar but not identical. For example, the addition of thiophenol or selenophenol to isolated FeMoco causes its rather broad S = 3/2 electron paramagnetic resonance signal to sharpen and more closely approach the signal exhibited by protein-bound FeMoco. The nature of this thiol/selenol binding site has been investigated by using Se-K edge extended x-ray absorption fine structure (EXAFS) to study selenophenol ligated to FeMoco, and the results are reported here. EXAFS data analysis at the ligand Se-K edge was performed with a set of software, GNXAS, that provides for direct calculation of the theoretical EXAFS signals and least-squares fits to the experimental data. Data analysis results show definitively that the selenol (and by inference thiol) binds to Fe at a distance of 2.4 A. In contrast, unacceptable fits are obtained with either Mo or S as the liganded atom (instead of Fe). These results provide quantitative details about an exchangeable thiol/selenol binding site on FeMoco in its isolated, solution state and establish an Fe atom as the site of this reaction. Furthermore, the utility of ligand-based EXAFS as a probe of coordination in polynuclear metal clusters is demonstrated.

Nitrogenase and biological nitrogen fixation

Biological nitrogen fixation is catalyzed by the nitrogenase enzyme system which consists of two metalloproteins, the iron (Fe-) protein and the molybdenum-iron (MoFe-) protein. Together, these proteins mediate the ATP-dependent reduction of dinitrogen to ammonia. Recent crystallographic analyses of Fe-protein and MoFe-protein have revealed the polypeptide fold and the structure and organization of the unusual metal centers in nitrogenase. These structure provide a molecular framework for addressing the mechanism of the nitrogenase-catalyzed reaction. General features of the nitrogenase system, including conformational coupling of nucleotide hydrolysis, aspects of the cluster structures, and the general spatial organization of redox centers within the protein subunits, are relevant to a wide range of biochemical systems.

Electron-paramagnetic-resonance and magnetic-circular-dichroism studies of the binding of cyanide and thiols to the thiols to the iron-molybdenum cofactor from Klebsiella pneumoniae nitrogenase

FeMoco, a low-M(r) metal cluster of probable composition Fe7MoS9 complexed with homocitrate, has been extracted with N-methylformamide from the MoFe protein of the nitrogenase enzyme from Klebsiella pneumoniae. The binding of cyanide and thiols to the FeMoco cluster in its paramagnetic S = 3/2 oxidation level has been studied by low-temperature e.p.r. and magnetic-circular-dichroism (m.c.d.) spectroscopies. Cyanide binds to isolated FeMoco at more than one site, and causes changes in the g values form g = 4.6, 3.2, 2.0 to g = 4.29, 3.82, 2.02 E.p.r. competition studies indicate that one cyanide can be displaced by thiolate from one type of site. The form of the low-temperature m.c.d. spectrum is little changed by ligand binding, thus the basic cluster structure remains intact. However, when benzenethiol is bound, a new intense band (lambda 387 nm) is observed, indicating the generation of an increased ligand-to-cluster charge-transfer interaction.

Molybdenum-independent nitrogenases of Azotobacter vinelandii: a functional species of alternative nitrogenase-3 isolated from a molybdenum-tolerant strain contains an iron-molybdenum cofactor

Nitrogenase-3 of Azotobacter vinelandii is synthesized under conditions of molybdenum and vanadium deficiency. The minimal metal requirement for its synthesis, and its metal content, indicated that the only transition metal in nitrogenase-3 was iron [Chisnell, Premakumar and Bishop (1988) J. Bacteriol. 170, 27-33; Pau, Mitchenall and Robson (1989) J. Bacteriol. 171, 124-129]. A new species of nitrogenase-3 has been purified from a strain of A. vinelandii (RP306) lacking structural genes for the Mo- and V-nitrogenases and containing a mutation which enables nitrogenase-3 to be synthesized in the presence of molybdenum. SDS/PAGE showed that component 1 contained a 15 kDa polypeptide which N-terminal amino acid sequence determination showed to be encoded by anfG. This confirms that nitrogenase-3, like V-nitrogenase, comprises three subunits. Preparations of the nitrogenase-3 from strain RP306 contained 24 Fe atoms and 1 Mo atom per molecule. Characterization of the cofactor centre of the enzyme by e.p.r. spectroscopy and an enzymic cofactor assay, together with stimulation of the growth of strain RP306 by Mo, showed that nitrogenase-3 can incorporate the Mo-nitrogenase cofactor (FeMoco) to form a functional enzyme. The specific activities (nmol of product produced/min per mg of protein) determined from activity titration curves were: under N2, NH3 formation 110, with concomitant H2 evolution of 220; under argon, H2 evolution 350; under 10% acetylene (C2H2) in argon, ethylene (C2H4) 58, ethane (C2H6) 26, and concomitant H2 evolution 226. The rate of formation of C2H6 was non-linear, and the C2H6/C2H4 ratio strongly dependent on the ratio of nitrogenase components.

Removal of the high-potential [4Fe-4S] center of the beta-subunit from Escherichia coli nitrate reductase. Physiological, biochemical, and EPR characterization of site-directed mutated enzymes

The beta-subunit of the nitrate reductase of Escherichia coli contains four groups of Cys residues (I-IV) which are thought to bind the single [3Fe-4S] center and the three [4Fe-4S] centers. The first or second Cys residue of group I was substituted by site-directed mutagenesis with Ala or Ser. Physiological, biochemical, and EPR studies were performed on the mutated enzymes. With small variations, the properties of these mutant enzymes do not differ from one another. They were found to be as abundant and as stably bound to the membrane as the native enzyme, provided the gamma-subunit was present. Although physiological activity was reduced, it was sufficient to allow growth on nitrate. The study of variations in EPR intensity as a function of the redox potential indicated that these enzymes only contained three iron-sulfur centers instead of the usual four in the native enzyme. Spectral EPR analysis showed that the [4Fe-4S] center of high redox potential (center 1, +80 mV) was missing. The loss of this center did not affect the stable integration of the other three centers. The data presented here are in total contrast to those we have reported for each of the other three centers (centers 2-4), the loss of which was detrimental to the integration of all centers and to the integration of the molybdenum cofactor (Augier et al., in press). Taken together, our results demonstrated that the first and second Cys residues of group I are the ligands of the [4Fe-4S] center (center 1, +80 mV) and that this center participates in electron transfer, but is dispensable. On the basis of these results, it is proposed that the [3Fe-4S] center (center 2, +60 mV) also plays a biological role and that in the native enzyme both high-potential centers, centers 1 and 2, contribute independently and in parallel to the electron transfer to the molybdenum cofactor.

Klebsiella pneumoniae nitrogenase MoFe protein: chymotryptic proteolysis affects function by limited cleavage of the beta-chain and provides high-specific-activity MoFe protein

Proteinase treatment with chymotrypsin has been used to probe the structure of native Klebsiella pneumoniae nitrogenase MoFe protein (Kp1). Reaction with chymotrypsin did not bleach Kp1, suggesting that it did not destroy the metal centres, and the Mo and Fe contents of Kp1 were unchanged. High ratios of chymotrypsin to Kp1 (1:1 by mass) cleaved the beta-chain of Kp1 to give 44 and 14 kDa polypeptides, which N-terminal amino acid sequence analysis showed to be derived from cleavage at residue beta-Phe124. A mutant MoFe protein, Kp1Met-124, in which beta-Phe124 is replaced by methionine, was not cleaved by chymotrypsin. Under non-denaturing conditions, the 'nicked' beta-chain of the wild-type protein remained associated with the alpha-chain. The alpha-chain was not cleaved by the proteinase treatment. Fission of the wild-type beta-chain was accompanied by loss of enzyme activity, loss of intensity of the g = 3.7 e.p.r. signal derived from dithionite-reduced FeMoco and by changes in the visible spectrum. The e.p.r. spectra of potassium ferricyanide-oxidized native and digested Kp1 show differences in the signals between g = 1.6 and 2.0. After prolonged treatment, the final specific activity of Kp1 was about 25 +/- 5% of the initial activity. This corresponded to 25 +/- 5% of the beta-chain which was resistant to proteolytic action. Brief treatment of Kp1 with a lower concentration of chymotrypsin (chymotrypsin/Kp1 ratio = 1:10 by mass, for 10 min) preferentially cleaved high-molecular-mass polypeptides that routinely contaminate preparations of Kp1 prepared by standard procedures. Treatment with chymotrypsin followed by gel filtration to remove the proteinase and cleaved protein fragments can therefore be used to increase significantly the specific activity of Kp1 preparations and remove contaminating activities, such as the ATPase activity of myokinase.

Site-directed mutagenesis of conserved cysteine residues within the beta subunit of Escherichia coli nitrate reductase. Physiological, biochemical, and EPR characterization of the mutated enzymes

We have used site-directed mutagenesis to alter the ligands to the iron-sulfur centers of Escherichia coli nitrate reductase A. The beta subunit of this enzyme contains four Cys groups which are thought to accommodate the single [3Fe-4S] center and the three [4Fe-4S] centers involved in the electron-transfer process from quinol to nitrate. The third Cys group (group III) contains a Trp at a site occupied by a Cys residue in typical ferredoxin arrangements or in the DmsB subunit of dimethyl sulfoxide (DMSO) reductase. In an attempt to determine the coordination site of the different iron-sulfur centers in the amino acid sequence, we have changed the Trp of group III to Cys, Ala, Phe, and Tyr and the first Cys residue of groups II-IV to Ala and Ser. Physiological, biochemical, and EPR studies were performed on the mutated enzymes. Substitution of Ala for either Cys184, Cys217, or Cys244 results in the full loss of all four iron-sulfur centers present in the wild-type enzyme. These inactive enzymes still possess the alpha,beta, and gamma polypeptides associated in a membrane-bound complex. These Cys have important structural roles and are very likely involved in the coordination of the iron-sulfur centers. Substitution of Cys184 with a Ser residue produces an enzyme containing the four iron-sulfur centers, but displaying reduced activity. EPR studies suggest that Cys184 is a ligand of the [4Fe-4S] center whose midpoint potential is -200 mV in the native enzyme. All substitutions performed in this study on Trp220 lead to mutant enzymes harboring the four iron-sulfur centers and a nitrate reductase activity close to that of the wild-type. In spite of the high similarity between the NarH and DmsB subunits, the Trp220-->Cys substitution does not allow the conversion of the [3Fe-4S] center of the nitrate reductase into a [4Fe-4S] center. Therefore, Trp220 does not seem to play any major role in the beta subunit.

Structural models for the metal centers in the nitrogenase molybdenum-iron protein

Structural models for the nitrogenase FeMo-cofactor and P-clusters are proposed based on crystallographic analysis of the nitrogenase molybdenum-iron (MoFe)-protein from Azotobacter vinelandii at 2.7 angstrom resolution. Each center consists of two bridged clusters; the FeMo-cofactor has 4Fe:3S and 1Mo:3Fe:3S clusters bridged by three non-protein ligands, and the P-clusters contain two 4Fe:4S clusters bridged by two cysteine thiol ligands. Six of the seven Fe sites in the FeMo-cofactor appear to have trigonal coordination geometry, including one ligand provided by a bridging group. The remaining Fe site has tetrahedral geometry and is liganded to the side chain of Cys alpha 275. The Mo site exhibits approximate octahedral coordination geometry and is liganded by three sulfurs in the cofactor, two oxygens from homocitrate, and the imidazole side chain of His alpha 442. The P-clusters are liganded by six cysteine thiol groups, two which bridge the two clusters, alpha 88 and beta 95, and four which singly coordinate the remaining Fe sites, alpha 62, alpha 154, beta 70, and beta 153. The side chain of Ser beta 188 may also coordinate one iron. The polypeptide folds of the homologous alpha and beta subunits surrounding the P-clusters are approximately related by a twofold rotation that may be utilized in the binding interactions between the MoFe-protein and the nitrogenase Fe-protein. Neither the FeMo-cofactor nor the P-clusters are exposed to the surface, suggesting that substrate entry, electron transfer, and product release must involve a carefully regulated sequence of interactions between the MoFe-protein and Fe-protein of nitrogenase.

Cys184 and Cys187 of NifL protein of Klebsiella pneumoniae are not absolutely required for inhibition of NifA activity

The sequence Cys184-Ala-Asp-Cys187 in the NifL protein of Klebsiella pneumoniae, for which a role in oxygen sensing and/or metal binding has been proposed, was altered by introducing two mutations, Cys184----Ala and Cys187----Ala, using oligodeoxyribonucleotide-directed mutagenesis. Neither mutation abolished ammonium or oxygen control of nif transcription, although some impairment of function was apparent. The two Cys residues are therefore unlikely to have a direct role in oxygen sensing or metal binding, but probably make some contribution to protein folding or stability.

Altered nitrogenase MoFe proteins from Azotobacter vinelandii. Analysis of MoFe proteins having amino acid substitutions for the conserved cysteine residues within the beta-subunit

The regions surrounding the three strictly conserved cysteine residues (positions 70, 95 and 153) in the beta-subunit of the Azotobacter vinelandii nitrogenase MoFe protein have been proposed to provide P-cluster environments [Dean, Setterquist, Brigle, Scott, Laird & Newton (1990) Mol. Microbiol. 4, 1505-1512]. In the present study, each of these cysteine residues was individually substituted by either serine or alanine by site-directed mutagenesis of the nifK gene, which encodes the MoFe protein beta-subunit. A mutant strain for which the codon for Cys-153 is removed was also isolated. Significant structural or functional roles are indicated for the cysteine residues at positions 70 and 95, where substitution by either serine or alanine eliminates diazotrophic growth of the resulting strains and abolishes or markedly decreases both MoFe-protein acetylene-reduction activity and the intensity of the whole-cell S = 3/2 e.p.r. signal. Changes introduced at position 153 have various effects on the functional properties of the enzyme. The strains produced either by deletion of the Cys-153 residue or its substitution by serine exhibit only a moderate decrease in diazotrophic growth and MoFe-protein activity and no loss of the whole-cell e.p.r.-signal intensity. In contrast, substitution by alanine eliminates diazotrophic growth and very markedly decreases both MoFe-protein activity and e.p.r.-signal intensity. These results are interpreted in terms of a metallocluster-driven protein rearrangement. After purification of the altered MoFe protein, in which serine replaces Cys-153, an investigation of its catalytic and spectroscopic properties confirms that neither the FeMo cofactor, i.e. the substrate-reduction site, nor the component-protein interaction site has been affected. Instead, these data indicate a disruption in electron transfer within the MoFe protein, which is consistent with a role for this residue (and region) at the P clusters.

Role for the nitrogenase MoFe protein alpha-subunit in FeMo-cofactor binding and catalysis

Two components constitute Mo-dependent nitrogenase (EC 1.18.6.1)--the Fe protein (a homodimer encoded by nifH) and the MoFe protein (an alpha 2 beta 2 tetramer encoded by nifDK). The MoFe protein provides the substrate-binding site and probably contains six prosthetic groups of two types--four Fe-S centres and two Fe- and Mo-containing cofactors. To determine the distribution and catalytic function of these metalloclusters, we and others are attempting to change the catalytic and spectroscopic features of nitrogenase by substituting specific amino acids targeted as potential metallocluster ligands, particularly those to the FeMo-cofactor, which is responsible for the biologically unique electron paramagnetic resonance signal (S = 3/2) of nitrogenase, and is believed to be the N2-reducing site. Here we describe mutant strains of Azotobacter vinelandii that have single amino-acid substitutions within the MoFe protein alpha-subunit. These substitutions alter both substrate-reduction properties and the unique electron paramagnetic resonance signal, indicating that the FeMo-cofactor is associated with both the alpha-subunit and the substrate-reducing site.