X-ray crystal structure of the Fe-only hydrogenase (CpI) from Clostridium pasteurianum to 1.8 angstrom resolution.

A three-dimensional structure for the monomeric iron-containing hydrogenase (CpI) from Clostridium pasteurianum was determined to 1.8 angstrom resolution by x-ray crystallography using multiwavelength anomalous dispersion (MAD) phasing. CpI, an enzyme that catalyzes the two-electron reduction of two protons to yield dihydrogen, was found to contain 20 gram atoms of iron per mole of protein, arranged into five distinct [Fe-S] clusters. The probable active-site cluster, previously termed the H-cluster, was found to be an unexpected arrangement of six iron atoms existing as a [4Fe-4S] cubane subcluster covalently bridged by a cysteinate thiol to a [2Fe] subcluster. The iron atoms of the [2Fe] subcluster both exist with an octahedral coordination geometry and are bridged to each other by three non-protein atoms, assigned as two sulfide atoms and one carbonyl or cyanide molecule. This structure provides insights into the mechanism of biological hydrogen activation and has broader implications for [Fe-S] cluster structure and function in biological systems.

Infrared spectroscopy of the nitrogenase MoFe protein under electrochemical control: potential-triggered CO binding.

We demonstrate electrochemical control of the nitrogenase MoFe protein, in the absence of Fe protein or ATP, using europium(iii/ii) polyaminocarboxylate complexes as electron transfer mediators. This allows the potential dependence of proton reduction and inhibitor (CO) binding to the active site FeMo-cofactor to be established. Reduction of protons to H2 is catalyzed by the wild type MoFe protein and ß-98Tyr→His and ß-99Phe→His variants of the MoFe protein at potentials more negative than -800 mV (vs. SHE), with greater electrocatalytic proton reduction rates observed for the variants compared to the wild type protein. Electrocatalytic proton reduction is strongly attenuated by carbon monoxide (CO), and the potential-dependence of CO binding to the FeMo-cofactor is determined by in situ infrared (IR) spectroelectrochemistry. The vibrational wavenumbers for CO coordinated to the FeMo-cofactor are consistent with earlier IR studies on the MoFe protein with Fe protein/ATP as reductant showing that electrochemically generated states of the protein are closely related to states generated with the native Fe protein as electron donor.

Aerobic, inactive forms of Azotobacter vinelandii hydrogenase: activation kinetics and insensitivity to C2H2 inhibition.

Azotobacter vinelandii hydrogenase (EC class 1.12), either purified or membrane-associated, was obtained aerobically in an inactive state. The kinetics of activation by treatment with a reductant (H2 or dithionite) were determined. Three distinct phases of the activation were observed. Aerobically prepared, inactive hydrogenase was insensitive to acetylene inhibition, but could be rendered acetylene-sensitive by reduction with dithionite. These findings indicate that acetylene inhibition of hydrogenase requires catalytically active enzyme.

Thermodynamics of nucleotide interactions with the Azotobacter vinelandii nitrogenase iron protein.

The nitrogenase iron (Fe) protein binds two molecules of MgATP or MgADP, which results in protein conformational changes that are important for subsequent steps of the nitrogenase reaction mechanism. In the present work, isothermal titration calorimetry has been used to deconvolute the apparent binding constants (K'a1 and K'a2) and the thermodynamic terms (delta H' degree and delta S' degree) for each of the two binding events of MgATP or MgADP to either the reduced or oxidized states of the Fe protein from Azotobacter vinelandii. The Fe protein was found to bind two nucleotides with positive cooperativity and the oxidation state of the [4Fe-4S] cluster of the Fe protein was found to influence the affinity for binding nucleotides, with the oxidized ([4Fe-4S]2+) state having up to a 15-fold higher affinity for nucleotides when compared to the reduced ([4Fe-4S]1+) state. The first nucleotide binding reaction was found to be driven by a large favorable entropy change (delta S' degree = 10-21 cal mol-1 K-1), with a less favorable or unfavorable enthalpy change (delta H' degree = +1.5 to -3.3 kcal mol-1). In contrast, the second nucleotide binding reaction was found to be driven by a favorable change in enthalpy (delta H' degree = -3.1 to -13.0 kcal mol-1), with generally less favorable entropy changes. A plot of the associated enthalpy (-delta H' degree) and entropy terms (-T delta S' degree) for each nucleotide and protein binding reaction revealed a linear relationship with a slope of 1.12, consistent with strong enthalpy-entropy compensation. These results indicate that the binding of the first nucleotide to the nitrogenase Fe protein results in structural changes accompanied by the reorganization of bound water molecules, whereas the second nucleotide binding reaction appears to result in much smaller structural changes and is probably largely driven by bonding interactions. Evidence is presented that the total free energy change (delta G' degree) derived from the binding of two nucleotides to the Fe protein accounts for the total change in the midpoint potential of the [4Fe-4S] cluster.

Docking of nitrogenase iron- and molybdenum-iron proteins for electron transfer and MgATP hydrolysis: the role of arginine 140 and lysine 143 of the Azotobacter vinelandii iron protein.

Docking of the nitrogenase component proteins, the iron protein (FeP) and the molybdenum-iron protein (MoFeP), is required for MgATP hydrolysis, electron transfer between the component proteins, and substrate reductions catalyzed by nitrogenase. The present work examines the function of 3 charged amino acids, Arg 140, Glu 141, and Lys 143, of the Azotobacter vinelandii FeP in nitrogenase component protein docking. The function of these amino acids was probed by changing each to the neutral amino acid glutamine using site-directed mutagenesis. The altered FePs were expressed in A. vinelandii in place of the wild-type FeP. Changing Glu 141 to Gln (E141Q) had no adverse effects on the function of nitrogenase in whole cells, indicating that this charged residue is not essential to nitrogenase function. In contrast, changing Arg 140 or Lys 143 to Gln (R140Q and K143Q) resulted in a significant decrease in nitrogenase activity, suggesting that these charged amino acid residues play an important role in some function of the FeP. The function of each amino acid was deduced by analysis of the properties of the purified R140Q and K143Q FePs. Both altered proteins were found to support reduced substrate reduction rates when coupled to wild-type MoFeP. Detailed analysis revealed that changing these residues to Gln resulted in a dramatic reduction in the affinity of the altered FeP for binding to the MoFeP. This was deduced in FeP titration, NaCl inhibition, and MoFeP protection from Fe2+ chelation experiments.(ABSTRACT TRUNCATED AT 250 WORDS)

Increasing nitrogenase catalytic efficiency for MgATP by changing serine 16 of its Fe protein to threonine: use of Mn2+ to show interaction of serine 16 with Mg2+.

MgATP-binding and hydrolysis are an integral part of the nitrogenase catalytic mechanism. We are exploring the function of MgATP hydrolysis in this reaction by analyzing the properties of the Fe protein (FeP) component of Azotobacter vinelandii nitrogenase altered by site-directed mutagenesis. We have previously (Seefeldt, L.C., Morgan, T.V., Dean, D.R., & Mortenson, L.E., 1992, J. Biol. Chem. 267, 6680-6688) identified a region near the N-terminus of FeP that is involved in interaction with MgATP. This region of FeP is homologous to the well-known nucleotide-binding motif GXXXXGKS/T. In the present work, we examined the function of the four hydroxyl-containing amino acids immediately C-terminal to the conserved lysine 15 that is involved in interaction with the gamma-phosphate of MgATP. We have established, by altering independently Thr 17, Thr 18, and Thr 19 to alanine, that a hydroxyl-containing residue is not needed at these positions for FeP to function. In contrast, an hydroxyl-containing amino acid at position 16 was found to be critical for FeP function. When the strictly conserved Ser 16 was altered to Ala, Cys, Asp, or Gly, the FeP did not support N2 fixation when expressed in place of the wild-type FeP in A. vinelandii. Altering Ser 16 to Thr (S16T), however, resulted in the expression of an FeP that was partially active. This S16T FeP was purified to homogeneity, and its biochemical examination allowed us to assign a catalytic function to this hydroxyl group in the nitrogenase mechanism. Of particular importance was the finding that the S16T FeP had a significantly higher affinity for MgATP than the wild-type FeP, with a measured Km of 20 microM compared to the wild-type FeP Km of 220 microM. This increased kinetic affinity for MgATP was reflected in a significantly stronger binding of the S16T FeP for MgATP. In contrast, the affinity for MgADP, which binds at the same site as MgATP, was unchanged. The catalytic efficiency (kcat/Km) of S16T FeP was found to be 5.3-fold higher than for the wild-type FeP, with the S16T FeP supporting up to 10 times greater nitrogenase activity at low MgATP concentrations. This indicates a role for the hydroxyl group at position 16 in interaction with MgATP but not MgADP. The site of interaction of this residue was further defined by examining the properties of wild-type and S16T FePs in utilizing MnATP compared with MgATP.(ABSTRACT TRUNCATED AT 400 WORDS)

Purification to homogeneity of Azotobacter vinelandii hydrogenase: a nickel and iron containing alpha beta dimer.

Azotobacter vinelandii hydrogenase has been purified to homogeneity from membranes. The enzyme was solubilized with Triton X-100 followed by ammonium sulfate-hexane extractions to remove lipids and detergent. The enzyme was then purified by carboxymethyl-Sepharose and octyl-Sepharose column chromatography. All purification steps were performed under anaerobic conditions in the presence of dithionite and dithiothreitol. The enzyme was purified 143-fold from membranes to a specific activity of 124 mumol of H2 uptake . min-1 . mg protein-1. Nondenaturing polyacrylamide gel electrophoresis of the hydrogenase revealed a single band which stained for both activity and protein. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis revealed two bands corresponding to peptides of 67,000 and 31,000 daltons. Densitometric scans of the SDS-gel indicated a molar ratio of the two bands of 1.07 +/- 0.05. The molecular weight of the native enzyme was determined by three different methods. While gel permeation gave a molecular weight of 53,000, sucrose density gradient centrifugation and native polyacrylamide gel electrophoresis gave molecular weights of 98,600 +/- 10,000 and 98,600 +/- 2,000, respectively. We conclude that the A. vinelandii hydrogenase is an alpha beta dimer (98,000 daltons) with subunits of 67,000 and 31,000 daltons. Analyses for nickel and iron indicated 0.68 +/- 0.01 mol Ni/mol hydrogenase and 6.6 +/- 0.5 mol Fe/mol hydrogenase. The isoelectric point of the enzyme was 6.1 +/- 0.01. In addition, several catalytic properties of the enzyme have been examined. The Km for H2 was 0.86 microM, and H2 evolution was observed in the presence of reduced methyl viologen. The pH profile of enzyme activity with methylene blue as the electron acceptor has been determined, along with the Km and Vmax for various electron acceptors.

The role of the MoFe protein alpha-125Phe and beta-125Phe residues in Azotobacter vinelandii MoFe protein-Fe protein interaction.

Site-directed mutagenesis and gene-replacement techniques were used to substitute alanine for the MoFe protein alpha- and beta-subunit phenylalanine-125 residues both separately and in combination. These residues are located on the surface of the MoFe protein near the pseudosymmetric axis of symmetry between the alpha- and beta-subunits. Altered MoFe proteins that contain an alanine substitution at only one of the respective positions exhibit proton reduction activities of about 25-50% when compared to that of the wild-type protein. The lower level of proton reduction also corresponds with decreases in the rates of MgATP hydrolysis. The MoFe protein which contains alanine substitutions in both the alpha- and beta- subunits did not exhibit any proton reduction activity or MgATP hydrolysis. Stopped flow spectrophotometry of the singly substituted MoFe proteins indicate primary electron transfer rate constants approximately an order of magnitude slower than what is observed for wild-type MoFe protein, while no primary electron transfer is observed for the doubly substituted MoFe protein. The doubly substituted MoFe protein is able to interact with the Fe protein as shown by chemical crosslinking experiments. However, this protein does not form a tight complex with the Fe protein when treated with MgADP-AlF4- or when using the altered 127delta Fe protein. Stopped flow spectrophotometry was also used to quantitate the first-order dissociation rate constants for the two component proteins. These results suggest that the 125Phe residues are involved in an early event(s) that occurs upon component protein docking and could be involved in eliciting MgATP hydrolysis.

Hydrolysis of nucleoside triphosphates other than ATP by nitrogenase.

The hydrolysis of ATP to ADP and P(i) is an integral part of all substrate reduction reactions catalyzed by nitrogenase. In this work, evidence is presented that nitrogenases isolated from Azotobacter vinelandii and Clostridium pasteurianum can hydrolyze MgGTP, MgITP, and MgUTP to their respective nucleoside diphosphates at rates comparable to those measured for MgATP hydrolysis. The reactions were dependent on the presence of both the iron (Fe) protein and the molybdenum-iron (MoFe) protein. The oxidation state of nitrogenase was found to greatly influence the nucleotide hydrolysis rates. MgATP hydrolysis rates were 20 times higher under dithionite reducing conditions (approximately 4,000 nmol of MgADP formed per min/mg of Fe protein) as compared with indigo disulfonate oxidizing conditions (200 nmol of MgADP formed per min/mg of Fe protein). In contrast, MgGTP, MgITP, and MgUTP hydrolysis rates were significantly higher under oxidizing conditions (1,400-2,000 nmol of MgNDP formed per min/mg of Fe protein) as compared with reducing conditions (80-230 nmol of MgNDP formed per min/mg of Fe protein). The K(m) values for MgATP, MgGTP, MgUTP, and MgITP hydrolysis were found to be similar (330-540 microM) for both the reduced and oxidized states of nitrogenase. Incubation of Fe and MoFe proteins with each of the MgNTP molecules and AlF(4)(-) resulted in the formation of non-dissociating protein-protein complexes, presumably with trapped AlF(4)(-) x MgNDP. The implications of these results in understanding how nucleotide hydrolysis is coupled to substrate reduction in nitrogenase are discussed.

Electron transfer from the nitrogenase iron protein to the [8Fe-(7/8)S] clusters of the molybdenum-iron protein.

The reduction of substrates catalyzed by nitrogenase requires electron transfer between the iron (Fe) protein and the molybdenum-iron (MoFe) protein in a reaction that is coupled to the hydrolysis of MgATP. The [4Fe-4S] cluster of the Fe protein transfers one electron ultimately to the M-clusters (FeMoco) of the MoFe protein for substrate reduction, with the P-clusters ([8Fe-(7/8)S]) of the MoFe protein as proposed electron transfer intermediates. This work presents direct EPR evidence for primary electron transfer from the [4Fe-4S] cluster of the Fe protein to the P-clusters of the MoFe protein in a reaction that requires the MgATP-bound state of the Fe protein. An oxidized state of the MoFe protein was prepared in which the P-clusters were oxidized by 2 equiv of electrons to the P2+ state. In this oxidation state, the M-clusters (S = 3/2) and the P(2+-clusters (S > or = 3) are paramagnetic and can be observed by perpendicular and parallel mode EPR, providing the opportunity to follow electron transfer from the Fe protein to either cluster type in the MoFe protein. Electron transfer from the reduced [4Fe-4S]1+ cluster of two different Fe proteins to the P2+ clusters of the MoFe protein was observed by the disappearance of the [4Fe-4S]1+ cluster EPR signal and the conversion of the MoFe protein P-clusters from the P2+ to the P1+ oxidation state. In the first case, stoichiometric quantities of the wild-type Fe protein transferred one electron to the P-clusters only in the presence of MgATP. MgADP would not support this electron transfer reaction. In the second case, an altered Fe protein (L127 delta) that is in a conformation resembling the MgATP-bound state was found to transfer an electron to the P-clusters in the absence of MgATP. These results suggest that the first electron transferred from the Fe protein goes to the P-cluster and that the MgATP-bound protein conformation of the Fe protein, not MgATP hydrolysis, is required for this electron transfer reaction.

The [4Fe-4S] cluster domain of the nitrogenase iron protein facilitates conformational changes required for the cooperative binding of two nucleotides.

MgATP binding and hydrolysis are central to all reduction reactions catalyzed by nitrogenase. The iron (Fe) protein component of nitrogenase is a homodimeric protein with a bridging [4Fe-4S] cluster and two nucleotide binding sites, one on each subunit. This work presents evidence that the [4Fe-4S] cluster domain of the nitrogenase Fe protein functions as a hinge region between the two nucleotide binding domains, participating in the cooperative binding of two nucleotides. Alanine residues at position 98 (located near the [4Fe-4S] cluster) of the Azotobacter vinelandii Fe protein were changed by means of site-directed mutagenesis to Val (V) and Gly (G), and the resulting altered proteins were purified and characterized. While the wild-type and A98G Fe proteins were found to bind two nucleotides (MgATP or MgADP) with strong cooperativity (Hill coefficient of 2), the A98V Fe protein was found to bind one nucleotide with no apparent cooperativity. The binding of two nucleotides to the wild-type Fe protein is known to induce protein conformational changes which are reflected as changes in the properties of the [4Fe-4S] cluster, including a change in the redox potential of the [4Fe-4S] cluster of -120 mV for MgATP binding (-300 to -420 mV) and of -160 mV for MgADP binding (-300 to -460 mV). The binding of one nucleotide to the A98V Fe protein was found to result in only half the lowering of the redox potential, with MgATP binding resulting in a -80 mV change (-280 to -360 mV) and MgADP binding resulting in a -50 mV change (-280 to -330 mV). Results from 1H NMR, EPR, and CD spectra, along with Fe chelation rates, were all consistent with the binding of a single nucleotide to the A98V Fe protein inducing a partial conformational change. Finally, the A98V Fe protein with one nucleotide bound, still bound to the molybdenum-iron protein but did not support MgATP hydrolysis, electron transfer, or substrate reduction. A model is discussed in which the [4Fe-4S] cluster domain can be viewed as a hinge region between the two nucleotide binding domains which facilitates conformational rearrangements required for the cooperative binding of a second nucleotide.

Reduction of thiocyanate, cyanate, and carbon disulfide by nitrogenase: kinetic characterization and EPR spectroscopic analysis.

Nitrogenase catalyzes the reduction of N2, protons, and a number of alternative substrates that contain C-C, C-N, N-N, and N-O double and triple bonds. Recently it has been shown that nitrogenase also reduces the C==S bond of COS and the C==O bond of CO2. The current work demonstrates that the COS analogs SCN-, CS2, and OCNH are novel substrates for nitrogenase and that the reduction of these substrates produces changes in the electron paramagnetic resonance (EPR) spectrum of nitrogenase, providing insight into the mechanism of substrate reduction by nitrogenase. CH4, HCN, H2S, and NH4+ were detected as products of the nitrogenase-catalyzed reduction of SCN-. CS2 was reduced by nitrogenase to H2S, providing the first demonstration of CS2 reduction catalyzed by a purified enzyme. CO was detected as a product of KOCN reduction by nitrogenase. Interestingly, the Km for KOCN reduction to CO decreased at lower pH values, suggesting that OCNH rather than OCN- was the substrate for nitrogenase. Analysis of the EPR spectra of nitrogenase under turnover conditions in the presence of KOCN, CS2, or KSCN revealed new EPR signals. Signals with g-values corresponding to those reported for CO bound to the iron-molybdenum cofactor of nitrogenase were detected during turnover of nitrogenase in the presence of KOCN. During SCN- and CS2 reduction by nitrogenase, novel EPR inflections were observed that appear to report the interaction between nitrogenase and a bound substrate or a transient intermediate produced during the reduction of SCN- and CS2.

Changes in the midpoint potentials of the nitrogenase metal centers as a result of iron protein-molybdenum-iron protein complex formation.

All nitrogenase-catalyzed substrate reduction reactions require the transient association between the iron (Fe) protein component and the molybdenum-iron (MoFe) protein component with concomitant intercomponent electron transfer and MgATP hydrolysis. Understanding the effects of Fe protein-MoFe protein complex formation on the properties of the nitrogenase metal centers is thus essential to understanding the electron transfer reactions. This work presents evidence for significant shifts in midpoint potentials for two of the three nitrogenase metal centers as a result of Fe protein binding to the MoFe protein. The midpoint potentials for the three nitrogenase metal centers, namely the [4Fe-4S] cluster of the Fe protein, and the [8Fe-7S] (or P-) clusters and FeMo cofactors (or M-centers) of the MoFe protein, were determined within a nondissociating nitrogenase complex prepared with a site-specifically altered Fe protein (Leu at position 127 deleted, L127Delta). The midpoint potential for each metal center was determined by mediated redox titrations, with the redox state of each center being monitored by parallel and perpendicular mode EPR spectroscopy. The midpoint potential of the Fe protein [4Fe-4S]2+/1+ cluster couple was observed to change by -200 mV from -420 mV in the uncomplexed L127Delta Fe protein to -620 mV in the L127Delta Fe protein-MoFe protein complex. The midpoint potential of the two electron oxidized couple of the P-clusters (P2+/N) of the MoFe protein was observed to shift by -80 mV upon protein-protein complex formation. No significant change in the midpoint potential of an oxidized state of FeMoco (Mox/N) was observed upon complex formation. These results provide insights into the energetics of intercomponent electron transfer in nitrogenase, suggesting that the energy of protein-protein complex formation is coupled to an increase in the driving force for electron transfer. The results are interpreted in light of the expected changes in the protein environments of the metal centers within the nitrogenase complex.

A mediated thin-layer voltammetry method for the study of redox protein electrochemistry.

A novel mediated thin-layer voltammetry technique that allows the rapid determination of midpoint potentials and electron transfer rate constants for small quantities of redox active proteins is described. Thin-layer voltammograms simulated for an electrolyte containing a redox active protein and an electron transfer mediator show that the rapid homogeneous electron exchange reaction between the protein and the mediator serves to mediate the charge transfer of the protein at the electrode, which does not take place in the absence of the mediator, and results in the observation of an apparently reversible redox couple. Both theoretical and experimental data are presented which suggest that the thin-layer voltammetry method will be generally applicable for the determination of protein redox potentials with the proper selection of mediators. Rate constants for the electron transfer between metalloproteins and mediators can be evaluated by comparing experimental voltammograms with theoretical data from simulations. The technique is demonstrated for the metalloproteins cytochrome c, ferredoxin, and the iron protein of nitrogenase.

A continuous, spectrophotometric activity assay for nitrogenase using the reductant titanium(III) citrate.

A continuous, spectrophotometric assay for determining electron transfer rates through nitrogenase during substrate reduction reactions was developed. The assay takes advantage of the facts that Ti(III) citrate can serve as a reductant for nitrogenase-catalyzed reduction reactions and that oxidation of Ti(III) citrate to Ti(IV) citrate results in a dramatic change in its absorption spectrum. Ti(III) citrate supported nitrogenase-catalyzed substrate (e.g., H+ or acetylene) reduction reactions at about the same rate as that supported by the reductant dithionite (S2O4(2-)). In addition, Ti(III) citrate had an absorption maximum centered at 325 nm, while oxidized Ti(IV) citrate had a much lower absorption in this wavelength region. An absorption coefficient for Ti(III) citrate of 0.73 mM-1.cm-1 at 340 nm was determined by titration with redox dyes with known absorption coefficients. Using this experimentally determined absorption coefficient, we developed an assay that provides a convenient way to determine electron transfer rates through nitrogenase in real time by spectrophotometrically following the oxidation of Ti(III) citrate to Ti(IV) citrate. Average electron transfer rates of 3749 +/- 218 nmol of electrons transferred.min-1.mg iron protein-1 for H+ reduction were determined using this assay which are directly comparable to the rates calculated from fixed time point, gas chromatographic assays of H2 formation. The utility of the Ti(III) citrate assay for nitrogenase is discussed and demonstrated using the nitrogenase inhibitors MgADP, CN-, and NO.

Cyanide inactivation of hydrogenase from Azotobacter vinelandii.

The effects of cyanide on membrane-associated and purified hydrogenase from Azotobacter vinelandii were characterized. Inactivation of hydrogenase by cyanide was dependent on the activity (oxidation) state of the enzyme. Active (reduced) hydrogenase showed no inactivation when treated with cyanide over several hours. Treatment of reversibly inactive (oxidized) states of both membrane-associated and purified hydrogenase, however, resulted in a time-dependent, irreversible loss of hydrogenase activity. The rate of cyanide inactivation was dependent on the cyanide concentration and was an apparent first-order process for purified enzyme (bimolecular rate constant, 23.1 M-1 min-1 for CN-). The rate of inactivation decreased with decreasing pH. [14C]cyanide remained associated with cyanide-inactivated hydrogenase after gel filtration chromatography, with a stoichiometry of 1.7 mol of cyanide bound per mol of inactive enzyme. The presence of saturating concentrations of CO had no effect on the rate or extent of cyanide inactivation of hydrogenases. The results indicate that cyanide can cause a time-dependent, irreversible inactivation of hydrogenase in the oxidized, activatable state but has no effect when hydrogenase is in the reduced, active state.

Elucidation of a MgATP signal transduction pathway in the nitrogenase iron protein: formation of a conformation resembling the MgATP-bound state by protein engineering.

The present work defines one MgATP signal transduction pathway in the nitrogenase iron (Fe) protein. Deletion of an amino acid (Leu 127) by site-directed mutagenesis in the protein chain between Asp 125, located in the ATP binding site, and Cys 132, a ligand to the [4Fe-4S] cluster, resulted in protein conformational changes resembling the MgATP-bound state in the absence of any bound nucleotides. Specifically, 1H nuclear magnetic resonance, electron paramagnetic resonance, and circular dichroism spectroscopic properties, along with Fe chelation assays, suggested that deletion of Leu 127 in the Fe protein resulted in changes in the electronic properties of the [4Fe-4S] cluster similar to those normally observed upon MgATP binding to the wild-type Fe protein. Deletion of Leu 127 of the Fe protein lowered the redox potential of of the [4FE-4S] cluster by 112 mV compared to the wild-typeFe protein (-412mV compared to -294 mV). A nearly identical lowering of the redox potential by 120 mV occurs in the wild-type Fe protein upon binding MgATP (-294 mV compared to 420 mV). The L127delta Fe protein did not contain bound nucleotides which could account for the observed conformational changes. The present results support a model in which the protein chain from ASP 125 to Cys 132 acts as one pathway for MgATP signal transduction and suggests a mechanism for this transduction to the [4Fe-4S] cluster. The L127delta Fe protein was found to still bind 2 MgATP or 2 MgADP molecules/Fe protein. Unlike the wild-type Fe protein, the L127delta Fe protein bound 2 ADP molecules/Fe protein in the absence of Mg2+. Finally, the L127delta protein was found to bind to the MoFe protein, although the complex did not catalyze MgATP hydrolysis or substrate reduction. In concurrence with previous models, homologies between the Asp 125 to Cys 132 transduction pathway in Fe protein and the switch II region of the broad class of GTPase signal transduction proteins (G-proteins) are discussed.

Redox-dependent subunit dissociation of Azotobacter vinelandii hydrogenase in the presence of sodium dodecyl sulfate.

Hydrogenases catalyze the reversible activation of dihydrogen. We have previously demonstrated that the purified hydrogenase from the nitrogen-fixing microorganism Azotobacter vinelandii is an alpha beta dimer (98,000 Da) with subunits of 67,000 (alpha) and 31,000 (beta) daltons and that this enzyme contains iron and nickel. The enzyme can be purified anaerobically in the presence of dithionite in a fully active state that is irreversibly inactivated by exposure to O2. Analysis of this hydrogenase by sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (PAGE) following boiling in SDS yields two protein staining bands corresponding to the alpha and beta subunits. However, when this enzyme was treated with SDS (25-65 degrees C) for up to 30 min under anaerobic/reductive conditions and then analyzed by anaerobic SDS-PAGE, a protein staining band corresponding to an apparent molecular mass of 58,000 Da was observed that stained for hydrogenase activity. Analysis of the 58,000-Da activity staining band by a Western immunoblot or a second aerobic SDS-polyacrylamide gel revealed that this protein actually consisted of both the alpha and beta subunits. Thus, the activity staining band (apparent 58,000 Da) represents the 98,000-Da dimer migrating abnormally on SDS-PAGE. Treatment of the anaerobically purified hydrogenase with SDS under aerobic conditions or under anaerobic conditions with electron acceptors prior to electrophoresis resulted in no activity staining band and the separated alpha and beta subunits. A. vinelandii hydrogenase was also purified under aerobic conditions in an inactive O2 stable form that can be activated by removal of oxygen followed by addition of reductant. This enzyme (as isolated), the activated form, and the reoxidized form were analyzed for their stability toward denaturation by SDS. We conclude that the dissociation of the A. vinelandii hydrogenase subunits in SDS is controlled by the redox state of the enzyme suggesting an important role of one or more redox sites in controlling the structure of this enzyme.

Modulating the midpoint potential of the [4Fe-4S] cluster of the nitrogenase Fe protein.

Protein-bound [FeS] clusters function widely in biological electron-transfer reactions, where their midpoint potentials control both the kinetics and thermodynamics of these reactions. The polarity of the protein environment around [FeS] clusters appears to contribute largely to modulating their midpoint potentials, with local protein dipoles and water dipoles largely defining the polarity. The function of the [4Fe-4S] cluster containing Fe protein in nitrogenase catalysis is, at least in part, to serve as the nucleotide-dependent electron donor to the MoFe protein which contains the sites for substrate binding and reduction. The ability of the Fe protein to function in this manner is dependent on its ability to adopt the appropriate conformation for productive interaction with the MoFe protein and on its ability to change redox potentials to provide the driving force required for electron transfer. Phenylalanine at position 135 is located near the [4Fe-4S] cluster of nitrogenase Fe protein and has been suggested by amino acid substitution studies to participate in defining both the midpoint potential and the nucleotide-induced changes in the [4Fe-4S] cluster. In the present study, the crystal structure of the Azotobacter vinelandii nitrogenase Fe protein variant having phenylalanine at position 135 substituted by tryptophan has been determined by X-ray diffraction methods and refined to 2.4 A resolution. A comparison of available Fe protein structures not only provides a structural basis for the more positive midpoint potential observed in the tryptophan substituted variant but also suggests a possible general mechanism by which the midpoint potential could be controlled by nucleotide interactions and nitrogenase complex formation.

Molecular and immunological comparison of membrane-bound, H2-oxidizing hydrogenases of Bradyrhizobium japonicum, Alcaligenes eutrophus, Alcaligenes latus, and Azotobacter vinelandii.

The membrane-bound hydrogenases of Bradyrhizobium japonicum, Alcaligenes eutrophus, Alcaligenes latus, and Azotobacter vinelandii were purified extensively and compared. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of each hydrogenase revealed two prominent protein bands, one near 60 kilodaltons and the other near 30 kilodaltons. The migration distances during nondenaturing polyacrylamide gel electrophoresis were similar for all except A. vinelandii hydrogenase, which migrated further than the other three. The amino acid composition of each hydrogenase was determined, revealing substantial similarity among these enzymes. This was confirmed by calculation of S delta Q values, which ranged from 8.0 to 26.7 S delta Q units. S delta Q is defined as sigma j(Xi,j-Xk,j)2, where i and k identify the proteins compared and Xj is the content (residues per 100) of a given amino acid of type j. The hydrogenases of this study were also compared with an enzyme-linked immunosorbent assay. Antibody raised against B. japonicum hydrogenase cross-reacted with all four hydrogenases, but to various degrees and in the order B. japonicum greater than A. latus greater than A. eutrophus greater than A. vinelandii. Antibody raised against A. eutrophus hydrogenase also cross-reacted with all four hydrogenases, following the pattern of cross-reaction A. eutrophus greater than A. latus = B. japonicum greater than A. vinelandii. Antibody raised against B. japonicum hydrogenase inhibited B. japonicum hydrogenase activity to a greater extent than the A. eutrophus and A. latus activities; no inhibition of A. vinelandii hydrogenase activity was detected. The results of these experiments indicated remarkable homology of the hydrogenases from these four microorganisms.