Purification and properties of the activating enzyme for iron protein of nitrogenase from the photosynthetic bacterium Rhodospirillum rubrum

The oxygen-labile, activating enzyme for iron protein from the photosynthetic bacterium, Rhodospirillum rubrum, was purified 11,800-fold using a combination of chromatophore washing, DE52-cellulose chromatography, hydroxylapatite chromatography, reactive red-120 cross-linked agarose chromatography, reactive red-120 cross-linked agarose chromatography, and Sephadex G-75 gel filtration. Activating enzyme appeared homogeneous on silver-stained sodium dodecyl sulfate-polyacrylamide gels, and the staining intensity of the activating-enzyme band was correlated with the activating-enzyme activity observed in in vitro assays. Either formaldehyde fixation or higher acrylamide concentration was required to accurately assess the purity of activating enzyme on silver-stained gels. Activating enzyme was stable for 30 days at 4 degrees C. Dithiothreitol was a necessary component for the stability of partially purified activating enzyme. NaCl inhibited the coupled assay for activating enzyme. The pI of activating enzyme was determined to be 6.5. Activating enzyme is composed of a minimum of 336 amino acids and a minimum calculated Mr is 32,032. The Mr of activating enzyme was estimated to be 21,700 by analytical gel filtration and 32,800 by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. An absorption maximum at 280 nm was observed for the activating enzyme.     

Reversible regulation of the nitrogenase iron protein from Rhodospirillum rubrum by ADP-ribosylation in vitro.

Nitrogenase activity in the photosynthetic bacterium Rhodospirillum rubrum is reversibly regulated by interconversion of the Fe protein between a modified and an unmodified form. Since the discovery of the activation process in 1976, investigators have been unable to demonstrate the inactivation (modification) reaction in vitro. In this study, NAD-dependent modification and concomitant inactivation of the Fe protein were demonstrated in crude extracts of R. rubrum. Activation of the in vitro-modified Fe protein by activating enzyme and structural similarity between the in vivo and in vitro modifications are presented as evidence that the in vitro modification is the physiologically relevant ADP-ribosylation reaction. Using a partially purified preparation, we showed that the inactivating enzyme activity is stimulated by divalent metal ions and ADP, that O2-denatured Fe protein will not serve as a substrate, and that dithionite inhibits the modification reaction.     

Studies on the activating enzyme for iron protein of nitrogenase from Rhodospirillum rubrum

Removal of ADP-ribose from the iron protein of nitrogenase by activating enzyme resulted in the activation of the inactive iron protein. A radioassay that directly measured the initial velocity of the activation was developed using iron protein radiolabeled with either [8-3H]- or [G-32P]ADP-ribose. The release of radiolabeled ADP-ribose by activating enzyme was linearly correlated with the increase in the specific activity of the iron protein as measured by acetylene reduction. Both ATP and MnCl2 were required for the activation of inactive iron protein. The optimal ratio of [MnCl2]/[ATP] in the radioassay was 2:1, and the optimal concentrations were 4 mM and 2 mM for [MnCl2] and [ATP], respectively. The Km for inactive iron protein was 74 microM and the Vmax was 628 pmol of [32P] ADP-ribose released min-1 microgram of activating enzyme-1. Adenosine, cytidine, guanosine, or uridine mono-, di-, or triphosphates did not substitute for ATP in the activation of native iron protein. Activating enzyme removed ADP-ribose from oxygen-denatured iron protein in the absence of ATP. ADP, ADP-ribose, pyrophosphate, and high concentrations of NaCl inhibited activating enzyme activity.     

Purification of diacylglycerol kinase from Microsporum gypseum and its phosphorylation by the catalytic subunit of protein kinase A

Diacylglycerol (DG) kinase (EC 2.7.1.107) was purified to homogeneity from the soluble extract of Microsporum gypseum, a dermatophyte. Purified enzyme showed a final specific activity of 2172 pmol/min/mg protein and its apparent molecular weight on SDS-PAGE was found to be 93 kDa. The activity of purified enzyme was inhibited in a dose-dependent manner in the presence of DG-kinase inhibitor (D5919, Sigma). DG-kinase activity was found to be stimulated in the presence of phosphatidylcholine, phosphatidylethanolamine, and cardiolipin while the activity was alleviated in the presence of phosphatidic acid and arachidonic acid. Kinase activity was partially inhibited when assayed after prior treatment with alkaline phosphatase. Treatment of DG-kinase with the catalytic subunit of protein kinase A (PKA)-stimulated DG-kinase activity in a dose-dependent manner. Incubation of DG-kinase with the catalytic subunit of PKA led to the phosphorylation of DG-kinase as revealed by autoradiography. The phosphorylated band disappeared completely in the presence of specific PKA inhibitor. Increased activity of DG-kinase on incubation with the catalytic subunit of PKA was possibly due to the phosphorylation of the former by the latter. Whether this in vitro phosphorylation and activation of DG-kinase occurs under physiological conditions remains to be elucidated.     

Regulation of nitrogen fixation in Rhodospirillum rubrum grown under dark, fermentative conditions.

Rhodospirillum rubrum was shown to grow fermentatively on fructose with N2 as a nitrogen source. The nitrogenase activity of these cells was regulated by the NH4+ switch-off/switch-on mechanism in a manner identical to that for photosynthetically grown cells. In vitro, the inactive nitrogenase Fe protein from fermenting cells was reactivated by an endogenous membrane-bound, Mn2+-dependent activating enzyme that was interchangeable with the activating enzyme isolated from photosynthetic membranes.     

Comparison of active and inactive forms of iron protein from Rhodospirillum rubrum.

The Fe protein of nitrogenase from Rhodospirillum rubrum was purified in its active and inactive forms. It is shown that the inactive form exists as a two-subunit modified form of the enzyme as previously reported [Ludden & Burris (1978) Biochem. J. 175, 251--259]. In contrast, the active form exists as a single-subunit unmodified form of the enzyme. The upper subunit (on sodium dodecyl sulphate/polyacrylamide-gel electrophoresis) of the inactive form was shown to contain at least the phosphate group of the covalently bound modifying group. The active and inactive forms of the enzyme were shown to be identical proteins on the basis of amino-acid composition, tryptic-digest pattern and immunological cross-reactivity.     

Expression of the activating enzyme and Fe protein of nitrogenase from Rhodospirillum rubrum.

Activating enzyme (AE) is responsible for the in vitro activation of inactive Fe protein of nitrogenase from Rhodospirillum rubrum cells cultured anaerobically with glutamate as the N source. The expression of Fe protein and AE was examined in R. rubrum cultured photosynthetically or aerobically on media containing malate as the carbon source. One of the following N sources was used in each culture: glutamate, glutamine, limiting ammonia, high ammonia, glutamate plus histidine, and high ammonia plus histidine. Chromatophores from every culture exhibited AE activity; activity was highest in glutamate-grown cells. Fe protein was observed by rocket immunoelectrophoresis in cultures with nitrogenase activity. Several Nif-, Gln-, and His- mutants of R. rubrum were assayed for AE activity, nitrogenase activity, and Fe protein. Every mutant expressed AE activity, and Fe protein was observed in those cultures with nitrogenase activity. AE from every preparation was O2 labile, and each O2-denatured AE preparation inhibited activation by active AE.     

Two Forms of Nitrogenase from the Photosynthetic Bacterium Rhodospirillum rubrum

Acetylene reduction by nitrogenase from Rhodospirillum rubrum, unlike that by other nitrogenases, was recently found by other investigators to require an activation of the iron protein of nitrogenase by an activating system comprising a chromatophore membrane component, adenosine 5'-triphosphate (ATP), and divalent metal ions. In an extension of this work, we observed that the same activating system was also required for nitrogenase-linked H(2) evolution. However, we found that, depending on their nitrogen nutrition regime, R. rubrum cells produced two forms of nitrogenase that differed in their Fe protein components. Cells whose nitrogen supply was totally exhausted before harvest yielded predominantly a form of nitrogenase (A) whose enzymatic activity was not governed by the activating system, whereas cells supplied up to harvest time with N(2) or glutamate yielded predominantly a form of nitrogenase (R) whose enzymatic activity was regulated by the activating system. An unexpected finding was the rapid (less than 10 min in some cases) intracellular conversion of nitrogenase A to nitrogenase R brought about by the addition to nitrogen-starved cells of glutamine, asparagine, or, particularly, ammonia. This finding suggests that mechanisms other than de novo protein synthesis were involved in the conversion of nitrogenase A to the R form. The molecular weights of the Fe protein and Mo-Fe protein components from nitrogenases A and R were the same. However, nitrogenase A appeared to be larger in size, because it had more Fe protein units per Mo-Fe protein than did nitrogenase R. A distinguishing property of the Fe protein from nitrogenase R was its ATP requirement. When combined with the Mo-Fe protein (from either nitrogenase A or nitrogenase R), the R form of Fe protein required a lower ATP concentration but bound or utilized more ATP molecules during acetylene reduction than did the A form of Fe protein. No differences between the Fe proteins from the two forms of nitrogenase were found in the electron paramagnetic resonance spectrum, midpoint oxidation-reduction potential, or sensitivity to iron chelators.     

Partial purification of the protein system controlling the breakdown of trehalose in baker's yeast.

The inactive form of trehalase as well as its activating protein have been partially purified from resting cells of baker's yeast using (NH₄)₂SO₄ fractionation and subsequent DEAE- and CM-cellulose column chromatography. For its activation by cyclic 3′,5′-AMP the system appeared to be dependent on the presence of ATP and a divalent cation such as Mg²⁺, Mn²⁺ or Co²⁺. No sensitivity towards the pH was observed in the range 6.0 – 7.5. The amount of active trehalase formed was determined by the preincubation time and the concentration of the proteins involved. The activating protein partly lost its dependence on cyclic 3′,5′-AMP during purification. The results presented suggest that this protein may be a protein kinase and that activation of trehalase is associated with phosphorylation of the enzyme protein.     

Phosphorylation of the RNA-dependent protein kinase regulates its RNA-binding activity

The RNA-dependent protein kinase (PKR) is an interferon-induced, RNA-activated enzyme that phosphorylates the α-subunit of eukaryotic initiation factor 2 (eIF2α), inhibiting the function of the eIF2 complex and continued initiation of translation. When bound to an activating RNA and ATP, PKR undergoes autophosphorylation reactions at multiple serine and threonine residues. This autophosphorylation reaction stimulates the eIF2α kinase activity of PKR. The binding of certain viral RNAs inhibits the activation of PKR. Wild-type PKR is obtained as a highly phosphorylated protein when overexpressed in Escherichia coli. We report here that treatment of the isolated phosphoprotein with the catalytic subunit of protein phosphatase 1 dephosphorylates the enzyme. The in vitro autophosphorylation and eIF2α kinase activities of the dephosphorylated enzyme are stimulated by addition of RNA. Thus, inactivation by phosphatase treatment of autophosphorylated PKR obtained from overexpression in bacteria generates PKR in a form suitable for in vitro analysis of the RNA-induced activation mechanism. Furthermore, we used gel mobility shift assays, methidiumpropyl-EDTA·Fe footprinting and affinity chromatography to demonstrate differences in the RNA-binding properties of phospho- and dephosphoPKR. We found that dephosphorylation of PKR increases binding affinity of the enzyme for both kinase activating and inhibiting RNAs. These results are consistent with an activation mechanism that includes release of the activating RNA upon autophosphorylation of PKR prior to phosphorylation of eIF2α.     

Regulation of nitrogenase activity by ammonium chloride in Azospirillum spp.

Ammonium chloride (greater than or equal to 0.05 mM) effectively and reversibly inhibited the nitrogenase activity of Azospirillum brasilense, Azospirillum lipoferum and Azospirillum amazonense. The glutamine synthetase inhibitor L-methionine-DL- sulfoximine abolished this "switch-off" in A. lipoferum and A. brasilense, but not in A. amazonense. Azaserine, an inhibitor of glutamate synthase, inhibited nitrogenase activity itself. This provides further evidence for glutamine as a metabolite of regulatory importance in the NH4+ switch-off phenomenon. In A. brasilense and A. lipoferum, a transition period before the complete inhibition of nitrogenase activity after the addition of 1 mM ammonium chloride was observed. The in vitro nitrogenase activity also was decreased after treatment with ammonium. During sodium dodecyl sulfate-polyacrylamide gel electrophoresis, a second dinitrogenase reductase (Fe protein) subunit appeared, which migrated in coincidence with the modified subunit of the inactive Fe protein of the nitrogenase of Rhodospirillum rubrum. After the addition of ammonium 32P was incorporated into this subunit of the Fe protein of A. brasilense. In A. amazonense, the inhibition of nitrogenase activity by ammonium was only partial, and no transition period could be observed. The in vitro nitrogenase activity of ammonium-treated cells was not decreased, and no evidence for a modified Fe protein subunit was found. Nitrogenase extracts of A. amazonense were active and had an Fe protein that migrated as a close double band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis.     

Activation of the particulate low Km phosphodiesterase of adipocytes by addition of cAMP-dependent protein kinase

The purified catalytic subunit (C) of cAMP-dependent protein kinase produced a 2-fold activation of the low Km phosphodiesterase in crude microsomes (P-2 pellet) of rat adipocytes. This activation was C subunit concentration-dependent, ATP-dependent, blocked by a specific peptide inhibitor, and lost if the C subunit was first heat denatured. The concentration of ATP necessary for half-maximal activation of the low Km phosphodiesterase was 4.50 +/- 1.1 microM, which was nearly the same as the known Km of C subunit for ATP (3.1 microM) using other substrates. The concentration of C subunit producing half-maximal activation of phosphodiesterase was 0.22 +/- 0.04 microM, slightly less than the measured concentration of total C subunit in adipocytes (0.45 microM). The activation of the low Km phosphodiesterase by C subunit was specific, since on an equimolar basis, myosin light chain kinase, cGMP-dependent protein kinase, or Ca2+/calmodulin-dependent protein kinase II did not activate the enzyme. The percent stimulation of phosphodiesterase by C subunit was about the same as that produced by incubation of adipocytes with a cAMP analog, and the enzyme first activated in vivo with the analog was not activated to the same extent (on a percentage basis) by in vitro treatment with C subunit. Treatment of the crude microsomes with trypsin resulted in transfer of phosphodiesterase catalytic activity from the particulate to the supernatant fraction, but the enzyme in the supernatant was minimally activated by C subunit, suggesting either loss or dislocation of the regulatory component. The C subunit-mediated activation of phosphodiesterase was preserved after either transfer of phosphodiesterase activity to the supernatant fraction by nonionic detergents or partial purification of the transferred enzyme. The present findings are consistent with the suggestion that protein kinase regulates the concentration of cAMP through phosphodiesterase activation and provide direct evidence that the mechanism of activation involves phosphorylation.     

Distinctive iron requirement of tryptophan 5-monooxygenase: TPH1 requires dissociable ferrous iron

A peripheral type of tryptophan 5-monooxygenase (EC 1.14.16.4), TPH1, is very unstable in vitro, but the inactivation was reversible and full reactivation occurs upon anaerobic incubation with a high concentration of dithiothreitol (DTT, 15 mM). In this study, distinctive iron requirement of TPH1 was revealed through analysis of the enzyme's inactivation and activation by DTT. For this purpose, all the glasswares, plastics, Sephadex G-25 gels, and reagents including protein solutions had been treated with metal chelators, and apo-TPH was prepared by treatment with EDTA. Apo-TPH thus prepared exclusively required free Fe2+ for its catalytic activity; 10(-8) M was enough under the strict absence of Fe3+ but 10(-12) M was too low. No other metal ions including Fe3+ were effective. It appeared that Fe3+ bound to the enzyme with a higher affinity than Fe2+, resulting in the inactivation. Ascorbate, a non-thiol reducing agent, did not substitute DTT in the activation of TPH1, but enhanced the Fe2+-dependent activity of apo-TPH as effectively as DTT. Thus, the DTT-activation was essentially substituted by preparation of apo-TPH by the EDTA treatment and the assay of apo-TPH in the presence of Fe2+ and ascorbate. The activation of TPH1 by incubation with DTT was accompanied by exposure of 9 sulfhydryls out of the total 10 cysteine residues, but the cleavage of disulfide bonds seemed not to be crucial, even if it occurred. The effect of DTT was substituted by some other sulfhydryls whose structure was analogous to that of commonly used metal chelators. Based on these observations, the following dual roles of DTT are proposed: (1) in the activation of TPH, DTT removes inappropriate bound iron (Fe3+) as a chelator, keeping Fe3+ away from the enzyme's binding site which needs to bind Fe2+ for the catalytic activity, and (2) in both the activation and reaction processes, DTT prevents oxidation of Fe2+ to Fe3+ as a reducing agent.     

Isolation and purification of a rat liver 3-hydroxy-3-methylglutaryl-coenzyme reductase activating protein (RAP)

A protein with an estimated subunit mass of 19 kDa was isolated and purified from perfused rat liver cytosol. This protein activates hydroxymethylglutaryl-coenzyme A (HMG-CoA) reductase (NADPH) (EC 1.1.1.34), the rate-limiting enzyme in the cholesterol biosynthetic pathway. The activation process by this HMG-CoA reductase activating protein (RAP) is time-dependent and requires NADPH. Maximal activity of HMG-CoA reductase induced by RAP is comparable to that obtained in the presence of thiols, such as GSH, and can exceed 100-fold the activity obtained when thiols are omitted. Purified RAP lacks ability to reduce 5,5'-dithiobis-(2-nitrobenzoic acid). RAP was purified to homogeneity utilizing DEAE- and phenyl-Sepharose CL-4B column chromatography. The purified RAP migrates as a single band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis and shows multiple interconvertible aggregational forms on native polyacrylamide gel electrophoresis. A monospecific antibody against RAP was prepared by immunization of hens and extracted from either their egg yolks or serum. The catalytic activity of RAP might be responsible for the physiological activation of HMG-CoA reductase and regulation of its activity.     

The alpha subunit of eucaryotic initiation factor 2 is phosphorylated in mengovirus-infected mouse L cells.

Infection of mouse L cells with mengovirus resulted in the activation of a protein kinase (PK) that selectively phosphorylated the small, 38,000-molecular-weight alpha subunit of eucaryotic initiation factor 2 (eIF-2) in vitro. The mengovirus-activated kinase was detected in vitro approximately 3 h after virus adsorption. The ratio of phosphorylated to unphosphorylated eIF-2 also increased in vivo between 3 and 7 h after adsorption. The virus-activated kinase fractionated with the ribosomal pellet and had a high affinity for DEAE-cellulose and Mono Q ion-exchange columns. Gel electrophoresis of the kinase activity eluting from the Mono Q column and silver staining of the gel revealed only one protein band with a molecular mass of 70 kilodaltons. The optimal assay conditions for the mengovirus-activated kinase paralleled those of the double-stranded RNA-activated PK (dsRNA-PK). Lysates from infected cells contained elements capable of activating partially purified dsRNA-PK. These elements were identified as double-stranded RNA by their sensitivity to double-stranded RNase. The phosphorylation of the alpha subunit of eIF-2 coincided with the synthesis of dsRNA in infected cells, suggesting that the mengovirus-activated kinase is the dsRNA-PK. The phosphorylation of the alpha subunit of eIF-2 correlated with the global inhibition of protein synthesis that occurs at late times after infection.     

Phosphorylation of the catalytic subunit of protein kinase A. Autophosphorylation versus phosphorylation by phosphoinositide-dependent kinase-1

The identification of phosphoinositide-dependent kinase-1 (PDK-1) as an activating kinase for members of the AGC family of kinases has led to its implication as the activating kinase for cAMP-dependent protein kinase. It has been established in vitro that PDK-1 can phosphorylate the catalytic (C) subunit (), but the Escherichia coli-expressed C-subunit undergoes autophosphorylation. To assess which of these mechanisms occurs in mammalian cells, a set of mutations was engineered flanking the site of PDK-1 phosphorylation, Thr-197, on the activation segment of the C-subunit. Two distinct requirements appeared for autophosphorylation and phosphorylation by PDK-1. Autophosphorylation was disrupted by mutations that compromised activity (Thr-201 and Gly-200) or altered substrate recognition (Arg-194). Conversely, only residues peripheral to Thr-197 altered PDK-1 phosphorylation, including a potential hydrophobic PDK-1 binding site at the C terminus. To address the in vivo requirements for phosphorylation, select mutant proteins were transfected into COS-7 cells, and their phosphorylation state was assessed with phospho-specific antibodies. The phosphorylation pattern of these mutant proteins indicates that autophosphorylation is not the maturation mechanism in the eukaryotic cell; instead, a heterologous kinase with properties resembling the in vitro characteristics of PDK-1 is responsible for in vivo phosphorylation of PKA.     

Iron-molybdenum cofactor biosynthesis in Azotobacter vinelandii requires the iron protein of nitrogenase

Nitrogenase is composed of two separately purified proteins called the Fe protein and the MoFe protein. In Azotobacter vinelandii the genes encoding these structural components are clustered and ordered: nifH (Fe protein)-nifD (MoFe protein alpha subunit)-nifK (MoFe protein beta subunit). The MoFe protein contains an ironmolybdenum cofactor (FeMo cofactor) whose biosynthesis involves the participation of at least five gene products, nifQ, nifB, nifN, nifE, and nifV. In this study an A. vinelandii mutant strain, which contains a defined deletion within the nifH (Fe protein) gene, was isolated and studied. This mutant is still able to accumulate significant amounts of MoFe protein subunits. However, extracts of this nifH deletion strain have only very low levels of MoFe protein acetylene reduction activity. Fully active MoFe protein can be reconstituted by simply adding isolated FeMo cofactor to the extracts. Fe protein is not necessary to stabilize or insert this preformed FeMo cofactor into the FeMo cofactor-deficient MoFe protein synthesized by the nifH deletion strain. Extracts of the nifH deletion strain can carry out molybdate and ATP-dependent in vitro FeMo cofactor biosynthesis provided Fe protein is added, demonstrating that they contain the products encoded by the FeMo cofactor biosynthetic genes. These data demonstrate that the Fe protein is physically required for the biosynthesis of FeMo cofactor in A. vinelandii.     

Is the first enzyme of the shikimate pathway, 3-deoxy-D-arabino-heptulosonate-7-phosphate synthase (tyrosine sensitive), a copper metalloenzyme?

3-Deoxy-D-arabino-heptulosonate-7-phosphate synthase (tyrosine sensitive) was purified from Escherichia coli carrying the plasmid pKB45. Enzyme of high specific catalytic activity (70 mu/mg) was obtained from cells grown only to the early log phase. The purified protein contained Cu(II) and showed an absorption band at 350 nm. Metal-free, catalytically inactive apoenzyme could be produced by dialysis against cyanide ion, and the holoenzyme could be reconstituted in terms of both catalytic activity and A350 by the binding of one Cu(II) ion per enzyme subunit. Zn(II) also reactivated the apoenzyme to about 50% of the level seen with Cu(II), although in this case no band appeared at 350 nm. In contrast to earlier reports that the enzyme contains substoichiometric levels of iron, insignificant amounts of iron were found in the isolated enzyme, and neither Fe(II) nor FE(III) regenerated either an absorption band at 350 nm or any catalytic activity from the apoenzyme. The evident preference of the enzyme as isolated for (Cu)II suggests that the synthase might naturally be a copper metalloenzyme.     

Regulation of nitrogenase activity by covalent modification in Chromatium vinosum

Nitrogenase in Chromatium vinosum was rapidly, but reversibly inhibited by NH ₄ ⁺ . Activity of the Fe protin component of nitrogenase required both Mn²⁺ and activating enzyme. Activating enzyme from Rhodospirillum rubrum could replace Chromatium chromatophores in activating the Chromatium Fe protein, and conversely, a protein fraction prepared from Chromatium chromatophores was effective in activating R. rubrum Fe protein. Inactive Chromatium Fe protein contained a peptide covalently modified by a phosphate-containing molecule, which migrated the same in SDS-polyacrylamide gels as the modified subunit of R. rubrum Fe protein. In sum, these observations suggest that Chromatium nitrogenase activity is regulated by a covalent modification of the Fe protein in a manner similar to that of R. rubrum.Abbreviation HEPES N-2-hydroxyethyl piperazine-N-2-ethanesulfonic acid     

Construction and characterization of a heterodimeric iron protein: defining roles for adenosine triphosphate in nitrogenase catalysis

One molecule of MgATP binds to each subunit of the homodimeric Fe protein component of nitrogenase. Both MgATP molecules are hydrolyzed to MgADP and P(i) in reactions coupled to the transfer of one electron into the MoFe protein component. As an approach to assess the contributions of individual ATP binding sites, a heterodimeric Fe protein was produced that has an Asn substituted for residue 39 in the ATP binding domain in one subunit, while the normal Asp(39) residue within the other subunit remains unchanged. Separation of the heterodimeric Fe protein from a mixed population with homodimeric Fe proteins contained in crude extracts was accomplished by construction of a seven His tag on one subunit and a differential immobilized-metal-affinity chromatography technique. Three forms of the Fe protein (wild-type homodimeric Fe protein [Asp(39)/Asp(39)], altered homodimeric Fe protein [Asn(39)/Asn(39)], and heterodimeric Fe protein [Asp(39)/Asn(39)]) were compared on the basis of the biochemical and biophysical changes elicited by nucleotide binding. Among those features examined were the MgATP- and MgADP-induced protein conformational changes that are manifested by the susceptibility of the [4Fe-4S] cluster to chelation and by alterations in the electron paramagnetic resonance, circular dichroism, and midpoint potential of the [4Fe-4S] cluster. The results indicate that changes in the [4Fe-4S] cluster caused by nucleotide binding are the result of additive conformational changes contributed by the individual subunits. The [Asp(39)/Asn(39)] Fe protein did not support substrate reduction activity but did hydrolyze MgATP and showed MgATP-dependent primary electron transfer to the MoFe protein. These results support a model where each MgATP site contributes to the rate acceleration of primary electron transfer, but both MgATP sites must be functioning properly for substrate reduction. Like the altered homodimeric [Asn(39)/Asn(39)] Fe protein, the heterodimeric [Asp(39)/Asn(39)] Fe protein was found to form a high affinity complex with the MoFe protein, revealing that alteration on one subunit is sufficient to create a tight complex.