SEARCH

SEARCH BY CITATION

Keywords:

  • activity regulation;
  • crystallography;
  • DraG;
  • DraT;
  • manganese;
  • metalloprotein;
  • nitrogenase;
  • nitrogen fixation;
  • post-translational modification;
  • Rhodospirillum rubrum

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. DraT
  5. DraG
  6. Regulation of DraT and DraG
  7. Concluding remarks
  8. Acknowledgements
  9. References

Nitrogen fixation is the vital biochemical process in which atmospheric molecular nitrogen is made available to the biosphere. The process is highly energetically costly and thus tightly regulated. The activity of the key enzyme, nitrogenase, is controlled by reversible mono-ADP-ribosylation of one of its components, the Fe protein. This protein provides the other component, the MoFe protein, with the electrons required for the reduction of molecular nitrogen. The Fe-protein is ADP-ribosylated and de-ADP-ribosylated by dinitrogenase reductase ADP-ribosyl transferase and dinitrogenase reductase activating glycohydrolase, respectively. Here we review the current biochemical and structural knowledge of this central regulatory reaction.


Abbreviations
ARH

ADP-ribosylhydrolase

DraG

dinitrogenase reductase activating glycohydrolase

DraT

dinitrogenase reductase ADP-ribosyl transferase

FeMoco

FeMo cofactor

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. DraT
  5. DraG
  6. Regulation of DraT and DraG
  7. Concluding remarks
  8. Acknowledgements
  9. References

Biological nitrogen fixation is the process by which atmospheric molecular nitrogen is converted into ammonia, and it only occurs in some species of bacteria and archaea. The process is catalysed by nitrogenase, a metalloprotein that consists of two components, the Fe protein (dinitrogenase reductase) and the MoFe protein (dinitrogenase) [1]. The overall reaction may be written as

  • display math

The MoFe protein has an α2β2 structure, with each αβ being a functional unit, harbouring one set of cofactors each (Fig. 1). The FeMo cofactor (FeMoco) constitutes the active site where the reduction of molecular nitrogen takes place, as well as the hydrogen evolution that is a reaction in the catalytic cycle. FeMoco is unique to nitrogenase, and has the overall composition MoFe7S9 homocitrate. In addition, the MoFe protein contains another metal cluster, the P cluster, Fe8S7 [1].

image

Figure 1. The nitrogenase complex (PDB ID 1N2C). The α subunit of the MoFe protein is shown in blue and the β subunit is shown in green. The Fe protein is shown in red, and metal clusters are indicated as spheres. The arginine (Arg101; R. rubrum numbering) of the Fe protein that is ADP-ribosylated in the inactive protein is shown in magenta.

Download figure to PowerPoint

The Fe protein is a homodimeric protein with a Fe4S4 cluster, contained at the interface between the two subunits, to which it is bound through two cysteine residues in each subunit (Fig. 1). In addition, there is one ATP/ADP binding site in each subunit. The Fe protein is extremely sensitive to oxygen, and all diazotrophs maintain an anaerobic environment around nitrogenase [1].

The Fe protein is the electron donor to the MoFe protein, and two ATP molecules are hydrolysed for each electron transferred. As seen in Fig. 1, the Fe4S4 cluster in the Fe protein is close to the P cluster in the MoFe protein when the two interact, making transfer of one electron possible [2]. Electrons are then passed on to the FeMoco, where the reduction of molecular nitrogen and production of hydrogen take place. For one dinitrogen molecule to be reduced to two ammonia molecules, the Fe protein and the MoFe protein have to interact eight times [1].

As the reaction catalysed by nitrogenase requires a minimum of 16 ATP molecules, it is not surprising that the process is tightly regulated at the transcriptional level. In Klebsiella pneumoniae, expression of 20 nif genes is required for active nitrogen fixation. Although the number of nif genes varies between diazotrophs, expression and activation of the regulatory gene nifA are required in most cases. These processes are in turn controlled in response to the availability of usable nitrogen and the presence of oxygen. Limited nitrogen supply and anaerobic conditions support the expression of nitrogenase [3].

In some nitrogen-fixing photosynthetic bacteria and some species of Azospirillum, nitrogenase activity is also regulated at the metabolic level [4]. This regulation was first discovered in the photosynthetic bacterium Rhodospirillum rubrum, from which the most detailed knowledge about the proteins involved has been derived. However, studies of the system operating in the aerobic bacterium Azospirillum brasilense have greatly contributed to a more detailed understanding [5].

In R. rubrum, the regulation is manifested as a ‘switch-off’ effect, i.e. loss of nitrogenase activity, when the light is turned off or an effector, e.g. ammonium ions or glutamine, is added to a culture of actively nitrogen-fixing bacteria. When the light is turned on or the added effector is metabolized, nitrogenase activity is regained (Fig. 2A). The system thus responds to both the nitrogen status and the energy status in the cell. In A. brasilense, the energy ‘switch-off’ is instead caused by removal of oxygen [6].

image

Figure 2. The ‘switch-off’ effect. (A) Schematic diagram showing the time course of changes in nitrogenase activity. (B) Reversible ADP-ribosylation of the Fe protein by DraT and DraG. Me2+ indicates a divalent metal ion, Gln indicates glutamine, and NH4+ indicates addition of ammonium ions.

Download figure to PowerPoint

At the molecular level, the loss of activity is due to reversible ADP-ribosylation of an arginine in one of the subunits of the Fe protein, Arg101 in R. rubrum [7]. As shown in Fig. 1, Arg101 is located within the area of interaction between the Fe protein and the MoFe protein, and therefore the modification sterically hinders the interaction between the two proteins and thus electron transfer cannot take place. The two enzymes catalysing the reversible modification are dinitrogenase reductase ADP-ribosyl transferase (DraT) and dinitrogenase reductase activating glycohydrolase (DraG) (Fig. 2B). The genes for DraT (draT) and DraG (draG) are within the same operon, together with a third gene, draB, the function of whose gene product is not known [8]. The genes are constitutively expressed, but the cellular concentrations of DraT and DraG are very low.

DraT

  1. Top of page
  2. Abstract
  3. Introduction
  4. DraT
  5. DraG
  6. Regulation of DraT and DraG
  7. Concluding remarks
  8. Acknowledgements
  9. References

DraT is a monomeric 30 kDa protein that catalyses ADP-ribosylation, with NAD+ as the donor of the ADP-ribose moiety [9]. In the reaction with the Fe protein from R. rubrum, MgADP is required and MgATP inhibits the reaction. DraT is specific for the Fe protein as an acceptor but in vitro, not only the one from R. rubrum [10]. However, some NAD+ analogues may also function as the donor of the ADP-ribose moiety in vitro [11]. No structure for DraT has so far been published, which possibly is due to its instability in vitro. The R. rubrum protein requires the presence of another protein, a P protein, for stability and activity (T. Selão, P. Teixeira and S. Nordlund, Department of Biochemistry and Biophysics, Stockholm University, unpublished results). The biochemical activity of DraT is similar to a number of ADP-ribosylating bacterial toxins for which the structure is known, such as iota toxin [12], diphtheria toxin [13] and Certhrax toxin [14]. However, the sequence similarity between DraT and any solved structure is very low (well below 20% identity), and it is not possible to determine whether DraT is structurally homologous to these with any certainty. Profile-based fold recognition methods [15] suggest that this may be the case, but any detailed comparisons or conclusions require a structure for DraT.

DraG

  1. Top of page
  2. Abstract
  3. Introduction
  4. DraT
  5. DraG
  6. Regulation of DraT and DraG
  7. Concluding remarks
  8. Acknowledgements
  9. References

R. rubrum DraG has also been purified and shown to be a monomeric protein with a molecular mass of 32 kDa [16, 17]. It catalyses hydrolysis of the ADP-ribosyl moiety from the Fe protein, which thereby is unhindered to interact with and transfer electrons to the MoFe protein. The hydrolysis reaction requires the presence of ATP and a divalent cation, with Mn2+ and Fe2+ being equally efficient [18]. However, Mn2+ is believed to be the physiological cation, and there is no activity with Mg2+. DraG accepts artificial substrates as well as denatured ADP-ribosylated Fe protein [19].

In contrast to the situation for DraT, there are crystal structures available for a number of DraG family proteins. The structures of DraG from R. rubrum [20] (Fig. 3) and A. brasiliense have been published [21, 22]. Structures are also available for the human and mouse ADP-ribosylhydrolase 3 (ARH3) proteins [23, 24]. In addition, structures of the human ADP-ribosylhydrolase 1 (ARH1) protein (PDB ID 3HFW) and two prokaryotic putative ADP-ribosylhydrolases from Thermus thermophilus (PDB ID 2CWC) and Methanococcus janaschii (PDB ID 1T5J) have been deposited in the Protein Data Bank.

image

Figure 3. Overall structure of DraG from R. rubrum. The N–terminus is shown in blue and the C–terminus is shown in red. Metal ions (Mn2+) are indicated in purple, and coordinating residues are shown as sticks. The inset shows a close up of the di-manganese active site. The bridging solvent-derived ligand, which is believed to act as a nucleophile in the reaction, is indicated as a red sphere. Asn100, which is involved in sequestration of DraG to the membrane, is indicated.

Download figure to PowerPoint

The lowest pairwise sequence similarities between these proteins are of the order of 20–25% sequence identity. Structurally, the proteins form very similar all-α–helical monomers (Fig. 3). The active site is positioned in a deep but solvent-accessible cleft surrounded by loops that also contribute a number of the active-site residues. When comparing the various structures, the pairwise core RMSDs generally do not exceed 2 Å. The largest overall structural differences are observed between human ARH1 and the mammalian ARH3 proteins, for which pairwise core RMSDs are of the order of 2.2–2.4 Å.

The proteins form dinuclear metal active sites and require divalent metal ions for activity. In most cases, ADP-ribosylhydrolase proteins appear to bind Mg2+. However, the metal requirement differs, and R. rubrum DraG shows highest activity with Mn2+ (Fe2+) [18, 25]. The two metal ions are coordinated by four aspartates, one glutamate and two threonine residues. These are conserved in all proteins with the exception of Thr59 (R. rubrum numbering), which is conservatively replaced by serine in human ARH1. An additional aspartate (Asp97) is conserved in the prokaryotic, but not the mammalian, proteins, and is involved in both metal and substrate binding [20]. Differences in metal coordination are observed between proteins, between different crystal forms of the same protein [23] and between different complex structures [20]. This suggests that the metal coordination is dynamic and may adjust to different structural states during the catalytic cycle, which most likely contributes to substrate positioning and catalytic control. The residues involved in ADP-ribose binding are also largely conserved, suggesting a common mechanism for catalysis and substrate recognition.

A structure-based mechanism for substrate binding and cleavage of the protein–glycosidic bond has been suggested based on ligand complex structures and a serendipitously obtained protein–ADP-ribose reaction intermediate analogue in R. rubrum DraG. In the presence of excess ADP-ribose, DraG catalysed the ADP-ribosylation (back reaction) of a neighbouring molecule in the crystal lattice (Fig. 4) [20]. The key features of the mechanism include ring-opening of the substrate ribose followed by nucleophilic attack by a metal-activated solvent-derived hydroxide that bridges the active-site metal ions (Fig. 5). The mechanism is consistent with biochemical and mutational studies [20] and studies of ADP-ribose–amine model conjugates [26, 27], as well as very recent computational studies (F. Himo and B. Manta, Department of Organic Chemistry, Stockholm University, personal communication).

image

Figure 4. Structure of a reaction intermediate (PDB ID 2WOD), showing a ring-opened ADP-ribosyl lysine complex formed by DraG (green) that, in the presence of an excess of ADP-ribose, catalysed the ADP-ribosylation (back reaction) of a surface lysine (Lys54*) from a neighbouring molecule in the crystal lattice (blue). The observed structure is most likely the result of Amadori rearrangement, yielding a stable analogue of the second intermediate depicted in Fig. 5. A more detailed description has been provided previously [20]. The interactions of the ADP-ribose moiety with the protein are indicated, as well as the substrate-coordinating manganese ion (magenta).

Download figure to PowerPoint

image

Figure 5. Suggested key steps in the activating de-ADP-ribosylation of the the Fe protein catalysed by DraG. The suggested catalytic mechanism is inferred from structural, biochemical and chemical data. The ADP-ribosylated substrate protein is shown in blue.

Download figure to PowerPoint

Regulation of DraT and DraG

  1. Top of page
  2. Abstract
  3. Introduction
  4. DraT
  5. DraG
  6. Regulation of DraT and DraG
  7. Concluding remarks
  8. Acknowledgements
  9. References

One of the remaining challenges is to understand the regulation of DraT and DraG at the molecular level. Less is known about DraT regulation, but it is clear that a PII protein is involved in this regulation, at least in responding to the nitrogen status of the cell. PII proteins constitute a family of trimeric proteins that regulate a number of processes in prokaryotic cells that are related to or part of nitrogen metabolism [28]. These proteins exert their action in response to binding of ATP/ADP and 2–oxoglutarate, but also by being modified by uridylylation of a conserved tyrosine residue. Uridylylation is catalysed by the bi-functional uridyltransferase/uridylyl-removing enzyme GlnD, which catalyses uridylylation when there is nitrogen limitation in the cell [29]. Correspondingly, the uridyl moiety is removed by the reverse activity of GlnD under conditions of nitrogen sufficiency. In summary, under nitrogen-limiting conditions, PII proteins are uridylylated and have ATP, a divalent cation (Mg2+ or Mn2+) and 2–oxoglutarate bound. Under energy limitation, ADP binds to PII proteins and promotes the unuridylylated state.

There are three PII paralogues in R. rubrum (GlnB, GlnJ and GlnK) and two in A. brasilense (GlnB and GlnZ). There are strong indications that binding of an unmodified PII protein (GlnB) leads to activation of DraT, and thereby ADP-ribosylation of the Fe protein in both R. rubrum (T. Selão, P. Teixeira and S. Nordlund, Department of Biochemistry and Biophysics, Stockholm University, unpublished results) and A. brasilense [30]. The details are not known, and there are no data indicating that a PII protein is directly involved in the energy response.

The regulation of DraG activity is better known. Even in the early studies of the ‘switch-off’ effect in R. rubrum, it was established that DraG is associated with the chromatophore (plasma) membrane in R. rubrum, and may be solubilized by a high-salt wash [4]. Further studies have shown that sequestration of DraG to the membrane is a means of inactivation. This was given additional support when it was shown that substituting Asn100  in R. rubrum DraG by lysine led to loss of ‘switch-off’, i.e. no loss of nitrogenase activity, and DraG was found in the soluble fraction, i.e. not associated with the chromatophore membrane (H. Wang, C. Berthold Sjöberg, M. Högbom and S. Nordlund, Department of Biochemistry and Biophysics, Stockholm University, unpublished results). From the structure of the R. rubrum enzyme, it may be concluded that, if Asn100 is involved in association with the membrane, the active site in DraG will not be accessible (Fig. 3) [20].

The identity of the protein in the membrane to which DraG binds is not known. AmtB, an ammonium transporter, has been suggested [31, 32], but we propose that an as yet unidentified protein is the binding partner. However, it is generally believed that a PII protein, (GlnJ in R. rubrum and GlnZ in A. brasilense) is central in the process leading to association of DraG with the membrane in response to ammonium ions as the effector. A model of the regulation of DraT and DraG is summarized in Fig. 6.

image

Figure 6. Model of the regulation of DraT and DraG in R. rubrum. The middle part shows the situation when nitrogenase is active, DraG is free in the cytoplasm, both GlnB and GlnJ are uridylylated, and DraT is inactive. The right part shows the situation under nitrogen ‘switch-off’, where unmodified GlnJ is bound to AmtB1, which then interacts with an unknown membrane protein (X), which in turn changes its conformation to make DraG binding possible. DraT is active due to forming a complex with unmodified GlnB, possibly leading to increased affinity for NAD+, the substrate for ADP-ribosylation. The left part shows the situation under energy ‘switch-off’. Under these conditions, it is unclear why the membrane protein X adopts its DraG-binding conformation, but we suggest that the change in membrane potential plays a major role. The mechanism for activation of DraT is not known; one suggestion is that the concentration of NAD+ is increased to a level where the affinity of DraT for this substrate is sufficient to result in activity.

Download figure to PowerPoint

Concluding remarks

  1. Top of page
  2. Abstract
  3. Introduction
  4. DraT
  5. DraG
  6. Regulation of DraT and DraG
  7. Concluding remarks
  8. Acknowledgements
  9. References

Understanding of the molecular details of the metabolic regulation of nitrogenase activity has increased dramatically since the first studies of this regulatory mechanism some 40 years ago. Clearly the crystal structure of R. rubrum DraG has been very important for understanding the catalytic mechanism, and one of the current major challenges is to obtain a high-resolution structure of DraT, preferably in complex with a PII protein, as this is believed to be the active form of DraT. Furthermore, identification of the binding partner for the membrane sequestration of DraG,will shed light on the mechanism behind sensing and responding to energy limitation.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. DraT
  5. DraG
  6. Regulation of DraT and DraG
  7. Concluding remarks
  8. Acknowledgements
  9. References

This work was supported by grants from the Swedish Foundation for Strategic Research and the Knut and Alice Wallenberg Foundation (to M.H.) and the Swedish Research Council (to S.N.).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. DraT
  5. DraG
  6. Regulation of DraT and DraG
  7. Concluding remarks
  8. Acknowledgements
  9. References
  • 1
    Seefeldt LC, Hoffman BH & Dean DR (2009) Mechanism of Mo-dependent nitrogenase. Annu Rev Biochem 78, 701722.
  • 2
    Schindelin H, Kisker C, Schlessman JL, Howard JB & Rees DC (1997) Structure of ADP·AlF4–stabilized nitrogenase complex and its implications for signal transduction. Nature 387, 370376.
  • 3
    Dixon R & Kahn D (2004) Genetic regulation of biological nitrogen fixation. Nat Rev Microbiol 2, 621631.
  • 4
    Nordlund S (2000) Regulation of nitrogenase activity in phototrophic bacteria by reversible covalent modification. In Prokaryotic Nitrogen Fixation: A Model System for Analysis of a Biological Process (Triplett EW, ed), pp. 149163. Horizon Scientific Press, Wymondham, UK.
  • 5
    Huergo LF, Pedrosa FO, Mueller-Santos M, Chubatsu LS, Monterio RA, Merrick M & Souza EM (2012) PII signal transduction proteins: pivotal players in post-translational control of nitrogenase activity. Microbiology 158, 176190.
  • 6
    Zhang Y, Burris RH, Ludden PW & Roberts GP (1997) Regulation of nitrogen fixation in Azospirillum brasilense. FEMS Microbiol Lett 152, 195204.
  • 7
    Pope MR, Murrell SA & Ludden PW (1985) Covalent modification of the iron protein of nitrogenase from Rhodospirillum rubrum by adenosine diphosphoribosylation of a specific arginyl residue. Proc Natl Acad Sci USA 82, 31733177.
  • 8
    Fitzmaurice WP, Saari LL, Lowery RG, Ludden PW & Roberts GP (1989) Genes coding for the reversible ADP-ribosylation system of dinitrogenase reductase from Rhodospirillum rubrum. Mol Gen Genet 218, 340347.
  • 9
    Lowery RG & Ludden PW (1988) Purification and properties of dinitrogenase reductase inactivating ADP-ribosyltransferase from the photosynthetic bacterium Rhodospirillum rubrum. J Biol Chem 263, 1671416719.
  • 10
    Grunwald SK, Ryle MJ, Lanzilotta WN & Ludden PW (2000) ADP-ribosylation of variants of Azotobacter vinelandii dinitrogenase reductase by Rhodospirillum rubrum dinitrogenase reductase ADP-ribosyltransferase. J Bacteriol 182, 25972603.
  • 11
    Lowery RG & Ludden PW (1989) Effect of nucleotides on the activity of dinitrogenase reductase ADP-ribosyltransferase from Rhodospirillum rubrum. Biochemistry 28, 49564961.
  • 12
    Tsuge H, Nagahama M, Nishimura H, Hisatsune J, Sakaguchi Y, Itogawa Y, Katunuma N & Sakurai J (2003) Crystal structure and site-directed mutagenesis of enzymatic components from Clostridium perfringens iota-toxin. J Mol Biol 325, 471483.
  • 13
    Bell CE & Eisenberg D (1996) Crystal structure of diphtheria toxin bound to nicotinamide adenine dinucleotide. Biochemistry 35, 11371149.
  • 14
    Visschedyk D, Rochon A, Tempel W, Dimov S, Park H-W & Merrill AR (2012) Certhrax toxin, an anthrax-related ADP-ribosyltransferase from Bacillus cereus. J Biol Chem 49, 4108941102.
  • 15
    Buchan DWA, Ward SM, Lobley AE, Nugent TCO, Bryson K & Jones DT (2010) Protein annotation and modelling servers at University College London. Nucleic Acids Res 38, W563W568.
  • 16
    Saari LL, Triplett EW & Ludden PW (1984) Purification and properties of the activating enzyme for iron protein of nitrogenase from the photosynthetic bacterium Rhodospirillum rubrum. J Biol Chem 259, 1550215508.
  • 17
    Zumft WG & Nordlund S (1981) Stabilization and partial characterization of the activating enzyme for dinitrogenase reductase (Fe-protein) from Rhodospirillum rubrum. FEBS Lett 127, 7982.
  • 18
    Nordlund S & Noren A (1984) Dependence on divalent cations of the activation of inactive Fe-protein of nitrogenase from Rhodospirillum rubrum. Biochim Biophys Acta 791, 2127.
  • 19
    Saari LL, Pope MR, Murrell SA & Ludden PW (1986) Studies on the activating enzyme for iron protein of nitrogenase from Rhodospirillum rubrum. J Biol Chem 261, 49734977.
  • 20
    Berthold CL, Wang H, Nordlund S & Högbom M (2009) Mechanism of ADP-ribosylation removal revealed by the structure and ligand complexes of the dimanganese mono-ADP-ribosylhydrolase DraG. Proc Natl Acad Sci USA 106, 1424714252.
  • 21
    Li X-D, Huergo LF, Gasperina A, Pedrosa FO, Merrick M & Winkler FK (2009) Crystal structure of dinitrogenase reductase-activating glycohydrolase (DRAG) reveals conservation in the ADP-ribosylhydrolase fold and specific features in the ADP-ribose-binding pocket. J Mol Biol 390, 737746.
  • 22
    Rajendran C, Gerhardt ECM, Bjelic S, Gasperina A, Scarduelli M, Pedrosa FO, Chubatsu LS, Merrick M, Souza EM, Winkler FK et al. (2011) Crystal structure of the GlnZ–DraG complex reveals a different form of PII–target interaction. Proc Natl Acad Sci USA 108, 1897218976.
  • 23
    Mueller-Dieckmann C, Kernstock S, Lisurek M, von Kries JP, Haag F, Weiss MS & Koch-Nolte F (2006) The structure of human ADP-ribosylhydrolase 3 (ARH3) provides insights into the reversibility of protein ADP-ribosylation. Proc Natl Acad Sci USA 103, 1502615031.
  • 24
    Mueller-Dieckmann C, Kernstock S, Mueller-Dieckmann J, Weiss MS & Koch-Nolte F (2008) Structure of mouse ADP-ribosylhydrolase 3 (mARH3). Acta Crystallogr Sect F Struct Biol Cryst Commun 64, 156162.
  • 25
    Ljungström E, Yates MG & Nordlund S (1989) Purification of the activating enzyme for the Fe-protein of nitrogenase from Azospirillum brasilense. Biochim Biophys Acta 994, 210214.
  • 26
    Cervantes-Laurean D, Minter DE, Jacobson EL & Jacobson MK (1993) Protein glycation by ADP-ribose: studies of model conjugates. Biochemistry 32, 15281534.
  • 27
    Jacobson EL, Cervantes-Laurean D & Jacobson MK (1994) Glycation of proteins by ADP-ribose. Mol Cell Biochem 138, 207212.
  • 28
    Forchammer K (2008) PII signal transducers: novel functional and structural insights. Trends Microbiol 16, 6572.
  • 29
    Jonsson A & Nordlund S (2007) In vitro studies of the uridylylation of the three PII protein paralogs from Rhodospirillum rubrum: the transferase activity of R. rubrum GlnD is regulated by α–ketoglutarate and divalent cations but not by glutamine. J Bacteriol 189, 34713478.
  • 30
    Moure VR, Danyal K, Zhi-YongYang Z–Y, Wendroth S, Müller-Santos M, Pedrosa FO, Scarduelli M, Gerhardt ECM, Huergo LF, Souza EM et al. (2013) The nitrogenase regulatory enzyme dinitrogenase reductase ADP-ribosyltransferase (DraT) is activated by direct interaction with the signal transduction protein GlnB. J Bacteriol 195, 279286.
  • 31
    Huergo LF, Merrick M, Pedrosa FO, Chubatsu LS, Araujo LM & Souza EM (2007) Ternary complex formation between AmtB, GlnZ and the nitrogenase regulatory enzyme DraG reveals a novel facet of nitrogen regulation in bacteria. Mol Microbiol 66, 15231535.
  • 32
    Javelle A, Lupo D, Ripoche P, Fulford T, Merrick M & Winkler FK (2008) Substrate binding, deprotonation, and selectivity atthe periplasmic entrance of the Escherichia coli ammonia channel AmtB. Proc Natl Acad Sci USA 105, 50405045.