HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 18, 15472-15481, May 3, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/18/15472    most recent M112262200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Little, R. Articles by Dixon, R. Search for Related Content PubMed PubMed Citation Articles by Little, R. Articles by Dixon, R. Social Bookmarking                      What's this?

Direct Interaction of the NifL Regulatory Protein with the GlnK Signal Transducer Enables the Azotobacter vinelandii NifL-NifA Regulatory System to Respond to Conditions Replete for Nitrogen^*

Richard Little, Victoria Colombo, Andrew Leech^§, and Ray Dixon^¶

From the  Department of Molecular Microbiology, John Innes Centre, Norwich NR4 7UH and the ^§ School of Biological Sciences, University of East Anglia, Norwich NR4 7TJ, United Kingdom

Received for publication, December 21, 2001, and in revised form, February 19, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The Azotobacter vinelandii NifL-NifA regulatory system integrates metabolic signals for redox, energy, and nitrogen status to fine tune regulation of the synthesis of molybdenum nitrogenase. The NifL protein utilizes discrete mechanisms to perceive these signals leading to the formation of a protein-protein complex, which inhibits NifA activity. Whereas redox signaling is mediated via a flavin-containing PAS domain in the N-terminal region of NifL, the nitrogen status is sensed via interaction with PII-like signal transduction proteins and small molecular weight effectors. The nonuridylylated form of the PII-like protein encoded by A. vinelandii glnK (Av GlnK) stimulates NifL to inhibit transcriptional activation by NifA in vitro. Here we demonstrate that the nonmodified form of Av GlnK directly interacts with A. vinelandii NifL and that this interaction is dependent on Mg²⁺, ATP, and 2-oxoglutarate. Differences were observed in the regulation of the Av GlnK-NifL interaction by 2-oxoglutarate compared with the role of this effector in modulating the interaction of enteric PII-like proteins with their receptors. We also report that the interaction between Av GlnK and NifL is abolished by site-directed substitution of a single amino acid in the T-loop region of Av GlnK and that uridylylation of the conserved tyrosine residue in the T-loop inhibits the interaction. No association was detected between Av GlnK and the N-terminal region of NifL and our results demonstrate that Av GlnK directly interacts with the C-terminal histidine protein kinase-like domain.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The NifL-NifA regulatory system in Azotobacter vinelandii controls transcription of nitrogen fixation (nif) genes in response to redox, carbon, and nitrogen status. The nif-specific transcriptional activator, NifA, activates transcription by ⁵⁴ (^N)-RNA polymerase holoenzyme at nif promoters under conditions appropriate for nitrogen fixation, and the regulatory protein NifL controls the transcriptional activation functions of NifA in response to environmental cues (1). The NifL protein utilizes discrete mechanisms to perceive these signals (2), leading to the formation of a protein-protein complex that inhibits NifA activity (3, 4). The binding of adenosine nucleotides to NifL plays a key role in transducing environmental signals to form the inhibitory protein complex (3, 5). NifL is a flavoprotein that senses redox status via an N-terminal FAD-containing PAS domain (6-8). The mechanism whereby the NifL-NifA system perceives the nitrogen status is less well understood although recent evidence implicates direct interaction with PII-like signal transduction proteins (9).

PII-like proteins are among the most widely distributed signal transduction proteins in nature and serve to integrate signals of carbon and nitrogen status to regulate nitrogen metabolism (10-12). Many proteobacteria encode two paralogous PII signal-transducing proteins, termed PII and GlnK, and representatives of this protein family have been most thoroughly characterized in the enteric bacteria. The Escherichia coli PII-like proteins are covalently modified in response to nitrogen limitation by the uridylyltransferase/uridylyl-removing enzyme (UTase/UR) encoded by glnD (13). This enzyme catalyzes the noncooperative uridylylation of PII with up to three UMP groups covalently attached per trimer. Glutamine is the primary signal for the fixed nitrogen status in enteric bacteria (14, 15), and this effector binds to UTase/UR to stimulate the uridylyl-removing activity, thus de-uridylylating PII under nitrogen-sufficient conditions. Conversely, under conditions of nitrogen limitation when the concentration of glutamine is low, the UTase/UR uridylylates PII (10-12). Covalent modification of the PII-like proteins influences their ability to interact with different receptors to control nitrogen assimilation, including adenylyltransferase (ATase)¹ (16, 17) and NtrB (NRII) (18). The Escherichia coli PII protein (Ec PII) activates the phosphatase activity and inhibits the kinase activity of NtrB (NRII) (19, 20). PII also stimulates ATase to adenylylate glutamine synthetase. In contrast, PII-UMP inhibits the activity of ATase to stimulate de-adenylylation of glutamine synthetase but does not interact with NtrB, thus enabling the latter to phosphorylate NtrC (nitrogen regulatory protein NRI) (10-12). In addition to control via covalent modification, the activity of the PII paralogues is also modulated by the synergistic binding of the low molecular weight effectors, ATP and 2-oxoglutarate (21, 22). The interaction between PII with the ATase and NtrB is allosterically regulated by the binding of 2-oxoglutarate to PII such that binding of a single molecule of 2-oxoglutarate enhances the interaction of PII with these effectors, whereas anti-cooperative binding of more than one 2-oxoglutarate molecule per PII trimer diminishes the interaction with NtrB and ATase (17, 18).

A. vinelandii apparently encodes only a single PII-like protein (called Av GlnK or formerly Av PII), which is expressed from the glnK amtB operon (23). This protein appears to be essential for cell survival, as knock-out mutations in the glnK gene are apparently lethal. Av GlnK is subject to covalent modification by a homologue of E. coli UTase/UR encoded by A. vinelandii glnD (24, 25). The UTase/UR appears to be involved in the regulation of nitrogen fixation in A. vinelandii because mutations in glnD give rise to a Nif phenotype that can be suppressed by secondary mutations in nifL (25, 26). One possible interpretation of this result is that the uridylylation status of Av GlnK mediates regulation of nitrogen fixation via interaction with the NifL-NifA system. We have recently demonstrated with purified components that Av GlnK, but not Av GlnK-UMP, stimulates the ability of NifL to inhibit transcriptional activation by NifA in the presence of 2-oxoglutarate and ATP (9). The inhibitory activity of NifL was also stimulated by Ec PII, consistent with our observation that PII-like proteins are required for NifL to inhibit NifA under conditions of nitrogen excess when the A. vinelandii nifL-nifA operon is expressed in E. coli (27). The role of 2-oxoglutarate in regulating the NifL-NifA system is further complicated by our observation that the activities of these proteins are modulated by 2-oxoglutarate in the absence of Av GlnK (9). Thus, 2-oxoglutarate controls NifL-NifA activity directly, in addition to its potential role in controlling the activity of Av GlnK. These observations suggest a model in which Av GlnK signals the nitrogen status by direct interaction with the NifL-NifA system under conditions of nitrogen excess and that inhibition of NifA by NifL is relieved by elevated levels of 2-oxoglutarate when Av GlnK is uridylylated under conditions of nitrogen limitation (Fig. 1).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Model for the regulation of the NifL-NifA system in response to Av GlnK and 2-oxoglutarate. Under conditions of nitrogen excess (+N), nonmodified Av GlnK stimulates the formation of a complex in which NifL inhibits NifA activity. Under conditions of nitrogen limitation (N), Av GlnK is uridylylated and is unable to maintain NifL in an inhibitory form. The level of 2-oxoglutarate increases under these conditions and stimulates dissociation of NifL and NifA.

To elucidate the mechanism of nitrogen sensing by the A. vinelandii NifL-NifA system, it is necessary to determine which protein component(s) interact with Av GlnK and to analyze the effectors required for this interaction. Potentially, Av GlnK could interact with either NifL or NifA, or both of these components, to modulate transcriptional activation. In this report we have investigated, with purified components, the interaction of Av GlnK with NifL and NifA. We observe a specific interaction between NifL and Av GlnK that is dependent on the presence of both 2-oxoglutarate and ATP. The interaction was not detectable with a mutant form of GlnK that is defective in stimulating inhibition of NifA activity by NifL in vitro. Furthermore, our data indicate that the C-terminal nucleotide-binding domain of NifL is sufficient for the interaction with Av GlnK.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plasmids for Protein Overexpression-- Plasmid pTJ42 expressing native A. vinelandii NifA with a C-terminal hexahistidine tag (NifA_6-His) was prepared from plasmid pDB737 (28). An EcoRI-BamHI fragment from pDB737 containing the 3' end of NifA was cloned into the vector pTE103 to give plasmid pTJ39 containing a single BglII site within the EcoRI-BamHI region. A double-stranded oligonucleotide with engineered BglII and BamHI sticky ends and encoding a hexahistidine tag and stop codon, respectively, was cloned into pTJ39 yielding plasmid pTJ41. The EcoRI-BamHI fragment from pTJ41 was subsequently cloned into pDB737 to give plasmid pTJ42. Plasmid pVCO1, expressing A. vinelandii GlnD with a N-terminal histidine tag, was obtained by cloning a NdeI-BamHI fragment from pYZ5 (9) into the expression vector pET28(a)+ (Novagen). Ec PII was expressed from plasmid pBOP1 in E. coli strain RB9060 (glnB) as described in Ref. 29. All other constructs for protein expression have been previously described (2, 6, 9).

Site-directed Mutagenesis-- Mutagenesis was carried out with the QuikChange site-directed mutagenesis kit (Stratagene) according to the manufacturer's instructions. Mutagenic primers were 5'-AAGGGCCATACCTGCCTGTACCGGGG-3' (E44C) and 5'-GTACCGGGGTGCGTGCTACGTAGTCGAC-3' (E50C). All constructs were sequenced commercially (MWG Biotech) over their entire length to ensure that only the desired mutations were introduced. Plasmids overexpressing Av GlnK mutants were derived from plasmid pYZ1 described previously (9).

Protein Purification-- Tagged proteins used in this work were purified by metal chelate affinity chromatography. Two 1-ml HiTrap Chelating HP columns (Amersham Biosciences) were connected in series and equilibrated with 20 mM Tris-HCl, 50 mM KSCN, 5 mM imidazole, 250 mM NaCl, pH 8.0. Purification was carried out using Biocad Sprint Perfusion chromatography (PerkinElmer Biosystems) using a linear gradient from 0 to 750 mM imidazole over a 40-ml elution volume.

Av GlnK, Ec PII, ⁵⁴, and IHF were purified as described previously (9). Molar protein concentrations were calculated on the basis that NifL_6-His purifies as a tetramer and the truncated NifL polypeptides NifL-(1-284)_6-His, NifL-(147-519)_6-His, and NifL-(360-519)_6-His are tetrameric, dimeric, and monomeric, respectively (2). The molar concentrations of NifA and Av GlnK were calculated on the basis that these proteins are a dimer and trimer, respectively.

Open Complex Assays-- NifA-promoted catalysis of open promoter complexes by ⁵⁴-RNA polymerase was used to assay NifA activity and its inhibition by NifL as described previously (5, 9, 30). Linearized template DNA was provided by digesting plasmid pNH8 with EcoRI and BamHI to yield a 260-bp fragment, containing the Klebsiella pneumoniae nifH promoter and upstream activator sequences, which was 3' end-labeled with [-³²P]dGTP at the BamHI site. Reactions (final volume of 15 µl) were carried out in TAP buffer (50 mM Tris acetate, 100 mM potassium acetate, 8 mM magnesium acetate, 3.5% polyethylene glycol 8000, 1 mM dithiothreitol, pH 7.9) and contained 5 nM template DNA, 3.4 µg/ml denatured salmon sperm DNA, 125 nM core RNA polymerase, 200 nM ⁵⁴, 100 nM IHF, and 3.5 mM ATP. GTP (final concentration 500 µM) was present prior to the heparin challenge to allow formation of initiated complexes, which are more stable than open promoter complexes (5). In some cases an ATP regenerating system was provided by adding creatine kinase (20 units/ml) and creatine phosphate (12 mM). The reaction components (including Av GlnK at concentrations indicated in Fig. 2) were preincubated for 2 min at 30 °C, and reactions were then initiated by the addition of either NifA alone or NifA plus NifL. After 20 min of incubation, reactions were mixed with 3 µl of dye mix containing 50% glycerol, 0.05% bromphenol blue, 0.1% xylene cyanol, and 2 µg of heparin and immediately loaded onto a 4% (w/v) polyacrylamide gel (acrylamide:bisacrylamide ratio 80:1) in 25 mM Tris, 400 mM glycine, pH 8.6, which had been prerun at 180 V at room temperature down to a constant power of 2 watts. Gels were run for 2.5-3 h at 100 V. Gels were dried down, and the percentage of radioactivity in open complexes quantitated with the Fujix BAS1000 phosphorimager.

Retention of Proteins on Ni-NTA-Magnetic Agarose Beads-- Ni-NTA-magnetic agarose beads (Qiagen) were equilibrated in buffer containing 10 mM HEPES, 150 mM NaCl, 25 mM MgCl₂, and 20 mM imidazole, pH 7.5. NifL polypeptides, at a concentration of 0.3 µM, were immobilized by preincubation in a total volume of 500 µl containing 50 µl of the magnetic bead suspension. After 30 min, the beads were washed in the above buffer and Av GlnK was added to a final concentration of 1 µM in a volume of 500 µl. Following an additional 60-min incubation, the beads were washed, the buffer subsequently removed, and elution performed with HEPES buffer as above but containing 500 mM imidazole. Aliquots of each sample were analyzed by SDS-polyacrylamide gel electrophoresis.

BIAcore Surface Plasmon Resonance Detection-- Surface plasmon resonance experiments were performed using a BIAcore X biosensor system (BIAcore AB). Hexahistidine-tagged derivatives of NifL and NifA were immobilized on nickel-NTA biosensor chip surfaces. Nickel was first bound to the NTA surface through injection of 20-µl volumes of 500 µM nickel chloride. Proteins for immobilization were introduced to the protein-chip surface at a concentration of 40 nM. Experiments were performed at 25 °C in buffer containing 10 mM HEPES, 150 mM NaCl, 25 mM MgCl₂, pH 7.4, at a flow rate of 20 µl/min. Most experiments were carried out in the presence of a control protein in the reference flow cell, as stated in the figure legends. In all cases it was ensured that the binding response of the control protein upon immobilization was at least equal to that of the sample protein.

Isothermal Titration Calorimetry (ITC)-- Experiments were performed in a VP-ITC isothermal titration calorimeter (MicroCal, Inc.) at 28 °C in a cell volume of 1.35 ml. Binding of 2-oxoglutarate to PII-like proteins was measured by titrating 2 mM ligand from a 250-µl injection syringe into the sample cell, which was stirred at 300 rpm. 5- or 10-µl injections were made to the stoichiometric excess given in Fig. 6. Buffer conditions were 10 mM HEPES, 50 mM NaCl, 25 mM MgCl₂, 1 mM ATP, pH 7.0. The PII proteins were dialyzed overnight prior to ITC, and protein concentration was determined by the Bradford method, with bovine serum albumin as the standard. The heat change for the dilution of the ligand in the absence of protein was measured for each experiment and was subtracted from the measured heat change of ligand binding to protein. Data analysis was performed with the Origin program, provided by MicroCal.

Uridylylation of Av GlnK-- The uridylylation method was based on previously published methods (22). The reaction mixture contained 100 mM Tris-HCl, 25 mM MgCl₂, 100 mM KCl, 1 mM dithiothreitol, 0.3 mg/ml bovine serum albumin, 0.5 mM ATP, 0.5 mM UTP, 500 µM 2-oxoglutarate, pH 7.5. Av GlnK was present at a concentration of 15 µM and GlnD_6-His at a concentration of 1 µM. Following incubation at 30 °C, aliquots of the reaction were removed at time intervals to monitor uridylylation. Glycerol and KCl were added to 10% (v/v) and 350 mM final concentrations, respectively, and the samples were heated to 60 °C for 15 min to inactivate UTase/UR. Samples were added to an equal volume of native sample buffer (125 mM Tris-HCl, 20% glycerol, 0.05% bromphenol blue, pH 8.0) and analyzed by nondenaturing polyacrylamide gel electrophoresis.

Limited Proteolysis-- Limited proteolysis was performed at 20 °C in Tris acetate buffer comprising 50 mM Tris acetate, 100 mM potassium acetate, 8 mM magnesium acetate, 1 mM dithiothreitol, 1 mM ATP, 500 µM 2-oxoglutarate, pH 7.0. NifL and GlnK were combined in a final reaction volume of 130 µl and preincubated for 10 min prior to initiation of digestion. A trypsin:NifL weight ratio of 1:200 was used. 15-µl samples were removed at the time intervals indicated in the figure legend to tubes containing a 2-fold weight excess of soybean trypsin-chymotrypsin inhibitor. To these samples was added 15 µl of gel loading buffer (125 mM Tris-HCl, 4% sodium dodecyl sulfate, 20% glycerol, 10% -mercaptoethanol, 0.05% bromphenol blue, pH 8.6). Samples were heated at 100 °C for 4 min prior to electrophoretic separation.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Mutant Forms of Av GlnK Defective in Regulating NifL-NifA Activity-- To provide controls for the protein-protein interaction experiments, we performed site-directed mutagenesis of the surface exposed T-loop region of Av GlnK, which contains the tyrosine residue that is the target site for uridylylation and, by analogy with the well characterized E. coli PII-like proteins, is likely to play a major role in the interaction with receptors (31-34). The T-loop sequence of the Av GlnK protein (residues 37-55) differs from Ec PII in only a single amino acid at position 52. Engineered derivatives of Ec PII with single cysteine substitutions at glutamate 44 and glutamate 50 have been used previously to probe interactions between Ec PII and NtrB (NRII) with heterobifunctional cross-linking reagents, which label thiol groups (35). Although these mutations do not perturb the PII-NtrB interaction, mutations in the T-loop frequently allow discrimination between receptors (31). We prepared the equivalent E44C and E50C mutations in Av GlnK and purified the mutant proteins following overexpression in E. coli (Fig. 2A). Both mutant proteins behaved similarly to the native protein on purification. However, in contrast to the native protein, neither of the mutant proteins could be uridylylated by purified A. vinelandii UTase/UR (data not shown), indicating that they are either defective in the interaction with the UTase or are defective as substrates in the catalytic step.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Activity of wild-type and mutant forms of Av GlnK. A, purification of mutant and wild-type forms of Av GlnK. The samples were run on 12.5% SDS-polyacrylamide gel and stained with Brilliant Blue R250. Lane 1 shows markers with molecular mass in kilodaltons. Lanes 2-4 show wild-type Av GlnK, Av GlnK E44C, and Av GlnK E50C, respectively. B, influence of wild-type and mutant GlnK proteins on the activity of NifL and NifA, as determined by the formation of open promoter complexes. The assay contained 200 nM NifA dimer, 150 nM NifL dimer, 0.1 mM 2-oxoglutarate, 3.5 mM ATP, 0.5 mM GTP, 12 mM creatine phosphate, and 20 units/ml creatine kinase. Activity is plotted relative to the extent of NifA activity (open promoter complex formation) in the presence of NifL and 2-oxoglutarate. Additions were wild-type Av GlnK (closed squares), Av GlnK E44C (closed triangles), and Av GlnK E50C (open triangles).

The ability of the mutant proteins to activate the inhibitory function of NifL was assayed by measuring the influence of NifL on transcriptional activation by NifA at the nifH promoter in vitro. These assays quantitate the formation of heparin-resistant open promoter complexes by ⁵⁴-RNA polymerase holoenzyme catalyzed by NifA in the presence of nucleoside triphosphates and IHF (5). To simplify the assays, we use a truncated form of NifL, NifL-(147-519)_6-His, which lacks the redox response but retains the response to nitrogen status in vivo (2) and the ability to respond to Av GlnK in vitro (9). This truncated protein thus allows reactions to be performed under aerobic conditions without activating the redox response of NifL. NifL-(147-519)_6-His inhibits NifA activity in the presence of ATP, but this inhibition is relieved by the addition of 2-oxoglutarate. However, the addition of native Av GlnK enhances the inhibitory activity of NifL-(147-519)_6-His when both ATP and 2-oxoglutarate are present (9) (Fig. 2B). Whereas Av GlnK E50C, like native Av GlnK, possesses this activity, the mutant Av GlnK E44C protein exerted little or no influence on inhibition of NifA activity by NifL (Fig. 2B). Neither the native nor the mutant proteins had any influence on NifA activity in the absence of NifL (data not shown). It would therefore appear that the GlnK E44C mutant is substantially defective in activating the inhibitory function of NifL, whereas the Av GlnK E50C mutant is not defective in this function.

Av GlnK Specifically Interacts with NifL in the Presence of ATP and 2-Oxoglutarate-- A "pull-down" assay using Ni-NTA-magnetic agarose beads was employed to probe for interactions between Av GlnK and immobilized NifL and NifA in the presence and absence of effectors. Purified derivatives of NifL and NifA bearing a hexahistidine tag at the C terminus were immobilized on the Ni-NTA bead surface, and the beads were then incubated in the presence of Av GlnK. A wash phase was carried out to remove adventitiously bound protein, and elution of the His-tagged proteins was subsequently performed with 500 mM imidazole.

By analogy with the well characterized interactions between Ec PII-like proteins and their receptors, we suspected that the binding of both 2-oxoglutarate and ATP to Av GlnK would be required to detect any interaction. Accordingly, no interaction was resolved in the absence of ATP or 2-oxoglutarate, or in the presence of only a single effector (Fig. 3, lanes 2-4). In the presence of both ATP (3.5 mM) and 2-oxoglutarate (2 mM), Av GlnK co-eluted with the native NifL_6-His protein (Fig. 3, lanes 5 and 6) but not with the NifA_6-His protein (Fig. 3, lane 9). The Av GlnK E44C mutant protein defective in stimulation of NifL inhibition in vitro did not co-elute with NifL (Fig. 3, lane 7). In contrast, the GlnK E50C mutant, which possesses comparable activity to wild-type Av GlnK in the in vitro inhibition assay (Fig. 2B), was also retained on the beads with NifL (Fig. 3, lane 8). Control experiments revealed that neither native Av GlnK nor the mutant forms were retained on the beads in the absence of NifL (Fig. 3, lanes 10-19).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Co-retention of GlnK with NifL6-His on nickel-NTA-agarose requires ATP and 2-oxoglutarate. Reactions were as described under "Experimental Procedures" and contained either NifL6-His (0.3 µM) or NifA6-His (0.3 µM) immobilized on nickel-NTA-magnetic agarose beads. GlnK proteins were added to a final concentration of 1 µM. ATP (3.5 mM) and/or 2-oxoglutarate (2 mM) were added as indicated. Samples were run on 12.5% SDS-polyacrylamide gels. The gels were stained with Brilliant Blue R250. Lanes 1 and 21 show markers with molecular mass in kilodaltons.

The C-terminal Kinase-like Domain of NifL Interacts with Av GlnK-- A. vinelandii NifL contains at least three domains: the N-terminal redox-responsive PAS domain containing FAD (residues 1-170), the central region (residues 170-285) with unknown function, and the C-terminal domain (residues 290-519), which resembles the conserved transmitter domain found in members of the histidine protein kinase family (1, 2, 36, 37). By analogy with other histidine protein kinase transmitter modules whose structure is known, the C-terminal region of NifL probably consists of two domains, an "H"-like domain (residues 285-350), which contains a conserved histidine residue and may form a dimerization interface in NifL, and a "kinase"-like domain (residues 360-519) containing conserved N, G₁, F, and G₂ motifs found in other histidine protein kinases (Fig. 4A). Although the kinase-like domain of NifL binds adenosine nucleotides (2), no autokinase or phosphotransferase activities have been detected in vitro (28). Nevertheless, the similarity between NifL and the histidine protein kinase family is of interest, in light of the recent demonstration that the C-terminal kinase domain of NtrB (NRII) interacts with enteric PII (35, 38, 39).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Interaction of native and truncated NifL proteins with wild-type Av GlnK on nickel-NTA-agarose. Reaction conditions and component concentrations were identical to those in Fig. 3. A, diagram of the domain structure of NifL and the truncated polypeptides analyzed in this study. B, co-retention of NifL6-His and NifL-(147-519)6-His with wild-type Av GlnK. Lane 1 shows markers with molecular mass in kilodaltons. C, incubation of wild-type Av GlnK with native NifL6-His, the NifL N terminus (NifL-(1-284)6-His), and NifL-(360-519)6-His. Panel D shows the result of running a duplicate gel for Western analysis using anti-Ec PII antibody.

We sought to delineate regions of NifL that are critical for interaction with Av GlnK. The truncated NifL-(147-519)_6-His protein (2),which lacks the flavin-containing PAS domain (Fig. 4A), gave similar results to full-length NifL_6-His in the magnetic bead assay and Av GlnK co-eluted with this protein in the presence of 2-oxoglutarate and ATP (Fig. 4B, compare lanes 2 and 3 with lanes 4 and 5). Hence, the redox-responsive N-terminal PAS domain is not required for the interaction, as expected, because the NifL-(147-519)_6-His protein is responsive to Av GlnK in the in vitro inhibition assay (Fig. 2B) and is also responsive to nitrogen regulation in vivo (2, 27). To resolve whether the central region of NifL might be sufficient for the interaction, we performed pull-down experiments with an N-terminal polypeptide (NifL-(1-284)_6-His), which corresponds to a stable proteolytic fragment containing both the PAS domain and the central region (Fig. 4A) (2). In the nickel-NTA magnetic bead assay, Av GlnK was not retained with this polypeptide in the presence of ATP and 2-oxoglutarate inferring a requirement for the C-terminal region of NifL (Fig. 4C, lanes 6 and 7). To elucidate the role of the C-terminal region of NifL, we purified a polypeptide corresponding to the kinase-like domain (NifL-(360-519)_6-His) (Fig. 4A), which has been demonstrated to bind adenosine nucleotides (2). This polypeptide also co-eluted with Av GlnK in the magnetic agarose bead assay, suggesting that the kinase-like domain is sufficient for specific recognition (Fig. 4C, lanes 4 and 5). To confirm that Av GlnK interacted with NifL_6-His and NifL-(360-519)_6-His, but not NifL-(1-284)_6-His, we performed Western immunoblotting of the gel shown in Fig. 4C with anti-Ec PII antibody (which cross-reacts with Av GlnK). Av GlnK clearly did not co-elute with NifL-(1-284)_6-His but did so with the isolated "kinase" domain and with full-length NifL (Fig. 4D, compare lanes 6 and 7 with lanes 1-4).

BIAcore Analysis: Role of ATP and 2-Oxoglutarate in the Interaction-- We employed BIAcore surface plasmon resonance (SPR) instrumentation to further investigate interactions between Av GlnK and the NifL and NifA proteins. Hexahistidine-tagged derivatives of NifL and NifA were immobilized on chelating NTA sensor chips, and Av GlnK was injected as analyte under continuous flow conditions in the presence of ATP (3.5 mM) and 2-oxoglutarate (2 mM). Whereas immobilized NifA_6-His did not respond to Av GlnK under these conditions, an interaction was observed with NifL-(147-519)_6-His (Fig. 5A). In contrast, NifL-(1-284)_6-His showed no interaction with Av GlnK (data not shown), as anticipated from the magnetic bead assay. Because neither NifA nor NifL-(1-284)_6-His is apparently able to interact with Av GlnK, they were used as background controls when immobilized in the control flow cell. The overlay plot in Fig. 5B shows sensorgrams obtained with immobilized NifL-(147-519)_6-His with increasing Av GlnK concentrations. Kinetic analysis of the interaction in four independent experiments gave a K_D of 1 ± 0.25 µM. In contrast to native GlnK, the Av GlnK E44C mutant protein failed to interact even when added at high concentration (5 µM) (Fig. 5B), thus demonstrating the specificity of the interaction.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Surface plasmon resonance detection of the Av GlnK-NifL interaction and the influence of effectors. A, original data showing sensorgrams derived from injecting 0.6 µM GlnK over the sensorchip surface alone, or with either NifA6-His or NifL-(147-519)6-His immobilized on the NTA surface. Conditions were as described under "Experimental Procedures." ATP and 2-oxoglutarate were present at concentrations of 3.5 and 2 mM, respectively. B, titration of Av GlnK with NifL-(147-519)6-His in the presence of ATP and 2-oxoglutarate. The control flow cell contained NifA6-His, and the sample flow cell contained NifL-(147-519)6-His. The Av GlnK concentration was varied from 0.0625 to 5 µM. The sensorgram at the bottom with zero response was obtained from the injection of 5 µM Av GlnK but lacked MgCl2 in the analyte buffer. The sensorgram above this (also with no significant response) was obtained following injection of 5 µM E44C GlnK. C, influence of ATP concentration on the interaction of Av GlnK with NifL-(147-519)6-His. The control flow cell contained NifL-(1-284)6-His. 2-Oxoglutarate

Because the co-chromatography experiments revealed a requirement for ATP for the NifL-GlnK interaction, we further examined this dependence using BIAcore SPR. The role of adenosine nucleotides in this interaction is complicated by the occurrence of ATP binding sites on both GlnK and NifL. Ec PII binds ATP with relatively high affinity (apparent K_d 0.25 µM) (21), whereas the affinity of NifL for ATP is lower (apparent K_d ~130 µM). In the presence of 2 mM 2-oxoglutarate, no interaction was observed between Av GlnK and NifL-(147-519)_6-His in the absence of ATP and titration of the ATP concentration from 0.0625 to 500 µM resulted in a significant increase in the protein-protein interaction without any apparent effect on the association and dissociation rates (Fig. 5C). Applying a global fitting routine to this data resolved an apparent K_act for ATP of 197 µM. Because this value represents significantly weaker binding of ATP at this concentration of 2-oxoglutarate than that observed with other PII-like proteins (13, 21, 40) and is similar to the binding constant of ATP for NifL (2), it is possible that the involvement of ATP in the interaction reflects a requirement for binding of this ligand to both Av GlnK and NifL. We also observed that the interaction is dependent on the presence of Mg²⁺ (Fig. 5B), as has been observed for the Ec PII-NtrB interaction (21).

As mentioned in the Introduction, 2-oxoglutarate is also an allosteric effector of PII-like proteins and influences the interaction between Ec PII and NtrB (18, 20, 21). In the presence of 0.6 µM Av GlnK and 3.5 mM ATP, the GlnK-NifL interaction was responsive to the concentration of 2-oxoglutarate with increased binding observed within the range 5 µM to 2 mM (Fig. 5D). Similar data were obtained using a lower concentration of Av GlnK (0.125 µM, data not shown). We have not detected binding of 2-oxoglutarate to NifL, so it is likely that the 2-oxoglutarate requirement reflects binding to Av GlnK. The response to 2-oxoglutarate is within the reported physiological range, but its influence on the interaction at high concentration was unexpected, because the interaction of NtrB and ATase with Ec PII decreases as more than one binding site on the PII trimer becomes occupied (17, 18). The apparent K_act for 2-oxoglutarate calculated from three independent experiments was 25 ± 3 µM, which is higher than the experimentally determined value for binding of the first 2-oxoglutarate molecule to E. coli PII (~5 µM) (13, 21). Hence, either Av GlnK has a lower affinity for 2-oxoglutarate compared with E. coli PII or the interaction between Av GlnK and NifL requires the binding of more than one molecule of 2-oxoglutarate.

Binding of 2-Oxoglutarate to Ec PII and Av GlnK-- In an attempt to distinguish between the above possibilities, we performed ITC to compare the binding of 2-oxoglutarate to E. coli PII and Av GlnK. Titration of 70 µM E. coli PII with 2-oxoglutarate (concentration range, 0-450 µM) in the presence of 1 mM ATP gave an apparent K_d for 2-oxoglutarate of 17 µM. Enthalpy values were negative, demonstrating that the binding of 2-oxoglutarate is an exothermic process. Data analysis gave a best fit to a single-site model with a binding stoichiometry of 0.8 ± 0.01 mol/mol trimer (Fig. 6A). The concentration of protein used in these experiments was presumably too low to permit detection of the binding of additional molecules of 2-oxoglutarate to PII. It has been suggested that the binding of a single molecule of 2-oxoglutarate to PII exerts negative cooperativity upon the binding of additional molecules (11, 13, 20). Two independent ITC experiments with Av GlnK gave substantially different results to those obtained with Ec PII, although in this case it was more difficult to fit the data to a particular binding model. With 48 µM Av GlnK in the presence of 1 mM ATP, the best fit to the data gave a binding stoichiometry of 3.3 mol/mol trimer and an apparent K_d for 2-oxoglutarate of ~23 µM, assuming that each site has equal affinity. At 64 µM Av GlnK and saturating ATP, the data fitted best to an n value of 2 and an apparent K_d of ~54 µM (Fig. 6B). Because these values are in the same order of magnitude as the K_act for the interaction of Av GlnK with NifL, it is likely that 2-oxoglutarate acts exclusively as an effector of GlnK in this interaction. The data also suggest that Av GlnK, has a lower affinity for 2-oxoglutarate than Ec PII, but is competent to bind more than one molecule of 2-oxoglutarate at relatively low effector concentrations. Thus, Av GlnK apparently exhibits decreased negative cooperativity for effector binding compared with Ec PII and displays a response to 2-oxoglutarate similar to that observed with PII-UMP (13). Thus, binding of more than one molecule of 2-oxoglutarate to Av GlnK may be necessary to promote the interaction with NifL.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Calorimetric titration of the binding of 2-oxoglutarate to PII-like proteins. Titrations were carried out in buffer containing 1 mM ATP as described under "Experimental Procedures." A, binding isotherm observed upon injecting 2-oxoglutarate into Ec PII (70 µM trimer). The integrated heats of reaction are plotted against the molar ratio of total ligand concentration to total protein concentration. The solid line shows the best fit to the data according to a equivalent site binding model with n = 0.807 (± 0.01) sites/mol, Kd = 17 (± 0.6) µM, and H = 7.8 (±0.2) kcal/mol. B, binding isotherm observed upon injecting 2-oxoglutarate into Av GlnK at a trimer concentration of 64 µM. The solid line shows the best fit to the data according to a equivalent site binding model with n = 2.076 (± 0.01) sites/mol, Kd = 54.2 (± 4.1) µM, and H = 5.6 (±0.2) kcal/mol. was present at a concentration of 2 mM. From bottom to top, the solid lines show the interaction of 0.6 µM Av GlnK at an ATP concentration of 0, 5, 25, 50, 200, and 500 µM, respectively. D, influence of 2-oxoglutarate concentration on the interaction of Av GlnK with NifL-(147-519)6-His. The control flow cell contained NifA6-His. ATP was present at a concentration of 3.5 mM. From bottom to top, the solid lines show the interaction of 0.6 µM Av GlnK at a 2-oxoglutarate concentration of 0, 5, 10, 25, 50, 200, 500, and 2000 µM, respectively.

Influence of Uridylylation of Av GlnK on the Interaction with NifL-- We have previously demonstrated that the uridylylated form of Av GlnK is unable to stimulate the inhibitory activity of NifL in vitro (9), and it was therefore of interest to determine whether the interaction between Av GlnK and NifL was influenced by covalent modification. Av GlnK was incubated with A. vinelandii UTase/UR and UTP, and progressive uridylylation was followed by native gel analysis. The fully uridylylated protein (bearing three UMP groups per trimer) is resolved as the band of highest mobility and, under our experimental conditions, is the sole species following 40 min of incubation (see "Experimental Procedures").

In the "pull-down" magnetic agarose bead assay, fully uridylylated GlnK did not co-elute with NifL and was removed in the wash phase (Fig. 7A, lanes 4 and 5) whereas nonmodified GlnK co-eluted with NifL as expected (Fig. 7A, lanes 2 and 3). When used as the analyte in BIAcore SPR analysis, GlnK-UMP gave rise to a diminished response when compared with equivalent concentrations of nonmodified GlnK, although we detected no major changes in the association or dissociation rates (Fig. 7B). Hence, covalent modification of GlnK significantly influences the interaction with NifL as expected from our previous experiments, which demonstrated that uridylylation of Av GlnK diminishes its ability to activate the inhibitory function of NifL in vitro (9).

View larger version (15K): [in this window] [in a new window]   Fig. 7.   Influence of uridylylation on the interaction of Av GlnK with NifL. A, incubation of wild-type Av GlnK and GlnK-UMP with NifL6-His on nickel-NTA-agarose. Conditions were as described in Fig. 3. Lane 1 shows markers with molecular mass in kilodaltons. B, surface plasmon resonance detection. Conditions were as described under "Experimental Procedures." ATP and 2-oxoglutarate were present at a concentration of 3.5 and 2.0 mM, respectively. The control flow cell contained NifL-(1-284)6-His. Repeat injections of 0.6 µM GlnK and of 0.6 µM GlnK-UMP were made.

Effect of Av GlnK on Limited Proteolysis of the NifL Protein-- We have shown previously that binding of adenosine nucleotides to NifL changes the pattern of trypsin digestion, suggesting that nucleotide binding may alter the conformation of the C-terminal domain (2). To investigate whether the interaction of Av GlnK with NifL can be detected by limited proteolysis, we subjected NifL_6-His to trypsin digestion in the presence of ATP and 2-oxoglutarate. In the absence of Av GlnK, 2-oxoglutarate did not alter the pattern of digestion of NifL, either in the presence or absence of ATP (data not shown). Av GlnK was not subject to digestion under the conditions used. When NifL and the GlnK E44C mutant were both present, the digestion pattern was unchanged from that observed with NifL alone (Fig. 8A), the major proteolysis products under these conditions being the 33-kDa N-terminal fragment and the C-terminal kinase-like domain as observed previously (2). However, in the presence of native GlnK, NifL was protected from digestion because the rate of cleavage of the full-length protein decreased (Fig. 8B), although no major changes were detected in the pattern of proteolysis. Control reactions showed no protection by Av GlnK when either ATP or 2-oxoglutarate were absent from the reaction (data not shown).

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Limited trypsin proteolysis of NifL6-His in the presence and absence of wild-type and mutant forms of Av GlnK. A, original data. Reaction mixtures contained 1.25 µM NifL6-His and, where indicated, either 1.5 µM wild-type Av GlnK or 1.5 µM GlnK E44C. Lanes 2-8, NifL6-His, ATP, and 2-oxoglutarate at t = 0, 5, 20, 40, 60, 80, and 100 min, respectively; lanes 9-15, NifL6-His, ATP, 2-oxoglutarate, and Av GlnK E44C (time points correspond to those of lanes 2-8); lanes 16-22, NifL6-His, ATP, 2-oxoglutarate, and wild-type Av GlnK (time points correspond to those of lanes 2-8); lanes 23 and 24, Av GlnK E44C, ATP, and 2-oxoglutarate at t = 0 and 100 min, respectively; lanes 25 and 26, wild-type Av GlnK, ATP, and 2-oxoglutarate at t = 0 and 100 min, respectively. Lanes 1 and 27 show markers with molecular mass in kilodaltons. B, quantitative analysis of the data. The percentage of NifL remaining undigested at the indicated times was quantified using the MacBas version 2.0 image analysis software (Fuji Photo Film Company Ltd.). Reactions contained NifL6-His (closed squares), NifL6-His and Av GlnK E44C (closed triangles), or NifL6-His and wild-type Av GlnK (closed circles).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The model we propose for nitrogen regulation of the A. vinelandii NifL-NifA system invokes a pivotal role for Av GlnK in signaling the nitrogen response. We have demonstrated that the nonuridylylated form of Av GlnK, which is expected to be the major form of the protein present under conditions of nitrogen excess, directly interacts with the NifL protein and stimulates its inhibitory activity, presumably by favoring the formation of the NifL-NifA complex. To maintain the inhibitory activity of NifL under conditions of nitrogen sufficiency, it is possible that a ternary complex is formed, which dissociates under nitrogen-limiting conditions as the level of 2-oxoglutarate is elevated and GlnK becomes covalently modified. This study demonstrates that uridylylation of Av GlnK inhibits the interaction with NifL, and we have shown previously that uridylylation of a single subunit within the GlnK trimer is sufficient to decrease the inhibitory influence of NifL on NifA activity (9). In vivo studies also suggest that the uridylylated form of Av GlnK is required to prevent NifL from inactivating NifA under conditions of nitrogen limitation, and interaction between Av GlnK and NifL has been demonstrated recently using a yeast two-hybrid system (41).

Biochemical studies have established that the binding of ATP and 2-oxoglutarate to PII-like regulatory proteins influences interactions with their receptors. This also appears to be the case for the Av GlnK-NifL binary interaction, because we have shown that the binding of GlnK to NifL requires Mg²⁺, ATP, and 2-oxoglutarate. The interaction of Ec PII and GlnK with ATase and NtrB (NRII) is regulated allosterically by 2-oxoglutarate, such that these interactions are stimulated by low concentrations of this effector but are inhibited at higher concentrations when additional ligand binding sites on the PII proteins become occupied (11, 22). The involvement of this small molecule effector in the interaction of Av GlnK with NifL is clearly different from that observed with Ec PII and its receptors (e.g. ATase and NtrB), in that the interaction is favored when the concentration of 2-oxoglutarate is saturating. This suggests that 2-oxoglutarate does not act as an allosteric effector of the Av GlnK-NifL interaction. The K_act for this effector (~23 µM) is similar to the dissociation constants obtained for Av GlnK using isothermal titration calorimetry, and the binding studies suggest that the Av GlnK trimer binds more than one molecule of 2-oxoglutarate under conditions in which Ec PII binds only a single molecule of the effector per trimer. This binding behavior is similar to that exhibited by Ec PII-UMP, which shows reduced negative cooperativity compared with nonmodified Ec PII and saturates at three molecules of 2-oxoglutarate per trimer (13). The stimulation of the Av GlnK-NifL interaction by 2-oxoglutarate also resembles the role of this effector in regulating de-adenylylation of GS by PII-UMP (17), where 2-oxoglutarate does not apparently play an allosteric role within the physiological range of this effector, which varies from 0.1 to 1 mM in intact cells (27, 42). At first sight, the stimulatory influence of 2-oxoglutarate on the Av GlnK-NifL interaction appears counterintuitive because it would facilitate binding of nonmodified Av GlnK to NifL under nitrogen-limiting conditions, when the level of 2-oxoglutarate increases. However, this condition is unlikely to be prevalent in vivo because uridylylation of Av GlnK would be favored. Moreover, because the NifL-NifA system responds independently to 2-oxoglutarate at concentrations within the physiological range (when Av GlnK is absent) (9), it is possible that the additional binding of this effector to a component other than GlnK (e.g. NifA) might help to destabilize any ternary interaction between GlnK, NifL and NifA under conditions of nitrogen limitation, thus freeing NifA to activate transcription. Under nitrogen excess conditions, the level of 2-oxoglutarate (~0.1 mM) is sufficient to promote the interaction of Av GlnK with NifL. Moreover, even in the absence of Av GlnK, this concentration of the effector is not sufficient to prevent NifL from inhibiting NifA in the presence of adenosine nucleotides (9), thus favoring formation of a complex in which NifA is inactivated (Fig. 1). The involvement of Mg ATP in the GlnK-NifL interaction is complicated by the potential presence of nucleotide binding site(s) on both NifL and GlnK. The apparent K_act for MgATP (197 µM) is similar to the apparent K_d for the binding of MgATP to the C-terminal domain of NifL, and this may imply that the conformation change induced by the binding of adenosine nucleotides to NifL (2) is a prerequisite for the interaction.

The role of Av GlnK in signaling the nitrogen status to NifL may be analogous to that of Ec PII in controlling the kinase and phosphatase activities of NtrB (NRII). Although A. vinelandii NifL does not apparently hydrolyze ATP and neither kinase nor autophosphorylation activities have been detected, the C-terminal region of NifL does show significant homology to the transmitter region of the histidine protein kinases and in particular contains the four conserved regions designated N, G₁, F, and G₂, which are implicated in nucleotide binding (36, 37). Because adenosine nucleotides bind to the kinase-like domain to stimulate the NifL-NifA interaction, it is possible that this domain of NifL represents either an ancient precursor of the histidine protein kinases or has evolved from the "classical" kinases with loss of catalytic function (nucleoside triphosphate hydrolysis), so that nucleotide binding is utilized to drive conformational changes that promote interaction with NifA (3, 4). The homology between the kinase-like domains of NifL and NtrB (NRII) is of particular interest in view of the recent finding that this domain of NtrB (residues 190-339) interacts with Ec PII (35, 38). The C-terminal region of NifL is implicated in the interaction with Av GlnK because the N-terminal region (residues 1-284) did not interact, whereas the central plus the C-terminal regions were competent to do so. Our data demonstrate that a fragment of NifL (residues 360-519) corresponding to the kinase domain of NtrB is sufficient to interact with GlnK and this polypeptide, like its homologues in the histidine protein kinase family, binds adenosine nucleotides and purifies as a monomer in solution.

X-ray crystallographic analysis of the homotrimeric PII and GlnK proteins from E. coli reveals a similar structural core, although the conformation of the apical T-loop can alter significantly (43-45). The T-loop region of Ec PII is necessary for interaction with ATase, UTase, and NtrB and includes Tyr-51, the site of uridylylation (31, 33). The T-loop sequence of Av GlnK (residues 37-55) differs from Ec PII in only a single amino acid at position 52 and therefore may be functionally homologous to PII rather than GlnK, because GlnK proteins commonly differ from PII at residues 43, 51, and 53 (46). We have previously shown that Ec PII, but not Ec GlnK, can substitute for Av GlnK to stimulate the inhibitory activity of NifL in vitro (9). The Av GlnK E50C mutation renders the protein inert to uridylylation (data not shown), although it retains the ability to interact with NifL. However, the Av GlnK E44C mutant is unable to associate with NifL and is not responsive to UTase. In contrast, the equivalent mutant form of Ec PII is competent to interact with NtrB (NRII) (35), although the response of this mutant protein to UTase has not been reported. These observations further reinforce the importance of the T-loop region and the key nature of residues in receptor discrimination (31).

The mechanism of nitrogen sensing by the A. vinelandii NifL-NifA system (Fig. 1) is clearly different from that in K. pneumoniae because, in the former system, Av GlnK increases the inhibitory activity of NifL under conditions of nitrogen excess (9, 27), whereas, in the latter system, K. pneumoniae GlnK is required to prevent NifL from inhibiting NifA under nitrogen-limiting conditions (46, 47). As mentioned above, Av GlnK is likely to be functionally homologous to Ec PII in terms of the specificity of interaction with A. vinelandii NifL. In contrast, Ec PII functions poorly with the K. pneumoniae NifL-NifA system in the absence of mutations in the T-loop that alter the specificity toward that of GlnK (34). Furthermore, the ability of enteric GlnK to prevent inhibition of K. pneumoniae NifA activity by NifL is not influenced by the uridylylation state of GlnK (47). In addition the C-terminal domain of K. pneumoniae NifL, unlike that of A. vinelandii does not show strong homology to the histidine protein kinase family (37). In light of these differences, it is possible that enteric GlnK interacts with K. pneumoniae NifA to prevent inhibition by NifL, in line with genetic studies that implicate interaction between PII-like proteins and NifA in other diazotrophs (10).

	ACKNOWLEDGEMENTS

We thank Augie Pioszak and Alex Ninfa for strains for overexpression of Ec PII and Phil Buckle (BIAcore AB) for advice on the SPR experiments. We also thank Gary Sawers and Mike Merrick for comments on the manuscript.

	FOOTNOTES

^* This work was supported by grants from the Biotechnology and Biological Sciences Research Council.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed. Tel.: 44-1603-450747; Fax: 44-1603-450778; E-mail: ray.dixon@bbsrc.ac.uk.

Published, JBC Papers in Press, February 20, 2002, DOI 10.1074/jbc.M112262200

	ABBREVIATIONS

The abbreviations used are: ATase, adenylyltransferase, product of glnE; Av GlnK, PII-like signal transduction protein, product of A. vinelandii glnK; Ec PII, PII signal transduction protein, product of E. coli glnB; IHF, integration host factor; ITC, isothermal titration calorimetry; NtrB, nitrogen regulator II or NRII, product of E. coli ntrB; SPR, surface plasmon resonance; UTase/UR, uridylyltransferase/uridylyl-removing enzyme, product of glnD.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES