Structure-function analysis of glutamine synthetase adenylyltransferase (ATase, EC 2.7.7.49) of Escherichia coli

Glutamine synthetase adenylyltransferase (ATase) regulates the activity of glutamine synthetase by adenylylation and deadenylylation in response to signals of nitrogen and carbon status: glutamine, alpha-ketoglutarate, and the uridylylated and unmodified forms of the PII signal transduction protein. ATase consists of two conserved nucleotidyltransferase (NT) domains linked by a central region of approximately 200 amino acids. Here, we study the activities and regulation of mutated and truncated forms of ATase. Our results indicate the following. (i) The N-terminal NT domain contained the adenylyl-removing (AR) active site, and the C-terminal NT domain contained the adenylyltransferase (AT) active site. (ii) The enzyme contained a glutamine binding site, and glutamine increased the affinity for PII. (iii) The enzyme appeared to contain multiple sites for the binding of PII and PII-UMP. (iv) Truncated versions of ATase missing the C-terminal (NT) domain lacked both AT and AR activity, suggesting a role for the C-terminal NT domain in both activities. (v) The purified C-terminal NT domain and larger polypeptides containing this domain had significant basal AT activity, which was stimulated by glutamine. These polypeptides were indifferent to PII and PII-UMP, or their ATase activity was inhibited by either PII or PII-UMP. (vi) Certain point mutations in the central region or an internal deletion removing most of this part of the protein eliminated the AR activity and eliminated activation of the AT activity by PII, while not eliminating the binding of PII or PII-UMP. That is, these mutations in the central region appeared to destroy the communication between the PII and PII-UMP binding sites and the AT and AR active sites. (vii) Certain mutations in the central region of ATase appeared to dramatically improve the binding of glutamine to the enzyme. (viii) While the isolated AT and AR domains of ATase bound poorly to PII and PII-UMP, these domains bound PII and PII-UMP significantly better when linked to the central region of ATase. Together, our results indicate a highly coordinated enzyme, in which the AT and AR domains participate in each other's regulation and distant regulatory sites are in communication with each other. A model for the regulation of ATase by glutamine, PII, and PII-UMP consistent with all data is presented.     

The domains carrying the opposing activities in adenylyltransferase are separated by a central regulatory domain

Adenylyltransferase is a bifunctional enzyme that controls the enzymatic activity of dodecameric glutamine synthetase in Escherichia coli by reversible adenylylation and deadenylylation. Previous studies showed that the two similar but chemically distinct reactions are carried out by separate domains within adenylyltransferase. The N-terminal domain carries the deadenylylation activity, and the C-terminal domain carries the adenylylation activity [Jaggi R, van Heeswijk WC, Westerhoff HV, Ollis DL & Vasudevan SG (1997) EMBO J16, 5562-5571]. In this study, we further map the domain junctions of adenylyltransferase on the basis of solubility and enzymatic analysis of truncation constructs, and show for the first time that adenylyltransferase has three domains: the two activity domains and a central, probably regulatory (R), domain connected by interdomain Q-linkers (N-Q1-R-Q2-C). The various constructs, which have the opposing domain and or central domain removed, all retain their activity in the absence of their respective nitrogen status indicator, i.e. PII or PII-UMP. A panel of mAbs to adenylyltransferase was used to demonstrate that the cellular nitrogen status indicators, PII and PII-UMP, probably bind in the central regulatory domain to stimulate the adenylylation and deadenylylation reactions, respectively. In the light of these results, intramolecular signaling within adenylyltransferase is discussed.     

Escherichia coli glutamine synthetase adenylyltransferase (ATase, EC 2.7.7.49): kinetic characterization of regulation by PII, PII-UMP, glutamine, and alpha-ketoglutarate

Glutamine synthetase adenylyltranferase (ATase, EC 2.7.7.49) catalyzes the adenylylation and deadenylylation of glutamine synthetase (GS), regulating GS activity. The adenylyltransferase (AT) reaction is activated by glutamine and by the unmodified form of the PII signal transduction protein and is inhibited by the uridylylated form of PII, PII-UMP. Conversely, the adenylyl-removing (AR) reaction is activated by PII-UMP and is inhibited by glutamine and by PII. Both AT and AR reactions are regulated by alpha-ketoglutarate, which binds to PII and PII-UMP. Here, we present a kinetic analysis of the AT and AR activities and their regulation. Both AT and AR reactions used a sequential mechanism of rapid equilibrium random binding of substrates and products. Activators and inhibitors had little effect on the binding of substrates, instead exerting their effects on catalysis. Our results were consistent with PII, PII-UMP, and glutamine shifting the enzyme among at least six different enzyme forms, two of which were inactive, one of which exhibited AR activity, and three of which exhibited AT activity. In addition to a site for glutamine, the enzyme appeared to contain two distinct sites for PII and PII-UMP. The PII, PII-UMP, and glutamine sites were in communication so that the apparent activation and inhibition constants for regulators depended upon each other. The binding of PII was favored by glutamine and its level reduced by PII-UMP, whereas glutamine and PII-UMP competed for the enzyme. alpha-Ketoglutarate, which acts exclusively through its binding to PII and PII-UMP, did not alter the binding of PII or PII-UMP to the enzyme. Rather, alpha-ketoglutarate dramatically affected the extent of activation or inhibition of the enzyme by PII or PII-UMP. A working hypothesis for the regulation of the AT and AR activities, consistent with all data, is presented.     

Expression, purification, crystallization, and preliminary X-ray analysis of the N-terminal domain of Escherichia coli adenylyl transferase

A soluble N-terminal domain of the Escherichia coli adenylyl transferase (ATase) is responsible for deadenylylation activity of the intact enzyme. Previous studies of the deadenylylation activity have involved a fragment, AT-N423 (residues 1 to 423), which was extended by 17 amino acids to give AT-N440. This new domain is truncated at the end of a predicted helix and prior to a Q-linker. The domain was found to be very soluble and stable so that it could be purified to homogeneity and crystallized. This construct has deadenylylation activity that is independent of the low nitrogen status indicator PII-UMP. The crystals belong to space group P3(1)21 or its enantiomorph P3(2)21 with a=b=116.6 A and c=67.6 A.     

Structure of the N-terminal domain of Escherichia coli glutamine synthetase adenylyltransferase

We report the crystal structure of the N-terminal domain of Escherichia coli adenylyltransferase that catalyzes the reversible nucleotidylation of glutamine synthetase (GS), a key enzyme in nitrogen assimilation. This domain (AT-N440) catalyzes the deadenylylation and subsequent activation of GS. The structure has been divided into three subdomains, two of which bear some similarity to kanamycin nucleotidyltransferase (KNT). However, the orientation of the two domains in AT-N440 differs from that in KNT. The active site of AT-N440 has been identified on the basis of structural comparisons with KNT, DNA polymerase beta, and polyadenylate polymerase. AT-N440 has a cluster of metal binding residues that are conserved in polbeta-like nucleotidyl transferases. The location of residues conserved in all ATase sequences was found to cluster around the active site. Many of these residues are very likely to play a role in catalysis, substrate binding, or effector binding.     

A source of ultrasensitivity in the glutamine response of the bicyclic cascade system controlling glutamine synthetase adenylylation state and activity in Escherichia coli

Glutamine synthetase (GS) activity in Escherichia coli is regulated by reversible adenylylation, brought about by a bicyclic system comprised of uridylyltransferase/uridylyl-removing enzyme (UTase/UR), its substrate, PII, adenylyltransferase (ATase), and its substrate, GS. The modified and unmodified forms of PII produced by the upstream UTase/UR-PII cycle regulate the downstream ATase-GS cycle. A reconstituted UTase/UR-PII-ATase-GS bicyclic system has been shown to produce a highly ultrasensitive response of GS adenylylation state to the glutamine concentration, but its composite UTase/UR-PII and ATase-GS cycles displayed moderate glutamine sensitivities when examined separately. Glutamine sensitivity of the bicyclic system was significantly reduced when the trimeric PII protein was replaced by a heterotrimeric form of PII that was functionally monomeric, and coupling between the two cycles was different in systems containing wild-type or heterotrimeric PII. Thus, the trimeric nature of PII played a role in the glutamine response of the bicyclic system. We therefore examined regulation of the individual AT (adenylylation) and AR (deadenylylation) activities of ATase by PII preparations with various levels of uridylylation. AR activity was affected in a linear fashion by PII uridylylation, but partially modified wild-type PII activated the AT much less than expected based on the extent of PII modification. Partially modified wild-type PII also bound to ATase less than expected based upon the fraction of modified subunits. Our results suggest that the AT activity is only bound and activated by completely unmodified PII and that this design is largely responsible for ultrasensitivity of the bicyclic system.     

Adenylylation and catalytic properties of Mycobacterium tuberculosis glutamine synthetase expressed in Escherichia coli versus mycobacteria

Bacterial glutamine synthetases (GSs) are complex dodecameric oligomers that play a critical role in nitrogen metabolism, converting ammonia and glutamate to glutamine. Recently published reports suggest that GS from Mycobacterium tuberculosis (MTb) may be a therapeutic target (Harth, G., and Horwitz, M. A. (2003) Infect. Immun. 71, 456-464). In some bacteria, GS is regulated via adenylylation of some or all of the subunits within the aggregate; catalytic activity is inversely proportional to the extent of adenylylation. The adenylylation and deadenylylation of GS are catalyzed by adenylyl transferase (ATase). Here, we demonstrate via electrospray ionization mass spectrometry that GS from pathogenic M. tuberculosis is adenylylated by the Escherichia coli ATase. The adenylyl group can be hydrolyzed by snake venom phosphodiesterase to afford the unmodified enzyme. The site of adenylylation of MTb GS by the E. coli ATase is Tyr-406, as indicated by the lack of adenylylation of the Y406F mutant, and, as expected, is based on amino acid sequence alignments. Using electrospray ionization mass spectroscopy methodology, we found that GS is not adenylylated when obtained directly from MTb cultures that are not supplemented with glutamine. Under these conditions, the highly related but non-pathogenic Mycobacterium bovis BCG yields partially ( approximately 25%) adenylylated enzyme. Upon the addition of glutamine to the cultures, the MTb GS becomes significantly adenylylated ( approximately 30%), whereas the adenylylation of M. bovis BCG GS does not change. Collectively, the results demonstrate that MTb GS is a substrate for E. coli ATase, but only low adenylylation states are accessible. This parallels the low adenylylation states observed for GS from mycobacteria and suggests the intriguing possibility that adenylylation in the pathogenic versus non-pathogenic mycobacteria is differentially regulated.     

Cascade control of E. coli glutamine synthetase. II. Metabolite regulation of the enzymes in the cascade.

Enzymes and regulatory proteins involved in the cascade control of glutamine synthetase activity of Escherichia coli have been separated from one another and the effects of numerous metabolites on each step in the cascade have been determined. The adenylyl transferase (ATase) -catalyzed adenylylation of glutamine synthetase, which requires the presence of the unmodified form of the regulatory protein PII is enhanced by glutamine and is inhibited by either α-ketoglutarate (α-KG) or the uridylylated form (PII·UMP) of the regulatory protein. PII·UMP and α-KG act synergistically to inhibit this activity. In contrast, the PII·UMP-dependent, ATase-catalyzed deadenylylation of glutamine synthetase requires α-KG and ATP and is inhibited by glutamine or PII and synergistically by glutamine plus PII. The capacity of uridylyl transferase (UTase) to catalyze the uridylylation of PII is dependent on the presence of α-KG and ATP and is inhibited by glutamine. The deuridylylation of PII·UMP by the uridylyl removing enzyme (UR) is enhanced by glutamine but is unaffected by α-KG. However, CMP, UMP, and CoA all inhibit activity at 10^?6m. High concentrations of ATase inhibit both UR and UTase activities, presumably by binding the regulatory protein. Of more than 50 substances that alter the activity of at least one enzyme in the cascade, only α-KG and glutamine affect the activity at every step. This accounts for the observation that glutamine synthetase activity in vivo is very sensitive to the intracellular ratio of α-KG to glutamine.     

Glutamine synthetase adenylyltransferase from Escherichia coli: purification and physical and chemical properties.

The glutamine synthetase adenylyltransferase (EC 2.7.7.42), WHIch catalyzes the adenylylation and deadenylylation of glutamine synthetase in E. coli, has been stabilized and purified 2200-fold to apparent homogeneity. Sedimentation and electrophoresis studies show that the native enzyme is a single polypeptide chain of 115,000 +/- 5000 molecular weight with an isoelectric pH (PL) OF 4.98, a sedimentation coefficient (S20.w0) of 5.6S, and a molar frictional coefficient (f/f0) of 1.52. An alpha-helical content of approximately equal to 25% and approximately equal to 28% beta-pleated sheet and approximately equal to 47% random coil structures were estimated from circular dichroism measurements. The amino acid composition of the protein has been determined. The intrinsic tryptophanyl residue flourescence of adenylyltransferase is two fold greater than that of L-tryptophan; this property has been used to monitor ligand-induced conformational changes in the enzyme. Activators of the adenylylation reaction (ATP, L-glutamine, or the E. coli PII regulatory protein) produced an enhancement of fluorescence; alpha-ketoglutarate, an inhibitor of adenylylation and an activator of deadenulylation, caused a net decrease in fluorescence. The adenylytransferase has separate interaction sites for L-glutamine and the regulatory PII protein.     

The GlnD and GlnK homologues of Streptomyces coelicolor A3(2) are functionally dissimilar to their nitrogen regulatory system counterparts from enteric bacteria

Glutamine synthetase I (GSI) enzyme activity in Streptomyces coelicolor is controlled post-translationally by the adenylyltransferase (GlnE) as in enteric bacteria. Although other homologues of the Escherichia coli Ntr system (glnK, coding for a PII family protein; and glnD, coding for an uridylyltransferase) are found in the S. coelicolor genome, the regulation of the GSI activity was found to be different. The functions of glnK and glnD were analysed by specific mutants. Surprisingly, biochemical assay and two-dimensional PAGE analysis showed that modification of GSI by GlnE occurs normally in all mutant strains, and neither GlnK nor GlnD are required for the regulation of GlnE in response to nitrogen stimuli. Analysis of the post-translational regulation of GlnK in vivo by two-dimensional PAGE and mass spectrometry indicated that it is subject to both a reversible and a non-reversible modification in a direct response to nitrogen availability. The irreversible modification was identified as removal of the first three N-terminal amino acid residues of the protein, and the reversible modification as adenylylation of the conserved tyro-sine 51 residue that is known to be uridylylated in E. coli. The glnD insertion mutant expressing only the N-terminal half of GlnD was capable of adenylylating GlnK, but was unable to perform the reverse deadenylylation reaction in response to excess ammonium. The glnD null mutant completely lacked the ability to adenylylate GlnK. This work provides the first example of a PII protein that is modified by adenylylation, and demonstrates that this reaction is performed by a homologue of GlnD, previously described only as a uridylyltransferase enzyme.     

New methods for the colorimetric assay of PIID regulatory protein, uridylyltransferase, and uridylyl-removing enzyme in glutamine synthetase cascade

Adenylylation and deadenylylation of glutamine synthetase (GS) are catalyzed by the same adenylyltransferase (ATase). The ability of ATase to catalyze adenylylation is markedly stimulated by the unmodified form of a regulatory protein, P_IIA, whereas its capacity to catalyze deadenylylation is stimulated by the uridylylated form (P_IID) of the regulatory protein. Interconversion between P_IIA and P_IID is catalyzed by uridylyltransferase (UTase) and uridylylremoving enzyme (UR). New colorimetric methods were developed for the assays of P_IID, UTase, and UR activities. The P_IID activity is monitored by its unique ability to stimulate the ATase catalyzed formation of unadenylylated subunits from adenylylated GS. The inerease of unadenylylated subunits is determined by measuring the γ-glutamyltransferase activity of GS under conditions where the activity of an unadenylylated subunit is about 15 times greater than that of an adenylylated subunit (i.e., at pH 8.0 in the presence of Mn²⁺). Assays for UTase and UR enzyme are derived by coupling the P_IID assay to the UTase and UR reactions. For the UTase reaction, the formation of P_IID from P_IIA is measured, whereas the decrease in P_IID is followed for the UR assay. These assays have been applied to follow the activities of these proteins during their purification procedures, to the mechanistic studies on the deadenylylation reaction, and to determine the activities of these proteins in mutants produced during the genetic study of glutamine synthetase cascade. The problems evolved from these assays are discussed.     

Escherichia coli PII signal transduction protein controlling nitrogen assimilation acts as a sensor of adenylate energy charge in vitro

PII signal transduction proteins are among the most widely distributed signaling proteins in nature, controlling nitrogen assimilation in organisms ranging from bacteria to higher plants. PII proteins integrate signals of cellular metabolic status and interact with and regulate receptors that are signal transduction enzymes or key metabolic enzymes. Prior work with Escherichia coli PII showed that all signal transduction functions of PII required ATP binding to PII and that ATP binding was synergistic with the binding of alpha-ketoglutarate to PII. Furthermore, alpha-ketoglutarate, a cellular signal of nitrogen and carbon status, was observed to strongly regulate PII functions. Here, we show that in reconstituted signal transduction systems, ADP had a dramatic effect on PII regulation of two E. coli PII receptors, ATase, and NRII (NtrB), and on PII uridylylation by the signal transducing UTase/UR. ADP acted antagonistically to alpha-ketoglutarate, that is, low adenylylate energy charge acted to diminish signaling of nitrogen limitation. By individually studying the interactions that occur in the reconstituted signal transduction systems, we observed that essentially all PII and PII-UMP interactions were influenced by ADP. Our experiments also suggest that under certain conditions, the three nucleotide binding sites of the PII trimer may be occupied by combinations of ATP and ADP. In the aggregate, our results show that PII proteins, in addition to serving as sensors of alpha-ketoglutarate, have the capacity to serve as direct sensors of the adenylylate energy charge.     

Purification and structural properties of adenylylated and deadenylylated glutamine synthetase from Rhodopseudomonas sphaeroides

Glutamine synthetase (GS) of Rhodopseudomonas sphaeroides is regulated by adenylylation and deadenylylation. The extent of adenylylation/deadenylylation of the enzyme in cell free extracts was influenced by inorganic phosphate (P _i), -ketoglutarate, ATP and other nucleotides. While P _i and -ketoglutarate stimulated deadenylylation, ATP and other nucleotides enhanced adenylylation of the GS. By using proper combinations of the effectors and incubation conditions, any desired adenylylation state of GS could be adjusted in vitro. The enzyme was purified to electrophoretic homogenity by three steps including affinity chromatography on 5-AMP-Sepharose. Adenylylated and deadenylylated enzyme showed different UV-spectra and isoelectric points. The native enzyme had a molecular weight of 600,000, deadenylylated subunits of 50,000±1,000. Electron microscopic investigations revealed a dodecameric arrangement of subunits in two hexameric planes.     

Uridylylation of the PII protein from Herbaspirillum seropedicae

The PII protein is apparently involved in the control of NifA activity in Herbaspirillum seropedicae. To evaluate the probable role of PII in signal transduction, uridylylation assays were conducted with purified H. seropedicae PII and Escherichia coli GlnD, or a cell-free extract of H. seropedicae as sources of uridylylating activity. The results showed that alpha-ketoglutarate and ATP stimulate uridylylation whereas glutamine inhibits uridylylation. Deuridylylation of PII-UMP was dependent on glutamine and inhibited by ATP and alpha-ketoglutarate. PII uridylylation and (or) deuridylylation in response to these effectors suggests that PII is a nitrogen level signal transducer in H. seropedicae.     

Reversible adenylylation of glutamine synthetase is dynamically counterbalanced during steady-state growth of Escherichia coli

Glutamine synthetase (GS) is the central enzyme for nitrogen assimilation in Escherichia coli and is subject to reversible adenylylation (inactivation) by a bifunctional GS adenylyltransferase/adenylyl-removing enzyme (ATase). In vitro, both of the opposing activities of ATase are regulated by small effectors, most notably glutamine and 2-oxoglutarate. In vivo, adenylyltransferase (AT) activity is critical for growth adaptation when cells are shifted from nitrogen-limiting to nitrogen-excess conditions and a rapid decrease of GS activity by adenylylation is needed. Here, we show that the adenylyl-removing (AR) activity of ATase is required to counterbalance its AT activity during steady-state growth under both nitrogen-excess and nitrogen-limiting conditions. This conclusion was established by studying AR⁻/AT⁺ mutants, which surprisingly displayed steady-state growth defects in nitrogen-excess conditions due to excessive GS adenylylation. Moreover, GS was abnormally adenylylated in the AR⁻ mutants even under nitrogen-limiting conditions, whereas there was little GS adenylylation in wild-type strains. Despite the importance of AR activity, we establish that AT activity is significantly regulated in vivo, mainly by the cellular glutamine concentration. There is good general agreement between quantitative estimates of AT regulation in vivo and results derived from previous in vitro studies except at very low AT activities. We propose additional mechanisms for the low AT activities in vivo. The results suggest that dynamic counterbalance by reversible covalent modification may be a general strategy for controlling the activity of enzymes such as GS, whose physiological output allows adaptation to environmental fluctuations.     

Probing interactions of the homotrimeric PII signal transduction protein with its receptors by use of PII heterotrimers formed in vitro from wild-type and mutant subunits.

The homotrimeric PII signal transduction protein of Escherichia coli interacts with two small-molecule effectors, 2-ketoglutarate and ATP, regulates two protein receptors, the kinase/phosphatase nitrogen regulator II (NRII) and the glutamine synthetase (GS) adenylyltransferase (ATase), and is subject to reversible uridylylation, catalyzed by the uridylyltransferase/uridylyl-removing enzyme (UTase/UR). The site of PII uridylylation, Y51, is located at the apex of the solvent-exposed T-loop (E. Cheah, P. D. Carr, P. M. Suffolk, S. G. Vasudevan, N. E. Dixon, and D. L. Ollis, Structure 2:981-990, 1994), and an internally truncated PII lacking residues 47 to 53 formed trimers that bound the small-molecule effectors but were unable to be uridylylated or activate NRII and ATase (P. Jiang, P. Zucker, M. R. Atkinson, E. S. Kamberov, W. Tirasophon, P. Chandran, B. R. Schefke, and A. J. Ninfa, J. Bacteriol. 179:4342-4353, 1997). We investigated the ability of heterotrimers containing delta47-53 and wild-type subunits to become uridylylated and activate NRII and ATase. Heterotrimers were formed by denaturation and renaturation of protein mixtures; when such mixtures contained a fivefold excess of A47-53 subunits, the wild-type subunits were mostly redistributed into trimers containing one wild-type subunit and two mutant subunits. The resulting population of trimers was uridylylated and deuridylylated by UTase/UR, stimulated the phosphatase activity of NRII, and stimulated adenylylation of GS by ATase. In all except the ATase interaction, the activity of the hybrid trimers was greater than expected based on the number of wild-type subunits present. These results indicate that a single T-loop region within a trimer is sufficient for the productive interaction of PII with its protein receptors. We also formed heterotrimers containing wild-type subunits and subunits containing the G89A alteration (P. Jiang, P. Zucker, M. R. Atkinson, E. S. Kamberov, W. Tirasophon, P. Chandran, B. R. Schefke, and A. J. Ninfa, J. Bacteriol. 179: 4342-4353, 1997). The G89A mutant form of PII does not bind the small-molecule effectors, does not interact with UTase or with NRII, and interacts poorly with ATase. Heterotrimers formed with a 10/1 starting ratio of G89A to wild-type subunits interacted with UTase/UR and ATase to a lesser extent than expected based on the number of wild-type subunits present but activated NRII slightly better than expected based on the number of wild-type subunits present. Thus, intersubunit interactions within the PII trimer can adversely affect the activity of wild-type subunits and may affect the interactions with the different receptors in a variable way. Finally, we formed heterotrimers containing delta47-53 and G89A mutant subunits. These heterotrimers were not uridylylated, did not interact with NRII, and interacted with the ATase only to the extent expected based on the number of G89A subunits present. Thus, the G89A subunits, which contain an intact T-loop region, were not "repaired" by inclusion in heterotrimers along with delta47-53 subunits.