Substrate Binding Pattern Research Articles

The structural and functional investigation into an unusual nitrile synthase

The biosynthesis of neurotoxin aetokthonotoxin (AETX) that features a unique structure of pentabrominated biindole nitrile involves a first-of-its-kind nitrile synthase termed AetD, an enzyme that shares very low sequence identity to known structures and catalyzes an unprecedented mechanism. In this study, we resolve the crystal structure of AetD in complex with the substrate 5,7-di-Br-L-Trp. AetD adopts the heme oxygenase like fold and forms a hydrophobic cavity within a helical bundle to accommodate the indole moiety. A diiron cluster comprising two irons that serves as a catalytic center binds to the carboxyl O and the amino N of the substrate. Notably, we demonstrate that the AetD-catalyzed reaction is independent of the bromination of the substrate and also solved crystal structures of AetD in complex with 5-Br-L-Trp and L-Trp. Altogether, the present study reveals the substrate-binding pattern and validates the diiron cluster-comprising active center of AetD, which should provide important basis to support the mechanistic investigations into this class of nitrile synthase.

Identification and structural analysis of a carbohydrate-binding module specific to alginate, a representative of a new family, CBM96

Carbohydrate-binding modules (CBMs) are the noncatalytic modules that assist functions of the catalytic modules in carbohydrate-active enzymes, and they are usually discrete structural domains in larger multimodular enzymes. CBMs often occur in tandem in different alginate lyases belonging to the CBM families 13, 16, and 32. However, none of the currently known CBMs in alginate lyases specifically bind to an internal alginate chain. In our investigation of the multidomain alginate lyase Dp0100 carrying several ancillary domains, we identified an alginate-binding domain denoted TM6-N4 using protein truncation analysis. The structure of this CBM domain was determined at 1.35Å resolution. TM6-N4 exhibited an overall β-sandwich fold architecture with two antiparallel β-sheets. We identified an extended binding groove in the CBM using site-directed mutagenesis, docking, and surface electrostatic potential analysis. Affinity analysis revealed that residues of Lys10, Lys22, Lys25, Lys27, Lys31, Arg36, and Tyr159 located on the bottom or the wall of the shallow groove are responsible for alginate binding, and isothermal titration calorimetry analyses indicated that the binding cleft consists of six subsites for sugar recognition. This substrate binding pattern is typical for type B CBM, and it represents the first CBM domain that specifically binds internal alginate chain. Phylogenetic analysis supports that TM6-N4 constitutes the founding member of a new CBM family denoted as CBM96. Our reported structure not only facilitates the investigation of the CBM-alginate ligand recognition mechanism but also inspires the utilization of the CBM domain in biotechnical applications.

Structure and Function of Piezophilic Hyperthermophilic Pyrococcus yayanosii pApase.

3’-Phosphoadenosine 5’-monophosphate (pAp) is a byproduct of sulfate assimilation and coenzyme A metabolism. pAp can inhibit the activity of 3′-phosphoadenosine 5′-phosphosulfate (PAPS) reductase and sulfotransferase and regulate gene expression under stress conditions by inhibiting XRN family of exoribonucleases. In metazoans, plants, yeast, and some bacteria, pAp can be converted into 5’-adenosine monophosphate (AMP) and inorganic phosphate by CysQ. In some bacteria and archaea, nanoRNases (Nrn) from the Asp-His-His (DHH) phosphoesterase superfamily are responsible for recycling pAp. In addition, histidinol phosphatase from the amidohydrolase superfamily can hydrolyze pAp. The bacterial enzymes for pAp turnover and their catalysis mechanism have been well studied, but these processes remain unclear in archaea. Pyrococcus yayanosii, an obligate piezophilic hyperthermophilic archaea, encodes a DHH family pApase homolog (PyapApase). Biochemical characterization showed that PyapApase can efficiently convert pAp into AMP and phosphate. The resolved crystal structure of apo-PyapApase is similar to that of bacterial nanoRNaseA (NrnA), but they are slightly different in the α-helix linker connecting the DHH and Asp-His-His associated 1 (DHHA1) domains. The longer α-helix of PyapApase leads to a narrower substrate-binding cleft between the DHH and DHHA1 domains than what is observed in bacterial NrnA. Through mutation analysis of conserved amino acid residues involved in coordinating metal ion and binding substrate pAp, it was confirmed that PyapApase has an ion coordination pattern similar to that of NrnA and slightly different substrate binding patterns. The results provide combined structural and functional insight into the enzymatic turnover of pAp, implying the potential function of sulfate assimilation in hyperthermophilic cells.

An affinity prediction approach for the ligand of E3 ligase Cbl-b and an insight into substrate binding pattern

Protein-protein interactions (PPIs) are essentially fundamental to all cellular processes, so that developing small molecule inhibitors of PPIs have great significance despite representing a huge challenge. Studying PPIs with the help of peptide motifs could obtain the structural information and reference significance to reduce the difficulty in the development of small molecules. Computational methods are powerful tools to characterize peptide-protein interactions, especially molecular dynamics simulation and binding free energy calculation. Here, we established an affinity prediction model suitable for Casitas B lymphoma-b (Cbl-b) and phosphorylated motif system. According to the affinity data set of multiple truncated peptides, the force field, solvent model, and internal dielectric constant of molecular mechanics/generalized Born surface area (MM/GBSA) method were optimized. Further, we predicted the affinity of the rationally designed new sequences through this model and obtained a new 6-mer motif with a 7-fold increase in affinity and the comprehensive structure–activity relationship. Moreover, we proposed an insight of unexpected activity of the truncated 5-mer peptide and revealed the possible binding mode of the new highly active 6-mer motif by extended simulation. Our results showed that the activity enhancement of the truncated peptide was caused by the acetyl-mediated conformation change. The side chain of Arg and pTyr in the 6-mer motif co-occupied the site p1 to form numerous hydrogen bond interactions and increased hydrophobic interaction formed with Tyr266, leading to the higher affinity. The present work provided a reference to investigate the PPI of Cbl-b and phosphorylated substrates and guided the development of Cbl-b inhibitors.

A Comparative Study to Determine Substrate-Binding Pattern in Eicosanoid-Generating Enzymes by Computational Mapping

A Comparative Study to Determine Substrate-Binding Pattern in Eicosanoid-Generating Enzymes by Computational Mapping

Negative net charge of EF-hand loop I can affect both calcium sensitivity and substrate binding pattern in mnemiopsin 2

Mnemiopsin 2 from Mnemiopsis leidy has three Ca2+-binding motifs and has luminescence properties in the presence of calcium and coelenterazine. It has been reported that the pattern of calcium binding among EF-hand loops of various photoproteins is different. Here, we designed and constructed two variants of mnemiopsin 2 (E50G/D47N and E50G/E53T mutants) with modified EF-hand I in which the negative charge in the first loop was reduced. According to the activity measurements, the initial intensity of mutants decreases; while the decay rate increases in E50G/D47N. We concluded that the presence of negative charge at positions 47 and 53 of mnemiopsin 2 is critical for both calcium coordination and the interaction of the substrate with the core structure of mnemiopsin 2. Structural studies accompanied by equilibrium denaturation experiments were also performed and it was found that negative charges at the aforementioned positions also have structural consequences which can affect the conformational stability of photoproteins.

Comparative analyses and structural insights of the novel cytochrome P450 fusion protein family CYP5619 in Oomycetes

Phylogenetic and structural analysis of P450 proteins fused to peroxidase/dioxygenase has not been reported yet. We present phylogenetic and in silico structural analysis of the novel P450 fusion family CYP5619 from the deadliest fish pathogenic oomycete, Saprolegnia diclina. Data-mining and annotation of CYP5619 members revealed their unique presence in oomycetes. CYP5619 members have the highest number of conserved amino acids among eukaryotic P450s. The highest number of conserved amino acids (78%) occurred in the peroxidase/dioxygenase domain compared to the P450 domain (22%). In silico structural analysis using a high-quality CYP5619A1 model revealed that CYP5619A1 has characteristic P450 structural motifs including EXXR and CXG. However, the heme-binding domain (CXG) in CYP5619 members was found to be highly degenerated. The in silico substrate binding pattern revealed that CYP5619A1 have a high affinity to medium chain fatty acids. Interestingly, the controlling agent of S. diclina malachite green was predicted to have the highest binding affinity, along with linoleic acid. However, unlike fatty acids, none of the active site amino acids formed hydrogen bonds with malachite green. The study’s results will pave the way for assessing CYP5619A1’s role in S. diclina physiology, including the nature of malachite green binding.

Lactoperoxidase, an Antimicrobial Milk Protein, as a Potential Activator of Carcinogenic Heterocyclic Amines in Breast Cancer.

Lactoperoxidase (LPO) is an antimicrobial protein secreted from mammary, salivary and other mucosal glands. It is an important member of heme peroxidase enzymes and the primary peroxidase enzyme present in breast tissues. In addition to the antimicrobial properties, LPO has been shown to be associated with breast cancer etiology. Heterocyclic amines, an important class of environmental and dietary carcinogens, have been increasingly associated with breast cancer etiology. Heterocyclic amines undergo activation in breast tissue as a result of oxidation by LPO. The current study includes three important heterocyclic amines, 2-amino-3-methylimidazo[4,5-f]quinoline (IQ), 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx) and 2-amino-1-methy-6-phenylimidazo[4,5-b]-pyridine (PhIP), that have carcinogenic activity. The structural binding characterization of IQ, MeIQx and PhIP with LPO was done using in silico approaches. Their binding pattern and interactions with LPO amino acid residues were analyzed. The three compounds bound in the distal heme cavity of LPO without replacing the important water molecule required for oxidation of substrate compounds. PhIP displayed lesser binding affinity for LPO in comparison to IQ and MeIQx. The binding mode of heterocyclic amines in distal heme cavity of LPO resembled to that of substrate binding pattern. The three heterocyclic amines are suggested to act as LPO substrate. The undisturbed water molecule present in distal heme cavity of the LPO is expected to facilitate the oxidation and activation of the three heterocyclic amines. These activated compounds may potentially bind with DNA in breast tissues forming DNA adducts and may subsequently lead to breast cancer initiation.

In silico identification of novel antagonists and binding insights by structural and functional analyses of guanylate kinase of Leishmania donovani and interaction with inhibitors

Guanylate kinase (GK; EC 2.7.4.8) is an important enzyme which catalyzed the ATP-dependent phosphorylation of GMP (guanosine monophosphate) into GDP (guanosine diphosphate) for survival of parasites; this suggests that it is a good target for design inhibitor(s). Knowledge of 3D structure of GKLdv is essential for elucidating interaction studies with the inhibitors and compared with substrate binding pattern in both the built model and template structures. The final reliable model of GKLdv has revealed α/β type protein with a G-X2-G-X-G-K that has a catalytic conserved domain in both template and model structure. The core domain consists of a central five-stranded parallel β-sheet (β5, β6, β3 and β4) whereas model has both α1-α10 and β1-β9 and active sites consist of flexible loop (FL1 to FL3). The superimposition structure of main-chain atoms of the GK model with closely homologue of GK (1EX7, chain A), has revealed better alignment and its root mean square deviation (RMSD) (43% similarity) is 0.602Ǻ. The virtual screening result has suggested that the binding pocket interacts with 5GP (5′-guanylic acid) with exhibited higher score in both GOLD and LigandFit scores and involve in amino acid interaction (H-bonding) are Arg40, Try80, Tyr52, Ser36, Glu71, and Asp102. The reported antileishmanial diospyrin (IC₅₀-7.2μM) has also exhibited higher binding activities (Dock score - 49.314 and Fitness score - 52.69) with frequently involved amino acids are Arg45, Glu102, Tyr35, Tyr54, Ser38, Tyr82, Ile103, Asp104, Lys106, Lys18, Pro13 and Ser14. The sigma pi interaction between Ile103 with diospyrin ring has revealed greater stable complex after molecular dynamic simulation studies. The best scoring compounds need for further validation in-vitro test.

Crystal structure of Pelagibacterium halotolerans PE8: New insight into its substrate-binding pattern

Lysophospholipase_carboxylesterase (LPCE) has highly conserved homologs in many diverse species ranging from bacteria to humans, as well as substantial biological significance and potential therapeutic implications. However, its biological function and catalytic mechanism remain minimally investigated because of the lack of structural information. Here, we report the crystal structure of a bacterial esterase PE8 belonging to the LPCE family. The crystal structure of PE8 was solved with a high resolution of 1.66 Å. Compared with other homologs in the family, significant differences were observed in the amino acid sequence, three-dimensional structure, and substrate-binding pattern. Residue Arg79 undergoes configuration switching when binding to the substrate and forms a unique wall, leading to a relatively closed cavity in the substrate-binding pocket compared with the relatively more open and longer clefts in other homologs. Moreover, the mutant Met122Ala showed much stronger substrate affinity and higher catalytic efficiency because less steric repulsion acted on the substrates. Taken together, these results showed that, in PE8, Arg79 and Met122 play important roles in substrate binding and the binding pocket shaping, respectively. Our study provides new insight into the catalytic mechanism of LPCE, which may facilitate the development of structure-based therapeutics and other biocatalytic applications.

QM/MM investigation of the catalytic mechanism of angiotensin-converting enzyme.

Angiotensin-converting enzyme (ACE) converts angiotensin I to angiotensin II and degrades bradykinin and other vasoactive peptides. ACE inhibitors are used to treat diseases such as hypertension and heart failure. It is thus highly desirable to understand the catalytic mechanism of ACE, as this should facilitate the design of more powerful and selective ACE inhibitors. ACE exhibits two different active domains, the C-domain and the N-domain. In this work, we systematically investigated the inhibitor- and substrate-binding patterns in the N-domain of human ACE using a combined quantum mechanical and molecular mechanical approach. The hydrolysis of hippuryl-histidyl-leucine (HHL) as catalyzed by the N-domain of human somatic ACE was explored, and the effects of chloride ion on the overall reaction were also investigated. Two models, one with and one without a chloride ion at the first binding position, were then designed to examine the chloride dependence of inhibitor-substrate binding and the catalytic mechanism. Our calculations indicate that the hydrolysis reaction follows a stepwise general base/general acid catalysis path. The estimated mean free energy barrier height in the two models is about 15.6kcal/mol, which agrees very well with the experimentally estimated value of 15.8kcal/mol. Our simulations thus suggest that the N-domain is in a mixed form during ACE-catalyzed hydrolysis, with the single-chloride-ion and the double-chloride-ion forms existing simultaneously. Graphical Abstract Superposition of ACE C- and N- domains.

Insights into the structure–function relationship of disease resistance protein HCTR in maize (Zea mays L.): A computational structural biology approach

The disease resistance gene Hm1 of maize encodes a NADPH-dependent reductase enzyme, HC-toxin reductase (HCTR) that detoxifies the HC toxin secreted by the race specific fungus Cochliobolus carbonum race 1. HCTR enzyme shares 29.6% sequence identity with dihydroflavonol reductase (DFR) of grape, a key enzyme involved in flavonoid biosynthesis. Here we report the comparative modelling, molecular dynamics simulation and docking studies to explain the structure–function relationship and the mode of cofactor (NADPH) binding in HCTR enzyme at the molecular level. The nucleotide binding domain of modelled HCTR adopts a classic Rossmann fold and possesses a consensus glycine rich GxGxxG motif. Molecular simulation studies suggested that HCTR model retained stability throughout the simulation in aqueous solution. HCTR model showed considerable structural identities with the cofactor binding site of DFR, but significant difference in the catalytic site might be the reason of functional divergence between these families of proteins. Similarly electrostatic surface potential analysis of both HCTR and DFR revealed profound variations in the charge distribution over the substrate binding site, which can be correlated with the sequence variability and may suggest distinct substrate-binding patterns and differences in the catalytic mechanism. Docking results indicated Phe19, Gly21, Arg40, Thr90, Gly208, Arg218, Glu221 and Thr222 are important residues for cofactor (NADPH) binding through strong hydrogen bonding and electrostatic interactions. Alanine scanning and analysis of docking energies of mutant proteins suggested that Phe19, and Arg40 are two critical residues for the cofactor binding. The result from the present study is expected to pave the way for exploration of similar genes in other economically important crop varieties.

Influence of substituent groups in regioselective acylation of nucleosides by Novozym 435 lipase

A comparative study on the substrate recognition was conducted successfully in Novozym 435-mediated acylation of various 2′- or 5-substituted nucleosides with acyl donors carrying different aliphatic chain lengths (C6, C10, and C14). The unexpected results revealed that the physicochemical property of the substituents (such as the size, hydrophobicity, and substitutional position) in nucleosides profoundly influenced the behavior of the enzyme. The different substrate-binding patterns derived from the existence of the substituents in 2′- or 5-position of the nucleosides could account for this. Moreover, another possible factor governing the regioselectivity might be ascribed to the interaction between the substituent and acyl donor in addition to the geometrical configuration of the lipase's active site.

Two high-resolution structures of potato endo-1,3-β-glucanase reveal subdomain flexibility with implications for substrate binding

Endo-1,3-β-glucanases are widely distributed among bacteria, fungi and higher plants. They are responsible for hydrolysis of the glycosidic bond in specific polysaccharides with tracts of unsubstituted β-1,3-linked glucosyl residues. The plant enzymes belong to glycoside hydrolase family 17 (GH17) and are also members of class 2 of pathogenesis-related (PR) proteins. X-ray diffraction data were collected to 1.40 and 1.26 Å resolution from two crystals of endo-1,3-β-glucanase from Solanum tuberosum (potato, cultivar Désirée) which, despite having a similar packing framework, represented two separate crystal forms. In particular, they differed in the Matthews coefficient and are consequently referred to as higher density (HD; 1.40 Å resolution) and lower density (LD; 1.26 Å resolution) forms. The general fold of the protein resembles that of other known plant endo-1,3-β-glucanases and is defined by a (β/α)(8)-barrel with an additional subdomain built around the C-terminal half of the barrel. The structures revealed high flexibility of the subdomain, which forms part of the catalytic cleft. Comparison with structures of other GH17 endo-1,3-β-glucanases revealed differences in the arrangement of the secondary-structure elements in this region, which can be correlated with sequence variability and may suggest distinct substrate-binding patterns. The crystal structures revealed an unusual packing mode, clearly visible in the LD structure, caused by the presence of the C-terminal His(6) tag, which extends from the compact fold of the enzyme molecule and docks in the catalytic cleft of a neighbouring molecule. In this way, an infinite chain of His-tag-linked protein molecules is formed along the c direction.

Diversity of Bisubstrate Binding Modes of Adenosine Analogue–Oligoarginine Conjugates in Protein Kinase A and Implications for Protein Substrate Interactions

Crystal structures of the catalytic subunit α of cAMP-dependent protein kinase (PKAc) with three adenosine analogue–oligoarginine conjugates (ARCs) are presented. The rationally designed ARCs include moieties that, in combination, target both the ATP- and the peptide-substrate-binding sites of PKAc, thereby taking advantage of high-affinity binding interactions offered by the ATP site while utilizing an additional mechanism for target specificity via binding to the peptide substrate site. The crystal structuresdemonstrate that, in accord with the previously reported bisubstrate character of ARCs, the inhibitors occupy both binding sites of PKAc. Further, they show new binding modes that may also apply to natural protein substrates of PKAc, which have not been revealed by previous crystallographic studies. The crystal structures described here contribute to the understanding of the substrate-binding patterns of PKAc and should also facilitate the design of inhibitors targeting PKAc and related protein kinases.

Structural and functional comparison of MIF ortholog from Plasmodium yoelii with MIF from its rodent host

Host-derived macrophage migration inhibitory factor (MIF) has been implicated in the pathogenesis of malaria infection, especially in malarial anemia. Although two Plasmodium parasite-derived MIF orthologs, Plasmodium falciparum MIF and P. berghei MIF were identified recently, the crystal structure and the precise roles of Plasmodium-derived MIFs, particularly in combination with the host MIF, remain unknown. In this study, we identified another MIF ortholog from a rodent-specific P. yoelii (PyMIF). This molecule shares a conserved three-dimensional structure with murine MIF (MmMIF), but with a different substrate binding pattern and much lower tautomerase activity. It could activate host cells via several signaling pathways in vitro, and inhibiting macrophage apoptosis, also similarly to MmMIF. However, we found that PyMIF and MmMIF acted synergistically to activate the MAPK-ERK1/2 signaling pathway at very low concentration but acted antagonistically at higher concentration. Furthermore, we detected PyMIF in the sera of infected mice and found that injection of recombinant PyMIF (rPyMIF) during infection could up-regulate several pro-inflammatory cytokines in vivo and slightly delay the death of infected mice. These data suggest that PyMIF modulates host immune responses together with host MIF and has potential to prolong parasitemia or the chronicity of malaria infection.

Crystal structure and molecular modeling study of N-carbamoylsarcosine amidase Ta0454 from Thermoplasma acidophilum.

A crystal structure of the putative N-carbamoylsarcosine amidase (CSHase) Ta0454 from Thermoplasma acidophilum was solved by single-wavelength anomalous diffraction and refined at a resolution of 2.35A. CSHases are involved in the degradation of creatinine. Ta0454 shares a similar fold and a highly conserved C-D-K catalytic triad (Cys123, Asp9, and Lys90) with the structures of three cysteine hydrolases (PDB codes 1NBA, 1IM5, and 2H0R). Molecular dynamics (MD) simulations of Ta0454/N-carbamoylsarcosine and Ta0454/pyrazinamide complexes were performed to determine the structural basis of the substrate binding pattern for each ligand. Based on the MD-simulated trajectories, the MM/PBSA method predicts binding free energies of -24.5 and -17.1 kcal/mol for the two systems, respectively. The predicted binding free energies suggest that Ta0454 is selective for N-carbamoylsarcosine over pyrazinamide, and zinc ions play an important role in the favorable substrate bound states.

Comparative study of the prereactive protein kinase A Michaelis complex with Kemptide substrate

In the present work we have modeled the Michaelis complex of the cyclic-Adenosine Monophosphate Dependent (cAMD) Protein Kinase A (PKA) with Mg(2)ATP and the heptapeptide substrate Kemptide by classical molecular dynamics. The chosen synthetic substrate is relevant for its high efficiency and small size, and it has not been used in previous theoretical studies. The structural analysis of the data generated along the 6 ns simulation indicates that the modeled substrate-enzyme complex mimics the substrate binding pattern known for PKA. The values of the average prereactive distances obtained from the simulation do not exclude any of the two limiting situations proposed as mechanisms in the literature for the phosphorylation reaction (dissociative and associative) because the system oscillates between configurations compatible with each of them. Furthermore, the results obtained for the average interaction distances between active site residues concord in suggesting the plausibility of an alternative third reaction mechanism.

Fucosyltransferases: structure/function studies.

Alpha3-fucosyltransferases (alpha3-FucTs) catalyze the final step in the synthesis of a range of important glycoconjugates that function in cell adhesion and lymphocyte recirculation. Six members of this family of enzymes have been cloned from the human genome, and their expression pattern has been shown to be highly regulated. Each enzyme has a unique acceptor substrate binding pattern, and each generates a unique range of fucosylated products. Results from a range of studies have provided information on amino acids in the FucT sequence that contribute to the differential acceptor specificity for the FucTs, and to the binding of the nucleotide sugar donor GDP-fucose. These results, in conjunction with results obtained from the analysis of the disulfide bond pattern, have provided useful clues about the spatial distribution of amino acids that influence or directly contribute to substrate binding. This information is reviewed here, and a molecular fold prediction is presented which has been constructed based on the available information and current modeling methodology.

Purification and characterization of glutamine synthetase from Clostridium pasteurianum.

Glutamine synthetase from Clostridium pasteurianum grown on molasses as the sole carbon source and ammonium chloride as the nitrogen source has been purified to homogeneity (45-fold) with 32% recovery. The procedure involves ammonium sulfate precipitation and chromatography on a combined Sepharose 4B/DEAE-Sephadex A-50 column. The purified enzyme being very unstable was stabilized by the addition of 25% (v/v) glycerol. The enzyme has an unusually high molecular weight of 1 X 10(6) and 20 subunits of Mr 50 000 each, as determined by gel filtration and sodium dodecyl sulfate gel electrophoresis, respectively. It has an absorption maximum at 280 nm and a fluorescence emission maximum at 380 nm when excited at 280 nm. Its substrate binding pattern as studied by fluorescence quenching studies is different from that of the Escherichia coli enzyme. Both the gamma-glutamyltransferase and synthetase activities reside in the same protein as the ratio of the two activities at each step of purification remains constant and the enzyme exhibits optimal transferase and synthetase activities at the same pH (7.2) and temperature (50 degrees C). The thermal stabilities of both activities were also similar, and decay of both the activities at 50 degrees C ran parallel. The enzyme shows stabilization by substrates, as L-glutamate, Mg2+, and ATP + Mg2+ protected both the synthetase and gamma-glutamyltransferase activities against thermal inactivation. Storage in 25% (v/v) glycerol enhanced the thermal stability of glutamine synthetase. Metal ion requirement and substrate specificity of the enzyme have been examined. Maximum synthetase activity occurs when [Mg2+]: [ATP] = 2. The Km app values are as follows (in parentheses): ATP (0.34 mM), NH2OH (0.4 mM in the synthetase reaction and 4.1 mM in the transferase reaction), glutamine (14.7 mM), ADP (3.8 X 10(-4) mM), arsenate (2.5 mM), and L-glutamate (3.4 mM, 22.2 mM). The enzyme exhibits negative cooperativity in the binding of glutamate. Amino acids such as L-serine, glycine, L-alanine, and L-aspartic acid inhibit the enzyme.