Structure and Kinetics of a Monomeric Glucosamine 6-Phosphate Deaminase
2005; Elsevier BV; Volume: 280; Issue: 20 Linguagem: Inglês
10.1074/jbc.m502131200
ISSN1083-351X
AutoresFlorence Vincent, G.J. Davies, J.A. Brannigan,
Tópico(s)Biochemical and Molecular Research
ResumoGlucosamine 6-phosphate is converted to fructose 6-phosphate and ammonia by the action of the enzyme glucosamine 6-phosphate deaminase, NagB. This reaction is the final step in the specific GlcNAc utilization pathway and thus decides the metabolic fate of GlcNAc. Sequence analyses suggest that the NagB "superfamily" consists of three main clusters: multimeric and allosterically regulated glucosamine-6-phosphate deaminases (exemplified by Escherichia coli NagB), phosphogluconolactonases, and monomeric hexosamine-6-phosphate deaminases. Here we present the three-dimensional structure and kinetics of the first member of this latter group, the glucosamine-6-phosphate deaminase, NagB, from Bacillus subtilis. The structures were determined in ligand-complexed forms at resolutions around 1.4 Å. BsuNagB is monomeric in solution and as a consequence is active (kcat 28 s-1, Km(app) 0.13 mm) without the need for allosteric activators. A decrease in activity at high substrate concentrations may reflect substrate inhibition (with Ki of ∼4 mm). The structure completes the NagB superfamily structural landscape and thus allows further interrogation of genomic data in terms of the regulation of NagB and the metabolic fate(s) of glucosamine 6-phosphate. Glucosamine 6-phosphate is converted to fructose 6-phosphate and ammonia by the action of the enzyme glucosamine 6-phosphate deaminase, NagB. This reaction is the final step in the specific GlcNAc utilization pathway and thus decides the metabolic fate of GlcNAc. Sequence analyses suggest that the NagB "superfamily" consists of three main clusters: multimeric and allosterically regulated glucosamine-6-phosphate deaminases (exemplified by Escherichia coli NagB), phosphogluconolactonases, and monomeric hexosamine-6-phosphate deaminases. Here we present the three-dimensional structure and kinetics of the first member of this latter group, the glucosamine-6-phosphate deaminase, NagB, from Bacillus subtilis. The structures were determined in ligand-complexed forms at resolutions around 1.4 Å. BsuNagB is monomeric in solution and as a consequence is active (kcat 28 s-1, Km(app) 0.13 mm) without the need for allosteric activators. A decrease in activity at high substrate concentrations may reflect substrate inhibition (with Ki of ∼4 mm). The structure completes the NagB superfamily structural landscape and thus allows further interrogation of genomic data in terms of the regulation of NagB and the metabolic fate(s) of glucosamine 6-phosphate. NagB (glucosamine-6-phosphate deaminase; 2-amino-2-deoxy-d-glucose-6-phosphate aminohydrolase (ketol-isomerizing); EC 3.5.99.6) performs the isomerization and deamination reactions that transform glucosamine 6-phosphate (GlcN-6-P) 1The abbreviations used are: GlcN-6-P, glucosamine 6-phosphate; F6P, fructose-6-phosphate; 6-PGL, 6-phospho-d-glucono-1,5-lactone lactonohydrolase. 1The abbreviations used are: GlcN-6-P, glucosamine 6-phosphate; F6P, fructose-6-phosphate; 6-PGL, 6-phospho-d-glucono-1,5-lactone lactonohydrolase. to fructose 6-phosphate (F6P), with the concomitant release of ammonia. This reaction follows the first step in the biosynthetic pathway of amino-sugar nucleotides, which is the formation of glucosamine 6-phosphate, and is the final specific step in the pathway of GlcNAc utilization (1Warren L. Gottschalk A. Biosynthesis and Metabolism of Amino Sugars and Amino Sugar-containing Heterosaccharide: Glycoproteins. Elsevier, Amsterdam1972Google Scholar) (Fig. 1). The amino-sugar nucleotides are subsequently used as precursors for the biosynthesis of bacterial cell wall peptidoglycans and teichoic acids.NagB homologs form a superfamily, whose quaternary and genomic organization is particularly interesting. In both Escherichia coli and Bacillus sphaericus (2Alice A.F. Perez-Martinez G. Sanchez-Rivas C. Microbiology. 2003; 149: 1687-1698Crossref PubMed Scopus (16) Google Scholar), nagB and nagA are in the same large operon with the GlcNAc-phosphoenolpyruvate-phosphotransferase system. In most bacilli, including Bacillus subtilis, nagA and nagB exist in a much shorter operon. In B. subtilis, the genes encoding NagA and NagB overlap by 2 bp, and in Bacillus halodurans they overlap by 13 bp, suggesting that these genes are co-transcribed and translated. The complexity of genomic organization is also reflected, to some extent, in the quaternary structures of the enzymes themselves. In humans (3Arreola R. Valderrama B. Morante M.L. Horjales E. FEBS Lett. 2003; 551: 63-70Crossref PubMed Scopus (32) Google Scholar), mice, and E. coli (perhaps the most well-characterized), NagA is a tetramer (4Ferreira F.M. Mendoza-Hernandez G. Calcagno M.L. Minauro F. Delboni L.F. Oliva G. Acta. Crystallogr. Sect. D. 2000; 56: 670-672Crossref PubMed Scopus (7) Google Scholar, 5Souza J.M. Plumbridge J.A. Calcagno M.L. Arch. Biochem. Biophys. 1997; 340: 338-346Crossref PubMed Scopus (28) Google Scholar) and NagB a hexamer regulated by an allosteric mechanism (6Oliva G. Fontes M.R. Garratt R.C. Altamirano M.M. Calcagno M.L. Horjales E. Structure. 1995; 3: 1323-1332Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar), whereas in B. subtilis, a much simpler organization exists: NagA is dimeric (7Vincent F. Yates D. Garman E. Davies G.J. Brannigan J.A. J. Biol. Chem. 2004; 279: 2809-2816Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar) and NagB is a monomer.Quaternary structure and enzyme specificity thus manifest themselves in a NagB superfamily (Pfam PF01182) of three clusters (Fig. 2). The first cluster contains both human and E. coli NagB enzymes. It is amazing that the primary sequence of EcoNagB is more similar to its human counterpart than to its Gram-positive bacterial cousins. This branch of the family is defined by the shared quaternary structure and mode of regulation by the allosteric effector GlcNAc-6-P. The second branch of the family comprises the simpler, monomeric NagB proteins exemplified by BsuNagB (described here). Thus, there is a clear split between the Gram-positive and Gram-negative eubacterial divide, exemplified by the first two major classes, which include galactosamine/glucosamine isomerase/deaminase proteins. The third cluster of related sequences are predicted to be phosphogluconolactonases (6-phospho-d-glucono-1,5-lactone lactonohydrolase (6-PGL); EC 3.1.1.31), which catalyze the reaction 6-phospho-d-glucono-1,5-lactone + H2O to 6-phospho-d-gluconate, the second step in the pentose phosphate pathway (8Berg J.M. Tymoczko J.L. Stryer L. Biochemistry. W. H. Freeman and Co., New York2002Google Scholar). The genes for these enzymes are always adjacent to those encoding an NAD-dependent glucose-6-phosphate dehydrogenase. In some organisms, such as Plasmodium, 6-PGL and glucose-6-phosphate dehydrogenase are fused to give a single, multifunctional enzyme (9Clarke J.L. Scopes D.A. Sodeinde O. Mason P.J. Eur. J. Biochem. 2001; 268: 2013-2019Crossref PubMed Scopus (46) Google Scholar). In bacilli, the annotated glucose-6-phosphate dehydrogenase gene is termed zwf, which is associated with a putative 6-P-gluconate dehydrogenase gene (yqjI), whereas genes encoding 6-PGL are conspicuous by their absence.Fig. 2Phylogenetic tree of NagB-related protein sequences. Selected bacterial protein sequences (see Supplementary Material) were aligned with those of human (in italics) 6-PGL and glucosamine-6-phosphate deaminase (GNP1). Proteins of known three-dimensional structure are underlined. The tree was generated using the protpars program (parsimony method) in the Phylip package, using default parameters and bootstrap. The sequences partition into three clusters, with NagB from B. subtilis in a monophyletic clade from Firmicutes bacteria. The bacterial 6-PGL proteins are mainly from cyanobacteria and actinobacteria, whereas proteobacterial sources predominate in the cluster with Gnp1. Codes are based on SwissProt species designations as follows. Anasp, Anabaena sp.; Bachd, B. halodurans; Bacsu, B. subtilis; Bactn, Bacteroides thetaiotaomicron; Borbu, Borrelia burgdorferi; Cloab, Clostridium acetobutylicum; Clope, C. perfringens; Clote, C. tetani; Cordi, Corynebacterium diptheriae; Ecoli, Escherichia coli; Entfa, Enterococcus faecalis; Fusnn, Fusobacterium nucleatum; Glovi, Gleobacter violaceus; Haein, Haemophilus influenzae; Lisin, Listeria innocua; Myctu, Mycobacterium tuberculosis; Oceih, Oceanobacillus iheyensis; Pasmu, Pasteurella multocida; Pholu, Photorhabdus luminescens; Rhoba, Rhodopseudomonas palustris; Salty, Salmonella typhimurium; Staam, Staphylococcus aureus; Strpn, Streptococcus pnemoniae; Synel, Thermosynechococcus elongatus; Synyl, Synechocystis sp.; Thema, Thermatoga maritima; Thetn, Thermoanaerobacter tengcongensis; Trowt, Tropheryma whipplei; Vibpa, Vibrio parahemeolyticus; Yerpe, Yersinia pestis.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Structural representatives for the oligomeric, allosterically activated NagB are reported, and a Protein Data Bank file for a putative 6-PGL has been deposited. Here we present the expression, purification, kinetic characterization, and the 1.5-Å resolution crystal structure of the "missing link" of the superfamily, the B. subtilis NagB. The three-dimensional structure of this monomeric enzyme has also been solved in complex with the products of the reaction with GlcN-6-P and glucose 6-phosphate. BsuNagB is active without the need for allosteric regulation, but its kinetic activity displays substrate inhibition at high concentrations of GlcN-6-P. BsuNagB completes the structural portfolio and thus allows further interrogation of genomic data in terms of the regulation of NagB and the metabolic fate(s) of glucosamine 6-phosphate.MATERIALS AND METHODSCloning and Protein Production—The nagB gene was amplified from chromosomal DNA of B. subtilis strain IG20 (168 trp-) by the PCR method. Oligonucleotides used to prime the PCR were designed to incorporate suitable restriction endonuclease sites (NdeI and XhoI) for convenient cloning into the T7-promoter-based expression vectors pET28a and pET26b, to yield three separate constructs. These encoded a native protein plus hexahistidine-tagged versions at the N or C terminus. These constructs were designed to test the hypothesis that NagA and NagB could interact; we were unable to demonstrate any such interaction either in vitro or through co-expression. The genes were expressed in Escherichia coli BL21(DE3), a strain with an inducible T7 RNA polymerase gene (Novagen).For protein production, cell cultures were grown in LB medium containing 35 μg/ml kanamycin to an optical density of 0.7 at 600 nm before induction of protein expression by the addition of isopropyl 1-thio-β-d-galactopyranoside to a final concentration of 1 mm. After a further 3 h of growth, the cells were harvested by centrifugation and disrupted by sonication. The lysate was clarified by centrifugation and applied to a Q-Sepharose column resolved using a linear gradient of increasing sodium chloride. NagB-containing fractions were concentrated and further purified by gel filtration (Superdex75 16/60; Amersham Biosciences) in a 20 mm Tris-HCl (pH 8), 200 mm NaCl buffer, with apparent molecular mass estimated by comparison with standard protein markers. Pure fractions of NagB were pooled, washed into 20 mm Tris-HCl (pH 8), and concentrated to 43 mg/ml using a 10-kDa cut-off ultracentrifugation membrane (Vivaspin).Crystallization, Data Collection, and Processing—Crystals of NagB were grown by vapor phase diffusion using the hanging drop method with an equal volume (1 μl) of protein (10 mg/ml) and reservoir solution composed of 20% polyethylene glycol 8000, 0.1 m Tris-HCl (pH 8), and 0.2 m calcium acetate. Substrate soaks were performed by incubating crystals in reservoir solution containing 1 mm glucosamine 6-phosphate or glucose 6-phosphate. A single crystal of NagB was transferred to a solution containing the mother liquor components with 25% polyethylene glycol 550 MME as a cryoprotectant. The crystal was suspended in a film of solution using a rayon loop and flash-cooled to 120 K. The diffraction quality of the crystals was assessed using a home source, after which the crystals were taken to a synchrotron. A single wavelength experiment was conducted on beamline ID29 at the European Synchrotron Radiation Facility (ESRF), Grenoble, at a temperature of 100 K using an ADSC CCD detector. Data were integrated, scaled, and reduced using HKL2000 and SCALEPACK (23Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref Scopus (38368) Google Scholar). The crystals are monoclinic, with unit cell dimensions a = 63, b = 48, c = 72 Å, and β = 91°. This corresponds to a solvent content of 35.9%, assuming 2 molecules per asymmetric unit. All further crystallographic computations were carried out using the CCP4 suite of programs (24Project Collaborative Computational Acta Crystallogr. Sect. D. 1994; 50: 760-763Crossref PubMed Scopus (19707) Google Scholar).Phasing, Model Building, and Refinement—The structure was solved by molecular replacement, using AMoRe (25Navaza J. Saludijan P. Methods Enzymol. 1997; 276: 581-594Crossref PubMed Scopus (368) Google Scholar), with the NagB from Escherichia coli (38% sequence identity with B. subtilis NagB) as a search model (Protein Data Bank code 1dea) (6Oliva G. Fontes M.R. Garratt R.C. Altamirano M.M. Calcagno M.L. Horjales E. Structure. 1995; 3: 1323-1332Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar). The rotation and translation functions gave two solutions, resulting in two molecules in the asymmetric unit. After a rigid body refinement performed in AMoRe, we obtained a correlation coefficient of 48.5% and an R-factor of 45.4%. The electron density map calculated from the model was of sufficient quality to allow tracing of the Cα chain of the molecule using the REFMAC (26Murshudov G.N. Vagin A.A. Dodson E.J. Acta Crystallogr. Sect. D. 1997; 53: 240-255Crossref PubMed Scopus (13779) Google Scholar)/ARP-wARP (27Perrakis A. Morris R. Lamzin V.S. Nat. Struct. Biol. 1999; 6: 458-463Crossref PubMed Scopus (2562) Google Scholar) programs. The side chains were built with the program QUANTA (Accelrys, San Diego, CA) using Xautofit (28Oldfield T.J. Acta Crystallogr. Sect. D. 2001; 57: 82-94Crossref PubMed Scopus (77) Google Scholar), and the model was refined using the CCP4 program REFMAC and QUANTA. The final model contains 3811 nonhydrogen protein atoms with 402 water molecules and two phosphate ions. The crystallographic Rcryst and Rfree values are 15.5 and 21%, respectively (see Table I).Table ICrystal, data, and refinement statisticsCrystal ParametersNative+F6P from GlcN-6-PSpace groupP21P21Cell dimensionsa (Å)63.063.0b (Å)48.048.0c (Å)71.871.8β (degrees)91.091.0Molecules/asymmetric unit22Data qualityWavelength (Å)0.9168 (ID29)0.934 (ID14-1)Resolution of data (Å)29-1.526-1.4Resolution of data (outer shell) (Å)1.55-1.501.45-1.40Unique reflections67,23178,148Rmerge (outer shell)aRmerge = (Σhkl Σi|Ihkl - (Ihkl)|/Σhkl Σi[Ihkl])0.107 (0.498)0.048 (0.299)Mean I/σI (outer shell)14.7 (4.6)15.8 (4.3)Completeness (outer shell) %98.8 (97.7)92.9 (61.6)bThis incompleteness reflects data from the corners of a square detector. The 1.5-Å shell data are 97% completeMultiplicity (outer shell)3.8 (3.5)3.6 (3.2)RefinementProtein atoms38113818Solvent waters402855LigandDisordered sugar-phosphateF6PRcryst0.1550.123Rfree0.2100.165Root mean square deviation 1-2 bonds (Å)0.0190.020Root mean square deviation 1-3 angles (degrees)1.7691.737a Rmerge = (Σhkl Σi|Ihkl - (Ihkl)|/Σhkl Σi[Ihkl])b This incompleteness reflects data from the corners of a square detector. The 1.5-Å shell data are 97% complete Open table in a new tab The structure of the F6P complex was solved using the native structure. A rigid body refinement was performed in REFMAC, maintaining an identical set of cross-validation reflections. The structure was subsequently refined using REFMAC, as above, with ligand target geometry defined using QUANTA. The final model has 3818 protein atoms, 855 waters, and a Fru-6-P molecule in each molecule of the asymmetric unit. The crystallographic R-factor and R-free values are 12.3 and 16.5% respectively.Enzyme Assays—NagB activity was measured in a coupled assay relying on the increase in absorbance at 340 nm due to the conversion of NADP to NADPH. The assays were performed on a GBC Cintra10 spectrophotometer equipped with a thermoequilibrated cellblock. Unless otherwise indicated, the reactions were performed at 22 °C, pH 8, in acrylic cuvettes. The assay mixture contained 20 mm Tris-HCl, pH 8, 5 mm MgCl2, 1 mm NADP, phosphoglucoisomerase (2 μg/ml), and glucose-6-phosphate dehydrogenase (3 μg/ml) in a total volume of 980 μl. Varying concentrations of substrate diluted in 20 mm Tris-HCl, pH 8 (from 0.0312 to 32 mm) were used in experiments for the studies of substrate inhibition kinetics. The reactions were initiated by the addition of 20 μl of NagB to give a final concentration of 0.3 μg/ml in the reaction mixture. Experimental data describing the dependence of NagB activity on substrate concentration were fitted to the substrate inhibition equation (Equation 1, where Vi is the initial rate of the reaction) by a nonlinear regression method using the program GraFit (29Leatherbarrow R.J. GraFit. 5th Ed. Erithacus Software, Ltd., London2001Google Scholar). Vi=Vmax[S]Km+[S]+[S]2/Ki(Eq. 1) RESULTS AND DISCUSSIONOverall Structure—BsuNagB is a monomer of 244 residues in solution, as evidenced by analytical ultracentrifugation, gel filtration, and dynamic light scattering results (data not shown). Crystals of NagB (see "Materials and Methods") contain two molecules in the asymmetric unit. The structure was solved by molecular replacement using a single monomer (from the hexamer) of the R-state E. coli NagB structure (Protein Data Bank code 1dea) (6Oliva G. Fontes M.R. Garratt R.C. Altamirano M.M. Calcagno M.L. Horjales E. Structure. 1995; 3: 1323-1332Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar). The structure of BsuNagB is indeed similar to that of a monomer from EcoNagB, with a root mean square deviation of 0.97 Å over 239 equivalent Cα atoms (calculated with LSQMAN (10Kleywegt G.J. Jones T.A. ESF/CCP4 Newsletter. 1994; 31: 9-14Google Scholar)). The secondary structure elements superimpose, almost identically, apart from the C terminus of EcoNagB, which is 18 residues longer than BsuNagB in the form of an additional α-helix that is involved in allosteric regulation and hexamer formation (Fig. 3).Fig. 3Structure-based sequence alignment. The sequences of B. subtilis NagB, T. maritima 6-PGL, and E. coli NagB are aligned, with the secondary structures of the top and bottom sequences presented. Note the additional α-helix at the C terminus of E. coli NagB and the absence of an α-helix in 6-PGL between strands β7 and β8. More comprehensive sequence alignments are supplied as Supplementary Material. This figure was drawn using ESPript (31Gouet P. Courcelle E. Stuart D.I. Metoz F. Bioinformatics. 1999; 15: 305-308Crossref PubMed Scopus (2505) Google Scholar).View Large Image Figure ViewerDownload Hi-res image Download (PPT)The BsuNagB protein fold is a three-layer (α/β/α) structure, with a main β-sheet of seven parallel β-strands flanked on both sides with α- and 310-helices, Fig. 4A. The central body (residues 1-66, 95-153, and 192-239) of the enzyme is essentially a "Rossmann-like" fold and encloses the active site. Such an arrangement is similar to that of aldose-ketose hydrolases, such as ribose-5-phosphate isomerase (11Hamada K. Ago H. Sugahara M. Nodake Y. Kuramitsu S. Miyano M. J. Biol. Chem. 2003; 278: 49183-49190Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar), and their mode of phosphate binding also appears to be conserved. Residues 158-178 form a helix-loop motif that has been suggested to represent an active site "lid" of NagB (6Oliva G. Fontes M.R. Garratt R.C. Altamirano M.M. Calcagno M.L. Horjales E. Structure. 1995; 3: 1323-1332Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar) and are part of a domain that constitutes a mixed β-sheet of three β-strands surrounded by three α-helices (Fig. 4).Fig. 4The three-dimensional structures of NagB from B. subtilis and E. coli. a, protein schematic diagram of the three-dimensional structure of the fructose-6-phosphate complex of NagB showing the α/β/α sandwich fold. The helices (red) are numbered from α1 to α7, and the β-strands (cyan) are numbered from β1 to β10. This figure is in divergent ("wall-eyed") stereo. b, hexamer of NagB E. coli shown in schematic diagram representation view along the 3-fold axis and at 90° along one of the three 2-fold axes. The structure is rainbow-colored. Both pictures were drawn using PyMOL (32DeLano W.L. PyMOL. DeLano Scientific, San Carlos, CA2002Google Scholar).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Substrate Binding and Catalysis—The architecture of the catalytic site is formed by the loops β2-α2, α5-β8, β6-β7, and β3-β4 and helix α5 (Fig. 4A). After refinement, the "native" structure of BsuNagB exhibited significant additional electron density near the catalytic center. This was modeled as a phosphate group plus two carbons and an oxygen atom, which probably represent the carbon atoms C-6 and C-5 and the oxygen O-5 of the extremity of a sugar phosphate compound. We were not able to build the rest of the molecule, and waters have been placed to match the remaining density. This ligand is probably a disordered sugar-phosphate that has been sequestered from E. coli during protein overexpression.The structure of NagB with GlcN-6-P was subsequently studied through crystal soaking experiments, and the structure was solved at 1.4-Å resolution. The "unbiased" electron density unambiguously shows a molecule of fructose 6-phosphate (F6P), in an extended open chain form, in the catalytic site of each monomer of the asymmetric unit, Fig. 5. Since F6P is the product of the reaction, BsuNagB has been able to perform the isomerization/deamination reaction both in the absence of added co-factors and "in-crystal." This is in marked contrast to EcoNagB, which requires both the allosteric activator GlcNAc-6-P and the subsequent conformational change from the T- to the R-state of the enzyme for activity (6Oliva G. Fontes M.R. Garratt R.C. Altamirano M.M. Calcagno M.L. Horjales E. Structure. 1995; 3: 1323-1332Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar, 12Horjales E. Altamirano M.M. Calcagno M.L. Garratt R.C. Oliva G. Struct. Fold Des. 1999; 7: 527-537Abstract Full Text Full Text PDF Scopus (35) Google Scholar, 13Rudino-Pinera E. Morales-Arrieta S. Rojas-Trejo S.P. Horjales E. Acta Crystallogr. Sect. D. 2002; 58: 10-20Crossref PubMed Scopus (23) Google Scholar). Superposition of the "native" BsuNagB structure with the BsuNagB·F6P complex reveals very similar structures with a root mean square deviation of 0.32 Å for 241 equivalent Cα atoms. This is in contrast with the drastic structural changes that occur in EcoNagB upon binding of allosteric activator. The BsuNagB structure is thus very close to that of the active R conformer of EcoNagB rather than the ligand-free T conformer, which undergoes large structural differences in the lid region (and also some quaternary rotation movement of subunits in the hexameric structure) upon ligand binding. The equivalent "lid" region in BsuNagB does not appear to move upon binding of substrate, although the native structure does contain a partial occupancy of disordered residual ligand and thus may not be a true native structure. In an attempt to obtain a complex of intact substrate, BsuNagB crystals were incubated with glucose 6-phosphate, whose lack of a 2-amino group should render it unreactive or, at least, a poor substrate. The observed electron density corresponded, however, not to glucose 6-phosphate, but to the product F6P, showing that in-crystal glucose 6-phosphate is indeed able to act as a substrate (data not shown).Fig. 5Electron density at the active center of B. subtilis NagB. The reaction product F6P is represented surrounded by the interacting residues of the active center. This figure in divergent ("wall-eyed") stereo was drawn using PyMOL (32DeLano W.L. PyMOL. DeLano Scientific, San Carlos, CA2002Google Scholar) and shows a REFMAC maximum likelihood/σA-weighted 2Fo - Fc electron density at ∼0.53 electrons/Å2. Asp67 is visible in two similar conformations in the electron density maps.View Large Image Figure ViewerDownload Hi-res image Download (PPT)A catalytic mechanism has been proposed for EcoNagB based upon the seminal kinetic observations by Rose and Midelfort (14Midelfort C.F. Rose I.A. Biochemistry. 1977; 16: 1590-1596Crossref PubMed Scopus (48) Google Scholar) and the three-dimensional structure of a competitive inhibitor in the active site (6Oliva G. Fontes M.R. Garratt R.C. Altamirano M.M. Calcagno M.L. Horjales E. Structure. 1995; 3: 1323-1332Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar). The same general reaction scheme may be applied to BsuNagB, since F6P interacts with the equivalent conserved residues as observed for the R-state EcoNagB (Figs. 6 and 7). Most authors believe that the reaction occurs on the open-chain form of the sugars. Midelfort and Rose (14Midelfort C.F. Rose I.A. Biochemistry. 1977; 16: 1590-1596Crossref PubMed Scopus (48) Google Scholar) proposed that the α-anomer of GlcN-6-P is the favored ring form substrate, despite the reaction occurring on the open form of the sugar, since the ring-closed substrate in 4C1 (chair) conformation is then consistent with a cis-enolamine intermediate, and kinetics indicated a preference for the α-anomeric form. Given that the conformational change from a ring-closed to extended ring-open species, as observed in-crystal, is enormous compared with the rotation around a single bond, such an interpretation seems unlikely. One can certainly model a ring form in the active center (although we and others (see Fig. 5 of Ref. 6Oliva G. Fontes M.R. Garratt R.C. Altamirano M.M. Calcagno M.L. Horjales E. Structure. 1995; 3: 1323-1332Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar)) find that a β-anomeric form docks with less steric clashes. His138 is perfectly poised to give general acid assistance to ring opening through protonation of O-5, as also observed on the recent ribose-5-phosphate isomerase structures (15Roos A.K. Burgos E. Ericsson D.J. Salmon L. Mowbray S.L. J. Biol. Chem. 2005; 280: 6416-6422Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar); indeed, it also hydrogen-bonds to the open chain O-5 in the F6P structure (Fig. 6).Fig. 6Schematic diagram of the active center interactions of NagB. Distances of direct hydrogen bonds, in Å, are given.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig. 7Putative reaction mechanism of NagB. The reaction catalyzed by 6-PGL is shown in the inset, and a small section of the fructose 6-phosphate product of NagB (drawn as in Fig. 5) is also included for reference. Formation of the enamine intermediate is followed by attack of water at C-2. It is likely that this passes through the more electrophilic iminium ion with subsequent departure of the ammonia (or the ammonium ion), yielding fructose 6-phosphate given appropriate proton transfer (not shown).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Whatever the mechanism or requirement for ring opening, proton abstraction from C-2 is most likely performed by Asp67 with the C-H σ-bond in plane with the p-orbital of the sugar carbonyl. Formation of the enamine intermediate would then be followed by attack of water at C-2. It is likely that this passes through the more electrophilic iminium ion with subsequent departure of the ammonia (or the ammonium ion) both driving the reaction and thus yielding fructose 6-phosphate (Fig. 7).Enzyme Activity—In the R-state of EcoNagB, two phosphates ions can be observed per monomer, in the catalytic site and at the subunit interface, the latter representing the allosteric site. To reach the ligand-free T-state, the enzyme needs to bind the allosteric activator GlcNAc-6-P. Our attempts to soak the crystals of BsuNagB in GlcNAc-6-P failed to indicate any binding of this compound. The lack of multimers and the absence of an activator-binding site suggest that BsuNagB is not regulated by an allosteric mechanism. Modified Michaelis-Menten kinetic analysis of NagB, including a model for substrate inhibition, generates kcat = 28 ± 6 s-1 and Km(GlcN-6-P) = 0.13 ± 0.02 mm. At high substrate concentrations, BsuNagB displays a decrease in activity in what
Referência(s)