Inhibition of Escherichia coliGlucosamine-6-phosphate Synthase by Reactive Intermediate Analogues

• DOI: 10.1074/jbc.275.1.135
• Publication Date: 2002-07-26T15:04:49Z
• Authors: Stephen L. Bearne; Christian Blouin
• Keywords: Amino sugar; Glutamine amidotransferase

Glucosamine-6-phosphate synthase (GlmS) catalyzes the formation of d-glucosamine 6-phosphate fromd-fructose using l-glutamine as ammonia source. Because N-acetylglucosamine is an essential building block both bacterial cell walls and fungal wall chitin, enzyme a potential target for antibacterial antifungal agents. The most potent carbohydrate-based inhibitor GlmS reported to date 2-amino-2-deoxy-d-glucitol 6-phosphate, analogue putative cis-enolamine intermediate formed during catalysis. interaction series structurally related analogues with described. Although arabinose oxime 5-phosphate identified good competitive inhibition constant equal 1.2 (±0.3) mm, presence amino function at 2-position shown be important inhibition. Comparison binding affinities 2-deoxy-d-glucitol indicates that contributes −4.1 (±0.1) kcal/mol free energy binding. Similarly, comparison 2-deoxy-d-glucose andd-glucosamine −3.0 product Interactions between 2-amino its ligands contribute uniform thecis-enolamine evidenced by similar contribution group ofd-glucosamine respectively. glucosamine-6-phosphate d-fructose (l-glutamine:d-fructose-6-phosphate aminotransferase (GlmS,1 EC 2.6.1.16)) first step in hexosamine biosynthesis, convertingd-fructose (Fru-6-P) intod-glucosamine (GlcN-6-P) glutamine source (Scheme 1) (1Ghosh S. Blumenthal H.J. Davidson E. Roseman J. Biol. Chem. 1960; 235: 1265-1273Abstract Full Text PDF PubMed Google Scholar, 2Badet-Denisot M.-A. René L. Badet B. Bull. Soc. Chim. Fr. 1993; 130: 249-255Google 3Massière F. Badet-Denisot Cell. Mol. Life Sci. 1998; 54: 205-222Crossref Scopus (168) Scholar). GlcN-6-P precursor uridine diphospho-N-acetylglucosamine from which other sugar-containing molecules are derived. One these products,N-acetylglucosamine, constituent peptidoglycan layer chitin. Accordingly, offers agents has attracted interest several research groups (2Badet-Denisot two coupled enzymatic reactions. hydrolysis yield glutamate nascent ammonia, transferred Fru-6-P. second reaction isomerization Fru-6-P ketose aldose, corresponding Heyns rearrangement (4Kort M.J. Adv. Carbohydr. Biochem. 1970; 25: 311-349Crossref (79) 5Golinelli-Pimpaneau Le Goffic Am. 1989; 111: 3029-3034Crossref (45) Like amidotransferases, organized into domains: NH2-terminal amidotransferase domain, glutamine, COOH-terminal (3Massière 6Mei Zalkin H. Bacteriol. 1990; 172: 3512-3514Crossref 7Zalkin Enzymol. 66: 203-309PubMed 8Denisot Arch. Biophys. 1991; 288: 225-230Crossref (42) been studied extensively utilizes cysteine thiol, forms γ-glutamyl thioester reaction. This catalytic role was confirmed conversion alanine site-directed mutagenesis abolished activity In general, amidotransferases inactivated affinity such 6-diazo-5-oxo-l-norleucine 6-chloro-5-oxo-l-norleucine (chloroketone), alkylate residue 9Buchanan J.M. 1973; 39: 91-183PubMed Indeed, many active irreversible inactivators developed contain electrophilic γ-position react irreversibly residue. More recently, attempts develop inhibitors have made hope developing more specificity (10Corizzi V. Commun. 1992; : 189-190Crossref 11Badet-Denisot Leriche C. Massière Bioorg. Med. Lett. 1995; 5: 815-820Crossref (35) 12Bearne S.L. 1996; 271: 3052-3057PubMed 13Leriche Eur. 1997; 245: 418-422Crossref (17) ketose/aldose isomerase proceeds abstraction C1 pro-R hydrogen fructosimine form acis-enolamine reactive that, upon reprotonation Re face C2, gives rise 2) (5Golinelli-Pimpaneau accord this mechanism, co-workers (14Leriche 118: 1797-1798Crossref (29) Scholar) GlmS, absence displays low phosphoglucoisomerase activity. Analogues expected (15Wolfenden R. Acc. Res. 1972; 10-18Crossref (546) 16Wolfenden 1974; 3: 111-207Crossref (87) 17Wolfenden Annu. Rev. Bioeng. 1976; 271-306Crossref (381) 18Wolfenden Frick Page M.I. Williams A. Enzyme Mechanisms. Royal Society Chemistry, London1987: 97-122Google indeed, (GlcNol-6-P) (11Badet-Denisot Identification those structural elements necessary tight part design. paper describes attempt probe requirements enzyme. energetic determined. d-Arabinose, d-arabinose 5-phosphate, 2-amino-2-deoxy-d-glucose (d-glucosamine 6-phosphate), (dGlc-6-P), d-Fru-6-P, d-ribose were purchased Sigma Chemical Company. All chemicals analytical grade used without further purification. NMR spectra (13C, 31P) obtained Bruker AC 250F spectrometer. shifts (δ) 13C 31P relative deuterium lock signal external H3PO4 (85% w/v D2O), Elemental analyses performed Canadian Microanalytical Service Ltd., 2-Deoxy-d-glucitol (dGlcol-6-P) synthesized reduction (dGlc-6-P) sodium borohydride purified ion-exchange chromatography following procedures outlined preparation GlcNol-6-P (12Bearne dGlc-6-P (250 mg, 0.868 mmol) dissolved 10 ml water cooled on ice min. Sodium (0.750 g, 19.83 added solution small portions over 20 During NaBH4 addition, stirred vigorously held ice. After addition complete, 1 h room temperature. Reduction complete indicated inability reduce Fehling's reagent. Undissolved removed filtration, filtrate remaining destroyed dropwise 6 m acetic acid 30 min until final pH approximately 4. (25 ml) allowed come temperature h. then filtered, applied Dowex 50 (H+ form) column (1.5 × 47 cm) eluted water. Fractions containing thin cellulose (99% EtOH, n-BuOH, 0.15m citrate buffer, 4.0; 10:1:6 v/v/v) iron-sulfosalicylic spray reagent (19Clark Switzer R.L. Experimental Biochemistry. 2nd Ed. W. Freeman Co., San Francisco1977: 167Google sensitive phosphates. testing positive phosphate pooled diluted 0.50 volume methanol. solvent rotary evaporation (⩽ 37 °C) freed borate repeated treatment methanol followed removal methylborate ester evaporation. clear syrup taken up (H+form) dGlcol-6-P combined lyophilized yielding hygroscopic glassy powder (0.1953 78% yield). (62.896 MHz, D2O) δ 38.00 (s, C2), 61.36 C1), 69.18 (d, J C, P = 4.77 Hz, C6), 69.43 C3), 72.77 (d,J 7.63 C5), 75.03 C4); (101.26 decoupled 1.93 (s) (t, H, 5.72 Hz). disodium salt converted less monocyclohexylammonium (cyclohexylammonium form). analysis follows. C18H41O8N2P(345.38)Calculated:C41.73%H8.19%N4.06%P8.97%Found:C40.63%H8.12%N3.79%P8.96%ANALYSIS d-Arabinose prepared treating hydroxylamine described Finch Merchant (20Finch P. Z. 1979; 76: 225-232Crossref (5) 21Finch Perkin Trans. I. 1975; 1682-1686Crossref (52) Hockett et al. (22Hockett R.C. 1935; 57: 2265-2268Crossref (9) 23Hockett Maynard C.W. 1939; 61: 2111-2115Crossref (11) hexokinase-catalyzed phosphorylation proton-coupled spectrum enzymatically isolated showed broad triplets centered 5.74 ppm (J 7.3 Hz) 4.90 4.4 These signals assigned phosphorus nuclei H-5a,5b E andZ phosphate, respectively, the31P assignments There also five very minor peaks (≤ 7%) present 3.58, −4.74, −4.92, −9.07, −9.29 possibly due contaminating phosphoenolpyruvate or nucleotides. actual concentration ofd-arabinose studies determined integration 1H spectrum. well resolved doublets 7.59 6.95 arising H-1 Z integrated internal pyrazine standard. ratio toZ 85:15, agreement 83:17 earlier 24Bearne Wolfenden 34: 11515-11520Crossref specific 0.25 unit/mg protein. modified Morgan-Elson procedure, validity assay established previously Inorganic organic assays conducted according Leloir Cardini (25Leloir L.F. C.E. Methods 1957; 840-850Crossref (219) Assays 0.1m potassium 7.5, mm EDTA. follows: 4.6, 9.1, 13.7 mm; 3.1, 6.2, 9.4 18.2, 27.3 dGlcol-6-P, 25.0 50.0 dGlc-6-p, 57.2 171.7 mm. contained (6.7 10−3unit/ml), (15 mm), andd-Fru-6-P concentrations 0.45, 0.89, 1.79, 4.47, 8.94 found relatively insensitive changes ionic strength, therefore no correct strength. Complete Michaelis-Menten plots constructed all d-Fru-6-P given above. Kinetic data analyzed nonlinear regression program Enzymekinetics(1990) Trinity Software. constants triplicate, average value reported. error S.D. Much attention focused utilizing inhibit goal Typically, possess glutamate, believed located domain. effective beenN 3-fumaroyl-l-2,3-diaminopropanoate derivatives (26Chmara Andruszkiewicz Borowski Biochim. Acta. 1985; 870: 357-366Crossref (51) 27Badet Vermoote 1988; 27: 2282-2287Crossref (65) 28Kucharczyk N. Denisot 29: 3668-3676Crossref (40) 29Andruszkiewicz Chmara Milewski Int. Peptide 1986; 449-453Crossref (39) 2-amino-3-((N-halomethyl)amino)propanoate (30Auvin Cochet O. Kucharczyk 19: 143-151Crossref (14) Scholar,31Milewski 1115: 225-229Crossref (15) follow ordered Bi kinetic mechanism (27Badet Scholar), glutamine-like bind inactivate effectively even substrate Scholar,27Badet Recently, there effort identify approach design compounds either transition states intermediates catalysis mechanisms phosphorylated sugar isomerases catalyze interconversion aldose involve proton give enzyme-bound cis-enediol (32Rose I.A. Relat. Areas 43: 491-517PubMed 33Rose Philos. Lond-Biol. 1981; 293: 131-143Crossref (43) Virtually enzymes show metal ions exchange (to solvent) intramolecular transfer ratios greater than (33Rose 34Farber G.K. Glasfeld Tiraby G. Ringe D. Petsko G.A. 28: 7289-7297Crossref (109) As expected, thecis-enediol (18Wolfenden For example, triose-phosphate inhibited 2-phosphoglycolate 2-phosphoglycolohydroxamate (35Wolfenden 9: 3404-3407Crossref (47) 36Collins K.D. 249: 136-142Abstract glucose-6-phosphate 5-phosphoarabinoate (37Chirgwin Noltmann E.A. 250: 7272-7276Abstract arabinose-5-phosphate ribose-5-phosphate epimerase 4-phosphoerythronate (38Woodruff W.W. 254: 5866-5867Abstract 39Bigham E.C. Gragg Hall W.R. Kelsey J.E. Mallory Richardson D.C. Benedict Ray P.H. 1984; 717-726Crossref To extent state rate-determining resembles enediol, may considered analogues. A deaminase (40Midelfort C.F. Rose 1977; 16: 1590-1596Crossref (48) where yields 2). 1, phosphosugar discussed GlcNol-6-P, intermediate, 40Midelfort fact, tightest reversible GlmS. work Escherichia coli makes ligand structures their (K i) Table cases, respect values should accurate estimates dissociation since theK i (19 μm) 34 (±7) μm protection experiments carbonyl functions open chain ribose mimic double bond cis-enolamine. That 1-hydroxyl GlcNol-6-P. Changing stereochemistry C2 causes 4-fold 5-phosphate. observation not unexpected because isomerize carbohydrate substrates generally stereochemical discrimination (41Noltmann Enzymes. 6: 271-354Crossref (112) Scholar).Table ICompetitive enolamine intermediatea Inhibition Bearne 25 independently Scholar).b Arabinose represented straight highlight similarity intermediate. K apparent combination species (cyclic, acyclic, aldehyde, hydrate forms) under "Results Discussion."c syn (E) isomer intermediate; however, equilibrium mixture anti (Z) (15%) andsyn (85%) isomers.d saturating b Discussion." c isomers. d I shows acyclic so evident. However, neutral aqueous solution, only comprise 2.4 0.6% total present, respectively (42Pierce Serianni A.S. Barker 107: 2448-2456Crossref (75) Under conditions, exists β-furanose (81.1%), α-furanose (16.1%), (2.2%) Whether cyclic known. argued need ring opening spontaneous rate (18–21 s−1; Ref. 42Pierce (19–23 s−1). proceed via enolization (43Salas M. Vinuela Sols 1965; 240: 561-568Abstract 44Dyson J.E.D. 1968; 243: 1401-1414Abstract 45Schray K.J. Benkovic S.J. P.A. 248: 2219-2224Abstract mannose-6-phosphate (46Gracy R.W. 5410-5419Abstract (47Oliva Fontes M.R. Garratt Altamirano M.M. Calcagno M.L. Horjales Structure. 1323-1332Abstract (72) prior deprotonation. crystal complexes domain (8Denisot different reported, including Glu-6-P (48Teplyakov Obmolova Polikarpov 1047-1055Abstract (93) (49Teplyakov Protein 1999; 8: 596-602Crossref (77) Based structures, Teplyakov 49Teplyakov proposed uses His-504 general base Thus it possible (57.3%), (40.4%), (2.2%), (⩽0.2%) If aldehyde inhibitory species, adjustment observed reflect would upper limit 17 μm. seems convincing reason adjust manner. contains hydroxyl but lacks function. binds 7-fold It note tested 15% theanti 85% syn(E) form, latter being analogous Adjusting apparentK theE 1.0 still 53-fold Is lack principally missing function? Replacement compound differs substitution nitrogen C1. 4.1 (see below), expectedK 1.3 15-fold might serve estimate enzyme's actualcis-enolamine Corizzi non-isosteric phosphonate poor 2.5 mm). much higher 0.2 slightly work. because, unlike derived oxime, isosteric putativecis-enolamine explanation difference predominately form. Unfortunately, amounts Zforms phosphonate. Despite affinities, nature oximes reflects proposedcis-enolamine interactions always substantial number groups. understanding dissect contributions each measuring change results when one (50Jencks W.P. Proc. Natl. Acad. U. 78: 4046-4050Crossref (848) 51Fersht A.R. Shi J.-P. Knill-Jones Lowe D.M. Wilkinson A.J. Blow Brick Carter Waye M.M.Y. Winter Nature. 314: 235-238Crossref (993) 52Fersht Structure Mechanism Science. New York1999: 340-345Google type removing (51Fersht 53Street I.P. Armstrong C.R. Withers S.G. 6021-6027Crossref (188) When used, k cat/K natural compared calculating effect modification ground state. Alternatively, subsequent interpreted moiety (54Kati W.M. 1591-1593Crossref (58) assess (GlcNol-6-P). 2-Deoxy-d-glucose 2-position, weakly bound 46 ± 7 corresponds (ΔG dGlc-6-P) −1.90 0.09 −4.85 GlcN-6-P) 2The (ΔG) calculated equation ΔG =RTlnK T 310.2 R gas constant. (ΔΔG) ΔΔG −RTln(Ki H/Ki NH2) Ki H andKi NH2 represent lacking function, 0.1 (ΔΔG − 15 2 dGlcol-6-P) −2.59 0.08 −6.70 0.02 GlcNol-6-P) energies participating protein-ligand interactions. −3.4 −6.7 interacting phenylalanyl-tRNA synthetase (55Santi D.V. Danenberg P.V. 1971; 10: 4813-4820Crossref 56Mulivor Rappaport H.P. 123-134Crossref (23) isoleucyl-tRNA (57Holler Rainey Orme Bennett E.L. Calvin 12: 1150-1159Crossref (41) aminoglycoside 3′-phosphotransferases, types Ia IIa, large −6 −11 stabilization −RTln((k m)H/(k m)NH2)) (58Roestamadji Grapsas Mobashery 117: 11060-11069Crossref (69) Thus, same amount product, GlcN-6-P, does analogue, therefore, theuniform (59Albery W.J. Knowles J.R. 15: 5631-5640Crossref (592) 60Albery Angew. Engl. 285-293Crossref (146) Selective must caused another determinant, likely complexed substrates, products, often useful delineating enzyme-ligand (61Lolis 59: 597-630Crossref (Fig. 1). bonded carbonyls Val-399 Ala-602, molecule. molecule γ-carboxylate Glu-488 Removal disrupt three bonds resulting loss 3.0 ligand. ε-amino Lys-603 bonding Therefore, although protein appear going intermediate-enzyme complex product-enzyme concomitant gain single complex, remains roughly unchanged. Once structure whole (glutamine domains) solved, will interesting see if interacts responsible significantly indicating included inhibitors.